EPRI TR-105261 Boiler Tube Failures V1-V3 [PDF]

Zitiervorschau

SINGLE USER LICENSE AGREEMENT M AT E R I A L

THIS IS A LEGALLY BINDING AGREEMENT BETWEEN YOU AND THE ELECTRIC POWER RESEARCH INSTITUTE (EPRI). PLEASE READ IT CAREFULLY BEFORE REMOVING THE WRAPPING MATERIAL.

BY OPENING THIS SEALED BOOK YOU ARE AGREEING TO THE TERMS OF THIS AGREEMENT. IF YOU DO NOT AGREE TO THE TERMS OF THIS AGREEMENT, PROMPTLY RETURN THE UNOPENED BOOK TO EPRI AND THE PURCHASE PRICE WILL BE REFUNDED.

1. GRANT OF LICENSE EPRI grants you the nonexclusive and nontransferable right during the term of this agreement to use this book only for your own benefit and the benefit of your organization. This means that the following may use this book: (I) your company (at any site owned or operated by your company); (II) its subsidiaries or other related entities; and (III) a consultant to your company or related entities, if the consultant has entered into a contract agreeing not to disclose the book outside of its organization or to use the book for its own benefit or the benefit of any party other than your company. This shrink-wrap license agreement is subordinate to the terms of the Master Utility License Agreement between most U.S. EPRI member utilities and EPRI. Any EPRI member utility that does not have a Master Utility License Agreement may get one on request. 2. COPYRIGHT This book including the information contained in it, is owned by EPRI and is protected by United States and international copyright laws. You may not, without the prior written permission of EPRI, reproduce, translate or modify this book, in any form, in whole or in part, or prepare any derivative work based on this book. 3. RESTRICTIONS You may not rent, lease, license, disclose or give this book to any person or organization, or use the information contained in this book, for the benefit of any third party or for any purpose other than as specified above unless such use is with the prior written permission of EPRI. You agree to take all reasonable steps to prevent unauthorized disclosure or use of this book. Except as specified above, this agreement does not grant you any right to patents, copyrights, trade secrets, trade names, trademarks or any other intellectual property, rights or licenses in respect of this book. 4. TERM AND TERMINATION This license and this agreement are effective until terminated. You may terminate them at any time by destroying this book. EPRI has the right to terminate the license and this agreement immediately if you fail to comply with any term or condition of this agreement. Upon any termination you may destroy this book, but all obligations of nondisclosure will remain in effect. 5. DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, NOR ANY PERSON OR ORGANIZATION ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS OR SIMILAR ITEM DISCLOSED IN THIS BOOK, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY’S INTELLECTUAL PROPERTY, OR (III) THAT THIS BOOK IS SUITABLE TO ANY PARTICULAR USER’S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS BOOK OR ANY INFORMATION, APPARATUS, METHOD, PROCESS OR SIMILAR ITEM DISCLOSED IN THIS BOOK.

6. EXPORT The laws and regulations of the United States restrict the export and re-export of any portion of this book, and you agree not to export or re-export this book or any related technical data in any form without the appropriate United States and foreign government approvals. 7. CHOICE OF LAW This agreement will be governed by the laws of the State of California as applied to transactions taking place entirely in California between California residents. 8. INTEGRATION You have read and understand this agreement, and acknowledge that it is the final, complete and exclusive agreement between you and EPRI concerning its subject matter, superseding any prior related understanding or agreement. No waiver, variation or different terms of this agreement will be enforceable against EPRI unless EPRI gives its prior written consent, signed by an officer of EPRI.

E Printed with soy inks on recycled paper (50% recycled fiber, including 10% postconsumer waste) in the United States of America.

Boiler Tube Failures: Theory and Practice Volume 1: Boiler Tube Fundamentals

R. B. Dooley Electric Power Research Institute and W. P. McNaughton Cornice Engineering, Inc.

i

About EPRI Electricty is increasingly recognized as a key to societal progress throughout the world, driving economic prosperity and improving the quality of life. The Electric Power Research Institute delivers the science and technology to make the generation, delivery, and use of electricity affordable, efficient, and environmentally sound. Created by the nation’s electric utilities in 1973, EPRI is one of America’s oldest and largest research consortia, with some 700 members and an annual budget of about $500 million. Linked to a global network of technical specialists, EPRI scientists and engineers develop innovative solutions to the world’s toughest energy problems while expanding opportunities for a dynamic industry. EPRI . POWERING PROGRESS

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THIS BOOK WAS PREPARED BY THE ORGANIZATIONS NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, THE ORGANIZATIONS NAMED BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS BOOK, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY’S INTELLECTUAL PROPERTY, OR (III) THAT THIS BOOK IS SUITABLE TO ANY PARTICULAR USER’S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS BOOK OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS BOOK. ORGANIZATIONS THAT PREPARED THIS BOOK: ELECTRIC POWER RESEARCH INSTITUTE CORNICE ENGINEERING, INC.

This book is EPRI Licensed Material and contains a single-user, shrink-wrapped license.

ISBN 0-8033-5058-9

ORDERING INFORMATION Requests for copies of this book should be directed to the EPRI Distribution Center, 207 Coggins Drive, P.O. Box 23205, Pleasant Hill, CA 94523, (510) 934-4212. Electric Power Research Institute and EPRI are registered service marks of Electric Power Research Institute, Inc. Copyright © 1996 Electric Power Research Institute, Inc. All rights reserved.

ii

Preface

Boiler tube failures (BTF) have been the number one availability problem for utilities with fossil plants for as long as reliable statistics have been kept in individual utilities and by nations. The majority of BTF have been repeat failures, indicating that the return to service of a unit has classically been more important than understanding the mechanism and root cause of each BTF. Failures have emanated from poor initial design, from poor operation and maintenance, harsh fireside and cycle chemistry environments, and lack of proper management support. Sometimes it’s amazing to consider that some tubes do last over 200,000 hours without failure. The aim of this book is to provide the guidance necessary to accomplish this for the majority of tubes. Over the last twenty years, so many people have influenced my thinking on this very diverse topic and it is appropriate to mention some of them in somewhat chronological order. My first serious interface with BTF was at Ontario Hydro in the early seventies, and it was Syd Featherby who encouraged the initial coordinated attack; the first part of this was to develop a BTF Reporting System, which was fully supported by the upper management, next was to prioritize where the maximum effort should be placed in terms of determining the mechanisms of each BTF and then providing solutions which overcame the root causes. This basic credo of “understanding the mechanism, understanding the root cause, and then providing permanent solutions” has permeated all my BTF efforts ever since. Three other people at Ontario Hydro were also key in developing and supporting this overall approach: Duncan Sidey, Graham Stephenson, and Jim Westwood. Because the Ontario Hydro approach was successful, the Canadian Electric Association supported the implementation of a similar BTF Reduction System in all the Canadian Utilities in 1979; this was preceded by the production of the first compilation of failure mechanisms with Jim Westwood, and resulted in the reporting of the statistics from a much larger data base than one individual utility. This momentum was carried forward into the development of the EPRI BTF projects with the initial BTF Manual and then a BTF Correction, Prevention, and Control Demonstration Program with 16 host utilities, that improved the availability loss due to BTF markedly from 1986 to the present. Mention here must be given to John Dimmer and Gerry Lamping who helped develop the overall coordinated, companywide approach to BTF, to Otakar Jonas who assisted with integrating the optimum cycle chemistry, and to Ron Niebo (NERC) who incorporated the BTF mechanism categories into the NERC/GADS reporting system. Over the last 10 years, tremendous international support has also been available and assisted me in developing solutions to most of the known BTF and in demonstrating the overall BTF approach. Particular mention should be made of some of these people: Jim Davison (PowerGen in England), Dave Barnett (Pacific Power in Australia), Allan Ellery and Peter Ford (State Electricity Commission, Victoria in Australia), Co van Liere (KEMA in The Netherlands), Yuri Shtromberg (ORGRES in Russia), and Yuri Hoffman (Sverdlovenergo, Russia).

iii

EPRI has also conducted a number of BTF Root Cause projects and has held three BTF Conferences, which have consolidated our thinking. Two important publications have also contributed to our understanding of BTF: David French’s Metallurgical Failures in Fossil Fired Boilers, and the NALCO Guide to Boiler Tube Failures. The recent Boiler Tube Failure Metallurgical Guide, developed by Steve Paterson and his colleagues at Aptech Engineering Services, has provided the distinct metallurgical differences between failure mechanisms. Vis Viswanathan, my colleague at EPRI, has dedicated his professional career to the better understanding of remaining life techniques, and we have used this unabashedly. John Stringer, my boss at EPRI suggested the development of this book and has enthusiastically supported all our efforts over the last 18 months. This book represents our attempt to bring together the information on all the mechanisms in a form which separates the theory and the prevention. Clearly the latter should be most useful to the operating engineers responsible for BTF within a utility, while the former provides the necessary background knowledge of all failure types. The BTF Reduction Programs have been very successful in reducing the availability loss due to BTF over the last 15 years by coordinated management approaches. We envision that this compilation of technical aspects should finally remove BTF as the number one fossil utility problem.

Barry Dooley

Palo Alto, California June, 1995

iv

Acknowledgments

This three-volume work is a compilation of what is currently known about boiler tube failures in fossil-fueled power plants, fluidized-bed combustion systems and wasteto-energy boilers. It is an integration of the work performed by literally hundreds of researchers over the past twenty years and the authors have drawn extensively from that work. We have had the rare privilege of working with an outstanding group of experts and consultants worldwide who have provided review, comment, supporting documentation, illustrations and figures for this book. We would like to acknowledge the following key contributors in those regards: Individual M. Ball W. Bakker D.A. Barnett J. Blough A. Bursik P. Daniel J. Davison X. Du L.B. Dufor T. Flatley D. French F. Gabrielli T. Healy M.B. Henry J.J. Hickey Y. Hoffman P. James O. Jonas D. Kalmanovitch S. Kihara D. Lopez Lopez L. McQueen E. Metcalfe S. Paterson F. Pocock R. H. Richman G.G. Stephenson Y. Shtromberg H. Takaku E. Tolksdorf F.H. van Zyl R. Viswanathan J. Westwood I.G. Wright K. Woolhouse

Organization Consultant EPRI Pacific Power Foster Wheeler Consultant Babcock & Wilcox PowerGen China Light & Power KEMA PowerGen David N. French, Inc. ABB/CE ESB Ireland AUSTA Electric ESB Ireland Sverdlovenergo PowerGen Jonas, Inc. Riley Stoker IHI IIE Ontario Hydro National Power Aptech Engineering Services, Inc. Consultant Daedalus Associates, Inc. Ontario Hydro ORGRES CRIEPI VGB ESKOM EPRI Ontario Hydro ORNL FCA

Country England U.S.A Australia U.S.A. Germany U.S.A. England Hong Kong The Netherlands England U.S.A. U.S.A. Ireland Australia Ireland Russia England U.S.A. U.S.A. Japan Mexico Canada England U.S.A. U.S.A. U.S.A. Canada Russia Japan Germany South Africa U.S.A. Canada U.S.A. Australia

All the figures were drawn by Marilyn Winans of EPRI’s Visual Communications Group.

v

vi

Table of Contents Volume 1: Boiler Tube Fundamentals

Chapter

Page

Acknowledgments

iii

Preface

iv

Table of Contents

vi 1-1 1-1 1-2 1-3

1.7 1.8

Introduction and Background Introductory Comments Objectives of this Book Organization of this Book and How to Use It Introduction to the Water-Steam Cycle and Primary Components in Conventional Fossil-Fuel Power Plants Historical Developments in the Identification, Correction, and Prevention of BTF Recent Developments in the Identification, Correction and Prevention of BTF Today’s Situation and Challenges that Remain References

1-19 1-20 1-22

2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8

The Boiler Tube Operating Environment and Its Breakdown Introduction Basic Function and General Design Considerations The Reaction of Iron and Water/Steam: Oxide Formation Oxide Development and Breakdown in Water-Touched Tubes Overview of Thermal-Hydraulic Regimes and Waterside BTF Oxide Development and Breakdown in Steam-Touched Tubes The Combustion Process, Coal Effects and Fireside BTF References

2-1 2-1 2-2 2-6 2-8 2-12 2-14 2-21 2-25

3 3.1 3.2 3.3 3.4 3.5 3.6

Cycle Chemistry and Boiler Tube Failures Introduction and Significance of the Challenge Boiler Water Treatment Feedwater Treatment Developing Unit-Specific Guidelines Instrumentation and Monitoring for Boiler Water References

3-1 3-1 3-3 3-8 3-12 3-14 3-15

4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8

The Effects of Unit and Boiler Operation and Maintenance on BTF 4-1 Introduction and Background 4-1 Chemical Cleaning of Waterwalls 4-1 Chemical Cleaning of Superheaters/Reheaters 4-5 Chemical Cleaning in FBC Units 4-8 Start-Up, Shutdown, Cycling, and Unit Transients 4-8 Lay-Up 4-9 Commissioning 4-9 References 4-12

1 1.1 1.2 1.3 1.4 1.5 1.6

1-16 1-16

vii

Table of Contents Volume 1: Boiler Tube Fundamentals (continued)

Chapter 5 5.1 5.2 5.3 5.4

Page Company-Wide Programs for the Correction, Prevention and Control of Boiler Tube Failures Introduction Formalizing a Program for Correction, Prevention and Control Does It Work? Results from Field Application References

6 6.1

Metallurgical Analysis The Role of Metallurgical Analysis for Analyzing Boiler Tube Failures 6.2-6.19 Step-by-Step Guide to Metallurgical Analysis 6.20 References 7

6-1 6-1 6-3 6-10

7.10 7.11

Distinguishing Features of Some Mechanisms with Similar Appearances Introduction Waterside Underdeposit Corrosion Mechanisms: Hydrogen Damage, Caustic Gouging, and Acid Phosphate Corrosion Water-Touched Tubing: Short-Term Overheating (Three Grades) Water-Touched Tubing: Corrosion Fatigue Versus OD-Initiated Mechanical Fatigue Failure Mechanisms in Economizer Inlet Header Tubes: Thermal Fatigue, Erosion-Corrosion, and Flexibility-Induced Cracking SH Tubing: Long-Term Overheating (Creep) Versus Fireside Corrosion SH/RH Tubing: Flyash Erosion Versus Sootblower Erosion SH/RH Tubing: Graphitization and Long-Term Overheating (Creep) SH/RH Tubing: Intergranular Stress Corrosion Cracking (IGSCC), Intergranular Creep, and Intergranular Corrosion Secondary (Steam Impingement) Versus Primary Failures References

7-10 7-11 7-11

8 8.1 8.2 8.3 8.4

Boiler Tube Remaining Life Assessment Introduction Assessment for Tubes Operating in the Creep Regime Assessment for Tubes Operating in the Non-Creep Regime References

8-1 8-1 8-1 8-7 8-8

9

Determining the Extent of Macroscopic Damage: Overview of Inspection Methods, Monitoring, and Sampling Introduction Codes and Standards Ultrasonic Testing (UT) Other Standard Inspection Methods Monitoring Temperatures Monitoring Heat Flux Monitoring Displacements and Strains Sampling Hydrostatic Testing References

9-1 9-1 9-3 9-3 9-7 9-8 9-9 9-10 9-10 9-10 9-11

7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9

9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 9.10

viii

5-1 5-1 5-1 5-3 5-6

7-1 7-1 7-1 7-5 7-6 7-6 7-6 7-9 7-9

Table of Contents Volume 1: Boiler Tube Fundamentals (continued)

Chapter

Page

10 10.1 10.2 10.3 10.4 10.5 10.6

Determining the Extent of Microstructural Damage Introduction Microstructural Changes in Ferritic Materials Microstructural Changes in Austenitic Stainless Steels Assessment of Creep Damage in Boiler Tube Materials Post-Exposure Testing References

10-1 10-1 10-2 10-4 10-5 10-6 10-8

11 11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8

Repair and Replacement of Boiler Tubes Introduction General Strategies for Damaged Tubes Pre-Repair: Confirm Materials to be Repaired Applicable Codes for Weld Repairs Specific Repair Procedures Documentation Welding Co-Extruded Tubing References

11-1 11-1 11-1 11-3 11-3 11-4 11-7 11-7 11-8

Index

I-1

ix

x

Chapter 1 • Volume 1

HP turbine Feed IP turbine Attemperation

LP turbine Condenser Makeup

Deaerator Boiler HP heaters Condensate polisher Impurity ingress Corrosion Deposition

Feed

Introduction and Background

1.1 Introductory Comments It is extraordinary that relatively simple materials can be designed and constructed to function so effectively as boiler tubes under high-temperature and high-pressure conditions, subject to potential degradation by a variety of mechanical and thermal stresses, and with the potential for environmental attack on both the fluid- and fire-side. If there are no breakdowns from the original design conditions, watertouched boiler tubes (waterwalls, economizers, etc.) are designed for, and should have, essentially infinite life. The case for steam-touched tubes, such as in the superheater (SH) and reheater (RH) sections of modern boilers, is somewhat different because of the inevitability of creep-limited lifetime; although lifetimes well in excess of 200,000 operating hours are achievable. Unfortunately, boiler tube failures (BTF) remain a significant and pervasive problem in the electric power industry. Historically, they have been a primary contributor to lost availability in fossil-fired power plants, ranking as the largest equipment problem during the thirty years that

reliability statistics have been kept in North America.1 The most recently available statistics indicate that BTF are responsible for between 2-3% of lost availability of U.S. fossil-fired power plants, representing lost power generation that has been estimated to be in excess of $1 billion per year. Despite a concerted effort by many organizations to prevent BTF, and with an improvement of more than 1% in availability loss, still more than 30,000 failures have occurred in the last decade. Failures have occurred in all boiler areas: economizers, waterwalls, superheaters, and reheaters. More than 80% of all BTF force a shutdown, and a typical outage lasting three days can cost a utility $1,000,000 for replacement power. The BTF mechanisms representing the leading causes of availability loss are shown in Table 1-1. Although these results come from a program involving 16 U.S. utilities, the ranking is believed to be similar for the whole of the United States and similar rankings have been recorded in many other countries worldwide including Canada2, the United Kingdom3, Australia4, and Russia5.

Volume 1: Boiler Tube Fundamentals

1-1

Table 1-1 Boiler Tube Failure Mechanisms Representing The Largest Losses of Availability for U.S. Units in Order of Importance Corrosion Fatigue Flyash Erosion Hydrogen Damage Long-term Overheating (Creep) Short-term Overheating Sootblower Erosion Waterwall Fireside Corrosion Falling Slag Erosion For period 1986-1992. Adapted from NERC data.

The pervasive nature of the problem can be seen in the results of a boiler tube failure survey conducted in 1991. Table 1-2 shows the percentage of utilities that had experienced boiler tube failures by some of the key failure mechanisms.

1.2 Objectives of this Book Over the past ten years much technical work has been undertaken to understand boiler tube failures. The primary objective of this book is to provide the most recent knowledge about how to identify boiler tube failure mechanisms, determine their root cause, and how to apply immediate solutions and longer-term strategies to prevent their reoccurrence. Additional objectives are: • To provide direct, easy-to-follow actions to be taken if a boiler tube failure has occurred and, perhaps as important, actions to be taken if a precursor has occurred in a unit that might lead to a future boiler tube failure.

Table 1-2 Results of 1991 BTF Survey Have your units experienced: Boiler Tube Failure Mechanism

Yes

No

Corrosion Fatigue High Temperature Creep Fly Ash Erosion Dissimilar Metal Welds Hydrogen Damage Supercritical Waterwall Cracking

84% 80% 73% 52% 41% 34%

5% 9% 16% 36% 48% 27%

Source: Conference Questionnaire, Proceedings: International Conference on Boiler Tube Failures in Fossil Plants, November 5-7, 1991.6

• To provide sufficient information so that a company-wide BTF correction, prevention, and control program can be established. This compilation includes all known boiler tube failure mechanisms. It is never possible to anticipate what future concerns might arise, but a final objective of the work reported here is to provide enough information about how the breakdown situations in boiler tubes develop to allow rational approaches to be formulated for the analysis of as yet unknown challenges.

1.3 Organization of this Book and How to Use It This book is organized in three Volumes. Volume 1 provides information that is applicable to many individual mechanisms. By placing this generic information in Volume 1, the stage is set for the detailed discussion of individual mechanisms found in Volumes 2 and 3

• To provide sufficient background information, so that the reader, if interested, can understand why the prescriptions are made.

Volume 2 is focused exclusively on BTF mechanisms in water-touched tubing of conventional fossil-fuel power plants, that is in waterwalls and economizers.

• To provide guidance about the interactions between boiler tubes and their failures with overall unit health and operation practices.

Volume 3 covers mechanisms that affect superheater/reheater tubing in conventional units, along with three

1-2

Introduction and Background

mechanisms that affect both waterand steam-touched tubing (maintenance damage, material flaws, and welding flaws) and chapters on BTF mechanisms in FBC units (bubbling bed and circulating bed) and wasteto-energy units. 1.3.1 Volume 1: Boiler tube fundamentals and "top-down" implications of BTF. Volume 1 begins with a look at the significance of boiler tube failures and a historical perspective on the understanding of them. This is followed by discussions of importance to the understanding and correction of BTF by all mechanisms including: • Why do BTF arise? (Chapter 2) • How are unit cycle chemistry, operation and maintenance tied to boiler tube failures? (Chapters 3 and 4) • How can utility-wide programs for BTF prevention and control be set up? (Chapter 5) • What are the basics of boiler tube metallurgical investigations? (Chapter 6), • How can mechanisms that appear similar be distinguished? (Chapter 7)

• What methods are available to conduct remaining life assessments of boiler tubes? (Chapter 8) • How can the extent of microscopic or macroscopic damage be determined? (Chapters 9 and 10) • What are repair and replacement methods for boiler tubes? (Chapter 11). 1.3.2 Organization and content of volumes 2 and 3 Volumes 2 and 3 provide detailed discussions of individual BTF mechanisms; Table 1-3 provides an index to those chapters. With only a few exceptions, mostly in the case of those mechanisms which occur relatively infrequently, or for which mechanism, cause and required actions are obvious, each chapter is organized in the same manner. The first half of the chapter provides Theory and Background information including: • 1.0 Features of Failure and Typical Locations • 2.0 Mechanisms of Failure • 3.0 Possible Root Causes and Actions to Confirm • 4.0 Determining the Extent of Damage • 5.0 Background to Repairs, Immediate Solutions and Actions • 6.0 Background to Long-Term Actions and the Prevention of Repeat Failures • 7.0 Case Studies (if any) A key part of each Theory and Background section is the development of a Table that shows the possible root causes, actions to confirm, immediate actions/solutions and long-term actions to prevent repeat failures. It is important that the root cause of a particular damage mechanism be clearly identified so that the proper short- and longterm actions can be initiated. To fail in any of these steps, is to open the door to probable repeat failures.

Table 1-3 Boiler Tube Failure Mechanisms in Water-Touched Tubes of Conventionally-Fueled Power Plants Mechanism Chapter of Volume 2 Corrosion fatigue 13 Flyash erosion 14 Hydrogen damage 15 Acid phosphate corrosion 16 Caustic gouging 17 Fireside corrosion in coal-fired units 18 Thermal fatigue in supercritical waterwalls 19 Thermal fatigue of economizer inlet headers 20 Erosion corrosion (economizer inlet headers) 21 Sootblower erosion 22 Short-term overheating 23 Low-temperature creep 24 Chemical cleaning damage 25 Fatigue in water cooled circuits 26 Pitting in water-cooled tubes 27 Coal particle erosion 28 Falling slag damage 29 Acid dewpoint corrosion 30 Boiler Tube Failure Mechanisms in Steam-Touched Tubes Mechanism Long-term overheating/creep Fireside corrosion in coal-fired units Fireside corrosion in oil-fired units Dissimilar metal welds Short term overheating Stress corrosion cracking SH/RH sootblower erosion Fatigue in steam-touched tubes Rubbing tubes/fretting Pitting (RH loops) Graphitization SH/RH chemical cleaning

Chapter of Volume 3 32 33 34 35 36 37 38 39 40 41 42 43

Mechanisms Affecting both Water-Touched and Steam-Touched Tubes Mechanism Maintenance damage Material flaws Welding flaws

Chapter of Volume 3 44 45 46

BTF in Non-conventionally-Fired Units Mechanism Bubbling fluidized bed combustion units Circulating fluidized bed combustion units Water-touched tubes of MSW/RDF units

Chapter of Volume 3 47 48 49

Volume 1: Boiler Tube Fundamentals

1-3

For many boiler tube failure mechanisms, choices for preventing a reoccurrence of the problem are limited to only one or two options that directly address the underlying root cause. For some mechanisms, fireside corrosion is a notable example; the optimal choice of a long-term strategy may be as much an economic decision as one driven by engineering considerations. The importance of economic evaluation when seeking long-term solutions to BTF cannot be over-emphasized. The second half of each chapter contains Actions to be followed by the investigator or BTF team if a boiler tube failure has occurred and a particular mechanism is suspected, or if a unit precursor has occurred that might lead to a future BTF by this mechanism. The Actions are numbered in a manner consistent with the Theory and Background section. That is, Action 2 corresponds to Section 2.0 of the Theory and Background which provides additional information about the mechanism, why these specified actions are to be taken, and how the mechanism develops. 1.3.3 Using volumes 2 and 3. Figure 1-1 provides a flowchart for the use of this book. This figure has also been reproduced in the introductory chapters in Volume 2 and 3 for reference. As shown, three avenues are open to the investigator or BTF team depending upon the status of the BTF event: • A: BTF with mechanism unknown. If a BTF has occurred for the first time or a number of repeat failures have occurred, and the mechanism is not known, then the charts of typical appearance and location, Tables 1-4 and 1-5, should be consulted. These can provide a first suggestion toward a specific chapter in Volumes 2 and 3 where confirmation can occur.

1-4

Introduction and Background

• B: BTF with known mechanism. If the BTF Team has knowledge from past failures that a particular mechanism is the likely cause, then Table 1-3, the index to all the mechanisms and their location in Volumes 2 and 3, can be used to go directly to the appropriate chapter. • C: Anticipating future BTF. The BTF Team should continually anticipate possible failures by reviewing key unit/boiler operating events that can lead to future BTF. Table 1-6, discussed in detail in the next section has been compiled to help in this regard. Figure 1-1 shows the actions which will occur once a specific mechanism has tentatively been chosen. All such actions, built into each individual chapter on a mechanism, are structured to: confirm that the postulated mechanism is appropriate, determine the underlying cause, determine the extent of damage, implement immediate solutions and actions, implement long-term actions to prevent repeat failures, and evaluate any ramifications to the rest of the unit. 1.3.4 Unit "precursors" and BTF - an anticipatory approach. Path C may seem initially difficult for the BTF Team to execute. There are several reasons for this. It is anticipatory. That is, without a BTF having occurred, it may be initially difficult to get the resources necessary to evaluate whether a unit event is cause for concern for future boiler tubing failures. However, the process is not unlike routine inspection of components; it should only take one identification of an incipient failure to justify the cost-effectiveness of the practice to even the most cost-conscious management.

Further, the number of unit events that can be cause for concern can seem overwhelming because of the interconnected nature of unit operation, maintenance, equipment status, cycle chemistry, and boiler tube concerns. However, by stepping through Table 1-6, along with any modifications introduced by the BTF Team based on unit experience, the BTF Team or investigator may identify potential BTF mechanisms for additional evaluation. Table 1-6 is organized as a series of "unit precursors". These are events or conditions that could have been identified during some aspect of inspecting, operating, maintaining, repairing, or fueling the unit, which, although no BTF has occurred, should act as a signal that the condition identified should be cause for review of the potential for a future BTF. In other words, this Table allows the reader to get to the discussion of a specific mechanism to anticipate future problems. The BTF Team or investigator may find that the best way to implement Table 1-6 is to work through each precursor and query: "Has this precursor occurred in our utility/unit?", or "Have we taken this action recently?" If the answer to either is "yes", then a review of the mechanism(s) indicated in the final column may be needed. In compiling Table 1-6, an attempt has been made to limit the "precursor" list to those which: (i) can be easily identified, (ii) are important observations and will be useful for indicating a potential BTF problem, (iii) are not direct indications of boiler tube damage (an inspection that finds cracks at the toe of a tube/attachment weld would be a direct indicator of a BTF), and (iv) are reasonably likely to lead to a BTF based on past historical evidence.

A:

B:

C:

BTF Mechanism Unknown

BTF Known Mechanism (Table 1-3)

Anticipating Future BTF (Table 1-6)

Compare Macroscopic Appearance to Table 1-4 (Water-touched) or Table 1-5 (Steam-touched) Tubes to identify candidate(s)

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes. Follow actions

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes.

Action 1a: Perform Screening Analysis: Is it possible that this boiler tube failure is caused by this mechanism?

Action 1b: Screening Analysis: ¥ Review precursor list in mechanism chapter ¥ Remove tube sample to determine extent of damage

No

Yes Action 2: Determine (confirm) mechanism

Yes

Are BTF likely to occur in the future by this mechanism?

No Action 3: Determine root cause

Action 4: Determine extent of damage or affected areas

Action 5: Implement repairs, immediate solutions and actions

Action 6: Implement long-term solutions to prevent repeat failures Action 7: Determine possible ramifications/ancillary unit problems

Figure 1-1. Flowchart of actions for identifying, evaluating and anticipating boiler tube failures.

Volume 1: Boiler Tube Fundamentals

1-5

Table 1-4 Screening Table for Water-Touched Boiler Tube Failures Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 2 (or 3 as noted)

Thick-Edged Fracture Surface Thick-edged (pinhole leak also possible)

Multiple, transgranular cracks that initiate on the inside of the tube.

Near attachments, particularly where high restraint stresses can develop.

Corrosion Fatigue

13

Thick-edged, leak or window blowout

Internal damage: gouging, wall thinning; tube deposits.

High heat flux areas; hot side of tube; horizontal or inclined tubing; pad welds; locations with local flow disruptions such as upstream of weld, backing ring or other discontinuities.

Hydrogen Damage

15

Thick-edged

Multiple, parallel cracks on the outside tube surface or on membrane; sharp, V-shaped oxide coated cracks; wall thinning from external surface when found with fireside corrosion.

Maximum heat flux locations; fireside or waterwall tubing or membranes between tubes.

Supercritical Waterwall Cracking

19

Thick-edged, leak or crack.

First sign as pin-hole leak at toe of stub weld; multiple, longitudinal cracks; bore hole cracking.

Economizer inlet header stub tubes nearest the feedwater inlet.

Thermal Fatigue

20

Thick-edged

Outside surface initiated, intergranular crack growth with evidence of grain boundary creep cavitation and creep voids.

Predominant in tube bends, particularly at intrados on outside surface, and other locations subject to high residual, forming, or service stresses.

LowTemperature Creep Cracking

24

Thick-edged

Transgranular cracking, OD-initiated and associated with tubing (at tube bends longitudinal or attachments - transverse) or headers (particularly at the ends).

Near attachments, particularly solid or jammed sliding attachments; at bends in tubing.

Fatigue

26

Thin-Edged Fracture Surface Thin-edged, longitudinal, "cod- or "fish-mouth"

Polishing of tube outside surface; very localized damage, wastage flats.

Near side and rear walls; near economizer banks; near plugged or fouled passages; where previous baffles have been installed.

Flyash Erosion

14

Thin-edged, leak or split

Internal damage: gouging, wall thinning; tube deposits.

As for hydrogen damage.

Acid Phosphate Corrosion

16

Thin-edged, leak or split

Internal damage: gouging, wall thinning; tube deposits.

As for hydrogen damage.

Caustic Gouging

17

Thin-edged, long "fish-mouth"

External wastage; probably affecting a number of tubes; maximum wastage at crown facing flame (maybe flame impingement); damage extending in 120° arc around tube; hard deposits on tube outside surface.

Areas with locally substoichiometric environment; side and rear walls near burners; highest heat flux areas.

Fireside Corrosion (coal-fired units)

18

1-6

Introduction and Background

Table 1-4 Screening Table for Water-Touched Boiler Tube Failures (continued) Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 2 (or 3 as noted)

Thin-Edged Fracture Surface (continued) Thin-edged rupture

Erosion, wall thinning from inside; "orange peel" appearance.

Economizer inlet header stub tubes nearest to point of feedwater inlet.

ErosionCorrosion

21

Thin-edged, "fishmouth"

Wastage flats on tube external surface at 45° around tube from sootblower direction, little or no ash.

Circular pattern around wall blowers.

Sootblower Erosion

22

Generally thinedged

Often shows signs of tube bulging or "fish-mouth": appearance; real keys will be transformation products in microstructure. May also be thick-edged under certain circumstances.

Highest heat flux locations above locations such as: the site of a tube or orifice blockage, or in horizontal tubing where a downcomer steam "slug" can occur.

Short-Term Overheating

23

Thin-edged

External wastage, little or no ash; location should be key.

Tubes near replaceable wear liners in cyclone burners; throat or quarl region of burners.

Coal Particle Erosion

28

Thin-edged

External erosion or mechanical impact damage features.

Sloping wall tubes and/or ash hopper near bottom.

Falling Slag Damage

29

Thin-edged

External, thinned or missing external oxide; generally in economizer.

Low temperature areas of economizer.

Acid Dewpoint Corrosion

30

Internal tube surface damage.

Locations where boiler water can stagnate during unit shutdown (pitting).

Chemical Cleaning Damage or Pitting

25 or 27

Maintenance Damage

Chap. 44, Volume 3

Materials Flaws

Chap. 45, Volume 3

Welding Flaws

Chap. 46, Volume 3

Pinhole Damage Pinhole

Various Other Damage Types Depends on underlying cause.

Usually obvious from type of damage and correspondence to past maintenance activity.

Depends on defect.

Usually thickedged.

Care required to separate weld defects from another problem located at a weld.

Note: This table is based on simple, macroscopic features of failure and should be used as a guide to a particular chapter for further analysis. The more detailed discussions starting with Actions can then be used for identification and confirmation of the actual mechanism.

Volume 1: Boiler Tube Fundamentals

1-7

Table 1-5 Screening Table for Steam-Touched Boiler Tube Failures Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 3 (or 2 as noted)

Thick-Edged Fracture Surface Thick-edged

Outside surface initiated, intergranular crack growth with evidence of grain boundary creep cavitation and creep voids.

Predominant in lower temperature regions in tube bends, particularly at intrados on outside surface, and other locations subject to high residual, forming, or service stresses.

LowTemperature Creep Cracking

Chap. 24 Volume 2

Thick-edged

Internal thick scales, may be accompanied by external wastage at 10 o'clock and 2 o'clock positions; generally longitudinal (axial) orientation; damage on heated side of tube.

Highest temperature locations: near material transitions, where there is a variation in gastouched length, in or just beyond cavities, in the final leg of tubing just prior to the outlet header.

Long-Term Overheating (Creep)

32

Thick-edged, leak

Usually fusion line cracking on low alloy side of weld, circumferential orientation.

At dissimilar metal welds.

Dissimilar Metal Weld Failure

35

Thick-edged (may manifest as a pinhole)

Cracking is transgranular or intergranular usually with significant branching; initiation can be at ID (most common) or on OD, circumferential or longitudinal orientation.

Bends and straight tubing with low spots; high stress locations are particularly susceptible at bends, welds, tube attachments, supports or spacers.

Stress Corrosion Cracking

37

Thick-edged

Transgranular cracking, OD-initiated and associated with tubing (at tube bends or attachments) or headers (particularly at the ends).

Tubing-related failures associated with attachments or bends in tubing; header-related generally at ends of header.

Fatigue

39

Thick-edged, leak

Most commonly in HAZ of C or C-Mo steel tubes; key is microstructure appearance of graphite particles or nodules.

Adjacent to weld fusion line at heat affected zone most common.

Graphitization

42

Thin-Edged Fracture Surface Thin-edged (unless creep-assisted)

External polishing of tube surface; very localized damage.

Most prominent in backpass regions; bends near to walls.

Flyash Erosion

Chap. 14 Volume 2

Thin-edged

External damage; wastage at 10 and 2 o'clock (flue gas at 12 o'clock); longitudinal cracking; perhaps "alligator hide" appearance; real key to identification will be the presence of low-melting point ash in external deposits.

Highest temperature tubes: leading tubes, near transitions, tubes out of alignment, tubes around radiant cavities.

Fireside Corrosion (coalfired units and oil-fired units)

33 (Coal-fired units)

Often shows signs of tube bulging or "fish-mouth" appearance, longitudinal orientation.

Most commonly near bottom bends in vertical loops of SH/RH; outlet legs, and near material transitions.

Short-Term Overheating

36

Thin-edged

1-8

Introduction and Background

34 (Oil-fired units)

Table 1-5 Screening Table for Steam-Touched Boiler Tube Failures (continued) Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 3 (or 2 as noted)

Sootblower Erosion

38

Rubbing/ Fretting

40

Chemical Cleaning Damage or Pitting

41 or 43

Maintenance Damage

44

Materials Flaws

45

Welding Flaws

46

Thin-Edged Fracture Surface (continued) Thin-edged, pinhole or "thin" longitudinal blowout

External wastage flats at 45° around tube from sootblower direction, little or no ash.

Thin-edged

External damage; obvious metal-tometal contact on tube surface.

First tubes in from wall entrance of retractable blowers; tubes in direct path of retractable blowers.

Pinhole Damage Pitting

Internal tube surface damage.

For pitting: Tubes where condensate can form and remain during shutdown: bottoms of pendant loops on either SH or RH, low points in sagging horizontal tubes.

Various Other Damage Types Depends on the underlying cause

Usually obvious from type of damage and correspondence to past maintenance activity.

Depends on defect Usually thick-edged or pinholes

Care required to separate weld defects from another problem located at a weld.

Note: This table is based on simple, macroscopic features of failure and should be used as a guide to a particular chapter for further analysis. The more detailed discussions starting with Actions can then be used for identification and confirmation of the actual mechanism.

Volume 1: Boiler Tube Fundamentals

1-9

Table 1-6 Unit Precursors and Potential Future BTF 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

1.1 Watertouched tubes (waterside)

Excessive waterside deposits ( >> 30 mg/cm2) for high-pressure boilers.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), short-term overheating (23,V2)

Excessive waterside deposits, such as ripple Fe3O4 in once-through (O/T) and supercritical units.

Supercritical waterwall cracking (19,V2)

Boiler water samples that appear black (high suspended solids).

Acid phosphate corrosion (16,V2)

Corrosion/erosion in feedwater system; fouling in boiler feed pump or orifices.

• For supercritical or O/T units: supercritical waterwall cracking (19,V2) • For subcritical or non-O/T units - hydrogen damage (15,V2), acid phosphate corrosion (16,V2), or caustic gouging (17,V2) • Erosion-corrosion of economizer inlet header (21,V2)

Pressure drop across circulation pumps (orifices are plugging).

Short-term overheating in waterwall tubing (23,V2)

Flame impingement due to burner change or misalignment, leading to excessive tube deposits.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), fireside corrosion (18,V2)

Excessive furnace slagging that could lead to overheating in convective passes (or fuel change).

Short-term in overheating SH/RH tubing (36,V3)

Fresh rust found on tubes after unit washing, external flat spots, burnishing or polishing.

Flyash erosion (14,V2), sootblower erosion - waterwalls (22,V2), coal particle erosion (28,V2)

Failed tubes, any upstream tube leaks, as a warning to scout for the potential short-term overheating.

Short-term overheating in waterwall tubing (23,V2)

Significant hardness or ovality, particularly associated with tube bends, found during routine inspection.

Low-temperature creep cracking (24, V2)

Excessive steamside oxide (detected by UT measure of oxide thickness, or analysis of removed tube samples, evidence of excessive exfoliation like solid particle erosion in turbine).

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33&34,V3), dissimilar metal weld failures (35,V3), short-term overheating (36,V3)

Steamside deposits in RH tubing - particularly of sodium sulfate, or high Na or SO4 levels in steam.

Pitting and failure in steam-touched tubes (41,V3)

1.2 Watertouched tubes (fireside)

1.3 Steamtouched tubes (steamside)

1-10

Introduction and Background

Table 1-6 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance (continued) 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

1.4 Steamtouched tubes (fireside)

Excessive flue gas temperature, displaced fireball, delayed combustion, periodic overfiring or uneven firing of burners.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33 & 34,V3)

High levels of excess oxygen.

SH/RH fireside corrosion: oil-fired units (34,V3)

Blockage or laning of boiler gas passages observed during boiler inspection.

Flyash erosion (14,V2), long-term overheating/creep (32,V3), SH/RH fireside corrosion: coal/oil units (33 & 34,V3)

Excessive temperatures measured by thermocouples in vestibule or header area.

Flyash erosion (14,V2), long-term overheating/creep (32,V3), dissimilar metal weld failures (35,V3)

Evidence of "alligator hide" appearance on external tube surface, observed during boiler inspection, associated with wall loss or thinning.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33 & 34,V3)

Fresh rust found on tubes after unit washing, external flat spots, burnishing or polishing.

Flyash erosion (14,V2), sootblower erosion in SH/RH (38,V3)

Significant hardness or ovality, particularly associated with tube bends, found during routine inspection.

Low-temperature creep cracking (24,V2)

Distortion or misaligned tube rows found during routine inspection.

Flyash erosion (14,V2), SH/RH fireside corrosion (33 & 34,V3), dissimilar metal weld failures (35,V3), fatigue of steam-touched tubing (39,V3), rubbing/fretting (40,V3),

Failed tube supports and lugs, location of dissimilar metal welds close to fixed supports.

Fatigue of steam-touched tubing (39,V3), dissimilar metal weld failures (35,V3)

Volume 1: Boiler Tube Fundamentals

1-11

Table 1-6 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

2.1 All units

Problem with high levels of feedwater corrosion products; operating ranges for pH, cation conductivity or dissolved oxygen consistently outside recommended ranges, including persistent reducing conditions or excessive use of oxygen scavengers.

Corrosion fatigue (13,V2), hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), waterwall fireside corrosion (18,V2), supercritical waterwall cracking (19,V2), erosion/corrosion in economizer inlet header (21,V2), short-term overheating in waterwall tubing (23,V2),

Carryover of volatile chemicals from boiler, such as NaOH for units on caustic treatment, or excess of Na, SO4, and/or chloride; steam limits exceeded.

Stress corrosion cracking (37,V3), pitting in steamtouched tubes (41,V3)

Major acid contamination event (pH < 8) when unit is at full load; condenser leak, or breakdown of makeup or condensate polisher regeneration chemical.

Hydrogen damage (15,V2)

Evidence of a persistent problem with phosphate hideout, particularly where mono-sodium and/or an excess of di-sodium phosphate has been added to the boiler.

Acid phosphate corrosion (16,V2)

Persistent phosphate hideout with phosphate return causing a pH depression (7-8).

Corrosion fatigue (13,V2)

Caustic level in excess of that necessary for optimal control (>> 2 ppm).

Caustic gouging (17,V2)

Caustic, used in excess of that necessary for optimal control of contaminant ingress (to counteract pH depressions on startup).

Caustic gouging (17,V2)

pH depression during shutdown and early startup (pH around 7-8). Hideout/return of sulfate.

Corrosion fatigue (13,V2)

Caustic, used in excess of that necessary for optimal control (>> 2 ppm).

Caustic gouging (17,V2)

2.2 Units on Phosphate Treatments

2.3 Units on AVT

2.4 Units on Caustic Treatment

1-12

Introduction and Background

Table 1-6 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

3.1 Chemical cleaning

Evidence of shortcoming in chemical cleaning process such as inappropriate cleaning agent, excessively strong concentration or long cleaning time, too high a temperature, failure to neutralize, breakdown of inhibitor, inadequate rinse.

Chemical cleaning damage in waterwalls (25,V2) or SH/RH (43,V3), short-term overheating (23,V2 & 36,V3).

Shortcoming in SH/RH cleaning process such as inadequate rinse, improper flow verification.

Short-term overheating in SH/RH tubing (36,V3)

Evidence that level of Fe in cleaning solution continued to increase instead of leveling out when cleaning process was ended.

Chemical cleaning damage in waterwalls (25,V2) or SH/RH (43,V3)

Need for excessive cleaning in supercritical units (interval < 2 years).

Supercritical waterwall cracking (19,V2)

Contamination in SH/RH (particularly by chlorides) during chemical clean of SH/RH (breakdown of inhibitors or improper flushing of solvents) or waterwalls (caused by poor backfill procedures that failed to protect SH circuits).

Stress corrosion cracking (37,V3)

In water-touched tubes: use of backing rings, pad welds, canoe pieces, weld overlay that penetrates to inside surface - as a source of flow disruption and excessive deposits.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2)

Application of shielding, baffles, palliative coatings to mitigate flyash erosion without the use of a cold-air velocity test.

Flyash erosion (14,V2)

In water-touched tubes, Cu in water-side deposits.

Hydrogen damage (15,V2), welding defects (46,V3)

3.2 Repairs

Volume 1: Boiler Tube Fundamentals

1-13

Table 1-6 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

4.1 Startup Procedures

Feedwater introduced intermittently into economizer inlet at high flow rates during startups and particularly during off-line top-ups.

Economizer inlet header thermal fatigue (20,V2)

Rapid unit startups that cause the reheater to reach temperature before full flow starts (no furnace exit gas temperature control).

SH/RH fireside corrosion (33 & 34,V3)

Heat flux change caused by change to higher BTU-value coal, dual firing with gas, changeover to oil- or gas-firing leading to excessive tube deposits in waterwalls; new burners causing impingement.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), fireside corrosion (18,V2)

Implementing low excess air strategies for NOx control and the potential for waterwall fireside corrosion (note that unlike the other precursors in this Table, this is a possibility based on understanding the mechanism; to date no failures have been directly attributed to this cause).

Waterwall fireside corrosion (18,V2)

Operation with high levels of excess oxygen in oil-fired units (> 1%).

SH/RH fireside corrosion in oil-fired units (34,V3)

Change to a fuel that either contains more ash or contains elements which are more erosive such as quartz.

Flyash erosion (14,V2)

Change to a more corrosively-aggressive coal, particularly one high in chlorine, Na, K, or S contents.

Waterwall fireside corrosion (18,V2), acid dewpoint corrosion (30,V2), SH/RH fireside corrosion (33 & 34,V3)

Use of Mg-based additives (oil-fired units) leading to coating of waterwalls, reflecting heat into convection passes.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion in oil-fired units (34,V3)

4.4 Cycling

Conversion of the unit to cycling operation or an increase in the number of cycles.

Corrosion fatigue (13,V2), economizer inlet header thermal fatigue (20,V2), fatigue in water-touched (26,V2) or steamtouched tubing (39,V3),dissimilar metal weld failures (35,V3)

4.5 Shutdown or layup

Evidence of a shortcoming during unit shutdown/layup such as uncertainty about water and/or air quality during period, insufficient nitrogen blanketing, insufficient N2H4, evidence of air inleakage.

Pitting in water-touched (27,V2) or steam-touched tubes (41,V3), and maybe corrosion fatigue (13,V2)

Indication that stagnant, oxygenated water may have rested in tubes during shutdown or layup particularly in economizer and RH.

Pitting in water-touched (27,V2) or steam-touched tubes (41,V3)

Evidence that condensate is forming in SH/RH bends during unit shutdown, exacerbated if steam purity is not good (as determined by elevated levels of SO4).

Short-term overheating in SH/RH tubes (36,V3), pitting in steam-touched tubes (41,V3)

Operation above the maximum continuous design rating, with excess air flow settings above design, with unbalanced fans or air heaters leading to nonuniform gas flows.

Flyash erosion (14,V2)

Low drum level.

Short-term overheating (23,V2)

4.2 Combustion conditions

4.3 Fuel choices and changes

4.6 Other

1-14

Introduction and Background

Table 1-6 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

5.1 Condensers

Major condenser leaks or minor leaks that have occurred over a long period of time.

Hydrogen damage (15,V2)

Condenser leak leading to condenser cooling water constituents in attemperator spray water.

Stress corrosion cracking (37,V3)

5.2 Water treatment plant/ condensate polisher

Upset in water treatment plant or condensate polisher regeneration chemicals leading to low pH condition in boiler (pH < 8).

Hydrogen damage (15,V2)

Upset in water treatment plant or condensate polisher regeneration chemicals leading to high pH condition.

Caustic gouging (17,V2)

5.3 Drum

Carryover test indicates high mechanical carryover.

Stress corrosion cracking (37,V3), pitting in steamtouched tubing (41,V3)

Operating with high drum level allowing excessive carryover into steam.

Pitting in steam-touched tubing (41,V3)

5.4 Sootblowers

Poor sootblower maintenance.

Sootblower erosion in waterwalls (22,V2), SH/RH sootblower erosion (38,V3)

5.5 Low temperature headers

Header has large number of operating hours, has experienced large thermal gradients, spacing of ligament holes is small (< 3.5 cm), header thickness is well above Code minimum, header-to-stub tube joints made with partial penetration welds.

Economizer inlet header thermal fatigue (20,V2)

5.6 High temperature headers

Excessive relative movement of header/ tube during unit transients, restricted movement, header is not allowed to expand freely (maybe ash-related), unit change to cycling.

Fatigue in steam-touched tubing (39,V3).

5.7 Turbine

A problem with solid particle erosion (SPE) in the turbine.

Short-term overheating SH/RH tubing (36,V3), long-term overheating /creep (32,V3)

5.8 SH/RH Circuit (redesign)

Redesign of the SH/RH circuit may change the absorption patterns through other SH/RH sections and increase tube temperatures.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33 & 34,V3), dissimilar metal weld failures (35,V3)

5.9 Supports/ Attachments (redesign)

Addition of supports without consideration of their impact on the stresses of dissimilar metal welds.

Dissimilar metal weld failures (35,V3)

Redesign of waterwall tube attachments to increase flexibility without analysis to determine whether solution is actually beneficial.

Corrosion fatigue (13,V2)

Volume 1: Boiler Tube Fundamentals

1-15

Clearly, it is not possible to put every precursor in this compilation, but by using this listing it is hoped that two objectives are achieved. First, forced outages by BTF are further prevented by anticipating the pre-conditions to the most common mechanisms. Second, that a first step will be taken to improve the understanding of the complex, interconnected nature of cycle chemistry, operating practice, combustion processes, and maintenance effects on BTF. As a final note, the list should not preempt good engineering judgment; if a precursor is found in the unit that you think should be an alert of a future problem, follow it up, even if it is not in this particular list. 1.3.5 For BTF mechanisms not covered in this book. If, having gone through the above procedure, it appears that the BTF experienced is not covered in this book or if multiple mechanisms appear to be operative, then the generic investigation procedure shown in Figure 1-1 is still applicable. Specifically, it is important that the following sequence be followed: Understand the mechanism ¯ Determine the root cause ¯ Apply proper long-term solution Removal of a tube sample and use of metallurgical techniques should provide an understanding of the underlying damage (erosion, corrosion, overheating, creep, fatigue, environmentally-assisted cracking, etc.) and may help move the investigator to one of the covered mechanisms for additional specific guidance.

1-16

Introduction and Background

1.4 Introduction to the WaterSteam Cycle and Primary Components in Conventional Fossil-Fuel Power Plants

are typically around 400°C (~750°F). In supercritical units, waterwall materials typically operate at slightly higher temperatures: 454°C (849°F).

There are several excellent reviews of the design and operation of conventional fossil-fueled power plants, see for example references 7 and 8. A very much simplified review of primary power plant components and the water-steam cycle is provided here.

The steam next flows into superheater/reheater sections. There is a distribution of increasingly higher temperatures as the fluid moves through the circuit. This results in more severe demands on the tube materials including the need for better creep and oxidation resistance. For final steam temperatures of 538 to 565°C (1000 to 1050°F), tube temperatures in excess of 600°C (~1110°F) can be reached during the final stages of the superheater and reheater sections.

Water is preheated to a relatively low temperature by passing through a series of low pressure (LP) and high pressure (HP) feedwater heaters. It then passes through the various parts of the boiler where it is heated to superheated steam. From the exit from the boiler the steam is allowed to expand through the high pressure (HP) turbine from which it may be reheated and passed through intermediate pressure (IP) and low pressure (LP) turbines where further work is extracted. Exit steam from the LP turbine is condensed and fed back into the boiler through the feedwater heaters.9 Figures 1-2 and 1-3 show a schematic of this cycle for a drum and a once-through cycle, respectively. Also shown in those figures are the locations in the cycles where ingress, corrosion and deposition can occur. The ingress of contaminants and its deposition and corrosion have major implications to the analysis of boiler tube failures. Boiler tube temperatures are another important factor that will determine when failures will occur. Tube metal temperatures depend on the heat flux from the fireside, the internal fluid flow and the condition of the working fluid. The lowest temperature fluid is in the economizer and waterwall sections. The fluid enters the economizer as a liquid and becomes a steam-liquid mixture in the waterwalls caused by heat transfer through the walls. Temperatures in the fluid for subcritical boilers are limited to saturation temperatures for the given boiler pressure, thus tube temperatures

Additional details of the design, constructional materials and demands placed on boiler tubes in the various sections of the boiler are provided in Chapter 2.

1.5 Historical Developments in the Identification, Correction and Prevention of BTF Internationally, extensive research into the causes and prevention of boiler tube failures has been ongoing since the early 1950s. The problems confronted, the solutions sought, and the progression of technology in different countries have proceeded along many paths. As a result different countries have had, and continue to have, different failure types and have instituted differing approaches to their resolution. Factors that have shaped the research on individual boiler tube failure mechanisms include: (i) fuel sources available, (ii) design philosophy, (iii) an outbreak of a serious failure or series of failures, (iv) available technology, (v) knowledge of prior shortcomings, (vi) the level of available manufacturing technology and quality control that could be achieved, (vii) what type of cycle chemistry control was chosen, and (viii) the flexibility to change once a decision about one of the above had been made.

HP turbine Feed IP turbine Attemperation

LP turbine Condenser Makeup

Deaerator Boiler HP heaters Condensate polisher Impurity ingress Corrosion Deposition

Feed

Figure 1-2. Major unit components and locations of impurity ingress, corrosion and deposition in drum cycles. Source: R.B. Dooley and A. Bursik10

HP turbine IP turbine

LP turbine Condenser Makeup

Attemperation Deaerator Boiler HP heaters Condensate polisher Impurity ingress Corrosion Deposition

Feed

Figure 1-3. Major unit components and locations of impurity ingress, corrosion and deposition in oncethrough cycles. Source: R.B. Dooley and A. Bursik10

Volume 1: Boiler Tube Fundamentals

1-17

1.5.1 On the general nature of protective oxides. Evaluations of the general nature of protective oxides and laboratory studies of various boiler tube materials, as well as the study of specific failure mechanisms and field performance has been extensively pursued since the 1950s. The growth of magnetite on tube materials and the attack caused by acidic or basic solutions was investigated in the early 1960s.11, 12 The earliest extensive U.K. research on the corrosion behavior of steels in steam environments was a collaborative exercise in the 1960s between the CEGB, Brown Firth Research Laboratories, Brown Bayley Steels Ltd., and United Steel Cos. Ltd.13 1.5.2 Influence of available fuel sources. Units in the United Kingdom differ from those in North America in burning coals containing levels of chlorine up to 0.7% from which serious problems with waterwall and SH/RH fireside corrosion have arisen. Because of the seriousness of the problem, much effort has been, and still is, being undertaken to understand the basic mechanisms and to develop the optimal mitigation strategies. The recognition of the effect of coal composition on the incidence of fireside corrosion has influenced U.S. research into this mechanism as well, such as the studies conducted for Eastern U.S. coals by Borio, et al.14 In a similar light, some countries have major problems with highly erosive (high ash content) coals. In these countries (such as India, South Africa, Australia) the problem of flyash erosion is of major importance.

1-18

Introduction and Background

1.5.3 Influence of operating conditions. The advanced steam conditions that were typical of early oncethrough supercritical units in the U.S. (following their introduction in Europe) led to a host of problems with corrosion and overheating in boilers operating at 621°C (1150°F) and 650°C (1202°F).15, 16 The recognition in the 1950s that high steam temperatures can cause significant problems with low-melting point ash constituents and extensive fireside corrosion led to limitations on main steam temperature in coal-fired units. These limitations are, for the most part, still followed in today’s units, generally to a maximum temperature of 566°C (1050°F) and pressure of 3600 psig. In contrast, many countries including Canada, Australia, New Zealand, and Hong Kong, continued to employ drum units with 548°C (1000°F) and 2500 psig cycles and as a result, avoided the problems with serious fireside corrosion that had resulted in the higher temperature units. 1.5.4 BTF and the choice of water chemistry control (evolution of caustic treatment contrasted with phosphate treatment). In the 1950s and 1960s both the U.S. and the U.K. had serious waterside corrosion problems.17 Careful laboratory work in both countries set the scene for operational limits to water chemistry and interestingly led to two drastically different approaches. In the U.S., a seminal study was conducted in the early 1960s by the American Society of Mechanical Engineers Research Committee on Boiler Feedwater Studies to identify the cause of this very severe attack of boiler tubes.18, 19 The cause was identified as high levels of sodium hydroxide and sodium phosphate in

the treatments which were standard practice at the time. As a result, the chemistry of choice became congruent phosphate treatment so as to move away from the free hydroxide regime. As discussed in more detail in Chapter 3 of this volume, this choice has led to problems with control because of phosphate hideout and subsequently to the occurrence of BTF caused by acid phosphate corrosion. In response, there has been an evolution of phosphate treatment options which continues in U.S. units. In the U.K., the choice in the 1970s was to use NaOH because of its superior buffering ability against the contamination caused by any ingress of seawater from condenser leaks. That choice required strict control of the level of NaOH to prevent the continuation of waterside corrosion by a caustic gouging mechanism. 1.5.5 BTF and the choice of water chemistry control method (evolution of all-volatile treatment contrasted with oxygenated treatment). Water chemistry for supercritical units in the U.S. was derived from that of existing European practice at the time (early 1950s) which was allvolatile treatment (AVT). The use of AVT in once-through supercritical units has subsequently led to problems including: (i) the production of an excess of feedwater corrosion products, (ii) excessive boiler pressure drops, (iii) deposit buildup, (iv) BTF by a supercritical waterwall cracking mechanism, and (v) condenser tube failures. In the early 1970s European utilities changed to an oxygenated treatment chemistry and have been able to avoid these problems; similar changes are now occurring in U.S. supercritical units.

1.6 Recent Developments in the Identification, Correction and Prevention of BTF Over the last fifteen years there have been major efforts to improve not only the understanding of the technical details of boiler tube failures, but also the management approaches and the overall influences of plant cycle chemistry, operation, and maintenance on their occurrence. These efforts have been concentrated on four primary areas of interest: • Identification of BTF mechanisms, root causes, and corrective actions • Setting unit cycle chemistry levels and understanding the interactions with BTF • Understanding the effects of unit and boiler operation and maintenance on BTF • Formulating utility-wide, coordinated, management-supported programs for BTF correction, prevention, and control. A brief overview of some of these more recent activities sets the stage for more detailed discussion in later parts of this book. 1.6.1 Identification of BTF mechanisms, root causes, and corrective actions. The first compilation of all known boiler tube failure mechanisms and their direct causes was prepared in the late 1970s and published by the Canadian Electric Association as Analysis and Prevention of Boiler Tube Failures.20 An updated version, Manual for Investigation and Correction of Boiler Tube Failures was published by the Electric Power Research Institute in the mid-1980s.21 These technical compilations were used as the bases for overall managementsupported, programmatic approaches for the evaluation and correction of BTF on a utility-wide basis.22

By the early 1980s, twenty-two different BTF mechanisms had been identified; nineteen of those were considered to be well understood, meaning that approaches and solutions to root causes were available. The three failure mechanisms for which understanding was incomplete included: (i) corrosion fatigue tube failures that initiate from the waterside of waterwall and economizer tubing; (ii) circumferential cracking that initiates from the fireside of the waterwall tubing in the highest heat flux regions of supercritical units, and (iii) flyash erosion. Extensive research into these mechanisms over the last ten years has resulted in essentially a complete understanding of them and solutions are now available that can be confidently applied by utilities for their prevention. As metallurgical and root cause analysis have become more widespread, additional specific BTF mechanisms have been identified. For example, during the past five years, a new mechanism, acid phosphate corrosion, related to the use of congruent phosphate treatment, has been identified in the U.S. Worldwide survey of experience and research has assisted in understanding this mechanism and is providing cycle chemistry guidance to overcome it. The publication of the Boiler Tube Failure Metallurgical Evaluation Handbook23 provided a key resource and a sharp focus on metallurgical analysis to determine the correct mechanisms responsible for degradation or failure of boiler tubing. The present book, in three volumes, expands upon the base provided by these prior compilations. It combines those efforts with the wealth of recent research into specific mechanisms and unit-wide influences. Most of the major boiler tube failure mechanisms that have been identified worldwide are included. Investigation of tube failures falling outside the major types can be conducted following the generic approach that is provided.

1.6.2 Setting unit cycle chemistry levels and understanding the interactions with BTF. Even a causal review of the mechanisms on Table 1-3 will confirm that the control of plant cycle chemistry and the reduction of boiler tube failures are intimately connected. Many of the most prevalent BTF mechanisms, as well as other components throughout the unit, are directly influenced by cycle chemistry. The most direct, most effective, and ultimately the least expensive means to prevent many serious BTF is through the appropriate choice, control, and monitoring of the steam and water purity. Following the successful experience in Germany (VGB Guidelines24), England (CEGB Guidelines25) and Japan (CRIEPI26) with reducing chemistry and corrosion incidences through formalized cycle chemistry guidelines, the U.S. introduced a similar comprehensive approach with the release of the Interim Consensus Guidelines for Fossil Plant Cycle Chemistry27 and the Guideline Manual for Instrumentation and Control for Fossil Plant Cycle Chemistry28. Recent advances have been made in the understanding of, and control of, all the major chemistries. As a result several documents which represent revisions to the Interim Consensus Guidelines have been or should soon be released including: Guidelines for Oxygenated Treatment29, Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment for Drum Units30, Sodium Hydroxide for Conditioning the Boiler Water of Drum-Type Boilers17 and a document that will cover all-volatile treatment31. Because of the importance of cycle chemistry, monitoring and control, Chapter 3 of this volume is devoted to a discussion of the key aspects of cycle chemistry and its importance to boiler tube failures.

Volume 1: Boiler Tube Fundamentals

1-19

1.6.3 Understanding the effects of unit and boiler operation and maintenance on BTF. Similarly, there has been an increasing recognition of the effect of both maintenance activities including inspection, repair, chemical cleaning, and of unit operations such as unit cycling and layup, on the incidence of a number of BTF. Recent advances in these areas are discussed in Chapter 4 of this volume. 1.6.4 Formulating utility-wide, coordinated, management-supported programs for BTF correction, prevention, and control. As noted above, at the time that early BTF manuals were produced, nineteen of the twenty-two mechanisms that had been identified were understood and permanent solutions to each existed. Why then did (and still do) BTF continue to occur? Additional analysis indicated that, at that time, the primary barrier to achieving major availability improvement was not a lack of technical knowledge, but (i) getting that knowledge into the hands of more utility personnel and (ii) dealing with a variety of management and economics issues in a formalized corporate-wide manner. Specifically, it was found that in order to address the management and economic issues surrounding BTF, a company-wide BTF prevention program was required that could address the multi-functional nature of the problem. Such a program was developed and a major demonstration project involving sixteen utilities demonstrated that such a company-wide program could result in significant availability improvements.32 The results are dramatically illustrated in Figure 1-4 where the results of equivalent availability loss (EAL) for BTF are shown for the two sets of participating utilities and compared with the national average.

1-20

Introduction and Background

EAL% 3.8 3.4

Second 6 utilities

3.0 2.6 2.2

Nation average for units above 200 MW First 10 utilities

1.8 1.4 1.0 1985

Target for first 10 utilities: 1.45% 1986

1987

1988

1989

1990

1991

1992

Figure 1-4. Availability improvements achieved through a formalized BTF reduction program. EAL is equivalent availability loss caused by BTF.

A main objective of this book is to provide a centralized, up-to-date source of information about the identification and prevention of boiler tube failures, but it is clear from historical precedence that this alone will be insufficient to end the occurrence of BTF. For that reason, Chapter 5 discusses the importance of formalized programs for the identification, prevention and control of BTF. In a similar manner to the BTF reduction program, a demonstration of the needed activities for improvements in cycle chemistry has also been undertaken with thirteen utilities.33 Discussion of the critical factors for such programs is provided in Chapter 3.

1.7 TodayÕs Situation and Challenges that Remain The technical knowledge of BTF has improved dramatically since the earliest compilations. Within the past five years, complete understanding of the basic mechanisms and mitigation options for flyash erosion, corrosion fatigue, and circumferential cracking in supercritical units have been gathered. The importance of, and the requirements to implement, on a practical basis, the

cycle chemistry controls to prevent the known chemistry-controlled damage mechanisms are now known. Further, large-scale, utility demonstration projects have shown that technical knowledge plus a company-wide BTF correction, prevention, and control program can demonstrably achieve significant availability improvements. However, there remain some challenges. Two “generic” challenges, the frustrating appearance of repeat failures, and the misdiagnosis of mechanisms are discussed briefly here, as is a vision for the future. 1.7.1 Repeat failures. Analysis has indicated that a disturbing feature of many outbreaks of BTF is that an alarming number of them are repeat failures.34 There will, of course, be random tube failures caused by errors during engineering, fabrication, construction, operation and maintenance. Examples include: wrong or defective tube materials, poor field welding, defective or malfunctioning tube supports or spacers, etc. However, the costs associated with random tube failure incidents are small compared to repeat failures. Within the past few years

repeat tube failures have caused, for example: replacement of 40,000 linear feet of waterwall tubing due to hydrogen damage, replacement of an entire superheater due to overheating, and replacement of entire economizers because of oxygen pitting (local corrosion).34 Repeat failures are those that continue to occur on the same tube, same material or in the same boiler sections. This happens for one of four reasons: (i) the root cause of the initial failure was not determined, (ii) the root cause was improperly determined and thus the wrong corrective action was taken, (iii) the root cause was identified, the proper corrective action was attempted, but was not performed correctly, or (iv) no corrective action was taken. Without a proper understanding of the mechanism of failure, the root cause, the appropriate corrective actions, and the proper execution of those actions, it is not possible to apply permanent engineering solutions. One of the purposes of this book is to match mechanism, root cause and corrective action to help avoid this problem. 1.7.2 Distinguishing between similar mechanisms. That there is a significant loss of availability from all boiler tube failure causes is certain as reflected by the statistics cited above. It should be noted however, that there is some uncertainty as to relative contributions to unavailability for several of the more prominent mechanisms. This is caused by an inability to distinguish between some of these mechanisms, usually as a result of inferior metallurgical analysis. This problem then leads to incorrect root cause identification followed by an inappropriate and ineffective solution to the problem. Chapter 7 of this volume looks specifically at a comparison of some commonly misdiagnosed mechanisms. Key defining characteristics for each are also given in the relevant discussions of individual mechanisms.

Table 1-7 Goals and Vision for BTF Prevention An active BTF reduction program with corporate philosophy signed by senior management. Availability loss due to BTF of less than 1%. No cycle chemistry related BTF. Cycle chemistry operating guidelines for all units including optimization of shutdown, layup and start up. No maintenance influenced BTF. Approved welding and quality control procedures. Established NDE procedures. Qualified metallurgical analysis. Established life assessment methodology. Comprehensive documentation including: location of failures, mechanism responsible, root cause determined, repair procedures, and applied solutions. Sources: J.P. Dimmer, G.A. Lamping, O. Jonas, and R. Niebo32; R.B. Dooley35,36

1.7.3 A vision for the future. The inter-relationship of the various factors influencing BTF and goals for the near term, developed by EPRI and participating utilities, are shown in Table 1-7. These goals are achievable now; there are no technical barriers to a target of less than 1% availability loss caused by BTF by the turn of the century.36 In addition the following should be implemented: ¥ New units. New units will be designed to minimize problems with dissimilar weld metals, and to minimize the attachment stresses that can lead to corrosion fatigue problems. During commissioning of new units, specific testing will be done to ensure that BTF are minimized, such as cold air velocity tests to determine potential for flyash erosion, installation of monitoring equipment such as thermocouples to keep tube temperatures below design and oxidation limits, and archives of tube samples for future reference use. ¥ New technologies. It is important that the lessons learned about unit operating, maintenance, and cycle chemistry, and boiler tube failures, should not be forgotten when new power sources are designed and built. Combined cycle and heatrecovery steam generators are a case in point; the need for optimum

cycle chemistry and the careful design of attachment details should be applied in order to avoid the problems that have produced corrosion fatigue in conventional plants. ¥ Systematic records. A means to record where failures are occurring, the incidence, mechanism, and the corrective actions used will be in place. Without such a process to record failures and as much data as is available, much of the systematic approach to preventing failures is stymied. ¥ Identify ÒdamageÓ not ÒfailuresÓ. The entire sequence of events will be executed with a goal to have no failures, that is, utilities are to be encouraged to identify damaged tubes, prior to “failure”, and to execute the steps needed to address the underlying root causes. ¥ No repeat failures. Repeat failures are unacceptable and are avoidable. Repeat failures are those by the same mechanism in the same location as a result of the same cause. ¥ Need for a utility-wide commitment. Correction, prevention and control requires utility-wide commitment and can result in significant savings and measurable reduction in system-wide BTF.

Volume 1: Boiler Tube Fundamentals

1-21

1.8 References 1Niebo,

R.J., “Implementing a Boiler Tube Failure Mechanism Reporting Program”, in B. Dooley, ed., Proceedings: International Conference on Boiler Tube Failures in Fossil Plants, held in San Diego, California November 5-7, 1991, Proceedings TR-100493, Electric Power Research Institute, Palo Alto, CA, April, 1992, pp. 1-31 through 1-46. 2McNabb,

D., D. Sidey, and R.W. Patterson, “Canadian Boiler Tube Failure Experience”, in B. Dooley and D. Broske, eds., Boiler Tube Failures in Fossil Power Plants: Conference Proceedings, Conference held in Atlanta, Georgia, November 10-12, 1987, CS-5500-SR, Electric Power Research Institute, Palo Alto, CA, 1988, pp. 1-19 through 1-35. 3Senior,

B.A., “An Integrated Approach to BTF Integrity R&D”, op. cit., reference 1, pp. 12-55 through 12-67. 4Platfoot,

R.A. and C.A.J. Fletcher, “Boiler Tube Failures Caused by Grit Erosion”, ibid., pp. 5-29 through 5-43. 5Shtromberg,

Y., “Analysis of Tube Failures of Boiling Heating Surfaces in Fossil Plants of the USSR”, ibid., pp. 9-15 through 9-21. 6Conference

Questionnaire, Appendix A, ibid., pp. A-1 through A-10. 7Stultz, S.C. and J.B. Kitto, eds. Steam: Its Generation and Use, 40th Edition, Babcock & Wilcox Company, Barberton, Ohio, 1992. 8Singer,

J.G., ed., Combustion Fossil Power: A Reference Book on Fuel Burning and Steam Generation, 4th Edition, Combustion Engineering, Inc., Windsor, Connecticut, 1991. 9Jaffee,

R.I., “Metallurgical Problems and Opportunities in Coal Fired Steam Power Plants”, Met. Trans., Volume 10A, May, 1979, pp. 139-165. 10R.B.

Dooley and A. Bursik, “State of the Art in Fossil Plant Cycle Chemistry”, 12th International Conference on Water and Steam, held in Orlando, FL, September, 1994, Begel House “Physical Chemistry of Aqueous Systems”. 11Potter,

E.C. and G. M. W. Mann, Proc. 1st Int. Cong. Metall. Corrosion, London, Butterworks, 1961, p. 417. 12Field,

E.M., R.C. Stanley, A.M Adams, D.R. Holmes, “The Growth, Structure and Breakdown of Magnetite Films on Mild Steel”, Proc. 2nd Int. Conf. Metallic Corrosion, New York, 1963, p. 829. 13King,

C.W., M.T. Robinson, H. Howarth, and R. Perry, “Oxidation Behaviour of Steels in High Temperature Steam”, CEGB Report SSD/MID/R244/71, 1971, cited in Armitt, J, et al., The Spalling of Steam-Grown Oxide from Superheater and Reheater Tube Steels, Technical Planning Study 76-655, Final Report FP-686, Electric Power Research Institute, Palo Alto, CA, February, 1978.

1-22

Introduction and Background

14Borio,

R., et al., “The Control of High-Temperature Fireside Corrosion in Utility Coal-Fired Boilers”, U.S. Office of Coal Research, Research and Development Department, No. 41, April ,1969. 15Reid,

W.T., External Corrosion and Deposits - Boilers and Gas Turbines, Elsevier, New York, 1971. 16Williams,

D.N., H. R. Hazard, H.H. Krause, L.J. Flanigan, and I.G. Wright, Fireside Corrosion and Fly Ash Erosion in Boilers, Research Project 2711-1, Final Report CS-5071, Electric Power Research Institute, Palo Alto, CA, February, 1987. 17Ball, M., Sodium Hydroxide for Conditioning the Boiler Water of Drum-Type Boilers, Research Project 9000-20, Final Report TR-104007, Electric Power Research Institute, Palo Alto, California, January, 1995. See also M. Ball, “Caustic Treatment for Drum Boilers”, in R.B. Dooley and R. Pate, eds. Fourth International Conference on Cycle Chemistry in Fossil Boilers, held in Atlanta, Georgia, September 7-9, 1994, Final Report TR104502, Elecrtic Power Research Institute, Palo Alto, CA, January, 1995. 18Goldstein,

P., “A Research Study on Internal Corrosion of High Pressure Boilers”, Trans. ASME 90(A), 1, 1968, pp. 23-37. 19Goldstein,

P. and C.L. Burton, “A Research Study on Internal Corrosion of High Pressure Boilers - Final Report”, Trans. ASME 91(A), 1969, pp. 75-101.

20Dooley, R.B. and H.J. Westwood, Analysis and Prevention of Boiler Tube Failures, Report 83/237G-31, Canadian Electrical Association, Montreal, Quebec, November, 1983. 21Lamping, G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985. 22Dimmer,

J.P., G.A. Lamping, and O. Jonas, Boiler Tube Failure: Correction, Prevention, and Control, Research Project 1890-7, Final Report GS-6467, Electric Power Research Institute, Palo Alto, CA, July, 1989. 23Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993.

24VGB,

“VGB Guidelines for Feedwater, Boiler Water, and Steam of Generators Exceeding 68 bar Operating Pressure” (in German) VGB Technische Vereinigung der Grosskraftwerksbetreiber, e.V. VGB-R 450 L, Essen, 1983. See also A. Bursik, “VGB Guidelines on Boiler Feedwater, Boiler Water and Steam of Water-Tube Boilers”, IWC-84-116, presented at the 45th Annual Meeting of the International Water Conference, Pittsburgh, PA, October 22-24, 1984. 25Central

Electricity Generating Board, Generation Operation Memorandum (GOM) 72: Part 1 - Introduction and General Aspects of Chemical Control of the Steam Water Circuit, Issue 5, April, 1983. Part 2 - Chemical Control of the Steam Water Circuit of Drum-Type Boilers, Issue 6, September 1985. Part 4 - Sampling, Analysis, Instrumentation and Chemical Dosing, Issue 5, September, 1983. 26Japanese

Industrial Standard JIS B 8223-1977, Water Conditioning for Boiler Feed Water and Boiler Water, Japanese Standards Association, 1977. 27Aschoff,

A.F., Y.H. Lee, D.M. Sopocy, and O. Jonas, Interim Consensus Guidelines on Fossil Plant Cycle Chemistry, Research Project 2712-1, Final Report CS4629, Electric Power Research Institute, Palo Alto, CA, June, 1986. 28Hopkins,

R.D., et al., Guideline Manual on Instrumentation and Control for Fossil Plant Cycle Chemistry, Research Project 2712-2, Final Report CS5164, Electric Power Research Institute, Palo Alto, CA, April, 1987.

30Dooley, R.B., A. Aschoff, and F. Pocock, Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment for Drum Units, Final Report TR-103655, Electric Power Research Institute, Palo Alto, CA, December, 1994. 31Dooley, R.B., A. Aschoff, and F. Pocock, Cycle Chemistry Guidelines for Fossil Plants: All-Volatile Treatment, Final Report, TR-105041, Electric Power Research Institute, Palo Alto, CA, 1996. 32Dimmer,

J.P., G.A. Lamping, O. Jonas, and R. Niebo, Boiler Tube Failure Reduction Program, Research Project 1890-7, Final Report GS-7454, Electric Power Research Institute, Palo Alto, CA, August, 1991. 33Dimmer,

J.P.,L. Ruby, K. Shields, and O. Jonas, Cycle Chemistry Corrosion and Deposition: Correction, Prevention and Control, Research Project 2712-11, Final Report TR-103038, Electric Power Research Institute, Palo Alto, CA, December, 1993. 34Dooley,

R.B., “Cycle Chemistry Related Boiler Tube Failures and Reduction”, Proceedings of the International Water Conference, 50th Annual Meeting, held in Pittsburgh, Pennsylvania, October 23-25, 1989. 35Dooley,

R.B., “A Vision for Reducing Boiler Tube Failures”, Power Engineering, March, 1992, pp. 33-37.

36Dooley,

R.B., “A Vision for Reducing Boiler Tube Failures: Part II”, Power Engineering, May, 1992, pp. 41-42.

29Bursik, A., B. Dooley, and B. Larkin, Guidelines for Oxygenated Treatment for Fossil Plants, , Research Project 1403-45, Final Report TR-102285, Electric Power Research Institute, Palo Alto, CA, December, 1994.

Volume 1: Boiler Tube Fundamentals

1-23

1-24

Introduction and Background

Discontinuous Fe2O3

Original steel surface

Steam

++

Fe

2-

O

Note: = voidage

Outer columnar porous oxide (Fe3O4)

Ferritic steel

Inner equiaxed spinel

Chapter 2 • Volume 1

The Boiler Tube Operating Environment and its Breakdown 2.1 Introduction With the exception of the limitations on SH/RH tube life introduced by long-term creep, boiler tube failures occur because of some deviation or breakdown from the design condition. This chapter begins with an examination of the function and general design considerations for both water- and steam-touching tubing (Section 2.2). The extraordinary ability of tube materials to function in the conditions on the fluid-side of the tube is due primarily to the formation of protective oxides. The reaction of ironbased materials with water and steam is reviewed briefly in Section 2.3. Boiler tube failures (BTF) originating on the fluid-side do so because of some breakdown (mechanical, chemical, or thermal) of the normally protective oxide. This process, the formation and breakdown of protective oxides, is central to the analysis and understanding of BTF. Therefore, two sections (Sections 2.4 and 2.6) look in detail at oxide formation and its breakdown for water-touched and steam-touched tubes respectively. Section 2.5 examines the demands placed on the fluid-side of tubes by looking at the thermal-hydraulic

regimes present in tubes. A distinction is made between “global” thermal-hydraulic regimes and the local conditions that are set up by flow disruption. This chapter ends with an overview of the combustion process and demands placed on the fireside of tubes (Section 2.7). Of particular interest for the analysis of fireside tube failures are both the chemical and the mechanical aspects of ash formation by the combustion process; these will manifest themselves as corrosion or erosion mechanisms. This discussion should provide background for the more complete look at fireside tube failure mechanisms as they are described in the appropriate chapters of Volumes 2 and 3. Although obviously important to the occurrence of BTF, the details of the overall design of boilers are not discussed because of the complexity and the number of varieties. Two comprehensive references that can be consulted for such information are Steam: Its Generation and Use1 and Combustion Fossil Power: A Reference Book on Fuel Burning and Steam Generation.2

Volume 1: Boiler Tube Fundamentals

2-1

2.2 Basic Function and General Design Considerations At its most basic, the purpose of a boiler tube is to transfer heat generated by the combustion process to the water and/or steam within the tube. The tube must have sufficient mechanical strength (a function of its diameter, wall thickness, and material of construction) to contain the internal pressure, as well as adequate resistance to corrosion on both fluid and fireside surfaces. These requirements must be continuously met over the required life of the boiler, usually well in excess of 100,000 operating hours. Finally, economic considerations require that the cheapest adequate materials be used.3 Worldwide these demands are met by a number of international materials and design codes. This section briefly overviews the requirements in the United States. Material choice is a function of expected temperature of operation. Economizers and waterwall sections are usually constructed with a mild or medium carbon steel while low alloy ferritic steels are used for most superheater and reheater sections, with austenitic stainless steels specified for the highest temperature circuits or corrosion performance. Alloys for use as boiler tube materials in the United States are identified by an ASME designation (SA-xxx) or an essentially equivalent ASTM designation (A-xxx). ASME material specifications are found in ASME Boiler & Pressure Vessel (B&PV) Code, Section II, Part A: “Materials Specifications for Ferrous Materials”. The specific grade of material is required as well as the specification number. Table 2-1 provides an overview of the typical boiler tube materials, their basic composition, alloy designation, and general properties.

Table 2-1 Summary of Typical Boiler Tube Materials of Construction Class of Material

Typical Alloy Designations

General Properties

Carbon steel

SA-178: Welded carbon steel SA-192: Seamless carbon-silicon steel SA-210: Seamless carbon-manganesesilicon steel

Hypoeutectoid steels. Mild corrosion resistance. Moderate strength up to 538°C (1000°F). Susceptible to graphitization above 427°C (800°F).

Carbon molybdenum steel

SA-209-T1, T1a and T1b: Seamless carbon-1/2 molybdenum steel SA-250-T1: Welded carbon1/2 molybdenum steel

Greater creep strength than carbon steels. Susceptible to graphitization with prolonged exposure above 468°C (875°F).

Chromiummolybdenum steel

SA-213-T2: Seamless 1/2 chromium 1/2 molybdenum steel. SA-213-T12: Seamless 1Cr-1/2Mo steel SA-213-T11: Seamless 11/4Cr-1/2Mo steel SA-213-T22: Seamless 21/4Cr-1Mo steel SA-213-T21: Seamless 3Cr-1Mo steel SA-213-T5: Seamless 5Cr-1/2Mo steel SA-213-T5b: Seamless 5Cr-1/2Mo-Si steel SA-213-T5c: Seamless 5Cr-1/2Mo-Ti steel SA-213-T7: Seamless 7Cr-1/2Mo steel SA-213-T9: Seamless 9Cr-1Mo steel SA-213-T91: Seamless 9Cr-1Mo1/2V-X (Cb/N/Ni/Al) steel

Most common boiler tube materials (particularly T22 and T11). Each increase in Cr content yields improved properties, particularly higher strength, creep properties, and improved corrosion resistance. Resistant to graphitization.

Austenitic stainless steel

SA-213-TP304/304H: Seamless 18Cr-8Ni austenitic stainless SA-213-TP316/316H: Seamless 16Cr12Ni-2Mo austenitic stainless SA-213-TP321/321H: Seamless 17Cr11Ni-Ti austenitic stainless SA-213-TP347/347H: Seamless 18Cr10Ni-Cb austenitic stainless

Excellent oxidation resistance and good elevated temperature strength. "H" following designation indicates higher carbon content and slightly higher solution heat treat temperature.

Ferritic stainless steel

For use in highly aggressive or high temperature environment.

Martensitic stainless steel

For use in highly aggressive or high temperature environment.

Nonferrous alloys

Nickel-chromium (Alloy 600)

For use in highly aggressive or high temperature environment.

Nickel-chromiumiron

Alloy 800 or 800 H

For use in highly aggressive or high temperature environment.

Source: S.R. Paterson, et al.4

2-2

The Boiler Tube Operating Environment and its Breakdown

Table 2-2 Minimum Tensile and Yield Strengths Tube Steel Type

ASME Specificationa

Grade

Minimum Tensile Strength (ksi)

Minimum Yield Strength (ksi)

Carbon Steel

A list of the minimum tensile and yield strengths of some commonly used boiler tube materials is provided in Table 2-2. The maximum tube metal temperatures for a selection of materials is provided in Table 2-3. Because several materials are usually used in the SH/RH sections, a schematic showing the locations of materials and the transitions is critical, both for tracking materials and for developing the appropriate weld procedures.

Electric resistance welded

SA-178

A C

47b 60

26b 37

Seamless

SA-192

–

47b

26b

Seamless

SA-210

A1 C

60 70

37 40

Electric resistance welded

SA-226

–

47b

26b

Electric resistance welded

SA-250

T1

55

30

Seamless

SA-209

T1 T1a T1b T2

55 55 55 60

30 30 30 30

T5c T9c T11 T12 T22 T91d

60 60 60 60 60 85

30 30 30 30 30 60

Further, some care is required to ensure that the “upgraded” material has the necessary properties to meet all the requirements of the location. For example, increased erosion resistance may not be found in conjunction with increased creep strength in a particular alloy. When looking for an upgraded property of one type, care must be taken to assure that deterioration of another key property is not incurred.

TP304H TP316H TP321 TP347 TP347H

75 75 75 75 75

30 30 30 30 30

An extensive catalog of properties for boiler tube materials including high temperature properties, and the microstructure of original and service-exposed material is available in the EPRI Boiler Tube Failure Metallurgical Guide.4

Ferritic Alloy

SA-213

Seamless

Austenitic Stainless Steel Seamless

SA-213

“Upgrading” materials is, for several key BTF mechanisms, the optimal strategy. For example, where persistent long-term overheating and creep damage is occurring in the lower grade of two materials near a transition, extending the use of the higher grade material can often be the most cost-effective solution. If a material change is made, careful documentation should be made of the new material and its location so that, if needed, repairs can be properly executed.

Notes: a ASME Boiler and Pressure Code, Section I, Power Boilers - Part PG-9, Pipes, Tubes and PressureContaining Parts b Not required by ASME material specification. For purposes of design, these tensile properties may be assumed. c Not commonly used in modern boilers. d Recently approved. Adapted from: G.A. Lamping and R.M Arrowood, Jr.5

Volume 1: Boiler Tube Fundamentals

2-3

2.2.1 Basic considerations for waterwalls and economizers. Under normal conditions in waterwall tubes for subcritical boilers the steam-water temperature is limited to the saturation temperature for the given boiler pressure and thus the tube temperatures are typically less than 400°C (~ 750°F). In supercritical units, the waterwall materials typically operate to slightly higher temperatures (454°C (849°F)). As a result, high temperature creep is not a consideration and waterwall tubes are designed on the basis of short term tensile strength properties and for indefinite life. In practice, however, this goal is not achieved; waterwall failures account for a large fraction of BTF. Also low temperature creep can occur in economizer tubing bends and in the low temperature sections of superheaters and reheaters. Plain carbon steel, such as SA210 or SA192, is most often used for subcritical boilers in North American units with some use of SA213 T11 or T22 (chromium-molybdenum steel); the latter material is usually the material of choice for supercritical units. Maximum allowable design stress as specified in the ASME Code is chosen to be no higher than the lowest of:

Table 2-3 Maximum Tube Metal Temperatures Tube Steel Type

ASME Spec. No.

ASME1 Max. °F (°C)

B&W2 Max. °F (°C)

C-E3 Max. °F (°C)

Riley4 Max. °F (°C)

Carbon steel

SA-178C

1000 (538)5,6

950 (510)

850 (454)

850 (454)

"

SA-192

1000 (538)5,6

950 (510)

850 (454)

850 (454)

"

SA-210 A1

1000 (538)5,6

950 (510)

850 (454)

850 (454)

Carbon Moly

SA-209 T1

1000 (538)7

--

900 (482)

900 (482)

"

SA-209 T1a

1000 (538)7

975 (524)

--

--

Chrome Moly

SA-213 T11

1200 (649)

1050 (566)

1025 (552)

1025 (552)

"

SA-213 T22

1200 (649)

1115 (602)

1075 (580)

1075 (580)

Stainless

SA-213 321H

1500 (816)

1400 (760)

--

1500 (816)

"

SA-213 347H

1500 (816)

--

1300 (704)

--

"

SA-213 304H

1500 (816)

1400 (760)

1300 (704)

--

(i) 1/4 of the specified minimum tensile strength at room temperature; (ii)1/4 of the tensile strength at design temperature; (iii) 5/8 of the specified minimum yield strength at room temperature; (iv) 5/8 of the yield strength at design temperature. Tube metal temperatures are an important factor in the manner in which BTF develop. Tube metal temperatures depend on the heat flux from the fireside, the internal flow rate and the condition of the working fluid.

1 2 3 4 5 6 7

From ASME Boiler and Pressure Vessel Code, Table PG-23.1. This is the highest metal temperature for which maximum allowable stress values are given. From reference 1, 1978 edition, p. 29-11, Table 3. From reference 2, 1981 edition, p. 6-43, Table IV. From reference 6, 1983 edition, p. 263, Table VI. Upon prolonged exposure to temperatures above about 800°F (427°C), the carbide phase of carbon steel may be converted to graphite. Only killed steels shall be used above 850°F (454°C) Upon prolonged exposure to temperatures above about 875°F (468°C), the carbide phase of carbon-molybdenum may be converted to graphite.

Note: Direct comparison of maximum metal temperature is not meaningful without information on design heat transfer analysis and actual material properties. Source: G.A. Lamping and R.M Arrowood, Jr.5

2-4

The Boiler Tube Operating Environment and its Breakdown

Heat transfer through the tube wall is mainly by conduction and involves several temperature gradients, as shown schematically for a subcritical waterwall in Figure 2-1. Furnace gas temperatures near to the wall are typically around 1200°C (~ 2200°F) but a massive gradient exists between this and the tube metal, mainly due to the cooling effect of the internal fluid, but also to low conductivity of the gas boundary layer and fireside scale. Under normal circumstances, without heavy internal deposits, the gradient through the tube metal is small, typically about 25°C (45°F), and a further drop occurs through the waterside magnetite layer and the boundary layer in the steam/water mixture. Important consequences are: (i) tube metal temperature in the steam generating tubes is normally below 400°C (~ 750°F); and (ii) excessive waterside deposit growth raises tube metal temperature by restricting heat flow. Excess fireside corrosion scaling reduces metal temperature in the affected area but can cause problems due to excess heat flux elsewhere in the boiler.3 2.2.2 Basic considerations for superheaters and reheaters. The design of superheater/reheater (SH/RH) sections is more involved. There will be a distribution of increasingly higher metal temperatures through the circuit, necessitating tubes with increasing wall thickness and/or material changes. Thus the primary stages may use carbon steel tubing, followed by progressive change to low alloy ferritic steels for increasing creep and oxidation resistance, and finally to the use of austenitic stainless tubes for the highest temperature sections. Figure 2-2 shows the range of materials and sizes used in the construction of a typical superheater. Each manufacturer specifies a maximum operating temperature for each material, based on laboratory oxidation experiments. Thus, during the design stage, a tube material is used up to the gas-touched length

Fireside scale

Waterside

Boundary film

Boundary film Tube wall

Furnace (~1200°C)

Water (366°C)

Figure 2-1. Schematic of a typical temperature profile through a water wall tube. Source: R.B. Dooley and H.J. Westwood3

(tube heated length within the furnace) at which it is estimated the maximum oxidation temperature is reached. Transition is then made to the next higher grade material. In superheater and reheater tubes, the dry steam temperature has no saturation limitation and is simply determined by the balance between heat flux and internal flow rate. Depending on the design, final steam temperatures of 538 to 565°C (1000 to 1050°F) can require tube metal temperatures in excess of 600°C (~1110°F) in the last stages of the SH and RH sections. Tube materials are initially selected by performing a heat balance analysis. This starts with the gas outlet temperature from the air heater and works backward through all the convection and radiant sections to achieve the desired furnace gas outlet temperature by inserting the appropriate amount of tubing in each section to achieve the rated steam conditions. This is achieved by having bank effectiveness factors which vary with the heat pickup for each individual section or bank of tubing.3

Boiler tubes up to 12.5 cm (5.0 in.) in diameter are designed in accordance with the ASME Boiler and Pressure Vessel Code, Section I. Essentially, the code sets limits on wall thickness based on boiler pressure and operating temperature of the specific tube. Tube diameter is selected by the manufacturer based on experience and is not governed by the Code.3 For SH and RH tubes, the stresses are based on 100% of the stress to produce a creep rate of 0.01%/1000 hr, as determined from the most appropriate available data. In addition, code stresses are limited to 67% of the average stress to produce rupture in 100,000 hours, or 80% of the minimum stress for rupture in 100,000 hours, whichever is lower. RH tubes are relatively thin walled compared with SH tubes since the pressure in the RH is substantially lower than the rated boiler pressure. For example, for boilers operating at 16.2 MPa (2350 psi), RH pressure is typically 3.4 MPa (500 psi).3

Volume 1: Boiler Tube Fundamentals

2-5

The following formulae are currently used for minimum wall thickness and maximum operating pressure: PD T = ÑÑÑÑÑ + 0.005D + e 2S+P

P=

(2T - 0.01D - 2e) S ÑÑÑÑÑÑÑÑÑÐ [D - (T - 0.005D - e)]

Platen section Outlet header Inlet header

(2-1) 2" O.D. x 0.188" min SA-213, T11

(2-2)

where T = minimum wall thickness (in.)

2 1/8" O.D. x 0.281" min SA-213, T11 2 1/8" O.D. x 0.340" min SA-213, T11

2" O.D. x 0.188" min SA-209, T1

2 1/8" O.D. x 0.400" min SA-213, T22

2 1/8" O.D. x 0.440" min SA-213, T11

2 1/8" O.D. x 0.340" min SA-213, T11 2 1/8" O.D. x 0.340" min SA-213, T22

2 1/8" O.D. x 0.420" min SA-213, T22 2 1/8" O.D. x 0.400" min SA-213, TP374H 2 1/8" O.D. x 0.360" min SA-213, TP374H

2" O.D. x 0.188" min SA-213, T11 2 1/8" O.D. x 0.400" min SA-213, T9

D = outer diameter (in.)

x x

S = maximum allowable stress value at operating temperature of metal as given in reference table (psi)

Material Specifications Carbon steel SA-192 Carbon moly SA-209, T1 1 1/4% chrome, 1/2% moly SA-213, T11 2 1/4% chrome, 1% moly SA 213, T22 9% chrome, 1% moly SA 213, T9 SA-213, TP347H Columbium stainless steel

Superheater and reheater tubes operate at temperatures ranging from around 400°C (~ 750°F) to over 600°C (~ 1110°F) depending on location and design. Because of the higher operating temperatures, and as a result of the progressive buildup of internal steamside oxide, which increases tube metal temperatures, they are subject to accumulation of damage by creep, which ultimately causes rupture. Such tubes therefore have a finite service life, in contrast to economizer and waterwall tubes which should in principle last indefinitely.

2 1/8" O.D. x 0.360" min SA-213, T22

2" O.D. x 0.281" min SA-213, 2" O.D. x T11 0.220" min SA-213, T11 2" O.D. x 0.203" min SA-209, T1 2" O.D. x 0.168" min SA-209, T1

x

There is no corrosion allowance, per se, in the ASME Code, but effectively the term 0.005D covers this factor. Also, for cold drawn seamless tubing, ASME SA450 allows +22% on the calculated wall thickness, hence the conservatism in the Code. The maximum allowable stresses are determined on the basis of the operating metal temperature.

Horizontal section Outlet Header

2 1/8" O.D. x 0.320" min SA-213, TP347H

P = maximum allowable working pressure (psi)

e = thickness factor (in.)

2-6

Rear pendant Outlet header Inlet header

Front pendant Inlet header Outlet header

Indicates a dissimilar metal weld Indicated a change in material or wall thickness Horizontal section Ring inlet header

2" O.D. x 0.240" min SA-192

Figure 2-2. Example of boiler superheater tubing materials and sizes per section. Source: G.G. Stephenson and J.W. Prince7

An additional consideration is the formation of wustite (FeO), a nonprotective form of oxide, at temperature in excess of 570°C (~ 1060°F). This can lead to very rapid oxidation rates in superheater/reheater tubes and has led to limiting allowable temperatures for carbon steels to about 454 to 510°C (~ 850 to 950°F); for T-11 to 552 to 566° (~ 1025 to 1050°F) and for T-22 to 580 to 602°C (~ 1075 to 1115°F).

2.3 The Reaction of Iron and Water/Steam: Oxide Formation There is a plethora of information discussing the most likely mechanisms, characteristics, rates of growth and features of the oxides that form on iron-based materials when exposed to air, or more importantly for boiler tube purposes, to

The Boiler Tube Operating Environment and its Breakdown

water or steam (see, for example, references 8 and 9). Why should the practicing utility operator, engineer or chemist be concerned? There are at least two reasons; first, it is the formation of the thin oxide layers that make it possible for these materials to be used in the demanding temperature and pressure conditions of modern boilers. When those oxide layers breakdown, the result is the variety of boiler tube failures described in Volumes 2 and 3. The second reason is that the condition (thickness, morphology, composition) of the oxide scales formed are valuable diagnostic tools. A brief overview of the topic is presented here with additional detail in the appropriate discussions of individual mechanisms.

There are three stable, solid forms of oxide that occur as a result of the reaction of iron and water/steam under temperature and pressure conditions relevant to boiler practice: wustite (FeO), magnetite (Fe3O4), and hematite (a-Fe2O3). Table 2-4 provides an introduction to some of the differences among the three. The conditions which will tend to favor the formation of one of the three are a function of a number of variables including: oxygen concentration, temperature and pH.18 The iron-oxygen phase diagram shown in Figure 2-3 illustrates the temperature-oxygen regimes for each. Since there is a gradient of available oxygen (highest in the water/steam, lowest nearest the inside tube surface), the conditions will often be such that more than one form of oxide can exist. Further complicating the development of oxide will be the formation of multiple layers and laminated structures, topics that are further explored in following sections. Wustite (FeO), the form that is stable with the lowest oxygen concentration, is not stable below temperatures that depend on the alloy content of the steel and range from 560 to 620°C (~ 1040 to 1150°F) on 1Cr1/2 Mo and 21/4Cr-1Mo steels, at which point it decomposes into iron and magnetite. If it forms on the steamside of tubes, it could be between the tube metal and the predominant magnetite layer. As noted above, wustite is of concern because of the potential for accelerated oxidation if it forms. This has been suggested for the growth of multilaminated oxides in steam; FeO is not seen in the scales after shutdown. Magnetite (Fe3O4) is the predominant form of oxide. It exists over a wide range of oxygen partial pressures and temperatures.4 Hematite (a-Fe2O3), stable in the highest oxygen concentrations will form in the outermost layer of the oxide.

Table 2-4 Overview of Oxides of Iron Characteristic

Magnetite

Hematite

Wustite

Composition

Fe3O4

a-Fe2O3

FeO

Structure

Face-centered cubic, spinel

Growth mechanisms

Both cations and anions diffuse.

n-type conductor. Growth involves mainly anions.

p-type conductor. Growth involves mostly cations.

Stability

Above 560°C (1040°F); below it decomposes to Fe3O4 and iron.

Position in oxide layers

Predominant layer in typical oxides throughout boiler.

If found, will be on the outermost layer of the oxide nearest the water/ steam.

If found, it will be on the steamside between the tube metal and the predominant magnetite. In ferritic alloys, FeO occurs between the alloy spinel and the Fe3O4.

Oxygen Levels at Formation

A wide range of oxygen partial pressures

Highest

Lowest

Hardness (HV)

450-550

> 1000

250-350

Density (g/cm3)

5 - 5.4 (Ref. 10)

5.24 (Ref. 10)

5.7

Thermal conductivity (W cm-1 K-1)

0.0423 - 1.37 x 10-5 T (Ref. 11)

0.0423 - 1.37 x 10-5 T (Ref. 11)

Tensile fracture strain (x 104)

5 - 30 (Refs. 12, 13)

Young's modulus ( x 10-10 Nm-2)

14 - 26 (scale) (Ref. 12)

Relevance to analysis of BTF

Protective form of oxide. Its breakdown by chemical and/or mechanical means is at the root of most BTFs.

1-3 (Refs. 14, 15) 12.2 (scale) (Ref. 16)

12.8 (bulk) (Ref. 17) Is a non-protective form; if found can lead to rapid oxidation of SH/RH tubes. Indicator of overheating.

Volume 1: Boiler Tube Fundamentals

2-7

The growth mechanism involves a flux of iron ions from the steel to the outside interface of the oxide and an equal flux of oxygen ions from the boiler water to the steel surface, as illustrated schematically in Figure 24. There has been much discussion about whether these fluxes of iron and oxygen occur by a solid state diffusion mechanism through the oxide or whether the growth consists of diffusion of the reactants in liquid (boiler water) filled pores.3

Atom % Oxygen 54

56

58

Liquid oxide

Iron + liquid oxide Iron + liquid oxide

2600

WŸstite

Temperature (°C)

1200

Iron + wŸstite

1000

WŸstite + magnetite

3000

Magnetite + oxygen

Liquid oxide + magnetite

Hematite + oxygen

Liquid iron + liquid oxide

1400

60

Liquid oxide + oxygen

Magnetite

1600

52

Magnetite + hematite

2200

1800

Temperature °F

50 Liquid iron

800 1400

Iron + wŸstite

600 1040°F FeO

400

0

0.4 22

Iron+ magnetite

24 26 Weight % Oxygen

1000

Fe3O4

Fe2O3

28

30

Figure 2-3. Iron-oxygen phase diagram.

2.4 Oxide Development and Breakdown in Water-Touched Tubes 2.4.1 Growth of protective magnetite. The growth of protective magnetite which occurs adjacent to the tube wall in water-touched tubing under typical boiler conditions was first characterized around thirty years ago.12, 19, 20 The protective layer of magnetite (Fe3O4) is formed by the reaction between iron and neutral, or alkaline water.3 3 Fe + 4 H2O ® (Fe3O4) + 4 H2

2-8

(2-3)

It grows as a tenacious and coherent film which then impedes transport and diffusion; as a result, the rate of transport decreases with time. The rate of oxide formation is initially high but decreases as the layer thickness increases (parabolic growth) and becomes self-limiting. In actual practice the outer layer is infrequently formed because as it does it becomes entrained in the boiler water flow and is redeposited with feedwater corrosion products in other regions, maybe of higher heat flux. Even after years, the oxide layer is normally only a few micrometers thick (10mm - 15mm)21, yet will still provide protection to the boiler tube until mechanically cracked or fluxed away.

The Boiler Tube Operating Environment and its Breakdown

A number of explanations for how the various species are transported through the oxide include8: • Vapor phase transport of water vapor inward and ‘volatile’ Fe(OH)2 outward through liquidfilled pores22; • Solid-state, countercurrent diffusion of iron and oxygen ions through lattice defects, the twolayer structure arising because the outer part of the magnetite behaves as a metal deficit semiconductor, and the inner part as a metal excess semiconductor23; • Outward cation solid-state diffusion coupled with accelerated inward transport of water vapor across disconnected pores via a redox-type reaction24; • Inner layer growth by transport of water along pores to the scalemetal interface with rate of corrosion limited by outward diffusion of cations along oxide grain boundaries. Outer layer growth limited by the amount of outwardbound iron ions which depends on various corrosion reactions involved (compensating current carried by electron or proton transfer through the oxide or an external circuit through the water)25; The growth of Fe3O4 thus occurs at the two interfaces and, because of the equal fluxes, the growth rates are the same. As a result the oxide grows in a stress-free situation, and is therefore protective. The mechanism described varies very little in boiler water treated with caustic,

phosphate, or in boiler treated by all-volatile or oxygenated treatment.

Boiler water B O2

-

Protective Fe3O4

Fe2+ A Tube wall

Figure 2-4. Schematic of counter-flux diffusion growth of protective Fe3O4 in boiler water. Most often the outer layer is not present. Source: R.B. Dooley and H.J. Westwood3

a) AVT Treatment Deposit Loading, mg/cm2 120 pH = 9.4 pH = 9.2 pH = 9.0 Ð 9.2 pH = 8.5 Ð 9.1 pH = 8.0 Ð 8.5

100 80 60 40 20 0 0

10000

20000 30000 40000 50000 Service Exposure (hours)

60000

b) Combined or Neutral Oxygenated Treatments Deposit Loading, mg/cm2 120 Combined, pH = 7.6 Ð 8.2 Combined, pH = 6.6 Ð 7.8 Neutral, pH = 6.3 Ð 7.2

100 80 60 40 20 0

0

10000

20000 30000 40000 50000 Service Exposure (hours)

60000

The growth of waterside Fe3O4 occurs in economizer tubing in a similar fashion to that in the waterwall, although there is much less feedwater corrosion product deposition which is a function of heat flux and heat transfer. Further, waterside surfaces are generally more uneven and contain more pits than those in waterwalls. 2.4.2 Scale in waterwall tubes. The total amount of scale/deposits on a boiler tube water wall (either subcritical or supercritical once-through units) consists of the protective Fe3O4 described above plus a layer/thickness of deposited material, primarily feedwater corrosion products. This material can consist of Fe3O4, Cu, Zn, Ni, and other metals and oxides which are transported from the feedwater system and deposit on the water wall surfaces. The total thickness controls the heat transfer and is also the reason that boilers have to be chemically cleaned. The rate at which such deposits accumulate is a function of: (i) local heat flux, (ii) local thermal-hydraulic features, and (iii) boiler water chemistry.4 The typical effects of unit chemistry are shown in Figure 2-5 which contrasts the rate of deposit buildup for all-volatile treatment (AVT) and oxygenated treatment (OT) for some units in Russia. The experience can vary markedly between units even operating under the same chemistry. For example, scale densities of only 25 mg/cm2 have been found after 70,000 hours of operation for units under AVT. The effects of cycle chemistry and treatment options are explored in more detail in Chapter 3 of this volume.

Figure 2-5. Water wall deposit weight expressed as function of service hours for several water treatment methods. Note the dramatic benefit of oxygenated water treatments in eliminating rapid waterside deposits. Source: I.I. Chudnovskaya, Central Boiler and Turbine Institute, St. Petersburg, Russia

Volume 1: Boiler Tube Fundamentals

2-9

In supercritical and once-through boilers the oxides and metals depositing in wall tubes can form ripples under some circumstances. Figure 2-6 shows the typical surface appearance, a magnification through the cross-section of the ripples is shown in Figure 2-7. Although a complete understanding of the phenomenon is not in-hand, the following observations have been made about the process26: • Rippled surfaces offer greater flow resistance than normal smooth magnetite profiles and are generally responsible for increasing boiler pressure drop. • Heat flux seems to have little effect on formation. Ripple dimensions are the same on the fireside and non-fired side of tubes, however, ripple spacings are related to the flow velocity in the tube. 2.4.3 Breakdown of protective magnetite and the resulting BTF. While the protective magnetite remains intact, the tube is generally protected. However, once this layer is cracked, removed, fluxed away, or grows at an accelerated rate, then the protection can be considered to be lost. The manner in which the magnetite scale is modified, chemically by various contaminant species, or mechanically by the application of locally excessive strain levels, leads directly to the manifestation of various damage types.

Figure 2-6. Photograph showing "ripple" deposits on the inside surface of a supercritical waterwall tube.

Table 2-5 lists some of these breakdown processes and the associated boiler tube damage mechanisms for water-touched tubing. Several are associated with the excessive buildup of internal deposits, primarily as a result of feedwater corrosion products. The chemical attack of magnetite under such deposits can be rapid at temperatures consistent with those in an operating boiler. The Figure 2-7. Cross sectional view through ripple magnetite.

2-10

The Boiler Tube Operating Environment and its Breakdown

Table 2-5 Protective Magnetite Breakdown and the Resulting BTF Mechanisms in Water-Touched Tubing Boiler Tube Failure Mechanism

Nature of Breakdown Conditions

Chapters for Additional Information

Normal Condition

Counter flux of O2- and Fe2+. Oxide grows stress-free by a parabolic growth law.

Main text, this section

•Caustic Gouging •Acid Phosphate Corrosion

Underdeposit fluxing of magnetite followed by diffusion of locally concentrated corrosion species to form characteristic deposits.

Chapters 16, 17, Vol. 2

Hydrogen Damage

Underdeposit modification of magnetite growth mechanism to linear accumulation with internal strains; followed by chemical attack by locally concentrated, low pH fluid.

Chapter 15, Vol. 2

Corrosion Fatigue

Mechanical strain applied to oxide causes microfissures that are susceptible to environmental attack from bulk conditions.

Chapter 13, Vol. 2

Pitting

Primarily caused by stagnant, oxygenated water formed during shutdown which attacks oxide as generalized corrosion.

Chapter 27, Vol. 2

Chemical cleaning damage

Manifested as generalized corrosion because of direct chemical attack of oxide

Chapter 25, Vol. 2

rate of attack under typical bulk boiler water operating conditions as well as under either high or low pH is shown in Figure 2-8. The figure also provides an indication of the solubility of Fe3O4 under these conditions. It is this accelerated attack at pH levels outside a narrow range, and the nature of changes to the protective magnetite (specifically mechanisms for concentration of deposits), that underlies many of the waterside, waterwall and economizer boiler tube failure mechanisms. Under locally high pH conditions, caused by a concentration of sodium hydroxide (leading to caustic gouging) or alternatively a concentration of low Na: PO4 ratio phosphate liquids or solutions (leading to acid phosphate corrosion), a fluxing reaction removes the magnetite. There is a subsequent transport of

corrosion species to deposit at the outer layer. In the case of caustic gouging, the presence of concentrated solutions of sodium hydroxide results in the formation of characteristic crystals of sodium ferroate within the corrosion product. The magnetite breakdown reaction for these two waterside, underdepositcorrosion mechanisms is thus primarily chemical in the presence of concentrated corrosive solutions. In contrast, in the presence of acidic chloride solutions, Potter and Mann20 found that the growth of the oxide scale itself is changed. Specifically the oxidation rate of mild steel becomes linear and the magnetite scale formed is stratified or laminated. Further, it has been shown that the oxide that forms is internally stressed. Thus the oxide growth process is affected both chemically and mechanically for the case of acidic chloride contamina-

tion.27 Additional detailed discussion is presented in Chapter 15, Volume 2, in the section on the hydrogen damage mechanism. Dissolved salt contaminants such as chlorides and sulfates are of concern as they can affect the morphology, formation rate, thickness, and strength of the protective layer.31 In addition to modification of the way in which the scales forms (with subsequent chemical attack), the magnetite film can be damaged directly either by chemical means (corrosion) or by mechanical means (strain), or by the synergistic effect of the two.28 Destabilization primarily by chemical means usually occurs at pre-existing active sites31, resembles pitting and has sometimes been termed stress-assisted pitting. When the film is fractured primarily by strain of the substrate tube, corrosion paths are produced, leading to an array of cracks which is generally termed corrosion fatigue in boiler tubes. Rupture of the protective oxide film leads to more rapid damage by corrosion fatigue because (i) additional base metal is exposed to corrosion damage and (ii) the rupture, which is a crack or crack-like, acts as a stress concentrator. That there is a lower bound or critical level of strain that is required to fracture the protective oxide film and begin the corrosion fatigue damage process is supported by the absence of corrosion fatigue in boilers where the design considers the magnetite strain tolerance, and from modeling studies of oxide. The critical strain to fracture magnetite at high temperatures is generally reported to be between 0.01 and 0.1% strain.15, 29, 30, 31 The German design standard TRD 301, for example, requires that the strain level in tube oxide be kept below a certain limit (about 0.1% strain) during operation to avoid rupturing the magnetite scale.

Volume 1: Boiler Tube Fundamentals

2-11

depends critically on the maintenance of satisfactory boiling conditions within the tube.

a) 100.0

Solubility (m.mol.kg-1) Corrosion (mm/year)

2-

Fe2+

10.0

Fe(OH)4

1.0 0.1 Fe(OH)-3

0.01 0.001

Corrosion

FeOH+

0.0001

Fe(OH)2 3

2

4

5

6

Rapid acid chloride attack with hydrogen damage

pH 300°C

7

8

9

10

Normal bulk boiler water operating region

11

12

13

Rapid caustic attack

b) Compound Common name

Approximate pH of a 1% solution @ 25°C

HCl or H3PO4

NaH2PO4

Na2HPO4

Na3PO4

NaOH

Hydrochloric acid or phosphoric acid

Monosodium phosphate

Di-sodium phosphate

Tri-sodium phosphate

Sodium Hydroxide

1.5

4.8

8.8

12.0

13.5

Figure 2-8 a) Corrosion of mild steel and solubility of magnetite at 300°C. b) Some common contaminants and boiler water treatment chemicals.

2.5. Overview of ThermalHydraulic Regimes and Waterside BTF In previous sections it has been noted that local mechanical and chemical conditions can lead to a breakdown in the protective magnetite of tubing, but how is it possible that such conditions can be set up within the normal flow conditions of a boiler? Many past studies of the basic hydraulic and thermal conditions are the key to understanding the mechanisms of deposition of feedwater corrosion products and of

2-12

the concentration of contamination that led to the breakdown of the protective magnetite (see for example, Masterson, Castle and Mann27). 2.5.1 Global thermal-hydraulic regimes. Figure 2-9 illustrates the different global regimes of fluid condition in an idealized boiler tube fed by a bottom header. Over the major portion of the boiler tube, there is an annular flow of water at the tube inside diameter and a core of steam; a condition that occurs at a steam quality of about 5% and greater. This is also called nucleate boiling. Control of metal temperatures in waterwall tubes to avoid BTF

The Boiler Tube Operating Environment and its Breakdown

The effect of increasing heat flux on metal temperatures is shown schematically in Figure 2-10. Up to point A in the figure, water heating without boiling takes place; this is the situation for economizer sections that perform the final stages of the feedwater preheating. Between A and B, the local heat flux is sufficient to cause local or “nucleate boiling” at the solid-liquid interface but the steam bubbles so formed condense in the bulk fluid, helping to raise its temperature. This condition is termed “sub-cooled nucleate” or local boiling. Beyond B, the bubbles do not collapse and nucleate boiling with net steam evolution occurs up to point C. High rates of heat transfer exist during the sub-cooled and saturated boiling stages so that the tube metal temperature does not greatly exceed the saturation value. Beyond C, bubble coalescence begins to form a superheated steam film over part or all of the heating surface, the condition known as film boiling. From D to E, the film boiling is unstable, beyond E stable film boiling exists. When the local heat flux exceeds that at D, the tube metal temperature may increase very rapidly to D', which may well result in tube rupture if this overtemperature condition persists. For example, the tube can reach over 850°C (~ 1560°F), at which temperature rupture will ensue in a matter of minutes. D is clearly a key point and the adjacent point C is the point of departure from nucleate boiling, (DNB), or the critical heat flux. Factors promoting DNB or steam blanketing are: increases in heat flux, high steam quality (percentage of steam in the two-phase mixture), tube geometry, and pressure. Curves are available giving limiting values for specific boiler conditions. From the tube design standpoint, the onset of DNB can be delayed if the internal surface acts to create

Drum

F Steam Quality 20%

Annular flow

D'

C Log Heat Flux

Boiling Mode

D

E

B A

Churn Slug Bubbly Sub-cooled nucleate

5%

Log Metal Temperature Bottom header

Figure 2-9. Schematic representation of two-phase flow in a heated vertical tube of a drum boiler. Source: H.G. Masterson, et al.27 after P.M.C. Lacey and G.J. Kirby32

turbulent conditions tending to disrupt steam film formation. This has been successfully achieved with rifled tubes, for example, which are more expensive than plain tubing, but may be specified for high heat flux areas in boilers operating above 15.2 MPa (2200 psi). 2.5.2 Local thermal-hydraulic conditions. The processes described above present a global view of the thermal hydraulic conditions. However, it is the understanding of local boiling conditions, tube metal temperatures and heat flux which provides insight to subsequent discussion of boiler tube failures that occur in association with underdeposit corrosion. Clearly, if the normal situation, as described above and by Figure 2-9, persists during all operating periods,

Figure 2-10. The relationship shown schematically between increasing heat flux and metal temperature on a water-touched tube. Source: R.B. Dooley and H.J. Westwood3

then boiler tube problems should not develop. It is when the annular flow regime breaks down that BTF can initiate. The key feature is that the flow of water and the nucleate boiling process adjacent to the tube wall is disrupted and a local steam “blanket” (an area of high steam quality) is formed. Steam blanketing and dryout are caused by any of several different sets of conditions, each requiring a different route to prevent repeat failures. These include: stratification, excessive deposits, crevices, reduced fluid flow, excessive heat flux, and local flow pattern obstructions. Stratification refers to segregation of water to the bottom and of less dense steam to the top of a sloped or horizontal tube. This is the most common type of steam blanketing where a blanket of steam separates an area of tube surface from the adjacent water. Porous internal tube deposits can obstruct free flow of water to the tube surface and conduct heat inward; consequently, where there are thick deposits, boiling occurs away from the tube sur-

face and only superheated steam (or high boiling point concentrates) reaches the tube/deposit interface. Local flow pattern obstructions can cause low flow nodes immediately downstream of the obstruction without substantially altering the total flow through the tube. Examples are weld backing rings, penetrations of weld from poor repairs (such as pad welding, weld overlay, or canoe pieces) or corroded tube material laps. Such areas can be quite small, down to a size of 6 mm (1/4 in.). Local conditions that exacerbate deposition are listed in Table 2-6. These lead to a cascading set of problems that ultimately result in waterside BTF. Whereas a clean tube surface manifests only a slight rise in tube temperature for an increase in heat flux, once a deposit forms on the surface (such as excess oxide or deposition of feedwater corrosion products) there may be a marked rise in the tube temperature. This is illustrated schematically in Figure 2-11.

Volume 1: Boiler Tube Fundamentals

2-13

Table 2-6 Local Tube Conditions Which Can Lead to Initial Deposits Locations where the water/fluid flow adjacent to the tube wall is disrupted • Welded joints such as: • Joints with backing rings • Poor repair welds such as pad welds, "canoe" pieces or window welds • Poor weld overlay (penetrating to the inside surface) • Locations with existing internal deposits caused by: • A deposition mechanism • Deposits left from improper chemical cleaning • Locally high heat flux • Locally high steam quality • Geometric features • Bends around burners or openings • Sharp changes of direction (such as the nose of the furnace) • Tubes bending off lower headers and drums Locations with a high heat flux Locations where boiling first initiates Locations with thermal-hydraulic flow disruptions • Locations with local very high steam quality • Locations with horizontal or inclined tubing heated from above or below

Metal Temperature

Localized overheating of the tube (fireside conditions) • Flame impingement • Burner misalignment • Operating conditions such as overfiring or underfiring, gas channeling, or inadequate circulation rates • Drastic change in fuel source, such as higher BTU value coal, dual firing with gas, changeover to oil or gas firing where heat flux increases.

Salt concentrated 1000-fold

(a) Clean surface

Salt concentrated 100-fold Salt concentrated 10-fold

Deposit on surface

Deposit concentration generally occurs by some form of “wick boiling” whereby the contaminants within the bulk boiler water are able to penetrate the porous tube deposits and the moisture is then driven off by heating of the tube surface locally leaving behind the concentrated contaminants or chemicals. Figure 2-12 illustrates one variety of this process. This process of flow disruption, formation of deposits, and concentration of chemicals or contaminants provides a means by which concentrated liquids can stay in contact with a susceptible tube surface for a sufficient length of time for corrosion to occur. It is the basic process by which three of the most pervasive boiler tube failure mechanisms (hydrogen damage, caustic gouging and acid phosphate corrosion) develop. There are distinctions among the three essentially involving different concentrating solutions; but the basics for all of them are the disruption of flow and nucleate boiling, generating locally high quality steam areas which promote deposition, and subsequent concentrating reactions.

2.6 Oxide Development and Breakdown in Steam-Touched Tubes

(b)

Heat Flux

Figure 2-11. Effect of tube deposit on the heat transfer surface as a function of heat flux. Curve (a) shows a clean heat transfer surface; Curve (b) a surface carrying a thick deposit. The salt concentration factor in the deposit is also shown. Source: H.G. Masterson, et al.27

2-14

Increased damage is caused by concentration in the forming deposits of the corrosive substances contained in the boiling water. Estimates of the potential concentrating factor range from 10,000 times27 to over 100,000 times.33 Masterson, et al.27, emphasize that to give comparable corrosion rates, sodium hydroxide must concentrate by a factor of ten to one hundred times more strongly than acid chloride. This explains why an ingress of caustic from a breakdown in the makeup system doesn’t result in caustic gouging, whereas an ingress in acidic chloride nearly always results in hydrogen damage.

The Boiler Tube Operating Environment and its Breakdown

A protective oxide forms on the carbon, low-alloy and austentitic steels used in SH/RH tubing exposed to steam that is analogous to that for water-touched tube materials. The initial formation is rapid and in the

absence of heat flux decreases with time in a “parabolic” manner. Eventually for the ferritic materials, the layers formed become laminated and multilaminated and the rate changes to linear. The scales formed are then susceptible to exfoliation. The processes of oxide formation, change to a laminated structure and subsequent exfoliation are all normal occurrences of SH/RH tubing operating in these regimes.

Steam liberation

Solution replenishment

Heat transfer section

An excess of steamside scale can lead to increased tube temperatures which is a contributor to damage in superheater/reheater tubes by longterm and short-term overheating, fireside corrosion, and graphitization mechanisms. Table 2-7 lists the BTF mechanisms that are associated with an excess of steamside scale and/or the breakdown of protective oxides from the steamside.

Applied heat

Figure 2-12. Schematic representation of boiler salt concentration in a thick porous deposit. Source: H.G. Masterson, et al.27

Table 2-7 Protective Oxide Breakdown and the Resulting BTF Mechanisms in Steam-Touched Tubing Boiler Tube Failure Mechanism

Nature of Breakdown Conditions

Chapters for Additional Information

Normal Condition

Counter flux of O2- and Fe2+. Initial oxide grows stress-free by a parabolic growth. Later multilaminated oxide structures are formed according to a linear growth law. It is normal for these oxides to exfoliate.

Main text, this section

Fireside Corrosion

Although a fireside process, result is exacerbated in tubes by overheating caused by excessive steamside oxide growth (usually multilaminated).

Chapters 33 & 34, Vol. 3

Short-Term Overheating

Primarily caused by exfoliation of steamside oxide leading to tube blockage and the resultant rapid overheating.

Chapter 36, Vol. 3

Long-Term Overheating (Creep)

Primary cause is overheating of tubes due to inadequate initial design. Tube temperatures are elevated as the steamside oxide increases in thickness.

Chapter 32, Vol. 3

Pitting

Caused by stagnant, oxygenated water formed during shutdown which attacks oxide as generalized corrosion or the stagnant water can be acidified by mechanical carryover of sulfate in steam.

Chapter 41, Vol. 3

Chemical cleaning damage

Manifested as generalized corrosion because of direct chemical attack of oxide.

Chapter 43, Vol. 3

Volume 1: Boiler Tube Fundamentals

2-15

The processes are examined in more detail in the following sections beginning with oxide development in ferritic materials. 2.6.1 Growth of steamside oxide in ferritic materials. Growth of steamside oxide in ferritic materials is illustrated schematically in Figure 2-13. As for water-touched tubing, a counter diffusion mechanism (oxygen ions to the tube surface and Fe ions to the outer layer) occurs. The mobility of the alloying elements is considerably less than that of iron in the oxide and thus they stay at the inner layer at a level above that in the original material.

Discontinuous Fe2O3

Original steel surface

Steam

Fe

2-

O

Note: = voidage

Ferritic steel

Outer columnar porous oxide (Fe3O4)

++

Inner equiaxed spinel

Figure 2-13. Schematic of general two layered oxide growth in steam. Fe++ ions move outwards and O2- ions move inwards. Sometimes the inner layer also becomes laminated with alternating Fe3O4 and chromium-containing layers. Fe2O3 is often seen at the oxide/steam interface.

The inner layer is an equiaxed Fe, Cr oxide spinel. It has a very fine grain size (< 0.1 mm) and the spinel structure contains the alloying constituents (Cr, Ni, Si, etc.) at about 1.5 times their concentration in the metal with the iron concentration reduced accordingly. Etching reveals that the inner layer is composed of finegrained equiaxial grains. In the case of high temperature overheating this normally dense layer may contain porosity.4 The outer layer of iron oxide consists of coarse columnar grains of Fe3O4, but the grains become smaller and equiaxed near the base of the outer layer. A layer of Fe2O3 may be present on the outer surface of the oxide and within the outer layer depending on the oxidizing conditions and the oxidation rate at that time. The Fe2O3 has a finer grain size than the Fe3O4 and appears as a lighter phase in optical micrographs of polished sections, Figure 2-14. The boundary between the two layers is at the original tube surface. Figures 2-14a and 2-14b show that the layers are of about equal thickness and parallel sided. Unlike water-touched tubing where the outer layer is generally missing, for steam-touched tubes it is always present.

Figure 2-14. Appearance of Fe3O4, and Fe2O3 (finer grain size and lighter appearance) iron oxides on a ferritic (T22) superheater tube after 90,000 hours of operation at 2500 psi. (a) Optical photomicrograph (MAG:200 X). (b) SEM Fractograph. Source: S.R. Paterson, et al.4

2-16

The Boiler Tube Operating Environment and its Breakdown

The formation of the oxide occurs initially parabolically but at a later stage can become linear. Any deviation from the parabolic growth is associated with multilaminated scale. A two duplex layer structure is shown schematically in Figure 2-15; the process simply repeats to form multilaminated oxides. Exfoliation, discussed in more detail below, will also only occur when a multilaminated structure forms, and it will occur at the interface between the two duplex structures. If more than two duplex layers form, which is typical, the exfoliation will still occur between layers 2 and 3 in Figure 215. In addition to the protection provided, the oxide layers formed on the steamside of carbon and lowalloy steel tubing are important because they provide useful information about the service temperatures to which the tube has been exposed. Steamside scale will increase the tube temperatures by approximately 0.28°C (0.5°F) per 0.025 mm (0.001 inch) of oxide thickness in typical reheater tubes and by 1.67°C (3°F) per 0.025 mm (0.001 inch) for typical superheater tubes.4 Oxide scale analysis is a powerful tool to predict remaining life of superheater/ reheater tubes. This subject, along with a discussion about the reaction kinetics of oxide growth in SH/RH tubes is included in Chapter 8 of this Volume. As an example of the detrimental effects of the scale, a 0.508 mm (0.020 inch) thick scale will increase the typical superheater tube metal temperature by about 33°C (60°F) which will result in an increase in damage accumulation by creep by a factor of more than five. It will also increase the propensity for fireside corrosion. 2.6.2 Growth of steamside oxide in austenitic materials. Austenitic stainless steels generally corrode more slowly than ferritics under the same steam conditions because of their higher chromium content. As a result the scales are somewhat thin-

Fe2O3

Steam

1 2 3 4

Exfoliation

Original duplex Second duplex Ferritic steel

Figure 2-15. Schematic of the start of multilaminated oxide growth on ferritic steels. Layers 1 and 3 are essentially pure Fe3O4, 2 and 4 are Fe, Cr spinel. Exfoliation occurs between 2 and 3. When more laminations are present the exfoliation occur at the same interface between layers. Fe2O3 is often found at the outermost interface.

Fe2O3

Steam Porous Fe3O4

Original steel surface

Compact spinel Austenitic steel Note:

= voidage

Figure 2-16. Schematic of two-layered oxide growth on austenitic materials. Note irregularity and keyed nature of inner Fe, Cr spinel. Also note the voidage at the oxide/oxide interface; exfoliation occurs at this interface.

ner. Two-layered scales do form but with some marked differences. The situation is shown schematically in Figure 2-16. The inner layer forms in a more irregular manner. Grain boundaries in the steel affect the penetration of the oxide and consequently the metal/inner oxide interface is uneven and ‘keys’ the inner layer to the metal. The outer layer is columnar as for ferritics but contains more voids; as the oxide increases in thickness these voids accumulate near the oxide/oxide interface. Figure 2-17 shows a typical example. As with ferritic materials, a small amount of Fe2O3 is usually observed at the outer interface of the Fe3O4 layer. The oxide/oxide interface again occurs at the position of the original steel interface and remains flat irrespective of the oxide irregularities. The initiation of exfoliation in

austenitic materials occurs at the oxide/oxide interface and results in a single oxide layer and a lower chromium (and alloy element) concentration. The importance of exfoliation in SH/RH tubes is discussed next. 2.6.3 Exfoliation of steamside oxide scale and its effects. Steamside oxide (duplex on austenitic stainless steels and multilaminated on ferritics) can exfoliate (spall) because of the difference in the coefficient of thermal expansion between the base material and the oxide and also because of natural growth stresses in laminated oxides. Exfoliated oxide can collect in the lower U-bends of superheater and reheater tubes, thus restricting

Volume 1: Boiler Tube Fundamentals

2-17

steam flow and causing BTF. Even more importantly, the exfoliated oxide can be carried into the turbine causing solid particle erosion damage to blades, nozzles, and control valves. A rating system for steamside scale exfoliation has been developed and is shown in Table 2-8.4 Such information can provide an estimate of the potential for downstream damage. Figures 2-18a through 2-18e illustrate a progression of boiler tubes with increasingly severe levels of exfoliation.4 Some key points about the exfoliation of SH/RH materials include:

Figure 2-17. Appearance of steam grown oxide on an austenitic (321H) superheater tube. Note the uneven penetration of the inner spinel into the steel and accumulation of voids near to the oxide/oxide interface.

Table 2-8 Rating System for Steamside Oxide Scale Exfoliation Exfoliation Rating

Percent of Surface Area Exfoliated

Approximate Weight of Magnetite Released (Exfoliated) Per Unit Length of Tube g/m (lbm/100 ft)

1 (Little to None)

40

> 125 ( > 8.5)

Note: Approximate weight estimate based on an assumed density of 5.18 g/cm3 for iron oxide, a steamside scale thickness of 0.038 cm (0.015 in), and a tube inside diameter of 5.08 cm (2.00 in.). Source: S.R. Paterson, et al.4

2-18

The Boiler Tube Operating Environment and its Breakdown

• Susceptible materials. Carbon steel, low alloy steel and austenitic stainless steel are all susceptible to spalling as shown in Table 2-9 which presents data from a reheater with 63,430 hours of service and 450 starts.34, 35 • Failure criterion. Figure 2-19 illustrates an oxide failure criterion for austenitic materials. Failure will occur when the energy stored in a strained scale is greater than the energy required to generate new surfaces by delamination of the scale. Sources of strain from operation include: cooling strains (developed due to differential thermal contraction of a layered structure when cooled from one uniform temperature to a lower uniform temperature), removal of steady heat flux, thermal shock (caused by rapid cooling), oxide/oxide transformation, and flexural and other strains imposed by the system.

a. Exfoliation Rating = 1 (Little to No Exfoliation) Exfoliated Area = 0% Weight of Magnetite Exfoliated Per Length of Tube = None Steamside Oxide Thickness = 0.069 cm (0.027 in)

b. Exfoliation Rating = 2 (Mild) Exfoliated Area = 4% Weight of Magnetite Exfoliated Per Length of Tube = 7 g/m Steamside Oxide Thickness = 0.0216 cm (0.0085 in)

c. Exfoliation Rating = 3 (Moderate) Exfoliated Area = 14% Weight of Magnetite Exfoliated Per Length of Tube = 38 g/m Steamside Oxide Thickness = 0.033 cm (0.013 in)

d. Exfoliation Rating = 4 (Severe) Exfoliated Area = 32% Weight of Magnetite Exfoliated Per Length of Tube = 100 g/m Steamside Oxide Thickness = 0.038 cm (0.015 in)

e. Exfoliation Rating = 5 (Very severe) Exfoliated Area = 41% Weight of Magnetite Exfoliated Per Length of Tube = 78 g/m Steamside Oxide Thickness = 0.023 cm (0.009 in)

Figure 2-18. Comparison of the appearance of the steamside oxide scale for reheater tubes with varying degrees of exfoliation. Source: S.R. Paterson, et al.4

Volume 1: Boiler Tube Fundamentals

2-19

Table 2-9 Oxide Thickness on Reheater Tubes and Percent Exfoliated Material

Average thickness, mm (mils)

Percent Exfoliated

T-11

70 ( 3)

< 10

T-22

270 ( 11)

T-9

150 ( 6)

2.8

PO4, ppm

> 2.5

< 2.5

Cycle Chemistry Parameter

NaOH, ppm

1.0 - 1.5

Note: Values are normal operating limits for a coal-fired 2500 psi boiler with reheat. Sources: For PT & EPT (R.B. Dooley, A. Aschoff, and F. Pocock13) For CT (M. Ball16) For AVT (A.F. Aschoff, Y.H. Lee, D.M Sopocy, and O. Jonas8)

If satisfactory values cannot be obtained, the reasons for the high values should be investigated and, if necessary, the concentration of impurities and conditioning chemicals in the boiler water should be reduced. The EPRI interim guidelines, issued in the mid-1980s, make allowance for up to 5 ppb sodium in the steam.8 More recently, a detailed study by Ball16 suggests that ideally to prevent deposition, the steam from high pressure boilers should contain no more than 2 ppb sodium; well operated units should achieve less than 1 ppb. Benefits of CT over AVT include: (i) a higher tolerance to chloride in the boiler water which is important for dealing with condenser leaks, particularly in units using brackish or sea water for cooling, (ii) a reduced

3-6

risk of acid corrosion, and (iii) removing the need for a condensate polishing plant.16 Compared to phosphate treatment, CT can: (i) offer reduced risk of corrosion due to the ingress of chloride, (ii) avoid phosphate hide-out and associated complications, (iii) avoid the complications of monitoring and chemical control associated with phosphate chemistry, and (iv) reduce the risk of producing either acidic or alkaline boiler water conditions, causes of several corrosionrelated boiler tube failure mechanisms, as well as reduce the risk of general problems throughout the cycle.16 3.2.4 Optimization of drum boiler treatment. The correct choice of boiler water chemistry is vitally important to the prevention of boiler tube failures. For any method of chemical conditioning, it is important that the boiler is kept as clean as

Cycle Chemistry and Boiler Tube Failures

possible, that the ingress of impurities is kept to a minimum, and that action is taken if impurities enter the boiler. There are now an increasing number of options for boiler water chemistry control and the optimized operating regimes for each are becoming increasingly clear. A comparison of the limits for key cycle chemistry parameters between the various common boiler chemistry control methods is shown in Table 3-3. This example is for normal operating conditions in a 2500 psi boiler. Note that in addition to normal operating limits, it is vital that “action levels” be set as described below. A suggested road map for utilities to use in selecting the optimum drum boiler chemistry is shown in Figure 3-5. The choice will depend on, among other considerations, whether the unit has a condensate

Review normal No or current PT Are there any problems?

Continue use of current treatment

Yes Base-line monitoring

Evaluation of phosphate hideout behaviour No

Yes

Consider changing to EPT, AVT or CT

Yes

Are feedwater contaminant events significant?

No

Is condensate polisher available or practical? Yes

Optimize PT

No Convert to CT

Yes

Convert to AVT

Chemically clean boiler

Convert to EPT

Develop specific unit PT, EPT or CT guidelines

Develop specific AVT guidelines

Monitor to compare with baseline data

Normal operation

Figure 3-5. Roadmap to optimize boiler water treatment for drum boilers. (EPT) - equilibrium phosphate treatment; (AVT) - all-volatile treatment; (CT) - caustic treatment; (PT) - phosphate treatment.

Volume 1: Boiler Tube Fundamentals

3-7

polisher and the severity of feedwater contaminant events in accordance with the following general guidelines: • If hideout or other chemistry problems such as: (i) the use of excessive amounts of mono- or di-sodium phosphate added to the boiler, (ii) underdeposit corrosion BTF in the waterwalls, (iii) corrosion fatigue BTF, (iv) excessive turbine deposits, (v) condenser inleakage, or (vi) makeup water contaminant excursions have not been experienced, continue with the current treatment which may be PT, AVT or even CPT.

Table 3-4 Worldwide Problems Relating to Feedwater Corrosion Product Generation, Transport and Deposit Generation

Transport/ Deposition

• Low pressure feedwater heater erosion/corrosion

• Boiler feedpump impeller fouling

• High pressure feedwater heater erosion/corrosion

• Boiler orifice fouling

• Deaerator erosion/corrosion

• Boiler wall deposits and pressure drop

• Economizer inlet erosion/corrosion

• Ripple magnetite formation

• Feedwater piping erosion/corrosion

• Boiler tube failures

• Improve current practice by monitoring the critical cycle chemistry parameters as a function of different unit loads, if not currently done. Section 3.5 below discusses typical monitoring campaigns.

Source: R.B. Dooley17

• If there have been problems, it is necessary to determine which alternative (generally among PT, EPT, AVT or CT) will be optimal.

3.3 Feedwater Treatment

• If a condensate polisher is available, the choice is generally AVT. • If a condensate polisher is not available, then a choice will have to be made between EPT, PT or CT. • If monitoring and experience indicate few incidences of contaminant ingress, the choice would generally be EPT. • If there have been frequent incidences of contaminant ingress, then the choice is either PT or CT. • Whichever choice is made, the final and continuing step is to develop unit-specific guidelines and to compare monitoring data with the established baseline. The foregoing discussion was for drum boilers. For once-through units, the boiler water is controlled by the feedwater, the subject of the next section.

3-8

• Chemical cleaning

The generation of feedwater corrosion products, usually originating in low-pressure (LP) and high-pressure (HP) feedwater heaters and the deaerators, and their deposition are directly responsible for a number of problems in the cycle. Table 3-4 lists some of these. The relationship between improvements in feedwater chemistry and the opportunity for preventing boiler tube failures is the focus of this section. However, the benefits that will accrue throughout the unit from these actions should not be overlooked, including benefits to the turbine, feedwater heaters, condensers and throughout the boiler. The classical approach to feedwater treatment in the U.S. and many other countries has consisted of adjusting pH with ammonia to 8.8 9.1 for mixed copper/iron systems and to 9.2 - 9.6 for all-ferrous systems, to deoxygenate the feedwater mechanically in the condenser and deaerator, and to deoxygenate chemically by the addition of an oxygen scavenger. The belief was that all oxygen should be eliminated to control corrosion.

Cycle Chemistry and Boiler Tube Failures

This is the basic approach taken for all-volatile treatment, and until 1969 was the only feedwater treatment applied worldwide in plant cycles with subcritical and supercritical once-through steam generators; in the U.S. this was the case until November, 1991.18 AVT can be applied to all units, and is still the method of choice for plants with mixed metallurgy (copper and iron) in the feedwater train and/or for units without condensate polishers. In deoxygenated feedwater systems, very low levels of oxygen (3 x O2) Fe 20 g/ft2), less so for moderate deposit loading (5 - 20 g/ft2). Factors influencing the measurement of deposits include12: (i) tube geometry and type (ribbed, metal-

9-6

ized, bends, eroded areas), (ii) tube surface condition (internal and external), (iii) metal condition (corroded, cracked, bulged), (iv) wet (flooded) versus dry tubes, (v) internal deposit loading, (vi) internal deposit type (composition, density, morphology, stratification), and (vii) transducer type (frequency, delays, contact surface area). Locations chosen for UT testing should be among the high risk locations such as those subject to high heat flux or local flow disruptions. Table 2-6 has a complete list of local conditions that can lead to initial deposits. As with other UT methods, careful tube surface preparation is required to obtain accurate results. For this method, the external surface condition, as well as the internal condition, for example the extent of pitting, etc. will dominate the determination of deposit buildup. Calibration via tube sampling is required to provide a reference point for field test results. A minimum of one tube prior to, and one tube post testing is recommended with additional samples to provide a range of

Determining the Extent of Macroscopic Damage

9.3.4 Ultrasonic detection of microstructural changes. UT has also been used in an attempt to detect subtle changes in microstructure induced by hydrogen damage. The wave velocity, attenuation, and backscatter signal are all affected by hydrogen damage. In the 1980s an approach was tried to measure attenuation associated with grain boundary cavitation.13 A number of successful applications have been reported13; however the method was not fully satisfactory because of its sensitivity to various factors including surface condition (roughness), corrosion, pitting and transducer coupling. In a similar manner, the backscattering method was also found to have shortcomings, notably the backscatter from hydrogen damage was difficult to separate from that due to inside surface corrosion.14 As a result of these shortcomings a UT method based on velocity changes has been developed.13,14,15 The method relies on the fact that the velocity of sound through a material is a function of its modulus of elasticity, a material property that is changed by the accumulation of hydrogen damage. A pair of transducers are operated in the pitchcatch mode. As the signal passes through a damaged region, the change in transit time is noted; that change can then be related to the

depth of hydrogen damage. Both laboratory and field tests have confirmed the usefulness of the method developed.13,14 Additional information about the use of this technique for detecting hydrogen damage can be found in the mechanism discussion, Chapter 15, Volume 2.

9.4 Other Standard Inspection Methods 9.4.1 Visual examination. Visual examination is a primary means to assess boiler tubing for conditions such as: • Fireside wastage caused by corrosion and erosion processes. • Significant wastage by recognizing the presence of rusted tube locations within a few hours of a boiler wash, indicating the removal of protective surface oxides. • Broken attachments, hangers, or supports can warn of higher imposed stresses on affected tubing. • Tubes misaligned out of the general platen can lead to overheating or may be symptomatic of excessive restraint stresses. • Convection pass fouling or blocked passages might induce localized high velocity conditions and provide the potential for flyash erosion. • Flame impingement, carbon particle impingement, or burner misalignment can be a sign of localized overheating of tubes and eventually can manifest damage via a variety of mechanisms. • Water-side scaling, corrosion, deposition and/or pitting can be detected using borescopes, fiber optics, mirrors, and miniature closed circuit television systems.16 • Measurement of tube diameters can detect tube bulging, a precursor to failure by short-term overheating due to blockage, for example.

9.4.2 Liquid penetrant testing (PT). Liquid penetrant testing (PT) is used to find cracks or pores in materials provided that the discontinuities are clean and open to the surface. This method is applicable to magnetic and nonmagnetic materials and is particularly useful where magnetic particle examinations cannot be used. The penetrant may be a visible dye used in normal light or a fluorescent dye for detecting smaller defects. Typical boiler tube damage detected using PT includes dissimilar metal weld cracking, stress corrosion cracking, fatigue and weld defects. 9.4.3 Magnetic particle testing (MT). Magnetic particle testing is used for detecting surface or near-surface discontinuities in ferromagnetic materials such as damage caused by fatigue, stress corrosion cracking or surface-related material defects. For stress corrosion cracking, the cracks are usually tight and the use of fluorescent magnetic particle may be useful. MT is used primarily in the tube manufacturing process, but may also be employed to detect welding defects. The magnetic fields produced by DC are more penetrating than those produced by AC and allow the detection of discontinuities deeper in the material. MT cannot be used to inspect nonferromagnetic materials such as austenitic stainless steels. Joints between steels with dissimilar magnetic characteristics may create magnetic particle indications even though the joints are sound. 9.4.4 Radiographic testing (RT). Radiographic testing is used to detect surface and subsurface discontinuities which can be aligned with the propagation direction of the radiation beam. Factors to be considered for RT include: accessibility, radioisotope source strength, geometry, exposure time and distance, material thickness, and the spatial relationship of the type of discontinuity. Selection of the radiation

source for a particular tube wall thickness is a critical factor since low contrast and poor radiographic sensitivity will result if the energy of the source is too high or too low. RT is based on differences in density in the material so that discontinuities must be more or less aligned with the radiation beam. Since RT involves the use of a radioactive source, radiographers must be specially trained and licensed, and safety procedures established. A specialized radiographic technique, termed the Union Electric technique for its developer, can detect well-developed damage in dissimilar metal welds (DMW) made with stainless steel filler metal.17,18,19 The basics of the method are illustrated in Chapter 35, Volume 3 which addresses DMW failures. Although the method can detect damage levels down to about 5% of the interface, film interpolation was found to be difficult below about 15% damage.19 Good correlation between the technique and actual damage was confirmed for over 50 DMWs that had damage levels ranging from 5 to 90%.19 RT can also be used to measure the extent of waterside corrosion fatigue damage without removing the lagging and insulation. 9.4.5 Eddy current. In eddy current testing, a search coil is used to induced electric currents (electromagnetic induction) in a part to be inspected. Discontinuities in the part create eddy currents that are detected by the search coil. The method is very sensitive, which can create interpretation difficulties. A constraint on the technique is that the part has to be electrically conducting; otherwise it has widespread application. Sollish20 reviewed a remote field eddy current (RFEC) technology for use in measuring boiler tube wall thickness. A probe containing an exciter coil and one or more receiver coils is inserted into the

Volume 1: Boiler Tube Fundamentals

9-7

tube to be inspected. Currents generated by exciter and detected at the receiver coils are changed in magnitude and delayed in phase depending on the thickness of the tube they travel through.20 The results can be obtained for the entire tube circumference and length, overcoming the “spot-check” nature of other sampling NDE schemes. Since internal scale and deposits do not appreciably affect the results, cleaning is not required prior to the test. The method requires skilled interpretation, particularly as changes in tube geometry such as bends generate the same patterns as wall loss. The results, obtained on several hundred boilers, indicated a system accuracy of ± 10% (worst case) to ± 1% (best case) with the greater wall loss having the greater accuracy.20 As a result, note was made only of wall loss greater than 30%. Good repeatability was found. 9.4.6 Acoustic monitoring. Acoustic systems monitor the jet noise generated by a leak.21 This sound propagates through the furnace gas to sensitive piezoelectric transducers mounted on the walls. Relatively few probes, typically 6 to 12, can be used to monitor an entire boiler. While these systems are very good for leak detection, they are somewhat poorer at leak location. This is particularly true in tightly-spaced tube bundles such as the superheaters and reheaters, and yet these are the areas where manually locating a leak is most difficult. In addition to air- or fluid-borne noise generated by leaks, there is also structure-borne sound which can be detected by high frequency piezoelectric transducers. Whenever plastic strains occur in a material, acoustic energy is emitted, commonly called acoustic emission. Typically, the acoustic emission can be detected earlier in the degradation process than air-borne noise.22 Acoustic emission technology has been applied for a number of years in the industry. These applications include both crack growth detection

9-8

Stainless steel pad, longitudinally mounted and welded on three sides

Stainless steel gear clamps

Figure 9-5. Pad-type thermocouple attachment to tube. The steel pad can be insulated.

Flue gas flow Weld metal

Hot junctions

Dutchman with chordal T/C

Grind out channel

Outlet

Vestibule

Flue gas flow

Peen over channel

Shielding tube Thermocouple wire Cross Section

Side View

Figure 9-6. Typical location of in-furnace chordal thermocouple within tube assembly. Source: H.J. Grunloh and R.H. Ryder7

and leak detection in components; the feasibility of on-line acoustic emission monitoring in hostile environments has been demonstrated.

9.5 Monitoring Temperatures The accurate determination of temperature, particularly in SH/RH tubing is critical to an accurate determination of remaining tube life. Methods can be either direct, primarily thermocoupling, which provide current temperature levels, or indirect. A few comments about thermocouples are provided here. Chapter 10 provides more complete discussion about indirect methods such as oxide scale analysis or

Determining the Extent of Macroscopic Damage

microstructural changes which provide indications of past (average) temperatures. Thermocouples should be installed so as to provide good analysis of outlet temperature profiles across the boiler and through assemblies. Typically 20-30 percent of the tube outlet legs and a half a dozen tube inlet legs are instrumented in the penthouse outside of the fireside heat flux region.11 These should be regarded as a primary resource in assessing SH/RH temperatures and life.

New tube (no ID oxide)

Flue gas flow

Chordal thermocouples (OD and midwall) Old tube (ID oxide present) Shielding tube T/C lead wires

Figure 9-7. Chordal thermocoupled tube sections reinserted into original location within the tube assembly. Note stainless steel shielding tube which surrounds thermocouple lead wires. Source: H.J. Grunloh and R.H. Ryder7

9.5.1 Pad-type thermocouples. Pad-type thermocouples are common in boilers both as original construction and installed for monitoring operation and maintenance. Figure 9-5 shows the typical construction. A thermocouple junction is embedded in a small metal plate (approximately 1 inch square) then welded to the tube of interest and covered with insulation. The lead should be attached to the tube about every 1 to 2 inches to prevent burnout. The thermocouple is attached to the tube to allow differential thermal expansion of the metal plate and the tube. 9.5.2 In-furnace chordal thermocouples. Chordal thermocouples are designed to measure in-furnace tube metal temperatures directly and to provide information about through-wall temperature gradients. Figure 9-6 (left side) shows a typical location that would be monitored with chordal thermocouples. The installation is accomplished by placing a thermocouple junction at midwall points such as locations at 25% and 50% from the outside surface. Tube sections are drilled to place

the thermocouple wires which are then tack welded in place and covered by peening-in weld metal. The process is shown in the right side of Figure 9-6. The tube section thus instrumented is then welded back into the circuit and the lead wires covered with a stainless steel shielding tube. The resulting installation is shown in Figure 9-7. Additional detail about thermocouple theory and application can be found in reference 23.

9.6 Monitoring Heat Flux Heat flux can be monitored using chordal thermocouples, absorbed heat flux meters (flux tubes or domes), and incident heat flux meters (flux probes). At least three chordal thermocouples from the crown of a tube to the membrane in the case of waterwall measurements are appropriate for heat flux characterization24; although heat flux is not measured directly, it can be inferred from the differences in reading from the thermocouples.

thus deduce heat flux. At least two types are available, a flux tube and a dome. The latter have the advantage that they can be attached as discrete devices to the surfaces of existing tubes without removal of the tubes from the boiler. The flux meter is designed to measure total (convective and radiative) heat flux. It consists of a watercooled probe that houses a cylindrical block. The block has a serrated, blackened front surface to give high radiation absorption. Front and rear thermocouples are used to calculate heat flux.25 It has an accuracy of 3 percent and a cost of about $10,000.25 The Fluxdome consists of two concentric cylinders. An inner, or “sensor” cylinder contains two thermocouples; surrounding it the outer or “guard” cylinder ensures that there is minimum unwanted loss of heat through the side of the sensor cylinder. The temperature gradient in the sensor cylinder is measured with two thermocouples placed inside the sensor cylinder and is proportional to the heat flux. The Fluxdome has a manufacturers specified range of operation of 0 -1200°C with a temperature resolution of better than 3°C. Experience at the former CEGB indicated that absorbed heat flux was generally between 50-75% of the incident flux depending on tube condition.26 The range of absorbed heat fluxes was from less than 100 kW/m2 in heavily slagged areas to over 500 kW/m2 on clean tubes in boilers at high loads; typically values on CEGB combustion chambers were 300-350 kW/m2.26 It should also be noted that accurate heat flux measurement requires that the device used should have the same layer of ash as on the tubes surrounding it.

Absorbed heat flux meters measure a thermal gradient in a cylinder of known thermal conductivity, and

Volume 1: Boiler Tube Fundamentals

9-9

9.7 Monitoring Displacements and Strains Displacements are most commonly measured with linear voltage digital transducers (LVDTs). There are a variety of sensors available to measure strain. Those most commonly applied to boiler tubes are weldable strain gages and capacitance strain gages. Conventional resistance strain gages can be used for low-temperature locations but are generally limited to temperatures up to about 300°C (~ 575°F). High-temperature gages have been extensively evaluated in laboratory testing but in-plant use, although promising27, is more limited. Weldable platinum-tungsten gages have temperature ratings up to about 650°C (~ 1200°F) with field confirmation of reliable static strain measures up to about 480°C (~ 900°F). Nickel-chromium gages used in the same field tests were found to give accurate dynamic strains at temperatures up to around 815°C (~ 1500°F). The latter were used on the fireside of waterwall panels as a result. The gages did not require shielding, although the leads were shielded. Capacitance strain gages were used on the cold side of the waterwall panels.

9-10

Strain levels will be highly localized, so that care must be taken to pick locations for analysis carefully considering the complexity of the local geometry, loading conditions and temperature ranges. For practical purposes, strain gages are best used to measure changes in strain at a particular location as a function of changes in operating conditions.

9.8 Sampling Tube sampling is performed for a variety of reasons: (i) as an aid to root cause determination or confirmation, (ii) to confirm the accuracy of nondestructive examination results, (iii) to provide both qualitative and quantitative information about the type, extent and thickness of deposits, (iv) to confirm oxide scale thickness, (v) to provide material needed to perform detailed visual and metallurgical examinations. Sampling practices and test methods are provided in ASTM Standards D88728 and D348329. Some general considerations about sampling include: • Samples should be taken for a specific purpose, for example to confirm the results of NDE measures or to bound the “worst” conditions: highest temperatures, thickest steamside oxide, thickest internal deposits, or worst wall thinning, for example.

Determining the Extent of Macroscopic Damage

• A plan for the use of samples should be prepared that specifically lays out what kinds of testing are to be done on samples removed: metallographic, strength, creep rupture, wall, steamside oxide and waterside deposits thickness, measures, etc.

9.9 Hydrostatic Testing A hydrostatic pressure test is good practice following weld repairs and as a means to ensure that all damaged tubes have been identified following an outbreak of a particular damage mechanism. Testing is normally performed with room temperature water to at least normal operating pressure, but preferably to 1.5 times the drum design pressure. Visual examination to detect leaking after testing is indicated. For safety reasons the pressure should be reduced prior to visual inspection.

9.10 References 1Stephenson,

G.G. and J.W. Prince, Guidelines on Fossil Boiler Field Welding, Research Project 2504-02, Final Report TR-101699, Electric Power Research Institute, Palo Alto, CA, January, 1993. 2Cohen, P., ed., The ASME Handbook on Water Technology for Thermal Systems, American Society of Mechanical Engineers, New York, NY, 1989. 3American

Society for Metals, Metals Handbook, Volume 17: Nondestructive Inspection and Quality Control, 9th edition, American Society for Metals, Metals Park, Ohio, 1989. 4Lamping,

G.A., L.A. LeJune, and W.R. Meredith, “Ultrasonic Examination of Boiler Tubing: Automated Data Acquisition and Computer-Aided Data Analysis”, in Failures and Inspections of Fossil-Fired Boiler Tubes: Conference and Workshop, Final Report CS-3272, Electric Power Research Institute, Palo Alto, CA, December, 1983. 5Allen,

C.C. and E. R. Reinhart, “Optimized NDE Methods for Preventive Maintenance of Power Boilers”, ibid., pp. 5-1 through 5-15. 6American

Society for Testing and Materials, Standard E797-90 (1990), “Standard Practice for Measuring Thickness by Manual Ultrasonic Pulse-Echo Contact Method”, 1994 Annual Book of ASTM Standards: Nondestructive Testing, Volume 03.03, Society for Testing and Materials, Philadelphia, PA, 1994. 7Grunloh,

H.J. and R.H. Ryder, Life Assessment of Boiler Pressure Parts, Volume 7: Life Assessment Technology for Superheater/Reheater Tubes Research Project 225310, Final Report TR-103377-V1/7, Electric Power Research Institute, Palo Alto, CA, November, 1993 8Bonin,

D.W., “Nondestructive Oxide Thickness Measurement in Superheater and Reheater Tubing”, in Proceedings of the EPRI Fossil Plant Inspections Workshop, San Antonio, Texas, September 9-11, 1986, Final Report CS-5320, Electric Power Research Institute, Palo Alto, CA, 1987. 9Beak,

W.E. and D.W. Bonin, “Nondestructive Technology to Evaluate Superheater Condition”, in B. Dooley, and D. Broske, eds., Boiler Tube Failures in Fossil Power Plants: Conference Proceedings, Conference held in Atlanta, Georgia, November 10-12, 1987, CS-5500-SR, Electric Power Research Institute, Palo Alto, CA, 1988, pp. 5-27 through 5-38. 10Adkins,

W.J. and D.F. Powell, “Boiler Tube Internal Scale Measurement Using Ultrasonic Inspection Techniques and Interpretation of Results”, ibid., pp. 5-1 through 5-25.

11Hara,

K., C. Lee, R. Moser, T. Rettig, and K. Clark, Improved Superheater Component Longevity by Steam Flow Redistribution, Research Project 1893-13, Final Report TR-101697, Electric Power Research Institute, Palo Alto, CA, December, 1992. 12Hicks,

P., A. Banweg, and M. Parker, “The Use of Ultrasonic Testing in Determining Waterside Deposit Buildup in Boiler Systems”, Corrosion 94, Paper No. 205, NACE International, 1994. 13Lamping,

G.A., and S. Gehl, “Hydrogen Damage Assessment Using Ultrasonic Velocity Measurement”, in B. Dooley, ed., Proceedings: International Conference on Boiler Tube Failures in Fossil Plants, held in San Diego, California November 5-7, 1991, Proceedings TR-100493, Electric Power Research Institute, Palo Alto, CA, April, 1992, pp. 8-23 through 8-35. 14Birring,

A.S., D.G. Alcazar, J.J. Hanley, G.J. Hendrix, and S. Gehl, “Detection of Hydrogen Damage by Ultrasonics”, op. cit. reference 9, pp. 5-59 through 5-67.

15Birring,

A.S., D.G. Alcazar, J.J. Hanley, and S. Gehl, “Ultrasonic Detection of Hydrogen Damage”, Materials Evaluation, March, 1989, pp. 345-350, 369. 16Parker,

J.D., A. McMinn, R.J. Bell, R.H. Richman, W.P. McNaughton, J.P. Dimmer, J.E. Dammon, and D.S. Galpin, Condition Assessment Guidelines for Fossil Fuel Power Plant Components, Research Project 2596-10, Topical Report GS-6724, Electric Power Research Institute, Palo Alto, CA, March, 1990. 17Gurnea,

R.F, “Radiographic Technique for Detecting Cracks in Dissimilar Weld Joints”, in R. Viswanathan and D.A. Roberts, eds., Proceedings: Seminar on Dissimilar Welds in Fossil-Fired Boilers, held in New Orleans, LA., February 23-24, 1984, Research Project 1874-1, Proceedings CS-3623, Electric Power Research Institute, Palo Alto, CA, July, 1985, pp. 4-47 through 4-60. 18Prager,

M., D.I. Roberts, H.J. Grunloh, and K.H. Holko, Dissimilar-Weld Failure Analysis and Development Program, Volume 5: Evaluation of Acoustic Emission and Enhanced Radiography, Research Project 1874-1, Final Report CS-4252, 8 Volumes, Electric Power Research Institute, Palo Alto, CA., November, 1985 19Grunloh,

H.J., R.H. Ryder, and R. Hellner, “Damage Assessment and Predictive Maintenance of Dissimilar Metal Welds in Superheater and Reheater Tubes”, op. cit., reference 13, pp. 7-51 through 7-76. 20Sollish,

D.B., “Field Experience Utilizing State-of the-Art NDE Techniques Applied to In-Service Boiler Tube Examinations”, ibid., pp. 8-1 through 8-23.

Volume 1: Boiler Tube Fundamentals

9-11

21Stulen,

F.B., B. Salisbury, and J.R. Scheibel, “Leak Detection in Boilers”, op. cit. reference 9, pp. 5-91 through 5-104.

Control of Fireside Corrosion in Power Station Boilers, Second edition, Central Electricity Generating Board, 1977.

22Perkul, P.J., and S. Perrnise, Vibration Signature Analysis and Acoustic Emission Monitoring at Brayton, Final Report CS-1938, Electric Power Research Institute, Palo Alto, CA, July, 1981.

27Cane,

23Stultz,

S.C. and J.B. Kitto, Steam: Its Generation and Use, 40th Edition, Babcock & Wilcox Company, Barberton, Ohio, 1992. 24Nowak, S., et al., Corrosion Fatigue Boiler Tube Failures in Waterwalls and Economizers, Volume 5: Field Validation, Research Project 1890-5, Final Report TR100455, V5, Electric Power Research Institute, Palo Alto, CA, 1996. 25Crim, H.G. and E.K. Levy, Power Plant Performance Monitoring and Improvement, Volume 1: Boiler Optimization, Research Projects 1681, 2153, Final Report CS/EL-4415, Volume 1, Electric Power Research Institute, Palo Alto, CA, February, 1986.

9-12

Determining the Extent of Macroscopic Damage

26The

B.J. and J.A. Williams, “Remaining Life Prediction of High Temperature Materials”, International Materials Review, Volume 32, Number 5, 1987, p.241. 28American

Society for Testing and Materials, Standard D887-82 (1989), “Standard Practice for Sampling WaterFormed Deposits”, 1992 Annual Book of ASTM Standards: Water, Volume 11.02, American Society for Testing and Materials, Philadelphia, PA, 1992. 29American

Society for Testing and Materials, Standard D3483-83(1990), “Standard Test Method for Accumulated Deposition in a Steam Generator Tube”, 1992 Annual Book of ASTM Standards: Water, Volume 11.02, American Society for Testing and Materials, Philadelphia, PA, 1992.

8

6

4

2

0

Chapter 10 • Volume 1

1.0 Relative Fraction of M6C Carbide

Area Rate of Creep Void (x 103 mm-2)

10

0.8

Relative fraction of M6C carbide (D)

0.6

0.4

0.2

Area density of creep void (❍)

0 10 20 40 60 100 80 Creep Rupture Life Consumption Rate (%)

Determining the Extent of Microstructural Damage 10.1 Introduction The microstructures of all boiler tube materials are thermodynamically unstable at elevated temperature and are thus subject to time-dependent changes. Under normal conditions, any changes to carbon steel economizer and waterwall tube structures are very slight and can be neglected. Low-alloy ferritic SH and RH tubes, however, do show perceptible structural changes after long-term service, while austenitic materials undergo both short-term and occasionally, long-term changes. When operating conditions, particularly temperature, exceed design levels, microstructural changes can be more significant and can in some cases be used as a measure of the deviation from design conditions.1 A careful analysis of the microstructure of a boiler tube can provide significant information about the service conditions to which it has been subjected. Microstructural analysis is a vital tool in confirming the mechanism of failure and underlying root cause in failed tubes. Although oxide scale is often the key determinant of tube temperature history, in those conditions where the oxide is missing or incomplete, an analysis of the tube microstructure can also provide valuable guidance about temperature history, needed for the assessment of remaining life in unfailed tubes.

Table 10-1 lists those failure mechanisms which are specifically accompanied by fundamental changes to the tube microstructures. It excludes those mechanisms for which only cracking without other microstructural change, or surface hardening, such as caused by the impact of erosive particles, are evident. This Chapter begins with brief discussions of microstructural changes in ferritic steels (Section 10.2), and continues with an overview of changes in austenitic stainless steels (Section 10.3). An overview of methods for assessing the extent of creep damage, a topic which has seen considerable effort in the past few years (Section 10.4), and a note about post-exposure testing (Section 10.5) are also included. Detailed discussion of specific microstructural changes that accompany the mechanisms listed in Table 10-1 is included in the chapters referenced in the table. Catalogues of tube microstructures as a function of time, temperature and stress history are available. They should be consulted for comparative micrographs of undamaged and service-degraded materials of all boiler tube material classes, or to compare samples removed from the field with specimens containing known amounts of accumulated exposure.2 This process will provide an estimate of average tube metal temperature over its service life.

Volume 1: Boiler Tube Fundamentals

10-1

Table 10-1 BTF Mechanisms Manifesting Significant Microstructural Change Chapter, Volume for Additional Information

Mechanism

Nature of Microstructural Change

Hydrogen damage

Decarburization and microfissuring.

Chap. 15, Vol. 2

Short-term overheating in waterwall tubing

Significant transformation of initial microstructure with final morphology depending upon temperature reached.

Chap. 23, Vol. 2

• Nucleation, growth, and interlinking of creep voids or cracks (creep cavitation) • For ferritic materials: spheroidization, formation of decarburized layer, graphitization. • For austenitic materials: sigma phase microstructure and formation of grain boundary cavities.

Chap. 32, Vol. 3

• Signs of creep damage. Specifics depend on filler metal. • Formation of planar array of globular carbides.

Chap. 35, Vol. 3

Short-term overheating in SH/RH tubes

May be similar to that for short-term overheating in waterwall tubes, or for long-term overheating in SH/RH tubes, or may be none.

Chap. 36, Vol. 3

Graphitization

• Graphitization.

Chap. 42, Vol. 3

Long-term overheating/creep

Dissimilar metal weld failures

50 40 30 20 10

28

30

32

36 34 LMP/1000

38

40

42

Figure 10-1. Oxide thickness as a function of the Larson-Miller Time/Temperature Parameter (LMP) for T22 tube material (21/4 Cr-1 Mo). Source: S.R. Peterson, et al.2

10-2

Determining the Extent of Microstructural Damage

LMP = (T + 460) (20 + log10 t )

(10-1)

where T = temperature in ¡F t = exposure, hours

Oxide Thickness (mils)

0 26

The Larson-Miller Parameter (LMP) is commonly used to combine time and temperature into a single measure. It is used in many of the plots in this chapter to display an integrated time/temperature variable and is given by:

10.2 Microstructural Changes in Ferritic Materials 10.2.1 Oxide Thickness. Although strictly not a microstructural change, the analysis of oxide scale in ferritic SH/RH tubes, when possible, is so central to the analysis of damage accumulation and remaining life assessment, that a few summary comments are included here. Analysis of steamside oxide scale buildup is a key parameter for analyzing accumulated damage in boiler tubes. For the sake of comparisons with other parameters that follow, Figure 10-1 shows oxide thickness versus LMP typical of 21/4 Cr-1 Mo material removed from service and measured on in-service tubes. The basics of steamside oxide scale formation can be found in Chapter 2, their use in remaining life analysis (Chapter 8), and measurement by non-destructive means (Chapter 9).

Table 10-2 Qualitative Description of Degrees of Spheroidization Stage Identification

Degree of Spheroidization

1

Typical of the structure of a new tube consisting of ferrite and fine pearlite. May be partially spheroidized if post-weld heat treatment was used on a fabricated structure.

2

The first stage of carbide spheroidization usually coinciding with the appearance of small particles of carbides at the grain boundaries

3

An intermediate stage of spheroidization showing more distinct signs of carbide spheroidization in the pearlite areas, but some carbide plates still evident. Increased carbide precipitation within the ferrite grains and at the grain boundaries.

4

Spheroidization of the carbides is virtually complete, but they are still grouped in the original pearlite pattern.

5

Spheroidization is complete and the carbides are dispersed leaving little trace of the original pearlite areas.

6

There is a marked increase in the size of some of the carbide particles, partly due to coalescence.

7

Carbides disappear.

Source: From S.R. Paterson, et al.2 based on L.H. Toft and R.A. Marsden3

Spheroidization Rating 7 6 5 4

10.2.2 Spheroidization. If the steamside oxide has exfoliated or been removed by chemical cleaning then changes to the microstructure in ferritic material can be examined to deduce service history. The typical starting structure is ferrite with either (i) fine pearlite composed of alternating lamellae of ferrite and iron carbide Fe3C or (ii) acicular bainitic carbides. Over time the carbide tends to agglomerate into spheres. The degree to which this process has proceeded was classified initially into six stages by Toft and Marsden.3 A seventh stage has been added; the stages are identified by number 1-7 or by letter A-G; Table 10-2 provides a qualitative description of the stages. Figure 102 shows typical data of the spheroidization rating versus LMP for 21/4Cr-1Mo. A potential drawback to such analysis is that the current state of the microstructure will depend on the original microstructure (prior to service), which is often not well known.2 Three means of dealing with this problem are: (i) use laboratory heat treatments to simulate the fabrication procedure if it is known and thus characterize the starting microstructure, (ii) obtain samples from the coldest regions of the boiler and assume that their microstructure is the starting microstructure, (iii) use a catalog of the microstructures of various starting points such as contained in reference 2 to obtain a scatterband of behavior.

3 2 1 0 26

28

30

32

36 34 LMP/1000

38

40

42

Figure 10-2. Spherodization rating as a function of the Larson-Miller Time/Temperature Parameter (LMP) for T22 tube material (21/4 Cr-1 Mo). Source: S.R. Paterson, et al.2

Volume 1: Boiler Tube Fundamentals

10-3

10.2.3 Hardness. The strength of low alloy steels changes with exposure to temperature. Hardness is an indirect measure of strength. Measurements of the change in hardness can, as a result, be used to estimate average temperatures for a component. The approach is most suited to changes occurring as a result of carbide precipitation and growth when strain softening strain effects do not interfere.4 Figure 10-3 shows the relationship between LMP and change in material hardness. Drawbacks to using hardness for damage assessments include4: (i) uncertainty about the initial hardness of the material, (ii) sensitivity to local microstructural variation, (iii) strainsoftening effects can lead to erroneous results.

Vickers Hardness 280 260 240 220 9 Cr - 1 Mo

200

2 1/4 Cr - 1 Mo

180 160 140 1Cr - 1/2 Mo

120 100 33

34

35 36 37 38 39 40 41 Larson-Miller Parameter (LMP) = T(°R) (20 + log10t [Hr] x 10-3)

9 Cr - 1 Mo HAZ laboratory data, 100 g load 2 1/4 Cr - 1 Mo HAZ laboratory data, 100 g load 2 1/4 Cr - 1 Mo HAZ exservice weldment data, 100 g load

1 Cr - 1/2 Mo normalized material laboratory data, 20 kg load 1 Cr - 1/2 Mo normalized material laboratory data, 30 kg load

LMP includes temper treatment - 3 h, 640 °C

Figure 10-3. Correlations between hardness and the Larson-Miller Time/Temperature Parameter (LMP) for 1 Cr-1/2 Mo, 21/4 Cr- 1 Mo, and 9 Cr-1 Mo steels. Source: R. Viswanathan, et al.5

With a Knoop indentor and a 300 gram load 80 HRB

Knoop Hardness

160 75 HRB

150

70 HRB

140

65 HRB

130

Cold side away from rupture

Rockwell Hardness Conversion

170

120 0

0.1

0.2 0.3 0.4 0.5 0.6 0.7 Distance from Rupture Line (inch)

0.8

0.9

Figure 10-4. Midwall microhardness traverse to illustrate the loss in hardness in the vicinity of a long-term overheat failure. Source: S.R. Paterson, et al.2

10-4

Determining the Extent of Microstructural Damage

Figure 10-4 shows the decrease in hardness in material near a location that failed by long-term overheating. Figure 10-5 shows the trend of hardness with increasing LMP number for several data bases of 21/4Cr-1Mo material. While in-situ measurements of tube surface hardness may be used to identify strength decreases, some care is required to estimate the original (or at least nondegraded) hardness as a baseline for comparison.

10.3 Microstructural Changes in Austenitic Stainless Steels The nature and extent of carbide precipitation is a key variable in determining the exposure of the austenitic stainless steels to stress and temperature. For example, high carbon grades are particularly susceptible to sensitization in which chromium carbides form at the grain boundaries, depleting the chromium content of adjacent areas and leading to intergranular attack.2 A key identifying factor is the presence and extent of the sigma phase, a brittle intermetallic FeCr compound. It forms during long-term exposures at temperatures between 600 and 900°C (~ 1110 to 1650°F). Metallographic examination for the presence of sigma phase can be used to estimate average tube operating temperatures. Figure 10-6

Hardness (HRB)

shows the growth of sigma phase for various austenitic stainless steels at 700°C (1292°F).

Hardness above HRB 100

100

10.4 Assessment of Creep Damage in Boiler Tube Materials

90 80 70 60 50 26

28

32

30

36 34 LMP/1000

38

40

42

Figure 10-5. Hardness and the Larson-Miller Time/Temperature Parameter (LMP) for 1 Cr- 1/2 Mo, 21/4 Cr-1 Mo, and 9 Cr-1 Mo steels. Source: S.R. Paterson, et al.2

KOH Etched Area, %

5 4

Note that at high stress levels, for example > 0.75sy, creep failures may be experienced without microstructural change.

347 321 A-1 316 304

3 2 1 0 102

103 104 Aging Time at 700°C, h

10.4.1 Overview of available methods. A number of investigations have established a quantitative relationship between microstructural features and time-temperature relationships for thick section components. Methods to detect developing creep damage and to predict remaining life of high temperature components once creep cracks initiate have been under extensive development during the past few years.4, 6 Table 10-3 summarizes the methods, including open issues. Table 10-4 provides a list of the applications and some additional notes for the key techniques. Several of these methods are reviewed below.

105

Figure 10-6. Growth of sigma phase at 700°C in various austenitic steels. KOH is potassium hydroxide etchant. Source: S.R. Paterson, et al.2

10.4.2 Analysis of carbides. Laboratory and field tests have shown that the percentage of M6C carbides in 21/4Cr-1Mo steel increases with time and temperature in a predictable way. Nakatani, et al.8 and Sugita, et al.,9 have outlined a procedure to estimate life fraction from these results. Figure 10-7 is a plot of % M6C carbides and of creep voids versus rupture life expended and shows the concept. During early creep life (up to 50% consumption of life), the relationship between % M6C carbides and life

Volume 1: Boiler Tube Fundamentals

10-5

Table 10-3 Life Assessment Techniques for Creep Damage Crack Initiation: Technique

Issues

Calculation

Inaccurate

Extrapolation of past experience

Inaccurate

Conventional NDE

Inadequate resolution

High-resolution NDE • Acoustic emission • Positron annihilation • Barkhausen noise analysis

Not sufficiently developed at this time

Strain (dimension) mesaurement

Uncertainty regarding original dimensions • Lack of clear-cut failure criteria • Difficulty in detecting localized damage

Rupture testing

• Doesn’t account for increasing stress/ temperature history • Difficulty in sample removal • Not suitable as a monitoring technique • Validity of life-fraction rule • Effects of oxidation and specimen size • Uniaxial-to-multiaxial correlations

Microstructural evaluation • Cavitation measurement • Carbide-coarsening measurements • Lattice parameter • Ferrite composition analysis • Hardness monitoring

Quantitative relationships with remaining life are lacking

10.4.3 Creep cavitation models. There has also been significant work to relate the evolution of the microstructural cavities that develop during the creep process to evaluations of remaining life, again for thick section components. Classification of the amount of damage into four stages was first performed by Neubauer and Wedel10: isolated cavities, oriented cavities, linked cavities (microcracks), and macrocracks. The classification was placed on a more quantitative basis through the development of the “A” parameter, a measure of the number fraction of cavitated boundaries and the life fraction consumed.11 Creep cavitation surveillance of the coarse-grain heat affected zone of SH and RH outlet header stub tube welds is frequently performed. However, in tubing away from welds, creep cavitation occurs very late in life, and thus has not been used to monitor damage.

10.5 Post-Exposure Testing

Oxide scale and wall thickness Replication strain monitoring

Crack Propagation: Technique

Issues

Analysis combines NDE, stress analysis and crack growth with an end-of-life criterion to predict remaining life

• Uncertainties in NDE results • Lack of adequate crack growth data in creep and creep-fatigue • Lack of a clear-cut end-of-life criterion under creep conditions • Difficulty in assessing toughness of inservice components

10-6

fraction can be used as an index to remaining life.4 After that point, creep cavitation was the more useful index. Both the % M6C carbides and the % void formation can be measured by careful replication. Note that these results pertain to the heat affected zone of stub tube welds.

Determining the Extent of Microstructural Damage

A destructive test of samples removed from selected tubes can be used as an adjunct to any of the above methods, such as to confirm the results of an oxide analysis. A variety of tests are available including tube burst tests under conditions of accelerated temperature or stress, and axial or chordal rupture samples. Problems with accelerated testing include: (i) choice of sampling location, (ii) interpretation of how that sample corresponds to the tubes remaining in the SH/RH, (iii) choice of acceleration means, and (iv) interpreting how the resultant value corresponds to the properties of the remaining tubes.

Table 10-4 Creep Detection and Assessment Techniques, Applications, and Additional Notes Technique

Typical Applications

Notes

Replication

• Header stubs • Girth welds (weld and HAZ)

Estimates of creep damage only.

Replication, “A” parameter

• High stress regions in base metal • Weld HAZ

Estimates of creep damage only.

Carbide spacing

• Header, steam pipe base metal

Probable use as a monitor of temperature or with other methods for life assessment.

Accelerated rupture testing

• Header OD base metal • Steam pipe base metal

Creep damage only. Need oxidation correction.

Mini-specimen creep testing

• Header OD base metal and steam pipe base metal when limited material is available.

Hardness

• Base metal

Currently limited to estimates of average temperature, not remaining life.

Replica strain monitoring

• Welds • HAZ • Local strain regions of interest

Used to establish strain rate by re-examination at periodic intervals.

Ultrasonic testing

• Long seam welds • Ligaments (for long cracks) • Locations with extensive creep damage

Crack growth analysis

• Wherever cracks or crack-like defects are found during inspections

Creep crack growth and creep-fatigue crack growth methods have been developed; materials properties available.

Analysis of steam side oxide.

• Tubes

Scale will develop cracks at very low creep levels and thus may be a qualitative indicator of creep deformation.

Source: R.B. Dooley, et al.7

Volume 1: Boiler Tube Fundamentals

10-7

8

6

4

2

0

1.0 Relative Fraction of M6C Carbide

10 Area Rate of Creep Void (x 103 mm-2)

Results from post-exposure testing have been used three ways4: (i) as a direct measure of remaining life, however, this is difficult because tubes experience a non-simple increase in stress and service temperature with life, (ii) as a means to estimate the properties of the virgin material, which are then used in an analytical approach such as that for the oxide methods like EPRI TUBELIFE or other commercially available codes, as described in Chapter 8, and (iii) predicting remaining life through a stress rupture algorithm and these same analytical methods, Chapter 8.

0.8

Relative fraction of M6C carbide (D)

0.6

0.4

0.2

Area density of creep void (❍)

0 80 10 20 40 60 100 Creep Rupture Life Consumption Rate (%)

Figure 10-7. Relation between nondestructive structure parameter and creep rupture life consumption rate evaluated by creep rupture test of removed tubes. Source: R. Viswanathan, et al.4

10.6 References 1Dooley, R.B. and H.J. Westwood, Analysis and Prevention of Boiler Tube Failures, Report 83/237G-31, Canadian Electrical Association, Montreal, Quebec, November, 1983.

A.E. Meligi, T.V. Narayanan, and C.B. Bond, eds., PVPVolume 208, Power Plant Systems/Components Aging Management and Life Extension, Book No. H00634, American Society of Mechanical Engineers, 1991.

2Paterson,

7Dooley,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 3Toft,

L.H. and R.A. Marsden, “The Structure and Properties of 1%Cr-0.5%Mo Steel After Service in CEGB Power Stations”, in Conference on Structural Processes in Creep, JISI/JIM, London, 1963, p. 275. 4Viswanathan,

R., S.R. Paterson, H. Grunloh, and S. Gehl, “Life Assessment of Superheater/Reheater Tubes, in B. Dooley, ed., Proceedings: International Conference on Boiler Tube Failures in Fossil Plants, held in San Diego, California November 5-7, 1991, Proceedings TR100493, Electric Power Research Institute, Palo Alto, CA, April, 1992, pp. 7-1 through 7-49. 5Viswanathan,

R., J.R. Foulds, and D.A. Roberts, “Methods for Estimating the Temperature of Reheater and Superheater Tubes in Fossil Boilers”, Proceedings of the International Conference on Life Extension and Assessment, The Hague, June, 1988. 6Viswanathan,

R. and S. Gehl, “Advances in Life Assessment Techniques for Fossil Power Plant Components Operating at Elevated Temperatures”, in

10-8

Determining the Extent of Microstructural Damage

R.B., W.P. McNaughton, and R. Viswanathan, “Life Assessment and Component Condition Assessment in the United States”, Proceedings: VGB Conference on Assessment of Residual Service Life, Mannheim, Germany, July 6-7, 1992, pp. 26-1 through 26-36. 8Nakatani,

H., T. Yokoyama, F. Masuyama, and N. Nishimura, “Metallurgical Damage Detection and Life Evaluation System For Boiler Pressure Parts”, in Proceedings of the EPRI Conference on Predictive Maintenance of Fossil Plant Components, Boston, October, 1990. 9Sugita,

Y., et al., “Evaluation of Creep Damage Progress by Metallurgical Examination in Aged Power Boiler Pressure Parts”, ISIJ International, Volume 30, Number 10, 1990, pp. 859-904. 10Neubauer,

B. and V. Wedel, “Restlife Estimation of Creeping Components by Means of Replicas”, in Advances in Life Prediction Methods, D.A. Woodford and J.R. Whitehead, eds., American Society of Mechanical Engineers, New York, 1983, pp. 307-324. 11Cane,

B.J., and M. Shammas, “A Method for Remnant Life Estimation by Quantitative Assessment of Creep Cavitation on Plant”, Report TPRD/L2645/N84, Central Electricity Generating Board, United Kingdom, June, 1984.

Chapter 11 • Volume 1

Repair and Replacement of Boiler Tubes 11.1 Introduction There are five maintenance-controllable activities associated with every tube failure and repair: (i) Pre-repair inspection (ii) Removal of the failed tube section (iii) Repair/replacement of the failed tube

assumes that a BTF has occurred, the BTF Team has identified the specific mechanism and underlying root cause, alternative immediate actions have been considered, and a repair has been identified as the appropriate option. Key steps in the process are discussed throughout this chapter. The sections in which particular steps are discussed are shown on the figure.

(iv) Post-repair inspection and tests (v) Future preventive/control actions

11.2 General Strategies for Damaged Tubes

The purpose of this chapter is to overview the repair and replacement of the failed tube, step (iii) in the process. Methods for determining the extent of damage: pre-repair inspection and post-repair inspection and tests are covered in Chapter 9. Removal of a failed section, step (ii) above, and general metallurgical examination of such sections are discussed in Chapter 6. Future prevention and control actions are discussed within the chapters of Volumes 2 and 3 that cover specific mechanism.

There are essentially four general options available for the repair and/or replacement of boiler tubes:

Welding is the most important maintenance activity for repairing or replacing boiler tubes. An overall approach to weld repairs of boiler tubes is shown in Figure 11-1. It

• For leaks or failures: (i) repair or replacement to restore to original condition or (ii) to a condition acceptable for service (meeting ASME Boiler & Pressure Vessel Code requirements and other applicable rules). • For damaged, but not “failed” components: (iii) remove damage and assess whether remaining material meets service and Code requirements, or (iv) perform integrity assessment (“fitnessfor-service”) to determine whether component can be placed back in service in damaged condition at least until replacement is possible.

Volume 1: Boiler Tube Fundamentals

11-1

Text Section Specific Chapters in Volumes 2 and 3 for each mechanism

Action 4.0 in each mechanism, Volumes 2 and 3 Sections 11.2, 11.5, this chapter

Action (➠) or Step in the Sequence A BTF has occurred, the BTF Team has followed the Actions for a specific mechanism. The mechanism has been confirmed, root cause identified, and it has been determined that a weld repair is among the appropriate immediate actions to be taken.

Text Section

Action (➠) or Step in the Sequence

Section 11.5.3, this chapter

➠Remove damaged material/flaws

➠Use NDE to confirm that defect has been removed

➠Determine Extent of Damage as described for each mechanism (in Action 4.0)

Section 11.5, this chapter

➠Perform repair using Qualified

➠Determine Repair Options

Sections 11.5.4, 11.5.5, this chapter

➠Use temporary repair technique, such as a pad weld or window piece, if absolutely necessary

Procedure

➠Obtain access/clearance

➠Note the location of temporary

Section 11.3, this chapter

➠Identify materials and dimensions

➠Perform final inspection/testing

Sections 11.4, 11.5 this chapter

➠Establish repair technique, or use

repairs, Remove at next outage

previously qualified procedures

➠Choose Qualified Welder

Figure 11-1. Roadmap for Boiler Tube Weld Repairs Adapted from: G.G. Stephenson and J.W. Prince1

Where repair is not possible, option (iv), analysis, may be the only choice. However, such methods are often complex, costly, and may require significant time. The following historical information should be considered in making the decision about repair1: • Was the original material selection in the design appropriate?

11-2

• Did all the original materials meet specified requirements? • Did the operating conditions contribute to the failure in any way: creep, fatigue, erosion, shock or thermal loading? • Are there any unusual metallurgical characteristics in the defective or suspect area? • Has the problem occurred in any other units at the same site or within the industry?

Repair and Replacement of Boiler Tubes

Each utility should have in place a set of guidelines about minimum wall thickness to expedite the repair decision. General guidance about when to replace or restore tubes is available from boiler manufacturers. Table 11-1, for example, provides such guidance from one manufacturer as a function of tube thickness and location.

Methods of Construction”. Section I requires written welding procedures qualified to Section IX requirements and inspection requirements per Section V.

Table 11-1 Guidelines for Tube Repair/Replacement

Location

Actual Tube Wall Thickness Relative to Percent Specified Wall Thickness, t

Furnace Support Tubes Economizer Stringer Support Tubes “

Tubes equal to or greater than 85% t

Monitor thickness

Tubes less than 85% t

Restore tube wall thickness or replace tube

Tubes equal to or greater than 70% t

Monitor thickness

Tubes less than 70% t

Restore tube wall thickness or replace tube

Tubes equal to or greater than 85% t

Monitor thickness

Tubes less than 85% t

Restore tube wall thickness or replace tube

Economizer, Furnace Wall and other Water-Cooled Tubes “

Superheater, Reheater, and Other Steam-Cooled Tubes “

Course of Action

Source: Babcock & Wilcox, cited in G.G. Stephenson and J.W. Prince1

11.3 Pre-Repair: Confirm Materials to be Repaired It is critical to confirm the material(s) to be repaired prior to developing an optimized procedure. In-situ or laboratory confirmation is possible, depending on access. Chapter 2 provides more information about the materials that are typically used in boiler tubes. The specific grade of material is required as well as the specification number. That is, ASME SA-213 includes a variety of material grades, SA-213 T11 will require a different repair procedure from SA-213 T22 for example. Because several materials are usually used in the SH/RH sections, a schematic showing the locations of materials and the transitions is critical, both for tracking materials and for developing the appropriate weld procedures. “Upgrading” materials

is, for several key BTF mechanisms, the optimal strategy. If a material change is made, careful documentation should be made of the new material and its location so that any needed repairs can be properly executed.

11.4 Applicable Codes for Weld Repairs Applicable U.S. codes for the welding of boiler tubes include: • National Board Inspection Code (NBIC), American National Standard, ANSI/NB-23, The National Board of Boiler and Pressure Vessel Inspectors, Columbus, Ohio, January, 1991 • ASME Boiler and Pressure Vessel (B&PV) Code, Section I. “Power Boilers”, particularly Part PW, “Requirements for Boilers Fabricated by Welding” and PG, “General Requirements of All

• ASME B&PV Code, Section IX, “Welding and Brazing Qualifications” includes requirements for weld procedures, “Welding Procedure Specification” (WPS), procedure records, “Welding Procedure Qualification Record” (PQR), and welder qualification records, “Welder/Welding Operator Qualification Record” (WQ). • ASME B&PV Code, Section V, “Nondestructive Examination” covers inspection requirements for components fabricated by welding. • ASME B&PV Code, Section II, Part A, “Material Specification for Ferrous Materials” and Part C, “Welding Filler Materials” cover materials used in welding. To reduce the number of weld procedures, the ASME Code groups materials into “P” groups. Table 11-2 provides a summary of the ASME, Section IX classifications. Note that these groupings are for weld procedure development and execution only; the materials are not interchangeable in their properties. It should be further noted that the Code is written for original design and construction, usually shopwelded, and not for repair situations. Because of the difficulties in assuring the highest quality welds in a field condition (access problems, weld positions possible, inability to post-weld heat treat, etc.), repairs may not have the same margin of safety as inherent original construction performed to the Code requirements. On the other hand, there is a margin of safety built into Code requirements. The main message here is that engineering assessment and judgment may be required for

Volume 1: Boiler Tube Fundamentals

11-3

Table 11-2 P-Number Groupings for ASME Specified Tube Materials P- Number 1

3

ASME Specification Number and Grade

Nominal Composition Carbon steel

SA-178A, C, D SA-192 SA-210A1, C

Carbon-molybdenum Chromium-molybdenum Manganese-molybdenum

SA-209T1, T1A, T1B SA-213T2

4

1% to 11/2% Cr - 1/2% Mo 13/4% to 2% Cr - 1/2% Mo

SA-213T11, T12 SA-213T3

5

21/4% Cr - 1% Mo 3% Cr - 1% Mo 5% Cr - 1/2% Mo 7% Cr - 1% Mo 9% Cr - 1% Mo 9% Cr - 1% Mo - V - Cb- N

SA-213T22 SA-213T21 SA-213T5 SA-213T7 SA-213T9 SA-213T91

8

Austenitic stainless steel

SA-213TP304 SA-213TP316 SA-213TP321 SA-213TP347

Adapted from: G.G. Stephenson and J.W. Prince1

specific repair conditions; such assessments might include gaining a knowledge of: (i) actual operating conditions (temperatures, stresses, cyclic and steady state), (ii) the actual integrity of components (actual material properties in the aged condition, extent of damage and defect sizes), and (iii) the actual accumulation of damage in the component.

11.5 Specific Repair Procedures 11.5.1 Overview and general comments. The following will be applicable to all boiler tube weld repair procedures1: • Every utility should use a qualified welder with qualified procedures for all repairs. • Where possible use gas tungsten arc welding (GTAW) for root

11-4

passes rather than shielded metal arc welding (SMAW). Either can be used for the filler passes.

water sootblowers. Boiler tube failures have resulted from the mechanical damage done during slag removal (Chapter 44, Volume 3) and from sootblower damage (Chapter 22, Volume 2; Chapter 38, Volume 3). • Access is critical, and may be difficult. It may be necessary to cut away a sound tube to gain access to a damaged tube. Detailed instructions for the repair of boiler tubes by: (i) tube section replacement, (ii) grinding out the defect and welding, (iii) pad welding, (iv) window welding, (v) boiler tube build-up, (vi) dissimilar welds between low alloy steels, (vii) dissimilar metal welds between ferritic steels and austenitic stainless steels, and (viii) for large-scale tube replacement, have been developed.1 Table 11-3 summarizes the use of repair strategies (i) through (v). 11.5.2 Tube section replacement. This is the preferred method for boiler tube repair. Several of the steps are key from the viewpoint of boiler tube failures: • The damaged section of the tube should be saved for metallurgical examination (see Chapter 6), particularly if the mechanism or root cause are unknown.

• Do not use backing rings for tubing in the water-touched circuits. They will cause flow disruption, a precursor to the underdeposit corrosion mechanisms of hydrogen damage, caustic gouging, and acid phosphate corrosion.

• During cutting steps, care must be taken to avoid getting debris into tubes that could lead to blockage and subsequent boiler tube failures by short-term overheating. Thus, cut the bottom first and block.

• Use backing rings in SH/RH tubing only if a high quality root pass by either GTAW or SMAW cannot be made without backing.

• The replacement tube section should be the same material and thickness of that removed. If matching thickness is not available a thicker tube can be used, but not thinner. If a tube of the same or thicker section is not available, a higher grade material may be substituted. However, an austenitic stainless steel tube should not be used in watertouched service.

• Some semi-automatic welding processes have been used in limited applications, mainly for skin casing and membrane replacement. • Slag removal is required prior to performing maintenance work; removal methods include water lancing, shotgun, rappers and

Repair and Replacement of Boiler Tubes

• The length of replacement should be shorter than the distance between the prepared ends by 1.6 mm to 3.2 mm (1/16 in. to 1/8 in.) to allow for a root gap and some shrinkage of the first weld. The replacement length is more critical for waterwall tube replacement where tubes are held in place by membranes.1 • All completed welds should be examined visually and inspected according to Code requirements. A hydrostatic test is also recommended. Radiography may be used instead of hydrostatic testing if approved by regulatory agencies. Although the details will vary according to location and local conditions, the fit-up details for a typical boiler tube replacement and the sequence for a welding repair are shown in Figures 11-2 and 11-3, respectively.

Table 11-3 Repair Techniques and Their Application Method

Applicability

Tube section replacement

• The preferred repair method. • Consists of replacing tube section with one of same material and dimensions.

Grind out and re-weld

• Applicable for small defects such as pin-hole leaks. • Vee- preparation and re-weld.

Pad welding

• Should only be used in an emergency as repeat failures are nearly always guaranteed. Should be replaced at next major outage. • Involves mechanically closing the rupture to original contour, veeing a weld groove and re-welding.

Window welding

• May be required if access all around the tube is limited. Should be used only as a temporary measure with full replacement at next major outage.

Boiler tube build-up

• For restoring thinned tubes to original thickness.

Source: G.G. Stephenson and J.W. Prince1

11.5.3 Grinding out the defect and welding. This procedure should only be used for small defects such as pinhole leaks caused by porosity. If internal damage is suspected, it will not be possible to limit the vee dimension at the root to an acceptable level. Although quicker than a tube replacement, there is a significant chance of repeat failures with this kind of repair.1 11.5.4 Pad welding. A temporary repair such as pad welding or window welding may be performed to return the unit to service in the shortest possible time, but only in exceptional circumstances if absolutely necessary to reduce the time of a forced outage. Under such circumstances a pad weld must be replaced with a permanent repair at the next available outage; as a result, careful documentation of pad weld locations is indicated. Pad welding sometimes involves mechanically restoring the tube to its original shape by bending the edges of the rupture back together, followed by sealing the root and filling the vee with a weld. Differences in the thickness in the weld area are possible so that care is required not to burn through the thin sections.

Second Weld Root gap before starting first weld 0.0"Ð 1/32" Note: This end must be free to move down 1/16" due to shrinkage in first weld

Weld Details 37 1/2° + 2 1/2°

0" to 3/32"

1/16" to 5/32"

I.D. 2

I.D. 1 I.D. 1 shall be within 1/16" of I.D. 2

First Weld Root gap 1/16" Ð 5/32" Complete this weld before starting on opposite end

Figure 11-2. Fit-up details for boiler tube replacement. Source: G.G. Stephenson and J.W. Prince1

Volume 1: Boiler Tube Fundamentals

11-5

Good tube

Replacement tube

process. Figure 13-22 in Volume 2 shows an example of an attempt to place a pad weld over a region of corrosion fatigue attack.

Good tube

Similar considerations occur in the repair of superheater/reheater tubing, where there is uncertainty about the tube condition such as the existence of creep cracks and their depths, and of the conditions of the internal tube surface.

➂ Tack weld ➃ Complete weld Alignment bar (optional)

Weld back ➄ membrane (The membrane will be cut out the full tube length if tube is to be pulled out of plane for repair)

➀ Tack weld ➁ Complete weld

2" min 1/2" min

Figure 11-3. Sequence of welding a replacement tube section. Source: G.G. Stephenson and J.W. Prince1

Pad welding of thinned tubes has been used as a temporary measure until a scheduled outage. It is not, however, recommended as a regular or permanent repair procedure, nor as a long-term “fix” of the problem for the following reasons:

• Heat input of pad welding can affect internal deposits. Particularly bad is the presence of Cu in the internal deposits on water wall tubes. These can, during the welding process, become molten and embrittle the grain boundaries of the tube wall. If Cu is likely to be present some organizations have adopted a rule that pad welds should not be applied where the tube thickness is less than 3 mm.2

11-6

• There is no critical determination of the minimum wall to which the process can be applied. If applied to a wall which is too thin, then the weld beads can penetrate the wall or make the internal surface protrude into the water or steam flow. This will cause a disruption of the flow and lead to deposition, with the associated and concomitant events eventually leading to an underdeposit corrosion mechanism: hydrogen damage, caustic gouging, or acid phosphate corrosion. • There is no indication of whether a crack is present on the internal surface, such as is caused by corrosion fatigue. The pad welding will not seal the crack, which can then act as an initiating center or exacerbate the cracking

Repair and Replacement of Boiler Tubes

11.5.5 Window welding (canoe piece repair). The sequence for a window welding repair is illustrated in Figure 11-4. As for pad welding, a window welding repair should only be used as a temporary measure and should be replaced at the next available outage. Its primary use is where accessibility is limited. The main problem with the procedure is making an acceptable root fit-up all around the opening. If the patch does not match the interior dimension of the tube, flow disruption may occur initiating increased deposition and associated underdeposit boiler tube failures. 11.5.6 Boiler tube build-up. Weld build-up can be used to increase the wall thickness of thinned tubes.1 It should not be used if the remaining wall thickness is less than 60% of design thickness or on tubes less than 1.59 mm (1/16 inch) thick. Ultrasonic testing should be used to ensure these criteria are met and to determine the extent of thinned wall. Either GTAW or SMAW can be used to make the deposit; the former is preferred on thinner (up to 0.090 in thick) tubes to prevent burn-through. Care should be taken to remove oxides and scale prior to welding. The arc should be started in the thicker areas. Overlap beads by 50-75%. For vertical welds, downhill progression should be used. Finally, the reinforcement should be limited to 1/8 inch above the original surface.

11.7 Welding Co-Extruded Tubing

Locate bad section

Drill window holes

Remove damaged section

Grind: A) Internal bevel B) Window bevel

Replacement tube

Weld: A) Tube section B) Window plugs

Figure 11-4. Welding sequence for replacing a defective tube section using the window welding technique. Source: G.G. Stephenson and J.W. Prince1

11.5.7 Dissimilar welds between low alloy steels. In SH/RH repairs, it may be necessary to join two different ferritic materials. Filler metals should be qualified in the repair procedure. Successful procedures have used filler metals that match one or other of the base materials or provide some intermediate value. 11.5.8 Dissimilar metal welds between ferritic steels and austenitic stainless steels. Repairs involving welds between ferritic and austenitic stainless steels are discussed at length in Chapter 35, Volume 3. In brief, the optimal procedure is to use a shop-welded “dutchman” so that field repairs involve only like material welds, e.g., ferritic-to-ferritic and stainless-to-stainless. For in-situ welds, repairs with nickel-base filler metals are preferred although they are somewhat more difficult to execute than welds with stainless steel filler.

11.5.9 Large-scale tube replacement. For the large scale replacement of tubing, the procedures are mostly similar to those described above, with some additional precautions such as1: (i) leaving sufficient tubes in place to provide support until the replacements are installed, (ii) capping off the tubes to ensure that no debris enters tubes once they are cut, (iii) tack welding and completing the bottom weld first, and (iv) tack welding and completing the top welds starting with the smallest gaps first.

11.6 Documentation Documentation of welding procedures should be maintained including1: (i) welding procedure specifications, (ii) procedure qualification records, (iii) weld qualifications, (iv) heat treatment procedures, (v) inspection procedures, and (vi) specific detailed procedures, such as covering specific boiler tube repair procedures.

A somewhat specialized topic related to welding is that of coextruded tubing. One of the options available for the mitigation of fireside corrosion in either waterwalls or superheater/reheater tubing is the use of co-extruded tubing. This solution uses a highly corrosionresistant outer material that is metallurgically bonded to a less-expensive inner layer. The former CEGB has been a leader in the experience base of these materials. They typically used an outer layer of AISI T310 over an inner layer of hightemperature, high-strength stainless steel for superheater/reheater application, and AISI T310 outer with carbon steel inner layer for waterwall applications.3 The experience base has indicated corrosion performance improvements of 2 to 5 times the original material for these materials. Through 1984, the CEGB had over 10 years of experience in these materials including approximately 70,000 welds without technical weld problems or any weld failures. The primary goal of the welding of these materials is to use two different filler metals so as to have the eventual weld properties match those of the corresponding base metal. Flatley and Thursfield3 reviewed the required weld procedures. Conventional weld techniques, mostly SMAW and GTAW, are suitable. Generally a root pass of GTAW is followed by runs using either of the above processes and a filler consistent with the inner material. Then the outer layer filler is laid in with a corrosion-resistant material such as T309 applied by SMAW. Neither preheat nor post-weld heat treatment have been required.

Volume 1: Boiler Tube Fundamentals

11-7

11.8 References 1Stephenson,

G.G. and J.W. Prince, Guidelines on Fossil Boiler Field Welding, Research Project 2504-02, Final Report TR-101699, Electric Power Research Institute, Palo Alto, CA, January, 1993. 2Personal

Communication from D. Barnett (Pacific Power, Australia) to R.B. Dooley, 2-12-95.

11-8

Repair and Replacement of Boiler Tubes

3Flatley,

T. and T. Thursfield, “Review of Corrosion Resistant Co-Extruded Tube Development for Power Boilers”, in R.D. Sisson, Jr., ed., Coatings and Bimetallics for Aggressive Environments, American Society for Metals, Metals Park, OH, 1985.

Index

Acid contamination, 15-10 through 15-12 Acid deposition, 30-2, 30-3 Acid dewpoint corrosion (economizer), 30-1 through 30-12 actions, 30-8 through 30-12 determining the extent of damage, 30-6, 30-11 features of failure, 30-2, 30-8 locations of failure, 30-2 long-term actions and the prevention of repeat failures, 30-6, 30-12 mechanism, 30-3, 30-4, 30-9 precursors, 30-8 ramifications/ancillary problems, 30-12 repairs and immediate solutions/ actions, 30-6, 30-12 root causes and actions to confirm, 30-5, 30-10 Acid phosphate corrosion, 16-1 through 16-28 actions, 16-22 through 16-28 case study, 16-16 through 16-20 deposit characteristics, 7-1 through 7-4, 16-2, 16-4, 16-6, 16-19, 16-20 determining the extent of damage, 16-13, 16-25 distinguishing from hydrogen damage or caustic gouging, 7-1, 7-2, 16-3 features of failure, 7-1, 7-2, 16-2 through 16-4, 16-18 through 16-20, 16-22 locations of failure, 16-4, 16-5 long-term actions and the prevention of repeat failures, 16-14, 16-15, 16-27 mechanism, 16-6 through 16-8, 16-23 precursors, 16-22 ramifications/ancillary problems, 16-28 repairs and immediate solutions/ actions, 16-13, 16-26 root causes and actions to confirm, 7-4, 16-9 through 16-12, 16-24 Additives, oil-fired units, 34-10, 34-11, 34-18, 34-19 Air inleakage, 13-24, 27-6, 27-7, 27-9, 30-4, 30-5, 41-6

Alkali iron trisulfates, 33-2, 33-7, 33-8 Alkali salts, 33-2, 33-7, 33-8 All-volatile treatment (see also Feedwater treatment), 1-18, 3-9, 3-13 “Alligator hide”, 32-2, 33-3, 33-4, 34-5 American Society of Mechanical Engineers (ASME) Codes design, 2-2 through 2-6 non-destructive examination, 11-3 welding 11-3, 11-4 Ammonia, 3-8, 3-9 Ash analysis, 33-12 Austenitic welds (in dissimilar metal welds), 11-7, 35-2 through 35-9, 35-15 Availability losses and improvement, 1-20 Backing rings, 2-14, 11-4 Baffles (erosion), 14-12 Bell-shaped corrosion curve, 33-7, 33-8 Black boiler water samples, 16-11 Boiler pressure drop losses, 19-5, 19-6 Boiler Tube Failure (BTF) Reduction Program, 1-20, 5-1 through 5-3 corporate directives for BTF reduction, 5-2 goals, 1-20, 1-21, 5-2 multidisciplinary teams for BTF reduction, 5-2 Boiler tube failures formalizing programs for reduction of, 1-20, 5-1 through 5-6 historical developments in identification, correction and prevention, 1-16, 1-18 importance, 1-1 importance of operation and maintenance procedures in preventing, 4-1 influence of cycle chemistry, 1-18, 3-1 through 3-2 influence of fuel options, 1-18 influence of operating conditions, 1-18 influence of unit lay-up, 4-9 influence of unit transients, 4-8, 4-9 influencing or influenced by chemical cleaning, 4-2

largest availability losses, 1-1, 1-2 precursors to, 1-4, 1-10 through 1-15, 1-16, 12-7 through 12-12, 31-7 through 31-13 repeat failures, 1-20, 1-21 reporting and report form, 5-3 through 5-5 resulting from breakdown of protective magnetite in water-touched tubing, 2-11 resulting from breakdown of protective oxide in steamtouched tubing, 2-15 resulting from fireside conditions, 2-21 screening table, steam-touched tubes, 1-8, 1-9, 31-4, 31-5 screening table, water-touched tubes, 1-6, 1-7, 12-4, 12-5 steps in generic investigation 1-4, 1-5, 1-16, 12-2, 12-3, 31-2, 31-3 with significant microstructural changes, 10-2 worldwide statistics, 1-1 Boiler tubes (see also Superheater/ reheater tubes and Waterwalls and economizer tubes) design considerations, 2-2 through 2-6 materials and alloys, 2-2, 2-3, 2-6 maximum design and oxidation temperatures, 2-4, 23-2, 23-3 Boiler water treatment, 3-1 through 3-8 all-volatile treatment (see also Feedwater treatment), 1-18, 3-9, 3-13 caustic treatment, and caustic gouging, 17-5, 17-6, 17-10 guidelines for, 3-5, 3-6, 3-13 historical development of, 1-18, 3-5, 17-5 success factors for use of, 3-5 comparison of options, 3-6 effect on boiler tube failures 3-1, 3-2 factors during unit transients, 4-8 optimization of, 3-6 through 3-8

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-1

phosphate treatments, 3-3 through 3-5, 16-6 and acid phosphate corrosion, 16-6 through 16-8, 16-11, 16-12,16-16, 16-17 effect of chemical additions on operating regimes, 3-4 guidelines for, 3-5, 3-13 historical development of, 1-18, 3-3 Borio index, 33-9 Bubbling-bed FBC units, 47-1 through 47-12 chromized tubes, 47-10 plasma coatings, 47-10 tube armoring, 47-10 Burner misalignment, 15-10, 16-11, 17-10 Carbides, 10-5, 10-6 Carryover, 37-5, 37-6 of Na2SO4, 41-2, 41-5, 41-6 Caustic gouging, 17-1 through 17-22 actions, 17-16 through 17-22 case study, 17-14 deposit characteristics, 7-1 through 7-3, 17-2, 17-3, 17-6, 17-7 determining the extent of damage, 17-11, 17-19 distinguishing from hydrogen damage or acid phosphate corrosion, 7-1, 7-2, 17-2 electrochemical corrosion cell, 17-6, 17-7 features of failure, 7-1, 7-2, 17-2, 17-3, 17-16 locations of failure, 17-2 through 17-4 long-term actions and the prevention of repeat failures, 17-12, 17-13, 17-21 mechanism, 2-11, 2-14, 17-5 through 17-7, 17-17 precursors, 17-16 ramifications/ancillary problems, 17-22 repairs and immediate solutions/ actions, 17-11, 17-20 root causes and actions to confirm, 7-4, 17-8 through 17-10, 17-18 Caustic treatment (see also Boiler water treatment),1-18, 3-5, 3-6, 3-13, 17-5,17-6, 17-10

Chemical cleaning (see also Chemical cleaning damage in super heater/reheater tubes and Chemical cleaning damage: waterwalls) as indicator of non-optimized feedwater chemistry, 3-2 boiler tube failures influenced by, 4-2, 36-6, 36-8 effect of changing to oxygenated treatment, 3-11, 3-12 FBC units, 4-8 superheaters/reheaters, 4-5 through 4-7, 32-21, 33-21, 34-19, 37-5 through 37-10 importance of sampling, 4-6 locations to clean, 4-6 monitoring, 4-7 process optimization, 4-6, 4-7 reasons to perform, 4-5 solvent choice, 4-6 typical operations for, 4-7 when to clean, 4-6 waterwalls, 4-1 through 4-5 assessing cleanliness and deposit levels, 4-2, 4-3 guidelines for, 4-1 importance, 4-1 inhibitor breakdown, 25-4 monitoring Fe levels to determine finish, 4-5 possible problems that could lead to damage, 25-4 solvent choice, 4-3, 4-4 typical operations for, 4-5 when to clean, 4-2 Chemical cleaning damage in superheater/reheater tubes, 43-1 through 43-8 actions, 43-5 through 43-8 determining the extent of damage, 43-3, 43-7 features of failure, 43-2, 43-5 locations of failure, 43-2 long-term actions and the prevention of repeat failures, 43-4, 43-8 mechanism, 43-2, 43-6 precursors, 43-5 ramifications/ancillary problems, 43-8 repairs and immediate solutions/actions, 43-4, 43-8 root causes and actions to confirm, 43-3, 43-7

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-2

Chemical cleaning damage: water walls, 25-1 through 25-9 actions, 25-6 through 25-9 determining the extent of damage, 25-4, 25-8 features of failure, 25-2, 25-3, 25-6 long-term actions and the prevention of repeat failures, 25-5, 25-9 mechanism, 25-4, 25-7 precursors, 25-6 ramification/ancillary problems, 25-9 repairs and immediate solutions/ actions, 25-5, 25-9 root causes and actions to confirm, 25-4, 25-7 Chlorine in coal, 18-5, 18-6, 33-10, 33-11, 47-6 Chordal thermocouples, 9-8, 9-9 Chromizing waterwalls, 19-15, 19-16 Circulating-bed FBC units, 48-1 through 48-4 coatings, 48-2 erosion/abrasion, 48-2 through 48-4 underdeposit corrosion, 48-2 Coal composition (see also Combustion process and/or Fireside scale/ash) and corrosiveness, 18-4 through 18-6, 33-8 through 33-11 and erosiveness, 14-5 through 14-7 effect of chlorine content on fire side corrosion in steamtouched tubes, 33-10, 33-11 effect of chlorine content on fireside corrosion in watertouched tubes 18-5, 18-6 effect of sulfur level on fireside corrosion in water-touched tubes, 18-4 Coal particle erosion, 28-1 through 28-5 actions, 28-3 through 28-5 description and manifestation, 28-1 Coal Quality Impact Model (CQIM) 2-22, 33-14, 33-20 Coatings, 22-4, 48-2 for fireside corrosion in steamtouched tubing, 33-18, 34-16, 34-17 for fireside corrosion in watertouched tubing, 18-12 through 18-14 for sootblower erosion, 38-5

Cold air velocity test (CAVT) (see also Flyash erosion), 14-12 through 14-18 Co-extruded tubing for fireside corrosion in steamtouched tubes, 33-20, 34-17 for fireside corrosion in watertouched tubes, 18-14 welding, 11-7 Cold bent tubes and lowtemperature creep, 24-4 Cold end corrosion, 30-1 Combustion process, ash formation, erosiveness, and deposition, 2-22 through 2-24 formation of gaseous species, 2-22, 18-4 Commissioning of units, activities to prevent future boiler tube failures, 4-9, 4-11 Concentration in deposits, 2-13, 2-14, 15-4 through 15-6, 15-8, 15-10, 16-5 through 16-7, 17-4, 17-6, 17-7 Condenser leaks, 15-10, 15-11, 37-6 Congruent phosphate treatment (see also Boiler water treatment), 3-4, 16-6 Coordinated phosphate treatment (see also Boiler water treatment), 3-3, 3-4 Core monitoring parameters for cycle chemistry, 3-14 Corporate commitment needed to solve boiler tube failures, 5-1, 5-2 Corporate directives for BTF reduction, 5-2 Corrosion indices, 18-5, 18-6, 33-8 through 33-11 rates as a determinant of repair choices, 18-11 Corrosion fatigue, 13-1 through 13-41 actions, 13-35 through 13-41 analysis of field experience, 13-13 through 13-15 breakdown of magnetite, 13-10 through 13-12 case study, 13-30 through 13-32 determining the extent of damage, 13-26, 13-38 distinguishing from OD-initiated fatigue, 7-6, 7-7 environmental effects on initiation and propagation, 13-16 through 13-20

features of failure, 13-2 through 13-5, 13-35 Influence Diagram for the analysis of corrosion fatigue, 13-24 through 13-26, 13-30 through 13-32 locations of failure, 13-6 through 13-9 long-term actions and the prevention of repeat failures, 13-28, 13-29, 13-40 mechanism, 2-11, 13-10 through 13-20, 13-36 oxygenated treatment, effect on corrosion fatigue, 13-20 phosphate treatment, effect on corrosion fatigue, 13-18, 13-20 precursors, 13-35 ramifications/ancillary problems, 13-41 repairs and immediate solutions/actions, 13-27, 13-39 root causes and actions to con firm, 13-21 through 13-26, 13-37, 13-38 stress effects on initiation and propagation, 13-15, 13-16 Corrosion products, 1-17, 3-1, 3-2 Creep (see also Long-term overheating and Low-temperature creep cracking), 6-8, 7-6, 7-8, 24-1 through 24-11, 32-1 through 32-32 Creep cavitation, 10-6 Creep damage assessment techniques, 10-5 through 10-8 Larson-Miller Parameter (LMP), 10-2 through 10-6 Cycle chemistry (see also Boiler water treatment and Feedwater treatment), core monitoring parameters, 3-14 developing unit-specific guidelines, 3-12 through 3-13 diagnostic parameters, 3-14 goals for improvement program, 3-1, 3-2 guidelines documents for, 3-13 instrumentation and monitoring, 3-14 setting action levels, 3-12, 3-13 Cycling of units, 4-8, 4-9, 13-24, 20-4, 26-5, 35-12, 39-5 effect on boiler tube failures, 4-8, 4-9 effect on thermal fatigue in economizer inlet header tubes, 20-2

Departure from nucleate boiling (DNB), 2-12, 2-13 Deposit density, 4-2 Deposit weight, 4-2, 4-3 Deposits (see Waterside deposits, Feedwater corrosion products, Concentration in deposits, Oxides internal in steam-touched tubes, Oxides internal in water-touched tubes, Fireside scale/ash), 15-2, 15-4, 16-2 through 16-5, 17-2 through 17-4, 19-5 Diffusion screens (erosion), 14-14, 14-16 through 14-18 Dissolved oxygen, 13-11, 13-12, 13-16 through 13-20, 21-3, 21-4 Dissimilar metal welds, 35-1 through 35-25 actions, 35-19 through 35-25 case study, 35-17 determining the extent of damage, 35-12, 35-13, 35-22 features of failure, 35-2 through 35-5, 35-19 influence of welding variables, 35-7 through 35-9 locations of failure, 35-3 long-term actions and the prevention of repeat failures, 35-15, 35-16, 35-24 mechanism, 35-6 through 35-9, 35-20 microstructural changes in service, 35-6, 35-7 precursors, 35-19 ramifications/ancillary problems, 35-25 repairs and immediate solutions/ actions, 35-14, 35-23 root causes and actions to confirm, 35-10 through 35-12, 35-21 Distorted or misaligned tubes, 14-3, 14-4, 14-11, 33-6, 33-15, 33-21, 34-5, 34-15, 34-19, 35-10, 39-5, 40-1, 40-3 Distribution screens (erosion), 14-14, 14-16 through 14-18 DMW LIFE code, 35-16 Drum boiler water treatment, 3-3 through 3-8 Drum level control, 23-6, 37-6, 41-6 “Dutchman” repair, 11-7, 35-23 Economizer inlet header tube failures (see Erosion-corrosion of economizer inlet header tubes and/or Thermal fatigue in economizer inlet header tubes)

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-3

Electric resistance flash welding, 45-1 through 45-3 Equilibrium phosphate treatment (see also Boiler water treatment), 3-3, 3-5, 16-14 Erosion (see also Flyash erosion, Coal particle erosion, Falling slag damage, Sootblower erosion in superheater/reheater tubes, Sootblower erosion in water-touched tubing, Fluidized-bed combustion (FBC) units) abrasion index, 14-6, 14-7 basics of damage mechanism, 14-5 erosiveness of ash constituents, 2-23, 2-24, 14-5, 14-6 wear propensity calculation, 14-6, 14-7 Erosion-corrosion, general 3-9, 21-3, 21-4 Erosion-corrosion of economizer inlet header tubes, 21-1 through 21-9 actions, 21-7 through 21-9 determining the extent of damage, 21-5, 21-8 distinguishing from thermal fatigue and flexibility-induced cracking, 7-6, 7-7 features of failure, 21-1, 21-2, 21-7 locations of failure, 21-2, 21-3 long-term actions and the prevention of repeat failures, 21-5, 21-9 mechanism, 21-3, 21-8 precursors, 21-7 ramifications/ancillary problems, 21-9 repairs and immediate solutions/ actions, 21-5, 21-9 root causes and actions to confirm, 21-4, 21-8 Excess oxygen, high excess air in oil-fired units, 34-14, 34-19 low excess air, 18-1, 18-7, 34-14 Exfoliation of SH/RH steamside oxide, 2-17 through 2-21, 36-5, 36-7 effect of unit chemistry on, 2-21 effects, 2-17, 2-18, 2-21 failure criterion, 2-18, 2-20 rating severity of, 2-18, 2-19 susceptible materials, 2-18, 2-20, 2-21

Failure mechanisms fluidized-bed units, Chapters 47 and 48 list, 1-3 steam-touched tubes, Volume 3 waste-to-energy units, Chapter 49, Volume 3 water-touched tubes, Volume 2 Falling slag damage, 29-1 through 29-6 actions, 29-3 through 29-6 description and manifestation, 29-1, 29-2 Fatigue in superheater/reheater tubes, 39-1 through 39-12 actions, 39-9 through 39-12 determining the extent of damage, 39-7, 39-11 features of failure, 39-2, 39-9 locations of failure, 39-3, 39-4 long-term actions and the prevention of repeat failures, 39-7, 39-12 mechanism, 39-5, 39-10 precursors, 39-9 repairs and immediate solutions/actions, 39-7, 39-12 root causes and actions to confirm, 39-5, 39-6, 39-11 Fatigue in water-touched tubes, 26-1 through 26-12 actions, 26-9 through 26-12 determining the extent of damage, 26-7, 26-11 distinguishing from corrosion fatigue, 7-6, 7-7, 26-3 features of failure, 26-2, 26-9 locations of failure, 26-3, 26-4 long-term actions and the prevention of repeat failures, 26-8, 26-12 mechanism, 26-5, 26-10 precursors, 26-9 repairs and immediate solutions/ actions, 26-8, 26-11 root causes and actions to confirm, 26-6, 26-7, 26-11 Feedwater corrosion products, 1-17, 3-1, 3-2, 15-4, 15-14, 16-4, 16-5, 16-14, 17-2 through 17-4, 17-12, 23-5 Feedwater treatment, 3-8 through 3-12 all-volatile treatment (AVT), 3-9 guidelines for, 3-13 historical development of, 1-18 comparing AVT and oxygenated treatment, 3-9, 3-11, 3-12

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-4

factors during unit transients, 4-8, 4-9 importance of proper choice of, 3-8 optimizing for all-ferrous feedwater trains, 3-9 through 3-11, 21-5, 21-6 optimizing for mixed metallurgy feedwater trains, 3-11, 21-5, 21-6 oxygen scavenger use, 3-8 through 3-10, 21-3, 21-6 oxygenated treatment (OT), 3-9 through 3-11 effect on corrosion fatigue, 13-20 effect on oxide growth and exfoliation, 2-21 guidelines for in once-through and drum units, 3-13 historical development of, 1-18 to reduce deposition in waterwalls, 19-5, 19-6, 19-11, 19-13, 19-15 problems with erosion-corrosion throughout unit, 3-9 Fe-Fe carbide equilibrium diagram, 7-5, 23-2 FeO, 2-7, 32-9 Fe2O3, 2-7, 2-16, 2-17, 2-20, 32-9 Fe3O4, 2-7, 2-16, 2-17, 2-20, 32-9 Ferric oxide hydrate (FeOOH), 3-10 Film boiling, 2-12, 2-13 Finite element analysis for analyzing corrosion fatigue, 13-29 Fireside corrosion in SH/RH tubes (coal-fired units), 33-1 through 33-30 actions, 33-24 through 33-30 case study, 33-22 determining the extent of damage, 33-15, 33-27 distinguishing from long-term overheating, 6-8, 7-6 through 7-8, 33-4, 33-5 features of failure, 7-6, 7-8, 33-2 through 33-5, 33-24 locations of failure, 33-6 long-term actions and the prevention of repeat failures, 33-17 through 33-21, 33-28, 33-29 mechanism, 33-7 through 33-11, 33-25 precursors, 33-24 ramifications/ancillary problems, 33-30 repairs and immediate solutions/ actions, 33-16, 33-28

root causes and actions to confirm, 33-12 through 33-15, 33-26, 33-27 use of indices to predict likelihood of, 33-8 through 33-10, 33-15 Fireside corrosion in SH/RH tubes (oil-fired units), 34-1 through 34-26 actions, 34-21 through 34-26 determining the extent of damage, 34-14, 34-24 distinguishing from long-term overheating, 6-8, 7-6 through 7-8, 7-9, 34-5, 34-6 features of failure, 7-6, 7-8, 34-2 through 34-5, 34-21 locations of failure, 34-5 long-term actions and the prevention of repeat failures, 34-16 through 34-19, 34-25 mechanism, 34-7 through 34-10, 34-22 precursors, 34-21 ramifications/ancillary problems, 34-26 repairs and immediate solutions/ actions, 34-15, 34-24 root causes and actions to confirm, 34-11 through 34-14, 34-23 Fireside corrosion in water-touched tubes, 18-1 through 18-24 actions, 18-18 through 18-24 case study, 18-16 determining the extent of damage, 18-11, 18-21 effect of coal chlorine content on, 18-5, 18-6 features of failure, 18-2, 18-3, 18-18 locations of failure, 18-2, 18-3 long-term actions and the prevention of repeat failures, 18-12 through 18-15, 18-23 mechanism, 18-4 through 18-6, 18-19 precursors, 18-18 ramifications/ancillary problems, 18-24 repairs and immediate solutions/ actions, 18-11, 18-22 root causes and actions to confirm, 18-7 through 18-10 summary of field experience, 18-16

Fireside scale/ash, compositional analysis of, 33-12, 33-15 development on SH/RH tubing, 32-10 metallurgical analysis of, 6-9 Flame impingement, 15-10, 16-11, 17-10 Fluidized-bed combustion (FBC) units boiler tube failures in bubblingbed units, 47-1 through 47-12 boiler tube failures in circulatingbed units, 48-1 through 48-4 chemical cleaning of, 4-8 Fluxdome, 9-9 Flux meter, 9-9 Flyash erosion, 14-1 through 14-29 actions, 14-23 through 14-29 case studies, 14-19 through 14-21 cold air velocity test (CAVT), 14-12 through 14-18 determining the extent of damage, 14-11, 14-26 distinguishing from sootblower erosion in SH/RH tubes, 7-9 estimating solids loading, 14-16 features of failure, 14-2, 14-3, 14-23 locations of failure, 14-3, 14-4 long-term actions and the prevention of repeat failures, 14-12 through 14-18, 14-28, 14-29 mechanism, 14-5 through 14-7, 14-24 precursors, 14-23 protection options, 14-16 through 14-18 ramifications/ancillary problems, 14-29 repairs and immediate solutions/ actions, 14-11, 14-27 root causes and actions to confirm, 14-8 through 14-10, 14-25 Forging laps, 45-1, 45-2, 45-3 Fossil-fuel power plants, primary components, 1-16 Fretting, 40-1 through 40-5 Fuel changing, blending, washing, 14-10, 18-10, 18-14, 30-3, 30-5, 30-6, 33-15, 33-20, 34-7 Gas-touched length (GTL), 32-8, 34-5 plotting as a diagnostic tool, 32-15, 33-12, 33-15

Gas tungsten arc welding (GTAW), 11-4, 11-6, 11-7 Gouging of tubes, 15-2, 15-3, 16-2, 16-3, 17-2, 17-3 Graphitization, 42-1 through 42-11 actions, 42-9 through 42-11 determining the extent of damage, 42-6, 42-11 distinguishing from dissimilar metal weld failures, 42-3 distinguishing from long-term overheating (creep), 7-9, 42-3 features of failure, 42-2, 42-3, 42-9 kinetics of growth, 42-4, 42-5 locations of failure, 42-2 long-term actions and the prevention of repeat failures, 42-8, 42-11 mechanism, 42-4, 42-5, 42-10 repairs and immediate solutions/ actions, 42-7, 42-11 root causes and actions to confirm, 42-6, 42-11 Hardness assessing changes in, 10-4, 10-5 metallurgical analysis, 6-7 Header flexibility, 39-4 Heat flux effects of high levels, 15-10, 16-11, 17-10, 18-9 measuring with Fluxdome, 9-9 measuring with a flux meter, 9-9 monitoring, 9-9 Heat recovery steam generators (HRSG), 30-1 Hideout of phosphate, 3-4, 3-8, 16-6, 16-11, 16-12 Hydrazine, 3-8, 3-9, 21-4, 27-7 Hydrogen damage, 15-1 through 15-30 actions, 15-21 through 15-30 case studies, 15-16 through 15-19 deposit characteristics, 7-1 through 7-4, 15-3 determining the extent of damage, 9-1, 9-6, 9-7, 15-13, 15-25 distinguishing from caustic gouging or acid phosphate corrosion, 7-1, 7-2, 15-3 electrochemical corrosion cell, 17-6 features of failure, 7-1, 7-2, 15-2, 15-3, 15-7, 15-21 locations of failure, 15-4

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-5

long-term actions and the prevention of repeat failures, 15-14, 15-15, 15-27 mechanism, 2-11, 2-14, 15-5 through 15-7, 15-22, 15-23, 17-6 microstructural changes, 7-3, 15-2, 15-3, 15-7 precursors, 15-21 ramifications/ancillary problems, 15-29 repairs and immediate solutions/ actions, 15-14, 15-26 root causes and actions to confirm, 7-4, 15-8 through 15-12, 15-24 Hydrogen sulfide, 18-4 Hydrostatic testing, 9-10 In-bed wastage (in bubbling-bed FBC units), 47-5 through 47-8, 47-11, 47-12 Induction pressure welds (in dissimilar metal welds), 35-2, 46-2 Influence diagram, 13-24 through 13-26, 13-30 through 13-32 Inhibitor breakdown during chemical cleaning, 25-4 Inspection, 9-1 through 9-12 acoustic monitoring, 9-8 codes and standards, 9-3 eddy current testing, 9-1, 9-2, 9-7, 9-8 importance, 9-1 liquid penetrant testing, 9-1, 9-2, 9-7 magnetic particle testing, 9-1, 9-2, 9-7 NDE for different weld types, 46-4 radiographic testing, 9-1, 9-2, 9-7 ultrasonic testing, 9-1 through 9-7 detecting microstructural changes, 9-6, 9-7, 15-13 measuring steamside oxide thickness, 9-4 through 9-6 measuring wall thickness, 9-4 measuring waterside deposits, 9-6 Union Electric technique for dissimilar metal welds, 35-12, 35-13 visual examination, 9-7 Instrumentation for cycle chemistry monitoring, 3-14 Intergranular stress corrosion cracking (see also Stress corrosion cracking), 37-1 through 37-3

Lack of fusion weld defect (see also Welding/repair defects), 45-1 through 45-3 Laning of gas passages, 14-8, 32-16, 33-15, 34-14 Larson-Miller Parameter (LMP), 10-2 through 10-5 Lay-up, 4-9 through 4-11, 27-7, 41-5, 41-6, 41-8 Lifetime, tubes (see also Remaining life of tubes and Boiler tubes, design considerations), 2-2 through 2-6, 4-5, 8-1 through 8-8, 18-12, 23-2, 23-3, 32-18, 32-19 Long-term overheating (creep), 32-1 through 32-32 actions, 32-24 through 32-32 case study, 32-22 determining the extent of damage, 32-16, 32-29 distinguishing from fireside corrosion, 6-8, 7-6, 7-8, 7-9, 32-2 through 32-6 distinguishing from graphitization, 7-9 distinguishing from short-term overheating, 32-5 features of failure, 7-6, 7-8, 32-2 through 32-6, 32-24 locations of failure, 32-6 through 32-7 long-term actions and the prevention of repeat failures, 32-18 through 32-21, 32-31, 32-32 mechanism, 32-8 through 32-10, 32-25, 32-26 precursors, 32-24 ramifications/ancillary problems, 32-32 repairs and immediate solutions/ actions, 32-17, 32-30 root causes and actions to confirm, 7-6, 7-8, 32-11 through 32-16, 32-27, 32-28 Low excess air for Nox control, 18-1, 18-7 Low melting point ashes (see Melting points of fireside ashes) Low-temperature corrosion, 30-1 Low-temperature creep cracking, 24-1 through 24-11 actions, 24-8 through 24-11 determining the extent of damage, 24-6, 24-10 features of failure, 24-1, 24-2, 24-5, 24-8 locations of failure, 24-3

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-6

long-term actions and the prevention of repeat failures, 24-7, 24-11 mechanism, 24-4, 24-5, 24-9 precursors, 24-8 repairs and immediate solutions/ actions, 24-6, 24-11 root causes and actions to confirm, 24-6, 24-10 Magnetite, strain tolerance, 1-18, 2-18, 2-20, 13-10 Maintenance, effects on boiler tube failures, 4-1 through 4-12 Maintenance damage, 44-1 through 44-6 actions, 44-3 through 44-6 as a possible cause of short-term overheating in waterwall tubing, 23-5 description of the mechanism and its manifestation, 44-1, 44-2 Maricite, 16-2 through 16-4, 16-7, 16-8 Material flaws, 45-1 through 45-6 actions, 45-4 through 45-6 description of the mechanism and its manifestation, 45-1 through 45-3 Melting points of fireside ashes coal-fired, 33-7, 33-8 oil-fired, 34-2, 34-3, 34-7 through 34-10 waste-to-energy units, 49-3 through 49-5 Membrane fins, failures associated with, 45-1 through 45-3 Metallurgical analysis, 6-1 through 6-10 fireside scale/ash analysis, 6-9 flowchart of steps for, 6-2 importance of, 6-1 metallographic samples, 6-6, 6-7 oxide scale thickness and morphology, 6-7, 6-8 required background information, 6-4 ring sampling for dimensional checks, 6-6 sample evaluation form, 6-5 sample removal and shipping, 6-4 waterside deposits/scale analysis, 6-8, 6-9 MgO - V2O5 phase diagram, 34-10

Microstructure assessing changes in austenitic stainless steels, 10-4 through 10-5 assessing changes in ferritic steels, 10-1 through 10-4 Microvoids, 10-6 through 10-8 Misaligned or distorted tubes, 14-3, 14-4, 14-11, 33-6, 33-15, 33-21, 34-5, 34-15, 34-19, 35-10, 39-5, 40-1, 40-3 Molten deposits, 2-22 through 2-24 Molten salt attack, 32-10, 33-7 Monitoring displacements and strains, 9-10 heat flux, 9-9 temperatures, 8-6, 9-8, 9-9 Multidisciplinary teams for BTF reduction, 5-2 Multilaminated oxides, 2-16, 2-17 Municipal solid waste (MSW) units, BTF issues in, 49-1 through 49-7 Nickel-based welds (in dissimilar metal welds), 11-7, 35-2, 35-3, 35-5 through 35-9, 35-15 Nitrogen blanketing (see Layup) Nucleate boiling, 2-12, 2-13 Oil composition and corrosiveness, 34-7, 34-8 effect of additives on corrosiveness, 34-9, 34-10, 34-15, 34-18, 34-19 Oil-fired boilers fireside corrosion in, 34-1 through 34-26 maintenance damage while washing, 44-1 Operation and maintenance, effects on boiler tube failures, 4-1 through 4-12 Orifice plugging, 23-5 Ovality of tubes, 24-4, 24-5 Over-fire air, 18-1, 18-7 Oxide notch, 35-3, 35-4, 35-6 Oxide thickness (see also Oxides, internal in steam-touched tubes), 2-14 through 2-21, 4-5, 4-6, 6-7, 6-8, 8-2 through 8-6, 9-4 through 9-6, 10-2, 32-9 Oxides internal in steam-touched tubes, development and breakdown, 2-14 through 2-21, 10-2, 32-2, 32-9 exfoliation, 2-17 through 2-21, 36-5 through 36-7 failure criterion, 2-18, 2-20

growth on austenitic materials, 2-17, 8-4, 8-5 growth on ferritic materials, 2-16, 2-17, 8-4, 8-5, 10-2 influence on tube metal temperatures, 4-6, 8-4, 8-5, 9-4, 9-5, 32-2 life assessment analysis of, 8-2 through 8-4 life improvement by chemical cleaning of, 4-5 measuring by ultrasonic testing, 9-4 through 9-6 metallurgical analysis of, 6-7, 6-8 spalling, 2-17 through 2-21, 36-5, 36-6 Oxides, internal in water-touched tubes, comparing most common forms, 2-7 formation, 2-6 through 2-12, 19-7 general nature of, 1-18 model explaining regular array of cracking, 13-10, 13-11 Pourbaix diagram, 13-11, 13-12 protective magnetite breakdown and resulting boiler tube failures, 1-18, 2-10, 2-11, 13-10 through 13-13 protective magnetite growth, 2-8 strain tolerance of magnetite, 2-11, 13-10 Oxygen (see also Dissolved oxygen) effect on corrosion fatigue, 13-16 through 13-20 Oxygen scavengers 3-8, 3-9, 3-10, 21-3 through 21-6 Oxygenated treatment (see also Feedwater treatment), 1-18, 3-9 through 3-11, 3-13, 19-5, 19-6, 19-11, 19-13, 19-15 effect on corrosion fatigue, 13-20 effect on growth and exfoliation, 2-21, 19-5, 19-6, 19-11, 19-13, 19-15 Pad-type thermocouples, 9-8, 9-9 Pad welding (see also Repair and replacement of boiler tubes), 11-5, 11-6, 13-27, 15-15, 16-13, 16-14, 17-11, 17-12, 22-4, 38-6, 46-2, 46-3 Personnel, importance of training, 5-2 pH depression, 13-16 through 13-20, 13-23, 13-24, 15-10 through 15-12, 15-14, 15-15 pH elevation, 17-5 Phosphate control, 3-3, 3-4, 16-6 through 16-8 Phosphate control diagrams, 3-3, 3-4, 16-7

Phosphate hideout, 3-4, 3-8, 16-6, 16-11, 16-12 Phosphate treatment (see also Boiler water treatment), 1-18, 3-3 through 3-5, 3-13, 16-6 through 16-8, 16-12, 16-14, 16-16, 16-17 effect on corrosion fatigue, 13-18, 13-20 Pitting in superheater/reheater tubes, 41-1 through 41-14 actions, 41-10 through 41-14 determining the extent of damage, 41-8, 41-12 features of failure, 41-2, 41-3, 41-10 locations of failure, 41-2 long-term actions and the prevention of repeat failures, 41-8, 41-13 mechanism, 41-4, 41-11 precursors, 41-10 ramifications/ancillary problems, 41-13 repairs and immediate solutions/ actions, 41-8, 41-12 root causes and actions to confirm, 41-6, 41-7, 41-12 Pitting in water-touched tubes (see also Chemical cleaning damage: waterwalls), 27-1 through 27-13 actions, 27-9 through 27-13 determining the extent of damage, 27-7, 27-12 features of failure, 27-2, 27-3, 27-9 initiation, 27-4 locations of failure, 27-2 long-term actions and the prevention of repeat failures, 27-7, 27-13 mechanism, 27-4, 27-5, 27-10 precursors, 27-9 ramifications/ancillary problems, 27-13 repairs and immediate solutions/ actions, 27-7, 27-12 root causes and actions to con firm, 27-6, 27-11 Plasma coating (see Coatings) PODIS (Prediction of Damage in Service) code, 35-15, 35-16 Polythionic acid, 37-5, 37-6 Post-exposure testing of tubes 10-6, 10-8 Pourbaix diagram, iron, high temperature, 13-11, 13-12 Pressure drop across circulation pumps (orifices plugging), 23-4, 23-5

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-7

Pressure drop losses in boiler, 19-5, 19-6 Protective oxide, 1-18, 2-6 through 2-21 Pyrites (effect on erosion), 2-23, 2-24, 14-5 through 14-7 Quartz (effect on erosion), 2-23, 2-24, 14-5 through 14-7, 47-6 Reducing fireside conditions, 18-1 through 18-5, 18-7 Reducing feedwater conditions, 21-3, 21-4 Refuse-derived fuel (RDF) units (see also Waste-to-energy units) 49-1 through 49-7 Remaining life computer codes, 8-3 through 8-6 NOTIS, 8-3 TUBECALC, 8-3 TUBELIFE, 8-3 through 8-6, 10-8 TUBEPRO, 8-3 Remaining life of tubes, accelerated creep rupture testing, 8-5, 8-6 assessment, 8-1 through 8-8, 32-18, 32-19 assessment methods for SH/RH tubes, 8-1 through 8-7, 32-18, 32-19, 33-17, 33-18, 34-16 assessment methods for waterwalls and economizer tubes, 8-7 assessment to optimize actions for fireside corrosion, 18-12 computer codes, 8-3 through 8-6 for graphitization in SH/RH tubes, 42-4, 42-5 improvement by chemical cleaning of SH/RH tubes, 4-5 roadmap for analysis of, 8-3 statistical analysis, 8-6, 8-7 Repair and replacement of boiler tubes (see also Welding/repair defects), 11-1 through 11-8 boiler tube buildup, 11-6 codes for weld repair, 11-3 dissimilar metal welds 11-7 general requirements, 11-4 pad welding, 11-5, 11-6, 13-27, 15-15, 16-13, 16-14, 17-11, 17-12, 22-4, 38-5, 46-2, 46-3 repair strategies, 11-1, 11-2 roadmap for weld repair, 11-2 tube section replacement, 11-4, 11-5 welding co-extruded tubes, 11-7 welding problems that can lead to boiler tube failures, 46-2

window welding (canoe piece repairs), 11-6, 11-7, 15-15, 16-13, 17-11 Residual oils, 34-7 high vanadium, 34-7 low vanadium, 34-8 Mexican, 34-8 Rifled tubes, 2-13, 15-15, 16-14, 17-12 “Ripple” magnetite, 2-10, 19-3 Root passes in welding repairs, 11-4, 11-5 Rubbing/fretting failures, 40-1 through 40-5 actions, 40-3 through 40-5 description of the mechanism and its manifestation, 40-1, 40-2 Rupture times, 23-2, 23-3 Rust on tubes following washing, 14-2, 22-1, 38-2 Sampling, 9-10 Secondary tube failures, identifying, 7-10, 7-11 Shielded metal arc welding (SMAW), 11-4, 11-6, 11-7 Shields for corrosion resistance, 33-18, 33-19, 34-16 for erosion resistance, 14-12, 22-4 Short-term overheating in superheater/reheater tubes, 36-1 through 36-16 actions, 36-12 through 36-16 case study, 36-10 determining the extent of damage, 36-9, 36-15 distinguishing from long-term overheating, 36-2, 36-3 features of failure, 36-2 through 36-4, 36-12 locations of failure, 36-3, 36-4 long-term actions and the prevention of repeat failures, 36-10, 36-16 mechanism, 36-4, 36-13 precursors, 36-12 ramifications/ancillary problems, 36-16 repairs and immediate solutions/ actions, 36-9, 36-15 root causes and actions to confirm, 36-5 through 36-8, 36-14 Short-term overheating in waterwall tubing, 23-1 through 23-14 actions, 23-9 through 23-14 determining the extent of damage, 23-7, 23-12

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-8

distinguishing among the three levels of, 7-5, 7-6, 23-2 through 23-4 features of failure, 23-2 through 23-4, 23-9 locations of failure, 23-4 long-term actions and the prevention of repeat failures, 23-8, 23-13 mechanism, 23-5, 23-10 precursors, 23-9 ramifications/ancillary problems, 23-14 repairs and immediate solutions/ actions, 23-7, 23-13 root causes and actions to confirm, 23-5, 23-6, 23-11 Shutdown of units, 4-8, 4-9, 27-6, 37-10, 41-4 through 41-6 Sigma phase of austenitic stainless steels, 10-4, 10-5, 32-3 Slagging, 2-22, 2-23, 19-6 through 19-8, 29-2, 36-12 Slagging propensity, 29-2 Solid particle erosion in the turbine, 32-32, 36-5, 36-12, 36-16 Solvent choice for chemical cleaning, 4-3, 4-4, 4-6 Sootblower erosion in superheater/ reheater tubes, 38-1 through 38-10 actions, 38-7 through 38-10 determining the extent of damage, 38-5, 38-9 distinguishing from flyash erosion, 7-9, 38-3 features of failure, 38-2, 38-3, 38-7 locations of failure, 38-3 long-term actions and the prevention of repeat failures, 38-5, 38-6, 38-10 mechanism, 38-4, 38-8 precursors, 38-7 repairs and immediate solutions/ actions, 38-5, 38-10 root causes and actions to confirm, 38-4, 38-9 Sootblower erosion in water-touched tubing, 22-1 through 22-9 actions, 22-6 through 22-9 determining the extent of damage, 22-3, 22-8 features of failure, 22-1, 22-6 locations of failure, 22-1 long-term actions and the prevention of repeat failures, 22-4, 22-9

mechanism, 22-2, 22-7 precursors, 22-6 repairs and immediate solutions/ actions, 22-4, 22-8 root causes and actions to confirm, 22-2, 22-3, 22-8 Sootblower operation and maintenance practices (see also Sootblower erosion in superheater/reheater tubes and Sootblower erosion in watertouched tubes), 22-2 Spacers, 26-3 Spalling of SH/RH steamside oxide, 2-17 through 2-21, 36-5, 36-6 Spray coatings (see Coatings) Spheroidization, 10-3, 32-3, 42-4 Stagnant water, 27-1, 27-4, 27-5, 41-2, 41-4, 41-5 Startup of units, 4-8, 4-9, 27-6 Steam blanketing, 2-12, 2-13, 15-5, 15-6, 16-6 through 16-8, 17-5 through 17-7 Steam flow redistribution, 32-19 through 32-21, 33-21, 34-19 Steam impingement, importance of identifying, 7-10, 7-11 Steam monitoring, 3-14, 41-8 Steamside oxide (see Oxides, internal in steam-touched tubes) Strain age embrittlement, 45-1 Strains, monitoring, 9-10 Stress analysis for analyzing corrosion fatigue, 13-29 Stress corrosion cracking, 37-1 through 37-16 actions, 37-12 through 37-16 case study, 37-10 determining the extent of damage, 37-8, 37-15 distinguishing from stress corrosion cracking and intergranular corrosion, 7-10 features of failure, 37-2, 37-3, 37-12 locations of failure, 37-3 long-term actions and the prevention of repeat failures, 37-9, 37-16 mechanism, 37-4, 37-5, 37-13 precursors, 37-12 ramifications/ancillary problems, 37-16 repairs and immediate solutions/ actions, 37-8, 37-15 root causes and actions to confirm, 37-6 through 37-8, 37-14

Substoichiometric fireside conditions, 18-1 through 18-5, 18-7 Sulfidation, 18-4, 33-7, 33-8 Supercritical steam properties, 19-6 Supercritical waterwall cracking, 19-1 through 19-22 actions, 19-19 through 19-22 case study, 19-16 chromizing waterwalls, 19-15, 19-16 determining the extent of damage, 19-14, 19-21 features of failure, 19-2, 19-3, 19-18 in oil-/gas-fired units, 19-10 international experience base, 19-5, 19-6, 19-16 locations of failure, 19-4 long-term actions and the prevention of repeat failures, 19-15, 19-16, 19-22 mechanism, 19-5 through 19-10, 19-19 precursors, 19-18 ramifications/ancillary problems, 19-22 repairs and immediate solutions/ actions, 19-14, 19-21 root causes and actions to confirm, 19-11 through 19-13, 19-20 Superheater/reheater chemical cleaning (see also Chemical cleaning), 4-5 through 4-7 solvent choice, 4-6 Superheater/reheater tubes, basics, 2-5, 2-6, 32-8 failure mechanisms screening table, 1-8, 1-9, 31-4, 31-5 maximum metal temperatures, 32-8, 32-9 temperature distribution in, 32-11, 32-14, 32-15 Supports, 26-3, 35-10, 35-11, 39-3, 39-4 Temperature measurements, in economizer inlet headers, 20-6, 20-7, 20-10 in SH/RH tubes, 32-11, 32-14 10 o’clock - 2 o’clock flats, 32-2, 32-10, 33-2, 33-3, Thermal-hydraulic regimes in boiler tubes, 2-12 through 2-14 conditions that lead to deposit formation, 2-13, 2-14 global, 2-12, 2-13 local, 2-13, 2-14

Thermal fatigue in economizer inlet header tubes, 20-1 through 20-19 actions, 20-14 through 20-19 assessment methodology, 20-9 case study, 20-12, 20-13 determining the extent of damage, 20-8, 20-17 distinguishing from erosioncorrosion and flexibilityinduced cracking, 7-6, 7-7, 20-4 features of failure, 20-2, 20-3, 20-14 locations of failure, 20-2 long-term actions and the prevention of repeat failures, 20-11, 20-19 mechanism, 20-4, 20-5, 20-15 precursors, 20-14 ramifications/ancillary problems, 20-19 repairs and immediate solutions/ actions, 20-9 through 20-11, 20-18 root causes and actions to confirm, 20-6, 20-7, 20-16 Thermocouples, 8-6, 9-8, 9-9 chordal thermocouples, 9-8, 9-9 pad-type thermocouples, 9-8, 9-9 Thermogravimetry analysis, 33-12, 33-15 Transgranular stress corrosion cracking (see also Stress corrosion cracking), 37-1 through 37-3 TUBELIFE, 8-3 through 8-6, 10-8 Tube blockage, 23-5, 36-5 Tube build-up, 11-6 Tube manufacturing laps, 45-1, 45-2, 45-3 Tube ovality, 24-4, 24-5 Tube temperatures increased by increasing oxide thickness, 4-5, 8-3, 8-4, 9-4 measuring via thermocouples, 8-6, 9-8, 9-9 predicted by oxide growth laws compared to thermocouple measurements, 8-4 through 8-6 Two phase flow, 2-12, 2-13 U-bends in tubes as fatigue site, 26-3, 26-4, 39-3 Ultrasonic measurement of oxide thickness, 4-6, 9-4 through 9-6, 32-11

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-9

Underdeposit corrosion, acid phosphate corrosion, 16-1 through 16-28 caustic gouging, 17-1 through 17-22 distinguishing among the types, 7-1 through 7-5 hydrogen damage, 15-1 through 15-30 in bubbling-bed FBC units, 47-3, through 47-5, 47-9 in circulating-bed FBC units, 48-2 Unit lay-up, as a cause of boiler tube failures, 4-9, 27-7, 41-1, 41-6, 41-8 options, 4-9 through 4-11, 27-7, 41-8 Unit startup and shutdown, effect on boiler tube failures, 4-8, 4-9 effect on pitting in water-touched tubes, 27-6 effect on pitting in SH/RH tubes, 41-6 effect on stress corrosion cracking in SH/RH tubes, 37-10

V2O5 - MgO phase diagram, 34-10 V2O5 - Na2O phase diagram, 34-3 Vanadates, 32-2, 32-3 Vibration in tubes as cause of fatigue, 26-6, 39-6, 39-11 Vortex shedding, 26-6, 39-6, 39-11 Waste-to-energy units, BTF issues in, 49-1 through 49-7 additives, 49-5 erosion, 49-2, 49-3, 49-7 fireside corrosion of SH/RH, 49-3 through 49-6 fireside corrosion of waterwalls, 49-3 through 49-6 high chlorides, 49-2 Water chemistry (see Boiler water treatment and/or Feedwater treatment) Waterside fireside corrosion (see Fireside corrosion in water-touched tubes) Water-steam cycle ingress, corrosion and deposition in drum units, 1-17 ingress, corrosion and deposition in once-through units, 1-17 introduction to 1-16

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-10

Waterwall deposits, effect on tube metal temperatures, 19-7, 19-8 local tube conditions that can cause, 2-13, 2-14, 15-4, 15-5, 15-6, 15-8, 15-10, 16-5, 17-4 measuring by ultrasonic testing, 9-6 metallurgical analysis of, 6-8, 15-2, 15-3, 16-2, 16-3, 16-18 through 16-20, 17-2, 17-3 rate of accumulation, 2-9 Waterwalls and economizer tubes, basics, 2-4, 2-5 failure mechanisms screening table, 1-6, 1-7, 12-4, 12-5 Weld build-up, 11-6 Welding/repair defects, 45-1, 46-1 through 46-7 actions, 46-5 through 46-7 description of the mechanism and its manifestation, 46-1 through 46-4 Welding repairs (see also Repair and replacement of boiler tubes), 11-1 through 11-8 Wick boiling, 2-13, 2-14 Window welds (canoe piece repairs), 11-6, 11-7, 15-15, 16-13, 17-11

Boiler Tube Failures: Theory and Practice Volume 2: Water-Touched Tubes

R. B. Dooley Electric Power Research Institute and W. P. McNaughton Cornice Engineering, Inc.

i

About EPRI Electricty is increasingly recognized as a key to societal progress throughout the world, driving economic prosperity and improving the quality of life. The Electric Power Research Institute delivers the science and technology to make the generation, delivery, and use of electricity affordable, efficient, and environmentally sound. Created by the nation’s electric utilities in 1973, EPRI is one of America’s oldest and largest research consortia, with some 700 members and an annual budget of about $500 million. Linked to a global network of technical specialists, EPRI scientists and engineers develop innovative solutions to the world’s toughest energy problems while expanding opportunities for a dynamic industry. EPRI—POWERING PROGRESS

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THIS BOOK WAS PREPARED BY THE ORGANIZATIONS NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, THE ORGANIZATIONS NAMED BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS BOOK, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY’S INTELLECTUAL PROPERTY, OR (III) THAT THIS BOOK IS SUITABLE TO ANY PARTICULAR USER’S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS BOOK OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS BOOK. ORGANIZATIONS THAT PREPARED THIS BOOK: ELECTRIC POWER RESEARCH INSTITUTE CORNICE ENGINEERING, INC.

This book is EPRI Licensed Material and contains a single-user, shrink-wrapped license.

ISBN 0-8033-5059-7

ORDERING INFORMATION Requests for copies of this book should be directed to the EPRI Distribution Center, 207 Coggins Drive, P.O. Box 23205, Pleasant Hill, CA 94523, (510) 934-4212. Electric Power Research Institute and EPRI are registered service marks of Electric Power Research Institute, Inc. Copyright © 1996 Electric Power Research Institute, Inc. All rights reserved.

ii

Table of Contents Volume 2: Water-Touched Tubes

Chapter

Page

12 12.1 12.2 12.3 12.4

Introduction and Use of Volume 2 Subject Matter and Objectives for This Volume Organization of Volume 2 Optimizing the Use of this Volume For BTF Mechanisms Not Covered by This Book

12-1 12-1 12-1 12-2 12-2

13

Corrosion Fatigue Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS

13-1 13-1 13-2 13-10 13-21 13-26 13-27

14

15

13-28 13-30 13-33 13-35

Flyash Erosion Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS

14-1 14-1 14-2 14-5 14-8 14-11 14-11

Hydrogen Damage Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS

15-1 15-1 15-2 15-5 15-8 15-13 15-14

14-12 14-19 14-22 14-23

15-14 15-16 15-20 15-21

iii

Table of Contents Volume 2: Water-Touched Tubes (continued)

Chapter

Page

16

16-1 16-1 16-2 16-6 16-9 16-13 16-13

17

18

iv

Acid Phosphate Corrosion Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS

16-14 16-16 16-21 16-22

Caustic Gouging Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS

17-1 17-1 17-2 17-5 17-8 17-11 17-11

Waterwall Fireside Corrosion Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS

18-1 18-1 18-2 18-4 18-7 18-11 18-11

17-12 17-14 17-15 17-16

18-12 18-16 18-17 18-18

Table of Contents Volume 2: Water-Touched Tubes (continued)

19

20

21

22

Supercritical Waterwall Cracking Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS Thermal Fatigue Economizer Inlet Header Tubes Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS Erosion/Corrosion in Economizer Inlet Headers Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS Sootblower Erosion (Water-Touched Tubes) Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

19-1 19-1 19-2 19-5 19-11 19-14 19-14 19-15 19-16 19-17 19-18 20-1 20-1 20-2 20-4 20-6 20-8 20-9 20-11 20-12 20-13 20-14 21-1 21-1 21-1 21-3 21-4 21-5 21-5 21-5 21-6 21-7 22-1 22-1 22-1 22-2 22-2 22-3 22-4 22-4 22-5 22-6

v

Table of Contents Volume 2: Water-Touched Tubes (continued)

23

24

25

26

vi

Short-Term Overheating in Waterwall Tubing Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS Low-Temperature Creep Cracking Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS Chemical Cleaning Damage: Waterwalls Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS Fatigue in Water-Touched Tubes Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

23-1 23-1 23-2 23-5 23-5 23-7 23-7 23-8 23-8 23-9 24-1 24-1 24-2 24-4 24-6 24-6 24-6 24-7 24-7 24-8 25-1 25-1 25-2 25-4 25-4 25-4 25-5 25-5 25-5 25-6 26-1 26-1 26-2 26-5 26-6 26-7 26-8 26-8 26-8 26-9

Table of Contents Volume 2: Water-Touched Tubes (continued)

27

Pitting in Water-Touched Tubes Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

27-7 27-8 27-9

28

Coal Particle Erosion Description of Coal Particle Erosion and Its Manifestation References ACTIONS

28-1 28-1 28-2 28-3

29

Falling Slag Damage Description of Falling Slag Damage and Its Manifestation References ACTIONS

29-1 29-1 29-2 29-3

30

Acid Dewpoint Corrosion (Economizer) Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

30-1 30-1 30-2 30-3 30-5 30-6 30-6

Index

27-1 27-1 27-1 27-4 27-6 27-7 27-7

30-6 30-7 30-8 I-1

vii

viii

A:

B:

C:

BTF Mechanism Unknown

BTF Known Mechanism (Table 12-2)

Anticipating Future BTF (Table 12-3)

Compare Macroscopic Appearance to Table 12-1 (Water-touched) or Table 31-1 Volume 3 (Steam-touched) Tubes to identify candidate(s)

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes. Follow actions

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes.

Action 1a: Perform Screening Analysis: Is it possible that this boiler tube failure is caused by this mechanism?

Action 1B: Screening Analysis: ¥ Review precursor list in mechanism chapter ¥ Remove tube sample to determine extent of damage

No

Yes Action 2: Determine (confirm) mechanism

Yes

Are BTF likely to occur in the future by this mechanism?

Chapter 12 • Volume 2

Introduction and Use of Volume 2

No Action 3: Determine root cause

Action 4: Determine extent of damage or affected areas

Action 5: Implement repairs, immediate solutions and actions

Action 6: Implement long-term solutions to prevent repeat failures Action 7: Determine possible ramifications/ancillary unit problems

12.1 Subject Matter and Objectives for This Volume The primary objective of this volume is to provide the most recent knowledge about boiler tube failures (BTF) in water-touched tubing of conventional fossil-fueled power plants. Constituent objectives are: • To provide sufficient theory and background information so that the reader can (i) identify boiler tube failure mechanisms, (ii) determine their root cause, and (iii) apply immediate solutions to correct the problem, and (iv) implement longer-term strategies to prevent their reoccurrence. • To provide direct, easy-to-follow actions to be taken in the event that a boiler tube failure or precursor has occurred.

12.2 Organization of Volume 2 Each chapter deals with a specific boiler tube failure mechanism. With only a few exceptions each chapter consists of two parts. The first half covers the Theory and Background about the mechanism; the second half addresses Actions to be taken.

12.2.1 Theory and Background - the first half of each chapter. The Theory and Background matter generally includes the following topics: • 1.0 Features of Failure and Typical Locations • 2.0 Mechanisms of Failure • 3.0 Possible Root Causes and Actions to Confirm • 4.0 Determining the Extent of Damage • 5.0 Background to Repairs, Immediate Solutions and Actions • 6.0 Background to Long-Term Actions and the Prevention of Repeat Failures • 7.0 Case Studies • 8.0 References A key part of each Theory and Background section is a Table that ties together the possible root causes, actions to confirm, immediate actions/solutions and long-term actions. It is crucial that the root cause of the damage be clearly identified so that the correct actions (short- and longterm) can be properly chosen. To fail in either identification or correction is to open the door to repeat failures.

Volume 2: Water-Touched Tubes

12-1

12.2.2 Actions - the second half of each chapter. The second half of each chapter contains Actions to be followed by the investigator or BTF team if (i) a boiler tube failure has occurred and a particular mechanism is suspected, or (ii) if a unit precursor has occurred that might lead to a future BTF by this mechanism. Note that throughout the three volumes, actions are generally marked with a special symbol, "➠". The Actions are numbered in a manner consistent with the Theory and Background section. That is, Action 2 corresponds to Section 2.0 of the Theory and Background section; the former details specific actions to be taken to confirm the mechanism, the latter provides additional information about the mechanism, why these specified actions are to be taken and how the mechanism develops.

12.3 Optimizing the Use of this Volume Figure 12-1 shows that three avenues are open to the investigator or BTF team depending upon the status of the BTF event:

¥ A: BTF with mechanism unknown. If a BTF in watertouched tubing has occurred and the mechanism is not known, then Table 12-1 should be consulted. The aim of this table is to provide a starting point for the investigation based on the macroscopic appearance of the failure and a description of typical locations. From it, a preliminary choice of mechanism can be made, then the Actions for that mechanism followed to confirm that the choice was correct. Note that as shown in Table 12-1, three BTF mechanisms (those caused by maintenance damage, materials flaws, and welding flaws), common to both water-touched and steam-touched tubing, are covered in Volume 3.

12-2

Introduction and Use of Volume 2

• B: BTF with known mechanism. If the BTF Team has knowledge from past failures that a particular mechanism is the likely cause, then Table 12-2, an index to Volumes 2 and 3, can be used to go directly to the appropriate chapter. • C: Anticipating future BTF. The BTF Team should continually anticipate possible failures by reviewing key unit/boiler operating events, that can lead to future BTF. Table 12-3 is a tool that can help to anticipate BTF. It is organized as a series of "unit precursors". These are events or conditions that experience has shown should be cause for detailed evaluation of the potential for future BTF, even though no BTF has yet occurred. The process is not unlike routine inspection of components; it may take only one identification of an incipient failure to justify the cost-effectiveness of the practice to even the most cost-conscious management. The table is organized in five sections: (1.0) inspection results or appearance, (2.0) cycle chemistry events, (3.0) maintenance-related, (4.0) operation-related, and (5.0) specific equipment events. The BTF Team or investigator may find that the best way to implement this table is to work through each precursor and ask: "Has this precursor occurred in our utility/unit?", or "Have we taken this action recently?" If the answer to either is "yes", then a review of the mechanism(s) indicated in the final column may be indicated. Note that this table includes both water-touched and steam-touched tubing. In compiling this table, an attempt has been made to limit the "precursor" list to those which (i) can be easily identified, (ii) are important observations and will be useful for indicating a potential BTF problem, (iii) are not direct indications of

boiler tube damage (an inspection that finds cracks at the toe of a tube/attachment weld would be a direct indicator of a BTF), and (iv) are reasonably likely to lead to a BTF based on past evidence. Clearly, it is not possible to put every possible precursor in Table 12-3, but it is hoped that two objectives are achieved. First, that forced outages by BTF are reduced by anticipating the pre-conditions to the most common mechanisms. Second, that a first step will be taken to improve the understanding of the complex, interconnected nature of cycle chemistry, operating practice, combustion processes, and maintenance effects on BTF. As a final note, the list should not pre-empt good engineering judgment. If a precursor is found that you think should be an alert of a future problem, follow it up, even if it is not in this particular list.

12.4 For BTF Mechanisms Not Covered in this Book If, having gone through the above procedure, it appears that the BTF experienced is not covered in this book, or if multiple mechanisms appear to be operative, then the generic investigation procedure shown in Figure 12-1 is still applicable. Specifically, it is important that the following sequence be followed: Understand the mechanism ¯ Determine the root cause ¯ Apply proper long-term solution Removal of a tube sample and use of metallurgical techniques should enable an understanding of the underlying damage process (erosion, corrosion, overheating, creep, fatigue, environmentally-assisted cracking, etc.) and may facilitate assignment of the BTF to one of the categories discussed here, which will then provide additional guidance to the investigator.

A:

B:

C:

BTF Mechanism Unknown

BTF Known Mechanism (Table 12-2)

Anticipating Future BTF (Table 12-3)

Compare Macroscopic Appearance to Table 12-1 (Water-touched) or Table 31-1 Volume 3 (Steam-touched) Tubes to identify candidate(s)

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes. Follow actions

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes.

Action 1a: Perform Screening Analysis: Is it possible that this boiler tube failure is caused by this mechanism?

Action 1b: Screening Analysis: ¥ Review precursor list in mechanism chapter ¥ Remove tube sample to determine extent of damage

No

Yes Action 2: Determine (confirm) mechanism

Yes

Are BTF likely to occur in the future by this mechanism?

No Action 3: Determine root cause

Action 4: Determine extent of damage or affected areas

Action 5: Implement repairs, immediate solutions and actions

Action 6: Implement long-term solutions to prevent repeat failures Action 7: Determine possible ramifications/ancillary unit problems

Figure 12-1. Flowchart of actions for identifying, evaluating, and anticipating boiler tube failures.

Volume 2: Water-Touched Tubes

12-3

Table 12-1 Screening Table for Water-Touched Boiler Tube Failures Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 2 (or 3 as noted)

Thick-Edged Fracture Surface Thick-edged (pinhole leak also possible)

Multiple, transgranular cracks that initiate on the inside of the tube.

Near attachments, particularly where high restraint stresses can develop.

Corrosion Fatigue

13

Thick-edged, leak or window blowout

Internal damage: gouging, wall thinning; tube deposits.

High heat flux areas; hot side of tube; horizontal or inclined tubing; pad welds; locations with local flow disruptions such as upstream of weld, backing ring or other discontinuities.

Hydrogen Damage

15

Thick-edged

Multiple, parallel cracks on the outside tube surface or on membrane; sharp, V-shaped oxide coated cracks; wall thinning from external surface when found with fireside corrosion.

Maximum heat flux locations; fireside or waterwall tubing or membranes between tubes.

Supercritical Waterwall Cracking

19

Thick-edged, leak or crack.

First sign as pin-hole leak at toe of stub weld; multiple, longitudinal cracks; bore hole cracking.

Economizer inlet header stub tubes nearest the feedwater inlet.

Thermal Fatigue

20

Thick-edged

Outside surface initiated, intergranular crack growth with evidence of grain boundary creep cavitation and creep voids.

Predominant in tube bends, particularly at intrados on outside surface, and other locations subject to high residual, forming, or service stresses.

LowTemperature Creep Cracking

24

Thick-edged

Transgranular cracking, OD-initiated and associated with tubing (at tube bends longitudinal or attachments - transverse) or headers (particularly at the ends).

Near attachments, particularly solid or jammed sliding attachments; at bends in tubing.

Fatigue

26

Thin-Edged Fracture Surface Thin-edged, longitudinal, "cod- or "fish-mouth"

Polishing of tube outside surface; very localized damage, wastage flats.

Near side and rear walls; near economizer banks; near plugged or fouled passages; where previous baffles have been installed.

Flyash Erosion

14

Thin-edged, leak or split

Internal damage: gouging, wall thinning; tube deposits.

As for hydrogen damage.

Acid Phosphate Corrosion

16

Thin-edged, leak or split

Internal damage: gouging, wall thinning; tube deposits.

As for hydrogen damage.

Caustic Gouging

17

Thin-edged, long "fish-mouth"

External wastage; probably affecting a number of tubes; maximum wastage at crown facing flame (maybe flame impingement); damage extending in 120° arc around tube; hard deposits on tube outside surface.

Areas with locally substoichiometric environment; side and rear walls near burners; highest heat flux areas.

Fireside Corrosion (coal-fired units)

18

12-4

Introduction and Use of Volume 2

Table 12-1 Screening Table for Water-Touched Boiler Tube Failures (continued) Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 2 (or 3 as noted)

Thin-Edged Fracture Surface (continued) Thin-edged rupture

Erosion, wall thinning from inside; "orange peel" appearance.

Economizer inlet header stub tubes nearest to point of feedwater inlet.

ErosionCorrosion

21

Thin-edged, "fishmouth"

Wastage flats on tube external surface at 45° around tube from sootblower direction, little or no ash.

Circular pattern around wall blowers.

Sootblower Erosion

22

Generally thinedged

Often shows signs of tube bulging or "fish-mouth": appearance; real keys will be transformation products in microstructure. May also be thick-edged under certain circumstances.

Highest heat flux locations above locations such as: the site of a tube or orifice blockage, or in horizontal tubing where a downcomer steam "slug" can occur.

Short-Term Overheating

23

Thin-edged

External wastage, little or no ash; location should be key.

Tubes near replaceable wear liners in cyclone burners; throat or quarl region of burners.

Coal Particle Erosion

28

Thin-edged

External erosion or mechanical impact damage features.

Sloping wall tubes and/or ash hopper near bottom.

Falling Slag Damage

29

Thin-edged

External, thinned or missing external oxide; generally in economizer.

Low temperature areas of economizer.

Acid Dewpoint Corrosion

30

Internal tube surface damage.

Locations where boiler water can stagnate during unit shutdown (pitting).

Chemical Cleaning Damage or Pitting

25 or 27

Maintenance Damage

Chap. 44, Volume 3

Materials Flaws

Chap. 45, Volume 3

Welding Flaws

Chap. 46, Volume 3

Pinhole Damage Pinhole

Various Other Damage Types Depends on underlying cause.

Usually obvious from type of damage and correspondence to past maintenance activity.

Depends on defect.

Usually thickedged.

Care required to separate weld defects from another problem located at a weld.

Note: This table is based on simple, macroscopic features of failure and should be used as a guide to a particular chapter for further analysis. The more detailed discussions starting with Actions can then be used for identification and confirmation of the actual mechanism.

Volume 2: Water-Touched Tubes

12-5

Table 12-2 Index to BTF Mechanisms Chapter In Volume 3

Water-Touched Tubes

Chapter In Volume 2

Acid dewpoint corrosion

30

Chemical cleaning damage in SH/RH tubes

43

Acid phosphate corrosion

16

Dissimilar metal weld failures

35

Caustic gouging

17

Fatigue in steam-touched tubes

39

Chemical cleaning damage

25

Fireside corrosion in SH/RH tubes (coal-fired units)

33

Coal particle erosion

28

Fireside corrosion in SH/RH tubes (oil-fired units)

34

Corrosion fatigue

13

Flyash erosion

14 (Volume 2)

Erosion-corrosion (economizer inlet headers)

21

Graphitization

42

Fatigue in water-cooled tubes

26

Long-term overheating/creep

32

Falling slag erosion

29

Low-temperature creep

Fireside corrosion (coal-fired units)

18

Maintenance damage

44

Flyash erosion

14

Material flaws

45

Hydrogen damage

15

Pitting in steam-touched tubes

41

Low-temperature creep

24

Short-term overheating

36

Steam-Touched Tubes

24 (Volume 2)

Maintenance damage

44 (Volume 3)

Sootblower erosion in SH/RH tubes

38

Material flaws

45 (Volume 3)

Stress corrosion cracking

37

Pitting in water-touched tubes

27

Rubbing tubes/fretting

40

Short-term overheating

23

Welding flaws

46

Sootblower erosion in water-cooled tubes

22

Supercritical waterwall cracking

19

Thermal fatigue of economizer inlet headers

20

Welding flaws

12-6

Introduction and Use of Volume 2

46 (Volume 3)

Table 12-3 Unit Precursors and Potential Future BTF 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

1.1 Watertouched tubes (waterside)

Excessive waterside deposits ( >> 30 mg/cm2) for high-pressure boilers.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), short-term overheating (23,V2)

Excessive waterside deposits, such as ripple Fe3O4 in once-through (O/T) and supercritical units.

Supercritical waterwall cracking (19,V2)

Boiler water samples that appear black (high suspended solids).

Acid phosphate corrosion (16,V2)

Corrosion/erosion in feedwater system; fouling in boiler feed pump or orifices.

• For supercritical or O/T units: supercritical waterwall cracking (19,V2) • For subcritical or non-O/T units - hydrogen damage (15,V2), acid phosphate corrosion (16,V2), or caustic gouging (17,V2) • Erosion-corrosion of economizer inlet header (21,V2)

Pressure drop across circulation pumps (orifices are plugging).

Short-term overheating in waterwall tubing (23,V2)

Flame impingement due to burner change or misalignment, leading to excessive tube deposits.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), fireside corrosion (18,V2)

Excessive furnace slagging that could lead to overheating in convective passes (or fuel change).

Short-term in overheating SH/RH tubing (36,V3)

Fresh rust found on tubes after unit washing, external flat spots, burnishing or polishing.

Flyash erosion (14,V2), sootblower erosion - waterwalls (22,V2), coal particle erosion (28,V2)

Failed tubes, any upstream tube leaks, as a warning to scout for the potential short-term overheating.

Short-term overheating in waterwall tubing (23,V2)

Significant hardness or ovality, particularly associated with tube bends, found during routine inspection.

Low-temperature creep cracking (24, V2)

Excessive steamside oxide (detected by UT measure of oxide thickness, or analysis of removed tube samples, evidence of excessive exfoliation like solid particle erosion in turbine).

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33&34,V3), dissimilar metal weld failures (35,V3), short-term overheating (36,V3)

Steamside deposits in RH tubing - particularly of sodium sulfate, or high Na or SO4 levels in steam.

Pitting and failure in steam-touched tubes (41,V3)

1.2 Watertouched tubes (fireside)

1.3 Steamtouched tubes (steamside)

Volume 2: Water-Touched Tubes

12-7

Table 12-3 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance (continued) 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

1.4 Steamtouched tubes (fireside)

Excessive flue gas temperature, displaced fireball, delayed combustion, periodic overfiring or uneven firing of burners.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33 & 34,V3)

High levels of excess oxygen.

SH/RH fireside corrosion: oil-fired units (34,V3)

Blockage or laning of boiler gas passages observed during boiler inspection.

Flyash erosion (14,V2), long-term overheating/creep (32,V3), SH/RH fireside corrosion: coal/oil units (33 & 34,V3)

Excessive temperatures measured by thermocouples in vestibule or header area.

Flyash erosion (14,V2), long-term overheating/creep (32,V3), dissimilar metal weld failures (35,V3)

Evidence of "alligator hide" appearance on external tube surface, observed during boiler inspection, associated with wall loss or thinning.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33 & 34,V3)

Fresh rust found on tubes after unit washing, external flat spots, burnishing or polishing.

Flyash erosion (14,V2), sootblower erosion in SH/RH (38,V3)

Significant hardness or ovality, particularly associated with tube bends, found during routine inspection.

Low-temperature creep cracking (24, V2)

Distortion or misaligned tube rows found during routine inspection.

Flyash erosion (14,V2), SH/RH fireside corrosion (33 & 34,V3), dissimilar metal weld failures (35,V3), fatigue of steam-touched tubing (39,V3), rubbing/fretting (40,V3),

Failed tube supports and lugs, location of dissimilar metal welds close to fixed supports.

Fatigue of steam-touched tubing (39,V3), dissimilar metal weld failures (35,V3)

12-8

Introduction and Use of Volume 2

Table 12-3 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

2.1 All units

Problem with high levels of feedwater corrosion products; operating ranges for pH, cation conductivity or dissolved oxygen consistently outside recommended ranges, including persistent reducing conditions or excessive use of oxygen scavengers.

Corrosion fatigue (13,V2), hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), waterwall fireside corrosion (18,V2), supercritical waterwall cracking (19,V2), erosion/corrosion in economizer inlet header (21,V2), short-term overheating in waterwall tubing (23,V2),

Carryover of volatile chemicals from boiler, such as NaOH for units on caustic treatment, or excess of Na, SO4, and/or chloride; steam limits exceeded.

Stress corrosion cracking (37,V3), pitting in steamtouched tubes (41,V3)

Major acid contamination event (pH < 8) when unit is at full load; condenser leak, or breakdown of makeup or condensate polisher regeneration chemical.

Hydrogen damage (15,V2)

Evidence of a persistent problem with phosphate hideout, particularly where mono-sodium and/or an excess of di-sodium phosphate has been added to the boiler.

Acid phosphate corrosion (16,V2)

Persistent phosphate hideout with phosphate return causing a pH depression (7-8).

Corrosion fatigue (13,V2)

Caustic level in excess of that necessary for optimal control (>> 2 ppm).

Caustic gouging (17,V2)

Caustic, used in excess of that necessary for optimal control of contaminant ingress (to counteract pH depressions on startup).

Caustic gouging (17,V2)

pH depression during shutdown and early startup (pH around 7-8). Hideout/return of sulfate.

Corrosion fatigue (13,V2)

Caustic, used in excess of that necessary for optimal control (>> 2 ppm).

Caustic gouging (17,V2)

2.2 Units on Phosphate Treatments

2.3 Units on AVT

2.4 Units on Caustic Treatment

Volume 2: Water-Touched Tubes

12-9

Table 12-3 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

3.1 Chemical cleaning

Evidence of shortcoming in chemical cleaning process such as inappropriate cleaning agent, excessively strong concentration or long cleaning time, too high a temperature, failure to neutralize, breakdown of inhibitor, inadequate rinse.

Chemical cleaning damage in waterwalls (25,V2) or SH/RH (43,V3), short-term overheating (23,V2 & 36,V3).

Shortcoming in SH/RH cleaning process such as inadequate rinse, improper flow verification.

Short-term overheating in SH/RH tubing (36,V3)

Evidence that level of Fe in cleaning solution continued to increase instead of leveling out when cleaning process was ended.

Chemical cleaning damage in waterwalls (25,V2) or SH/RH (43,V3)

Need for excessive cleaning in supercritical units (interval < 2 years).

Supercritical waterwall cracking (19,V2)

Contamination in SH/RH (particularly by chlorides) during chemical clean of SH/RH (breakdown of inhibitors or improper flushing of solvents) or waterwalls (caused by poor backfill procedures that failed to protect SH circuits).

Stress corrosion cracking (37,V3)

In water-touched tubes: use of backing rings, pad welds, canoe pieces, weld overlay that penetrates to inside surface - as a source of flow disruption and excessive deposits.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2)

Application of shielding, baffles, palliative coatings to mitigate flyash erosion without the use of a cold-air velocity test.

Flyash erosion (14,V2)

In water-touched tubes, Cu in water-side deposits.

Hydrogen damage (15,V2), welding defects (46,V3)

3.2 Repairs

12-10

Introduction and Use of Volume 2

Table 12-3 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

4.1 Startup Procedures

Feedwater introduced intermittently into economizer inlet at high flow rates during startups and particularly during off-line top-ups.

Economizer inlet header thermal fatigue (20,V2)

Rapid unit startups that cause the reheater to reach temperature before full flow starts (no furnace exit gas temperature control).

SH/RH fireside corrosion (33 & 34,V3)

Heat flux change caused by change to higher BTU-value coal, dual firing with gas, changeover to oil- or gas-firing leading to excessive tube deposits in waterwalls; new burners causing impingement.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), fireside corrosion (18,V2)

Implementing low excess air strategies for NOx control and the potential for waterwall fireside corrosion (note that unlike the other precursors in this Table, this is a possibility based on understanding the mechanism; to date no failures have been directly attributed to this cause).

Waterwall fireside corrosion (18,V2)

Operation with high levels of excess oxygen in oil-fired units (> 1%).

SH/RH fireside corrosion in oil-fired units (34,V3)

Change to a fuel that either contains more ash or contains elements which are more erosive such as quartz.

Flyash erosion (14,V2)

Change to a more corrosively-aggressive coal, particularly one high in chlorine, Na, K, or S contents.

Waterwall fireside corrosion (18,V2), acid dewpoint corrosion (30,V2), SH/RH fireside corrosion (33 & 34,V3)

Use of Mg-based additives (oil-fired units) leading to coating of waterwalls, reflecting heat into convection passes.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion in oil-fired units (34,V3)

4.4 Cycling

Conversion of the unit to cycling operation or an increase in the number of cycles.

Corrosion fatigue (13,V2), economizer inlet header thermal fatigue (20,V2), fatigue in water-touched (26,V2) or steamtouched tubing (39,V3),dissimilar metal weld failures (35,V3)

4.5 Shutdown or layup

Evidence of a shortcoming during unit shutdown/layup such as uncertainty about water and/or air quality during period, insufficient nitrogen blanketing, insufficient N2H4, evidence of air inleakage.

Pitting in water-touched (27,V2) or steam-touched tubes (41,V3), and maybe corrosion fatigue (13,V2)

Indication that stagnant, oxygenated water may have rested in tubes during shutdown or layup particularly in economizer and RH.

Pitting in water-touched (27,V2) or steam-touched tubes (41,V3)

Evidence that condensate is forming in SH/RH bends during unit shutdown, exacerbated if steam purity is not good (as determined by elevated levels of SO4).

Short-term overheating in SH/RH tubes (36,V3), pitting in steam-touched tubes (41,V3)

Operation above the maximum continuous design rating, with excess air flow settings above design, with unbalanced fans or air heaters leading to nonuniform gas flows.

Flyash erosion (14,V2)

Low drum level.

Short-term overheating (23,V2)

4.2 Combustion conditions

4.3 Fuel choices and changes

4.6 Other

Volume 2: Water-Touched Tubes

12-11

Table 12-3 Unit Precursors and Potential Future BTF (continued) 1.0 Inspection/Appearance 2.0 Cycle Chemistry 3.0 Maintenance Related 4.0 Operation Related 5.0 Specific Equipment Category

Precursor

Mechanism(s) of Concern (Chapter,Volume)

5.1 Condensers

Major condenser leaks or minor leaks that have occurred over a long period of time.

Hydrogen damage (15,V2)

Condenser leak leading to condenser cooling water constituents in attemperator spray water.

Stress corrosion cracking (37,V3)

5.2 Water treatment plant/ condensate polisher

Upset in water treatment plant or condensate polisher regeneration chemicals leading to low pH condition in boiler (pH < 8).

Hydrogen damage (15,V2)

Upset in water treatment plant or condensate polisher regeneration chemicals leading to high pH condition.

Caustic gouging (17,V2)

5.3 Drum

Carryover test indicates high mechanical carryover.

Stress corrosion cracking (37,V3), pitting in steamtouched tubing (41,V3)

Operating with high drum level allowing excessive carryover into steam.

Pitting in steam-touched tubing (41,V3)

5.4 Sootblowers

Poor sootblower maintenance.

Sootblower erosion in waterwalls (22,V2), SH/RH sootblower erosion (38,V3)

5.5 Low temperature headers

Header has large number of operating hours, has experienced large thermal gradients, spacing of ligament holes is small (< 3.5 cm), header thickness is well above Code minimum, header-to-stub tube joints made with partial penetration welds.

Economizer inlet header thermal fatigue (20,V2)

5.6 High temperature headers

Excessive relative movement of header/ tube during unit transients, restricted movement, header is not allowed to expand freely (maybe ash-related), unit change to cycling.

Fatigue in steam-touched tubing (39,V3).

5.7 Turbine

A problem with solid particle erosion (SPE) in the turbine.

Short-term overheating SH/RH tubing (36,V3), long-term overheating /creep (32,V3)

5.8 SH/RH Circuit (redesign)

Redesign of the SH/RH circuit may change the absorption patterns through other SH/RH sections and increase tube temperatures.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33 & 34,V3), dissimilar metal weld failures (35,V3)

5.9 Supports/ Attachments (redesign)

Addition of supports without consideration of their impact on the stresses of dissimilar metal welds.

Dissimilar metal weld failures (35,V3)

Redesign of waterwall tube attachments to increase flexibility without analysis to determine whether solution is actually beneficial.

Corrosion fatigue (13,V2)

12-12

Introduction and Use of Volume 2

Chapter 13 • Volume 2

Corrosion Fatigue

Introduction Corrosion fatigue occurs by the combined actions of cyclic loading and a corrosive environment. The primary occurrence is on the waterside in waterwall and economizer tubing, usually located adjacent to attachments or restraints.

Corrosion fatigue in boiler tubes has been a major source of availability loss in fossil-fueled power plants for over ten years. It is one of the last major boiler tube failure mechanisms to be characterized to the point that root cause analysis and solutions to prevent recurrence can be defined. Although not as common as in subcritical boilers, the same damage has also been found in supercritical boilers.

Volume 2: Water-Touched Tubes

13-1

1. Features of Failure and Typical Locations Corrosion Fatigue: Identification Keys 1. Failures are initiated at the inside surface, at multiple initiation sites, which can be associated with pits or other surface discontinuities. 2. Damage at the outside tube surface can appear as a pinhole, a thick-edge crack that is usually axial but may be circumferential, or a thick-edge section “blow-out”. Particular care must be taken to distinguish corrosion fatigue from outside surface-initiated mechanical fatigue. 3. Cracks are usually wide, oxide filled, with irregular profiles and evidence of discontinuous growth. 4. Failures are nearly always associated with tube attachments or other locations where significant constraint stresses develop. A list of 24 generic failure locations has been developed.

1.1 Features of failure Corrosion fatigue failures occur in water-touched tubes, usually in waterwall tubing, but also in economizer tubing under some conditions. Corrosion fatigue cracks initiate on the inside surface. Although they are predominantly located on the coldside of the inside surface such damage may also form on the fireside. General features of corrosion fatigue cracks are listed in Table 13-1. A typical failure is shown in Figure 13-1; key features illustrated here are the initiation at multiple sites on the inside of the tube, the longitudinal orientation of the cracks and the association of the failure with an external tube attachment. Figure 13-2 shows schematically the general features of this damage type; these features are shown in actual tubes in Figures 13-3 and 13-4. As shown, cracks initiate from multiple sites at the inside surface, are oriented axially, can grow through the wall, and although they may be associated with pits and other internal defects, they are not necessarily associated with weld defects. Sometimes damage is also seen on the outside surface. Corrosion fatigue cracks have been identified with three types of macroscopic appearance: (i) a pinhole leak, (ii) a thick-edged crack oriented axially (usually) or circumferentially, or (iii) a thick-edge blow-out or rupture which usually follows the membrane weld line either on the hot or cold side. The pinhole leak on the tube outside surface is the most commonly observed form. Figure 13-4 would be typical of the cross section through such a defect type. This damage manifestation can often be confused with an OD-initiating mechanical fatigue crack. There are three means to distinguish the two. First, the surface-initiated fatigue

13-2

Corrosion Fatigue

crack is generally associated with a discontinuity or stress riser, such as the toe of a weld, whereas the ID-initiated corrosion fatigue crack would only rarely break through to the outside surface exactly at the toe. An exception would be cracks that linked—one initiated on the inside surface and the other initiated on the outside surface. Second, when grinding the flaw prior to weld repair, the outside-surface initiated damage will decrease with grinding depth into the tube, the inside-surface initiated damage will become more widespread. Third, externally initiated fatigue cracks tend to show up earlier in the life of the boiler. The second damage manifestation, the thick-edged crack such as shown in Figure 13-5. This damage type is generally associated with attachments, but may be of considerable length and extend beyond the attachment area. The third macroscopic manifestation of damage is the thick-edge blowout or rupture and is characterized by cracking down both sides of the tube along the weld lines of the membrane; this causes an entire section of tube to fail (Figure 13-1). This third form is rare but has the potential to cause catastrophic damage and can be a safety problem if it occurs on the cold side of the tube in an area with heavy personnel traffic. Microscopically, corrosion fatigue cracks are characterized by features such as multiple, wide, transgranular cracks with irregular profiles, usually filled with oxide and showing signs of discontinuous growth such as crack arrest marks. Figure 13-6 shows cross sections through corrosion fatigue cracks illustrating these features.

Table 13-1 Common Features of Corrosion Fatigue Damage Macro-features • Initiation from inside (waterside) of tube. • “Typical” development on “cold” side of the tube but can be on fireside. • Cracks usually oriented longitudinally with respect to tube axis, i.e., normal to the predominant stress field, which in the typical case are tensile hoop stresses. • Cracks can also be circumferential or any direction that is normal to the major applied stress. • Can be initiated from pits or other surface discontinuities. • Not OD-initiated. • Not specifically related to presence of weld defects. Micro-features • Multiple, transgranular cracks. • Cracks usually wide. • Cracks usually oxide filled and blunt tipped. • Crack profiles usually irregular. • Signs of discontinuous growth, re-initiations.

Figure 13-1. An example of corrosion fatigue cracking in SA-210 A1 waterwall tubing. Note that the main failure is coincident with an attachment welded onto the tube and that there are multiple longitudinal cracks on the tube inside surface. Source: G.I. Ogundele, et al.1b

Volume 2: Water-Touched Tubes

13-3

Protective Fe 3O4 Secondary crack

Water/steam side

Deposited particles

Slip band activity Spherical corrosion pits

Extensive oxidation Deformed particles

Onset of final fracture

Figure 13-2. Schematic showing the general features of corrosion fatigue cracks. Source: M.D.C. Moles and H.J. Westwood2

Figure 13-4. Throughwall penetration of a corrosion fatigue crack. Note how the crack decreases in width towards the outside surface. Source: H.J. Westwood and W.K. Lee3

Figure 13-3. Typical multiple corrosion fatigue cracks in a boiler waterwall tube. Source: D. Sidey, et al.1d

13-4

Corrosion Fatigue

Figure 13-5. Thick-edged failure by corrosion fatigue. Source: D. Sidey, et al.1d

Figure 13-6. Cross-sections of corrosion fatigue cracks showing typical features: oxide coating of the fracture surface, corrosion within the crack, wide crack mouths and tips, and a transgranular fracture path. Source: S.R. Paterson, et al.10

Volume 2: Water-Touched Tubes

13-5

Steam drum 17

2 10 10

Buckstay attachments (a) Corner (b) Tie-bars (c) Stirrups

Furnace gas exit scallop plate

Penthouse floor scallop plate

18

Rear wall arch

1

Top windbox casing attachment Burner throat region

1

Windbox

Bottom windbox casing attachment

Burner elevations

9

22

Gas recirculation duct attachment

21

Front wall S-bends

Buckstay elevations

Windbox extension flat bar 19

Boiler water seal 3 + 4 5

Side wall buckstay connection to slope wall

6

15

Side wall/slope wall connection

Side wall gusset plate

24

(a) Division wall penetration of slope

Slope region

24

(b) Lower division wall tube ties

Note: Buckstay corner failures occur at Buckstay elevation ( ) other than in Slope region and Burner elevation

Figure 13-7a. Typical locations for tube failures by corrosion fatigue. Locations in tangentially-fired boilers. Numbers refer to additional description given Table 13-2. Source: D McNabb, et al.1a

1.2 Locations of failure A major effort has gone into identifying the most common initiation sites for corrosion fatigue. The predominant locations are near tube attachments: locations where large stresses develop during transient operating conditions as thermal expansion has been constrained by the attachment. Typical locations include windbox casing attachments, buckstay attachments, and scallop bar attachments. In economizer tubing, failures have been reported in bends or the heataffected zone of welds. A list of twenty-four susceptible locations was developed following the detailed survey of ten operating units with a history of corrosion fatigue failures.1a The units were all subcritical drum units but otherwise were chosen to reflect a variety of

13-6

Corrosion Fatigue

factors: size, mode of operation, fuel type, quality of cooling water and cycle chemistry. Table 13-2 presents those locations in order of the frequency with which corrosion fatigue was found for the units surveyed. These locations are illustrated schematically in Figures 13-7a and 13-7b for two generic boiler designs - a tangentially-fired radiant boiler and a front/rear fired radiant boiler, respectively. As a general rule, any failure associated with constraint (attachment) should be examined carefully for evidence of a corrosion fatigue mechanism. Table 13-2 summarizes the key aspects of corrosion fatigue at each location including: a description of the design, nature of the failures, and recommended modifications. Table 13-2 also presents a “stress

14

Scallop plate penthouse floor connection

13

20

12

Side wall buckstay connection to baffle wall

Upper gas offtake gusset plates

Gas offtakes

Gas outlet Upper windbox casing attachment

7 8

(a) Burner throat region (b) Burner mount

Windbox and burners

Lower windbox casing attachment 23

Furnace floor connection between front and rear walls

Side wall/ slope wall connection

Slope region

16

6

End of waterwall membrane region

Division wall penetration of slope wall

Figure 13-7b. Typical locations for tube failures by corrosion fatigue. Locations in front/rear-fired radiant boilers. Numbers refer to additional description given Table 13-2. Source: D. McNabb, et al.1a

rank” for the location, and possible modifications that can be used in corrective strategies; these two topics are discussed in considerably more detail in sections that follow. The problem can also arise in economizer tubing. The general industry experience has indicated the most likely failure locations are (i) at bends, (ii) in welds with the potential for high residual stresses such as fin welds, and (iii) at attachments similar in nature to those outlined above for waterwalls. Parts of economizer circuits can form walls of the back pass.

A common denominator among nearly all of the most common failure locations is the presence of significant thermal gradients, induced by either (i) high heat flux typical, for example, of the sites in the combustion zone of the furnace, or (ii) a variety of other causes such as where tubing carrying different media (e.g. steam and water) are connected. Failures have also been found at locations without significant thermal gradient. These are generally the result of: (i) poor boiler water chemistry conditions, (ii) an unusual structural loading, or both.

Volume 2: Water-Touched Tubes

13-7

Table 13-2 Corrosion Fatigue Failure Site List with Descriptions, Stress Ranking, and Potential Modifications Stress Rank for Use in Influence Diagram

Location

Description

1. Windbox casing

a) Continuous scallop plate–primarily corner tubes affected b) Filler bars

B

No modification derived

B

Replace cast filler bars with plate formed filler bars

a) Rigid corner scallop plate connected to buckstay b) Lug mounted tie-bar connected to tubes at corner c) Tangent/membrane wall with filler bar connections

B A

Remove or relieve rigid corner Same as for case (a)

D

Remove filler bar

2. Buckstay corners*

Applied Modification

3. Boiler ash hopper seal plate

Continuous scallop plate

B

Change to U-bolt arrangement

4. Boiler seal heat shield (slag screen)

a) Continuous scallop plate

B

b) 6-8 tube tangential bar

C

Short tangent bar (3-4 tubes), or a U-bolt arrangement Same as for case (a)

5. Side wall gusset plate

Triangular plate between redirected tubes

A

Change to peg membrane

6. Division wall penetration of slope

a) Refractory box rigidly connected at the top and bottom b) Continuous scallop plate

D

Remove rigid connections

B

Use refractory box without rigid connections

a) Short bars welded between redirected tangent tubes b) Short bars welded between tubes in tangent tube wall

C

Replace tube ties with membrane bar Weld bar on hot side to restore neutral bending axis to geometric axis of tube

8. Burner barrel mounts

Direct connection from burner barrel to waterwall

C

Use mounting plate between burner and wall Increase the number of attachment lugs

9. Windbox extension vertical seal

Windbox extension duct welded directly to vertical flat bar–flat bar is on outside of windbox, but could also be on inside

D

Install expansion plate between windbox casing and flat bar, remove flat bar on inside

10. Buckstay connections to waterwalls

a) Continuous scallop tie-bar

C

Use stirrups or lugs on membrane walls Tack weld to alternate tubes on tangent tube wall

b) Continuous tangent bar tack welded to tubes • membrane wall • tangent tube wall

D B

Same as for case (a) Same as for case (a)

7. Burner throat/gas off-take tube ties

13-8

Corrosion Fatigue

B

Table 13-2 Corrosion Fatigue Failure Site List with Descriptions, Stress Ranking, and Potential Modifications (continued) Stress Rank for Use in Influence Diagram

Location

Description

Applied Modification

11. Scallop tie-bars

Tangent tube waterwalls – most failures at corners or associated with abnormally high loads

D

Address source of stress Remove weld from every other tube

12. Miscellaneous waterwall penetration gusset plates*

a) Sootblower penetrations b) Burner throat and gas off-takes

D C

Replace with peg membrane

13. Miscellaneous filler bar attachments*

a) Windbox strut attachment b) Side wall buckstay/baffle wall connection c) Slope wall support I-beam at side wall

D D B

Replace solid filler bars with formed plate filler bars

14. Penthouse floor attachments

Continuous scallop plate a) problems most common in corners b) more serious if connecting tubes carrying different media

D B

15. Side wall/slope wall seal

a) Scallop bar b) Rod welded between tubes

D B

Replace with refractory box

16. End of membrane

More serious adjacent to redirected tube

A

Cut back membrane

17. Furnace gas exit scallop plate

Continuous scallop plate • adjacent to redirected tubes

C

Move scallop plate further from redirected tubes and cover with refractory

18. Rear waterwall arch

Continuous scallop bar • adjacent to separation of hanger tubes

D

Cut scallop bar at intervals to make discontinuous

19. Side wall buckstay connection to slope wall

a) Tangent bar tack welded to tubes

C

b) Scallop bar tack welded on alternate sides of bar

D

Replace with scallop bar Evaluate necessity of attachment Same as for case (a)

20. Side wall buckstay connection to baffle wall

Flat bar connection to baffle wall seal welded with filler bars at side wall • lowest connection affected

C

No modification derived

21. Lower front/rear waterwall S-bends

Immediately downstream of mud drums, with locating scallop bars between tubes

B

Remove scallop bars and replace affected bends

22. Gas recirculation duct scallop plate attachment

Continuous scallop bar

D

No modification derived

23. Furnace floor connection between nose tubes

Direct connection between nose tubes in opposite walls • filler bars used • natural gas-fired boiler only

C

Replace solid filler bars with formed plate filler bars No other modification derived

24. Division wall tube ties*

First set of tube ties above slope wall

D

No modification derived

No modification determined

*Listed stress rank applies to locations within the combustion or radiant sections of the boiler. Source: D. McNabb, et al.1a

Volume 2: Water-Touched Tubes

13-9

2. Mechanism of Failure Corrosion Fatigue: Mechanism 1. Corrosion fatigue is caused by the synergistic effects of stress and environment. This leads to a breakdown of the protective magnetite on a tube surface by both mechanical (stress) and chemical (environment) means. 2. Corrosion fatigue is a discontinuous process with cracks initiating and growing during transient periods such as starts and stops, and full load operation. 3. Transient operations result in cyclic strains driven by temperature differences between attachments and the tube. During peak strain range periods reinitiation or initial cracking of the protective oxide will occur. Full load operation can result in a corrosive environment which allows crack growth. 4. Both initiation and propagation of the corrosion fatigue are influenced by the interactions of: operating factors, chemical factors, and strain factors. A probabilistic approach, termed the Influence Diagram, has been developed to assess the impact of these three factors on the accumulation of corrosion fatigue damage.

2.1 Introduction Corrosion fatigue is one of a number of failure mechanisms that consist of synergistic effects of stress and environment. Among the boiler tube failure mechanisms, other combinations have variously been termed stress corrosion cracking and stress-assisted pitting. In many ways the distinction among various stress/environment-driven failure mechanisms is artificial, characterized by whether the stress or the environmental aspect seems to be predominant. The end result in any case is the accumulation of damage through the interaction of the two basic contributors. A significant effort has gone into characterizing corrosion fatigue damage, differentiating it from damage caused by other mechanisms, and to targeting solutions specific to corrosion fatigue. The balance of this section reviews the following aspects of corrosion fatigue in boiler tubes: (2.2) breakdown of magnetite, (2.3) overview of mechanistic models, (2.4) analysis of trends in the field experience, (2.5) stress effects on initiation and propagation including both field measurements and finite element analysis, and (2.6) environmental effects including laboratory and field results.

2.2 Breakdown of magnetite during corrosion fatigue The use of carbon steel for boiler tubes in the high temperature and high pressure boiler environment depends on the formation of the protective layer of magnetite (Fe3O4) on the waterside of the tube.2-4 Corrosion fatigue presents perhaps the clearest example of the problems that develop once that film is damaged.

13-10

Corrosion Fatigue

In general, the magnetite layer can be damaged either by chemical means (corrosion) or by mechanical means (strain), or by the synergistic effect of the two5. Destabilization primarily by chemical means usually occurs at pre-existing active sites, resembles pitting and has sometimes been termed stress-assisted pitting. When the film is destabilized primarily by strain, corrosion paths are produced, leading to an array of cracks and is generally termed corrosion fatigue in boiler tubes. Rupture of the protective oxide film leads to more rapid damage by corrosion fatigue because (i) additional base metal is exposed to corrosion damage and (ii) the rupture, which is a crack or crack-like, acts as a stress concentrator. The critical strain to fracture magnetite at high temperatures is generally reported to be between 0.01 and 0.1% strain.1b, 6-8 That there is a lower bound or critical level of strain that is required to fracture the protective oxide film and begin the corrosion fatigue damage process is supported by the German field experience and from modeling studies of oxide. In the former instance, the German design standard TRD 301 requires that the strain level in oxide be kept below a certain limit, about 0.1% strain, during operation to avoid rupturing the magnetite scale. The lower level of damage from corrosion fatigue in those boilers seems to confirm the effectiveness of this limit. Modeling and analysis of oxide provides another indication of the effect of strain level on the appearance of corrosion fatigue damage. Damage

is often manifested as a surprisingly regular array of cracks, see Figure 13-3, for example. Grosskreutz and McNeil9 proposed an explanation for this phenomenon while analyzing layers of Al2O3. They suggested that regularly spaced cracks would form in a layer under strain, that the separation between cracks would be a function of the strain level, and that the separation would decrease until some minimum was reached. The stress relaxation model that they proposed is illustrated in Figure 13-8. The cracking process will result in stress relaxation, with the highest relaxation located immediately adjacent to the crack formed. Therefore, the maximum remaining (unrelaxed) stress will be centered between cracks. With an increasing strain applied to the coating or oxide layer, the next crack will then form at the center between the existing cracks assuming a layer with uniform properties. This model was used to explain the observation of cracks in silica coatings for 9 Cr 1 Mo in experimental work by Crouch and Dooley4; their experimental results were used by Hay and Meadowcroft11 to improve upon the original model. The same model appears to apply to the initiation of corrosion fatigue cracks and explains nicely the regular array of parallel cracks on the inner surface. In addition to the geometry of cracking, the type of oxide that will form depends on the electrochemical potential of the material, which is in turn controlled by the oxygen content of the water, see Figure 13-9, and by the pH. A potential-pH map, also known as a Pourbaix diagram, is used to identify the stable oxide species for selected conditions.

d

b

w

Ds

Stress in film (s)

( d2 )

s¦

x

o

Figure 13-8. Schematic representation of the development of a regular array of evenly-spaced cracks caused by corrosion fatigue. (sf) is fracture stress; (Ds(x)) is a measure of stress relaxation. Source: A.G. Crouch and R.B. Dooley4

Potential (V SHE) 0.2 200°C 275°C 0

-0.2 250°C 150°C

-0.4

Potentials at 275°C

-0.6

-0.8

Oxygen (ppb)

Potential (VSHE)

5

-0.780

320

-0.190

1000

-0.090

-1.0 1

5

10 100 1000 Oxygen Concentration (ppb)

10,000

Figure 13-9. Electrochemical potential of carbon steel in water as a function of dissolved oxygen content. Source: P.M. Scott and W.H. Bamford12

Volume 2: Water-Touched Tubes

13-11

Figure 13-10 shows a Pourbaix diagram for iron in high temperature water. Plotted on this figure are results of laboratory tests using artificially high levels of dissolved oxygen from a program described below; corrosion fatigue most readily occurs outside the region of magnetite stability. Once the film is breached, damage accumulates until the surface is repassivated, thus repairing the film. Dissolved salt contaminants such as chlorides and sulfates are of concern as they can affect the morphology, formation rate, thickness, and strength of the surface oxide. As will be described in more detail below, the contribution to corrosion fatigue damage in laboratory tests on boiler tube materials was not as significant for these environmental variables as for either pH or dissolved oxygen levels.

2.3 Mechanistic models of corrosion fatigue A review was undertaken of the current thinking on alternative mechanistic models that attempt to predict from fundamentals the process that leads to corrosion fatigue damage.1b A brief description of four such mechanistic models is included here: film rupture/stabilization, mechanical/chemical dissolution, hydrogen embrittlement, and/or strain-induced corrosion cracking. • Film rupture/stabilization. There are several variations of this model which ascribes accelerated crack growth to the rupturing of protective films and subsequent re-oxidation or corrosion when the bare metal is exposed to the environment.5, 14-17 A variation of this model explains the onset of corrosion fatigue or stress corrosion cracking as being controlled by crack tip effects that can be explained by the superposition of an environmental effect and a strain effect.18 • Mechanically-assisted chemical dissolution. Figure 13-11 illustrates the basics of this model. Vacancies, caused by dissolution of the metal surface in a corrosive

13-12

Corrosion Fatigue

Potential (Volt vs. SHE) 2.5 2.0 1.5

No corrosion fatigue Corrosion Fatigue

9 Fe 3+ 21

1.0

4

Fe42-

Fe(OH) 2+

0.5 Fe2+

28

+

Fe(OH) 2

0

30

-0.5

26 23

17

Fe2O3

13 Fe3O4

-1.0

Fe

-1.5 -2.0 -2

27

29

HFeO224

0

2

8 10 12 6 4 pH (e) @ 250°C

14 16

Figure 13-10. Pourbaix diagram showing the stable oxide film as a function of electrochemical potential and pH. Stability of a protective Fe3O4 scale is believed to be related to the severity of corrosion fatigue. Source: C.M. Chen, et al.13

Material

Migra

Active crack tip strain assisted dissolution

tion o f towar vacancies d c ra c k tip

Corrosive solution

Vacancies

Dissolution of metal atoms creates vacancies Atoms Highly strained region Vacancies migrate toward strained region Coalescence causes crack to grow

Figure 13-11. Schematic of the mechanical/chemical dissolution model with the feature of the possibility of corrosion-generated surface vacancies migrating to the crack tip.

environment are driven by a stress field and accumulate at the crack tip; such coalescence results in incremental crack growth.19 • Hydrogen assisted (or embrittlement) cracking. Hydrogen is produced by reaction of carbon steel with water. Absorption of free hydrogen into the tube metal at the crack tip has been suggested by a number of researchers as being at the root of corrosion fatigue and stress corrosion cracking mechanisms.20-23 A schematic of the process is shown in Figure 13-12.

Localized corrosion reactions (e.g. 3Fe + 4H2O - > Fe3O4 + 4H2) Hydrogen generation and absorption into crack tip Material Oxide bare scale or surfac e

Oxide free crack tip

Corrosive solution

a ck e cr v i s Pas

wa l l

s

Highly strained region

• Strain-induced corrosion cracking. Similar to the film rupture model, this concept involves the local disruption of protective oxide.6, 22, 24 Destabilization of the oxide can occur by the environment (dissolved oxygen content, conductivity and temperature of the water), mechanical means (strain rate and strain level), or by material characteristics (such as sulfur content).24

2.4 General trends in corrosion fatigue failures: result of the analysis of field results.

Hydrogen migration into hightly strained region ahead of crack tip

Figure 13-12. Schematic of the hydrogen embrittlement model showing how hydrogen is generated at an active crack tip and then absorbed into the material.

Tube Failures 140 120

A number of factors have an influence on the incidence of corrosion fatigue in boiler tubes: boiler water and feedwater quality, chemical cleaning, tube replacement, boiler modifications, and operating mode. General observations from field studies include:

100

1. Unit operation can have a significant effect on the incidence of corrosion fatigue as shown in Figures 13-13 and 13-14. Peaking units have a large number of starts and relatively few operating hours; cycling units are those that tend to load follow and have relatively few starts and a larger number of operating hours.

20

Individual boilers Range Cycling Boilers

80 60 Peaking Boilers 40

0

0

200

400

600

800 1000 1200 1400 1600 1800 Total Starts

Figure 13-13. Total boiler starts versus tube failures by corrosion fatigue. Source: D. McNabb, et al.1a

Volume 2: Water-Touched Tubes

13-13

2. Chemical cleans by hydrochloric acid were shown to aggravate corrosion fatigue as illustrated in Figure 13-15. For example, boiler 1 shows a step function increase in the number of tube failures by corrosion fatigue directly following hydrochloric acid cleans. No such dramatic increase in corrosion fatigue failures was found after cleans done with EDTA or citric acid. Two competing effects are in progress: crack tip blunting caused by acid cleaning may slow propagation of modest sized cracks, but in the case of preexisting severe damage may lead to through-wall failure in a much shorter period. 3. Boilers that have had problems maintaining boiler water and feedwater limits generally experience more boiler tube failures by corrosion fatigue. Units that have experienced extensive corrosion fatigue have had indications that large swings in pH may have occurred in the boiler. Such indicators include: hydrogen damage, or caustic gouging, or, for those units using congruent phosphate boiler water treatment, evidence of phosphate hideout. Particularly important are pH depressions during shutdown and startup where pH drops to less than 8 such as caused (i) in phosphate treatment units when there is a phosphate hideout return, or (ii) in AVT units caused by CO2 ingress and slippage through the condensate polishers. It is interesting that corrosion fatigue occurs much less frequently in caustic-treated boilers; these units do not usually experience a pH depression during shutdown or early startup. Hideout and hideout return analysis of sulfate can be very informative for boilers on AVT. Up to 1500 ppb of sulfate has been observed during the overnight shutdown of boilers.25

13-14

Corrosion Fatigue

Tube Failures 140

Range Individual boilers

Peaking Boilers

120 100 80

Cycling Boilers 60 40 20 0 0

20

40

60 80 100 120 140 Operating Hours (thousands)

160

180

Figure 13-14. Operating hours versus tube failures by corrosion fatigue. Source: D. McNabb, et al.1a

Tube Failures 140 Boiler 8

120

HCL clean EDTA or citric acid clean

100 Boiler 1 80 60 40 Boiler 7 20 10

0

10

20

30 40 50 60 70 Operating Hours (thousands)

80

90

100

Figure 13-15. Operating hours versus corrosion fatigue tube failures illustrating the effects of different chemical cleaning solutions. Source: D. McNabb, et al.1a

2. A methodology is not yet fully established that can (i) predict stress levels at locations susceptible to corrosion fatigue, (ii) predict the rate at which damage will accumulate, or (iii) confirm, a priori the degree to which a proposed attachment modification will result in lower tube stresses. 3. However, guidelines for the application of finite element analysis and field confirmation have been developed and confirmed in field testing. These should guide utilities in the careful use of such techniques.

As noted above, the state of stress is a primary consideration in the analysis of corrosion fatigue. Predicting where corrosion fatigue will occur, predicting how quickly damage will accumulate, and predicting the effectiveness of proposed modifications to mitigate a high stress condition will all ultimately require a knowledge of how to measure or analyze the relevant state of stress. This section reviews recent work toward that goal. Sources of loads acting on tubes include (i) boiler pressure, (ii) thermal gradient through the tube (heat flux), (iii) constraint during thermal

The strains measured in units were lower than expected and lower than those which were thought, from lab-

The second largest level of strains developed as a result of subcooling during warm starts. Subcooling occurs in natural circulation boilers when there is a top to bottom temperature stratification during “bottled” cool down periods. The magnitude of the subcooling effect first increases, peaks at about 50 hours after shutdown for most natural circulation units, then starts to decline as temperatures over the height of the boiler start to equalize. The largest thermal transient in the subsequent start was observed in the lower furnace. It is believed that strains resulting from subcooling may not be an important contributor to corrosion fatigue in boiler tubes although a more important factor in thicker walled components. The strains associated with subcooling occur early in the start of a unit when boiler pressure and heat fluxes are low, and thus overall tube stresses are low.

Peak Strain (me) Cold start

Steady state

»

1. Local stresses and water chemistry are considered to be the two major factors that promote corrosion fatigue.

oratory investigations, to be needed to develop corrosion fatigue damage. Refinements to the measurement process have been made and subsequently applied to other units and have resulted in measured strain levels on the order of 0.2 to 0.3%, more consistent with that needed for the accumulation of corrosion fatigue damage.

Steady state

»

Corrosion Fatigue: Summary of Key Stress Factors

An idealized view of the levels of strain that would be expected during various load changes are shown in Figure 13-16. The results of strain gauge tests performed on three different types of units confirmed that cold starts resulted in the highest measured tube strains; warm starts, hot starts, and load changes, both sliding pressure and constant pressure, resulted in lesser levels. The typical trends for a cold start are shown in Figure 13-17. The peak strain occurs at the point when the boiler reaches full operating pressure. This peak strain corresponds to the maximum thermal gradient through the tube as shown by the difference between the tube temperature and the attachment temperature shown in the top part of Figure 13-17. Hoop strains then decline to steady state levels as the thermal gradient decreases.

»

2.5 Stress effects on initiation and propagation of corrosion fatigue cracks

expansion, and (iv) weight of the attachment (for some locations). Field tests have identified two major sources of cyclic strain: strains resulting from the pressure and temperature ramp during boiler starts, and strains from subcooling in natural circulation boilers.

Partial sliding pressure shutdown Zero pressure shutdown

Sliding pressure load change

»

4. Boiler layup conditions were also suspected to aggravate damage by corrosion fatigue, particularly if the pH was depressed and dissolved oxygen levels were not controlled.

Shutdown

Time

Figure 13-16. Schematic of an idealized strain cycle for a cold start, sliding pressure load change and shutdown. Source: D. Sidey, et al.1c

Volume 2: Water-Touched Tubes

13-15

Guidelines for utility application of stress analysis in attacking corrosion fatigue problems are presented in Section 6 below on long-term actions to deal with corrosion fatigue damage.

2.6 Environment effects on the initiation and propagation of corrosion fatigue cracks

500 Hoop strain (me) 400 300

Tube temperature (°C)

Attachment temperature (°C)

200 Axial strain (me)

100 0

Corrosion Fatigue: Summary of Key Environment Factors 1. There is a clear effect of environmental parameters on the severity of corrosion fatigue initiation and propagation. The effects are independent of the basic treatment chosen, for example AVT compared to congruent phosphate treatment. 2. Most obvious effects are for pH excursions and the presence of high levels of dissolved oxygen. 3. pH is particularly important as low pH excursions, such as associated with phosphate hideout, can occur concurrently with high strain in susceptible tubes.

There is no comprehensive model that can predict the effect of various environmental factors on the rate of initiation and propagation of corrosion fatigue cracks in the boiler environment. A number of past experimental programs have been concerned with isolating the effect of one or more major parameters such as dissolved oxygen, pH and cation conductivity. A broad outline of those studies and some of the implications are presented in the first part of this section (2.6.1) There have also been a few field studies that have tried to correlate the results of the laboratory investigations with field measurements of similar factors, and with the inci-

13-16

Corrosion Fatigue

-100 500

Unit load (MW)

400

Feedwater flow (kg/s)

Coal fires Roll turbine Unit synchronization

300 Lost coal fires

200 100

Purge

Coal fires Oil fires

Drum pressure (MPa X 10)

0 80

10.8

CBD O2 (ppb)

CBD phosphate (ppm x 10)

10.2

60 CBD - pH

9.8 pH

40 CBD cation conductivity(mS/cm)

20 0 22

0

2

4 Time (hours)

6

9.4

8

9.0 10

Figure 13-17. Strain, temperature and cycle chemistry information collected on the cold start of a 500 MW unit. (CBD) is continuous blowdown. Source: D. Sidey, et al.1d

dence or extent of corrosion fatigue damage. Unfortunately there has not been good correlation between what was thought to be important in the laboratory and what was measured in the field. In the broad outline in section 2.6.2 is a summary of what is presently known, what is thought to be correct and is undergoing confirmation, and what is presently unknown. The discussion here is limited to specific knowledge about boiler tube materials, the boiler environment and stress levels, and actual field investigations in working boilers.

A knowledge of cycle chemistry particularly during starts and transients is essential. For example, at one unit the major parameters (dissolved oxygen, pH and cation conductivity) were well maintained throughout transients. This was the result of specific operating procedures which allowed up to 1 ppm free hydroxide to counteract a phosphate hideout problem. Prior to making that change in procedure, pH would drop to around 7 during shutdown and remain around that level until restart.26 Similar hideout was observed during load changes as

well. This hideout phenomenon has been well documented27, 28 and has important implications for units on congruent phosphate control. See additional discussion on this topic in Chapter 3, Volume 1.

Cycles to Initiation 104

103

In the final part of this section (2.6.3) a brief note clarifies the use of oxygenated feedwater treatment, and how its use in drum boilers can be consistent with the observation that increased levels of dissolved oxygen have increased the propensity for corrosion fatigue damage in laboratory tests.

102

10

Air Data (274°C) Clean AVT water at 274°C Contaminated AVT water at 274°C

1 10-5

10-4

10-3 10-2 10-1 Frequency (Hz)

1

10

Figure 13-18. Effect of cycle chemistry and frequency on the initiation of corrosion fatigue cracking. Source: G.I. Ogundele, et al.1b

Dissolved Oxygen (DO) ppb 104

103

LOG (N) = 3.3883 - 0.5256 LOG (DO) R = -0.9842 Phosphate test data AVT test data

102

10

1 10

102 103 Cycles to Initiation, N

104

Figure 13-19. Effect of oxygen on corrosion fatigue initiation at 274°C (525°F) and a frequency of 0.0005 Hz. Source: G.I. Ogundele, et al.1b

2.6.1 Laboratory results. The following results have been obtained from laboratory studies1b; it is important to note that there have been some differences from those obtained in field tests: 1. The number of cycles to crack initiation by corrosion fatigue are lower in simulated boiler water, either high-quality or contaminated, than in air. Figure 13-18 shows this result. The number of cycles to initiation is markedly influenced by the presence of contaminants (lower curve). However, there is very little difference in these results between the normal base chemistries for AVT and phosphate. 2. Figure 13-18 also indicates that the frequency of strain reversal is important for initiation. Lower frequencies increase the corrosion fatigue initiation potential. Frequency effects are also evident for propagation rates. Higher frequency tests in laboratory experiments are not thought to allow the full effects of the corrosion aspect of the damage mechanism to influence the crack growth. Long periods at lower strains between cycles allow full influence of corrosion effects. 3. Dissolved oxygen was shown to have a strong influence on cycles to initiate corrosion fatigue as illustrated in Figure 13-19.

Volume 2: Water-Touched Tubes

13-17

4. pH level was also shown to have a strong effect. A series of specimens was tested in solutions with a constant levels of 5 ppb dissolved oxygen and with varying phosphates to simulate acidic phosphate hideout return. The number of cycles to initiation was reduced to about 1/3 at both 204°C (400°F) and 274°C (525°F) when the pH level was approximately 6 as compared to the base condition of pH 9. Figure 13-20 illustrates this important result. 5. The presence of chlorides and sulfate had less noticeable effects, with the effect of chlorides up to 3000 ppb the most significant in reducing cycles to initiation. 2.6.2 Implications of laboratory tests to operating boilers and comparison to field results. What are the implications of laboratory test results, particularly for dissolved oxygen levels and pH on the potential for corrosion fatigue in boilers? The effect of oxygen levels are reviewed first. It is clear that in the laboratory environment an increase in dissolved oxygen levels from 5 to 1000 ppb will significantly decrease the number of cycles to initiate corrosion fatigue cracks (Figure 13-19). There is also no doubt that during a shutdown period the oxygen level in boiler water can reach high, perhaps saturation, levels; but upon first firing and circulation of the boiler water the dissolved oxygen levels decrease once the deaerator and drum start to separate oxygen. As a result, the experience from field testing has indicated that oxygen at high levels (> 20 ppb) does not generally occur at the same time as the peak in applied strain level (Figure 3-17); peak strains occur well after the time when oxygen levels have approached a low level. It is important to know the time dependency of the elevated oxygen levels on startup and especially important to monitor the oxygen levels in the downcomer as compared to the boiler drum.

13-18

Corrosion Fatigue

Cycles to Initiate Cracks 1200 400°F (204°C)

1000

525°F (274°C) 800

600

400

200

0 4

8

6

10

pH @ 25°C

Figure 13-20. The influence of pH on cycles to initiate corrosion fatigue cracks in deaerated boiler water (< 5 ppb oxygen). The pH was controlled with phosphate solutions of different Na:PO4 molar ratios. Source: R.B. Dooley and L.D. Paul32

This situation is contrasted with that for pH depressions on shutdown/ startup. Such depressions can be caused by phosphate hideout return, leakage from condensers, or ingress of carbon dioxide. In these cases, particularly the first, the pH level does not return to the normal range until the unit pressure has risen considerably and/or phosphate or caustic has been added to the unit. Thus, the pH of the boiler water can be depressed during the period of peak strain at locations susceptible to corrosion fatigue. A pH level of 8 and below is not unusual in these cases as shown in Figure 167. As indicated in Figure 13-20, this can reduce the number of cycles to initiate or re-initiate corrosion fatigue cracks. Note that care is required to judge field test results. Table 13-3 shows chemistry, strains and temperatures measured for a variety of operating

conditions in three units that had historical problems with corrosion fatigue-induced tube failures.1c At the time of testing, as shown in the table, boiler water chemistry for units A and B was generally well within acceptable values when tube strains were highest during both cold starts and during warm starts. What then was at the root of the problem with corrosion fatigue? Subsequent investigation indicated that each unit had been subject to considerable past problems with phosphate hideout and concurrent possibility of low pH excursions when operating with congruent phosphate treatment. As a result, both units had changed to equilibrium phosphate treatment before the field tests and had not been experiencing hideout, hideout return, or the low pH excursions which result.

Table 13-3 Field Test Results Temperature Range (Tube OD) DT (˚C)

Temperature Rate of Change (Tube OD) (˚C/hr)

Hoop Strain DeH (µe)

Axial Strain DeA (µe)

Unit A

170

70

280

Unit B

145

120

Unit C1

260

Unit C2 (trip)

pH

Dissolved Oxygen (ppb)

Cation conductivity (µS/cm)

-700

8.6 - 9.5

> 1,000 to < 1

2 - 10

500

50

10.0 - 9.5

> 1,000 to < 5

20 - 40

250

150

—

8.5 - 7.2

> 1,000 to < 1

1 - 11

160

1,750

2,000

—

8.5 - 7.2

> 1,000 to < 1

1 - 11

Unit A1

60

50

200

-500

9.2 - 9.4

0.2%), there is often intergranular attack of the tube material. • Laboratory tests in support of the field experience initially used N2 - O2 - H2O - SO2 gas mixtures to simulate the gas composition near the walls. The corrosion rates were parabolic and the addition of 400 - 2,000 ppm HCl did not greatly influence the cor-

Volume 2: Water-Touched Tubes

18-5

• Scales were multi-layered, loosely adherent, large grained and consisted of a porous outer layer of FeS surmounting several inner bands of FeS and Fe3O4.10 Although the scale morphology and corrosion rates obtained in those laboratory tests were comparable to plant behavior, no chloride was detected at the scale-metal interface and the intergranular penetrations, often a feature of plant scale, were not evident. work11

where • Further laboratory the SO2 was replaced by H2S showed that the corrosion rates were linear whether HCl was present or not. • Because of the inability to reproduce plant behavior, particularly in high alloy materials, work continues on laboratory investigations of the effects of heat flux and of free radicals in combusting gases. • An optimal preventive strategy would be a priori to prevent the development of reducing conditions since they play such a dominant role in fireside corrosion, without or with a contribution from chlorine. However, as that is often not possible, change to more resistant materials in the bulk, or as coatings or co-extruded tubes as discussed below, is often used.

18-6

Waterwall Fireside Corrosion

a)

Metal Loss (mm)

200

100 0 HCl 400 ppm HCl 2000 ppm HCl 500

1000 Time (hours)

1500

b) 1000 Metal Loss (mm)

rosion rate. When the oxygen in the gas was replaced by CO, the corrosion rates were parabolic in the absence of HCl. When 400 ppm HCl was introduced, linear kinetics were observed (Figure 18-5). Once transition to linear kinetics had occurred, there was little subsequent dependence of corrosion rate when the HCl was increased to 2,000 ppm.

800

0 HCl 400 ppm HCl 2000 ppm HCl

600 400 200 0 500

1000 Time (hours)

1500

Figure 18-5. Metal loss data for mild steel in N2-10% CO-10%H2O-0.5% SO2 with HCl concentrations as shown (a) 400°C (750°F). (b) 500°C (930°F). Source: S. Brooks and D.B. Meadowcroft10

It should be repeated that this work and the understanding of the effects of chlorine on high temperature corrosion is only applicable to U.K. coals at the moment. A number of U.S. utilities have been burning coals with elevated chlorine levels (up to 0.4 wt. %) from the Illinois basin and have not, to date, experienced the extensive corrosion that has been observed in U.K. plants.

Further field testing and laboratory work is currently being initiated to determine if plant operating parameters specific to UK plant may have been responsible for the emphasis on the role of chlorine, and to better understand the chlorine effects in U.S. coals. There may be a difference in the way that chlorine is bound in the coal and released in the flame between U.S. and U.K. coals.

3. Possible Root Causes and Actions to Confirm Waterwall Fireside Corrosion (Coal-Fired Units): Root Causes The most common cause is a local “reducing” environment. Additional causes are (ii) the occurrence and deposit of pyrosulfates, (iii) fuel factors, primarily burning coals with unusually aggressive ash, and (iv) direct impingement of carbonaceous particles.

3.1 Introduction Table 18-1 lists the root causes, actions to confirm, and general strategies for dealing with waterwall fireside corrosion.

3.2 Influence of a substoichiometric environment The most common cause of fireside corrosion in waterwalls is the presence of a “reducing” (substoichiometric) environment as characterized by high CO or low O2 in the flue gases adjacent to the wall. This can also be inferred by monitoring at the economizer outlet. This condition can result from a number of root causes as noted below. To evaluate which fireside corrosion mechanism is active, the following action is recommended: (a). Energy dispersive x-ray and/or x-ray dot mapping of metallographic cross sections through damaged tubes can be used to detect the presence and distribution of S, C, Na, K, and Cl. Such an analysis can determine whether the corrosion process is following one of the mechanisms discussed above. 3.2.1 Poor combustion conditions will lead to a reducing condition particularly in localized regions near walls around burners. A typical location is in the highest heat flux zone from near the bottom burner level to about 10 feet or so above the top burners. It could also occur as a result of changing combustion conditions to burn a new fuel, for example. Actions to confirm include: (b). Monitor for levels of O2, H2S, and CO. High levels of CO (> 1%) and low levels of oxygen (< 0.1%) near tube walls are of particular concern.9 The level of CO can also be

measured in the flue gas at the economizer outlet or after the ID fans. Oxygen levels can also be sampled at the economizer exit. It is important to note that measurements at the economizer outlet only provide an indication of the overall combustion in the boiler and not of the local environment at the corroding waterwall locations. (c). Field testing may be required to determine the combustion conditions in regions that are experiencing severe fireside corrosion. Such tests might include the use of waterwall deposition probes to collect deposits or the use of devices to determine the heat fluxes in the corrosion areas. 3.2.2 Where air distribution has been modified such as for implementation of low excess air strategies, where burner modifications have been made to meet low NOx requirements, or where low NOx burners have been installed with over-fire air. Actions to confirm include: (d). As in (b) and (c) above. 3.2.3 Poorly adjusted or worn burners. Actions to confirm: (e). Visual examination to ensure that flame impingement is not occurring, which could lead to a local reducing condition as well as possible local overheating of the tube. (f). Monitor for slagging conditions, as a sudden increase in local slagging in the furnace chamber can indicate a faulty burner.9 Again, this could be identified by using a waterwall deposition probe.

Volume 2: Water-Touched Tubes

18-7

Table 18-1 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

(a). Metallographic analysis to detect key contaminants in deposits to confirm which fireside corrosion mechanism is active.

• Choose repair strategy based on severity of corrosion rate. • Implement long-term strategy from choices in Figure 18-6 in conjunction with on-going program of remaining life assessment and monitoring.

• Choose long-term strategy from those shown in Figure 18-6. • Implement on-going program of remaining life assessment and monitoring. See Chapter 8, Volume 1 for additional detail.

3.2.1 Poor combustion conditions (general)

(b). Monitor for levels of O2 (< 0.1%), H2S, and CO (> 1%). (c). Field testing to detect combustion conditions in susceptible areas with waterwall deposition probes to collect deposits or heat flux measurement in corrosion areas.

• As above.

• As above, plus • Develop a fireside testing program. Guidance is provided in reference 20.

3.2.2 Where air distribution has been modified

(d). As in (b) and (c).

• As above.

• As above.

3.2.3 Poorly adjusted or worn burners

(e). Visual examination to detect localized flame impingement. (f). Monitor for change in furnace slagging conditions. Use waterwall deposition probe, as needed.

• As above.

• As above, plus • Alternate burner designs, as needed.

Actions to Confirm

3.2 Influence of a substoichiometric environment

3.3 Influences of the deposition of salts

3.3.1 Overheated tubes

18-8

(g). Analytical techniques to identify melting • As above. points of compounds in deposits. (h). Metallographic analysis to detect key contaminants in deposits.

• Choose long-term strategy from those shown in Figure 18-6. • Implement on-going program of remaining life assessment and monitoring. See additional detail in Chapter 8, Volume 1.

(i). Measure pressure drop across waterwall • As above, plus circuits that would be indicative of • Chemically clean waterincreasing deposits on waterside. walls waterside surfaces. (j). Direct metal temperature measurements See Chapter 4, Volume 1 with thermocouples or heat-flux meters. for additional guidance (k).Selective sampling to determine internal on chemical cleaning. deposit amount and composition. (l).Analysis of internal oxide and deposits. (m).Analysis of cycle chemistry monitoring devices.

• As above, plus • Develop optimum feedwater chemistry strategy so as to minimize ingress of feedwater corrosion products. • Investigate the use of oxygenated treatment as a means to eliminate the internal boiler deposits.

Waterwall Fireside Corrosion

Table 18-1 Major Root Cause Influences, Confirmation and Corrective Actions (Continued) Major Root Cause Influences

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

• As above.

• As above.

• As above.

• As above, plus • Develop a fireside testing program. Guidance is provided in reference 20. • Investigate coal changes with Coal Quality Impact Model (CQIM) or equivalent, including economics evaluation.

• As above.

• Choose repair strategy based on severity of corrosion rate. • Implement long-term strategy from choices in Figure 18-6 in conjunction with on-going program of remaining life assessment and monitoring.

• As above, plus • Adjust mill classification.

• As above, plus • Develop a fireside testing program. Guidance is provided in reference 20.

3.4 Influences related to fuel factors

3.4.1 Change in fuel to coal with high corrosivity

(n). Analysis of propensity for coal corrosivity via available index methods.

3.5 Root causes of direct carbon deposition

3.5.1 Carbon particle impingement

(o). Visual and metallographic examination to detect key contaminants in deposits to confirm which fireside corrosion mechanism is active. (p). Periodic sampling from mills to ensure proper level of coal fineness.

3.3 Influences of the occurrence and deposition of salts This cause of waterwall fireside corrosion is less common than the problem caused by a reducing atmosphere, mostly because of the limitations on temperatures and SO3 found on waterwall tubes which normally puts them outside the susceptible temperature range. The melting points of typical salts can be found in the discussion of fireside corrosion in superheater/reheaters of coal-fired units (Chapter 33, Volume). Actions to confirm may include: (g). If liquid ash corrosion is suspected, thermogravimetry (ASTM E1131)12 or differential thermal analysis (ASTM E794)13 can identify melting points of compounds in the waterwall deposits or those collected on deposition probes.2

(h). Energy dispersive x-ray and/or x-ray dot mapping of metallographic cross sections through damaged tubes can detect the presence and distribution of S, C, Na, K, and Cl.

(j). Direct measurement of tube temperatures. Metal temperatures can be measured directly with thermocouples or by using heat-flux meters.

3.3.1 Overheated tubes. There are a number of underlying causes of overheated tubes, any one of which could create temperatures high enough to allow local attack by molten alkali salts. Such conditions might include excessive internal deposits such as rippled magnetite deposits in supercritical units, overheating from other sources of restricted water flow, direct flame impingement, excessive flue gas temperature or heat flux.

(k). Selective tube sampling to determine whether internal deposit buildup has been significant, and to determine the composition and morphology of deposits. Here formation of “ripple” magnetite is very important in once-through units in elevating tube temperatures. Deposits present in excess of 20 mg/cm2 (18.7 g/ft2) should be considered potentially harmful in elevating tube temperatures.

Actions to confirm may include:

(l). Analysis of internal oxide and deposits to estimate tube temperatures.

(i). Direct measurement of pressure drop across waterwall circuits to detect an increase in waterside deposits.

Volume 2: Water-Touched Tubes

18-9

(m). Analysis of cycle chemistry monitoring devices, particularly to detect high levels of feedwater corrosion products at the economizer inlet.

3.4 Influences related to fuel factors 3.4.1 Change of fuel to a coal with unusually corrosive ash. The composition of the coal can have a dominant effect on fireside corrosion. Although no quantitative index to coal corrosivity exists, corrosion potential increases with higher levels of Cl, Na, S, and K, and decreases with an increase in alkaline earth oxides (CaO and MgO). Also co-firing of municipal or industrial waste can accelerate fireside corrosion. Corrosive constituents of these

18-10

Waterwall Fireside Corrosion

wastes can include lead, zinc, sodium, potassium and chlorine. Actions to Confirm: (n). Evaluate changes in coal composition using available corrosivity indices as a judge of relative propensity for increased fireside corrosion.

3.5 Root causes of the direct deposition of carbon. The impingement of incompletely combusted coal particles can cause corrosive attack directly by creating a locally reducing environment. This could be caused by defective pulverization or classification.9

3.5.1 Carbon particle impingement Actions to confirm include: (o). Visual and metallographic examination. (p). Analysis of pulverized fuel (PF) exiting the pulverizers will provide an indication that the fineness is acceptable. The usual procedure involves sampling the PF and analyzing the amount of residue that remains after passing a sample through 60 (213mm) and 200 (74mm) mesh sieves. Such an analysis has the ability to indicate whether a large percentage of large particles (> 200mm) is the result of poor mill classification or inadequate grinding. The distribution riffles are also important in ensuring even flow to the burners and their adjustment should be checked.

4. Determining the Extent of Damage The primary means to establish the extent of damage will be to use ultrasonic techniques to detect wall thinning. Surveys should be conducted in conjunction with remaining life assessment techniques to determine the rate of wastage. The acceptable rate of wastage will depend on an analysis of the desired remaining life. However, as a rule of thumb, any wastage rate detected over 40 nm/hr (~ 14 mils/yr) will require further consideration. Chapter 9, Volume 1 reviews

the use of ultrasonic testing for wall thinning measurements. Locations should be chosen among those which are most susceptible as shown in Figure 18-4 and described above in Section 1.2. Emphasis should be on finding the locations of maximum damage, not an average rate of wastage. Surveys should be conducted prior to, and after, change of fuel supply or retrofitting low NOx burners.

5. Background to Repairs, Immediate Solutions and Actions Waterwall Fireside Corrosion (Coal Fired Units): Immediate Solutions and Actions Actions to prevent fireside corrosion will generally fall into two major categories: (i) materials strategies, i.e., providing protection or replacing the tubing, and (ii) design or operating strategies. For the most part both are longer term options.

5.1 Importance of linking repairs and actions to an assessment of remaining life. For many of the boiler tube failure mechanisms, the choice of corrective measures is limited to a single option, directly tied to the root cause. Fireside corrosion, in waterwalls and for superheaters/ reheaters, is one of the mechanisms where there are multiple options and a careful evaluation of the economics, as well as the engineering is critical. It is likely that the engineering and economic reality because of available fuel, required combustion conditions, etc. is that a certain level of corrosion wastage will occur. A remaining life analysis will help determine the implications of such constraints, including the development of a plan for periodic replacement of the most seriously affected tubes. Chapter 8, Volume 1 discusses remaining life assessment methods in more detail.

5.2 Repairs Repairs intended to be short-term can use either the same material in a direct replacement or can use a palliative coating. Pad welding of thinned tubes has been used as a temporary measure until the next scheduled outage. It is not recommended as a regular repair proce-

dure for waterwalls nor as a longterm “fix” of the problem because it will almost certainly result in repeat failures. Chapter 11, Volume 1 reviews boiler tube repair procedures including a discussion of the problems that can be introduced by pad welding. Over the longer term, the choice of repair strategy will depend on the severity of the corrosion problem. Rules of thumb are:

• If the corrosion rate is on the order of 40 nm/hr (~ 14 mils/year) the tubing can be replaced with the same alloy and will last approximately the same length of time. • If the corrosion rate is > 40 nm/hr (~ 14 mils/year) then the use of an alternate alloy or a faceted tube containing a thicker crown should be considered. • If the corrosion rate is >> 40 nm/hr, for example, as high as 300-500 nm/hr (~ 100 to 175 /mils/year), then the use of a coextruded tube should be considered. In such a composite, the inner material should be mild steel to provide waterside protection and pressure containment, with an outer coating of a resistant material such as Alloy 310. Additional discussion of material changes is given below.

Volume 2: Water-Touched Tubes

18-11

6. Background to Long-Term Actions and Prevention of Repeat Failures Waterwall Fireside Corrosion (Coal Fired Units): Long-Term Actions 1. Actions to reduce fireside corrosion will generally fall into two major categories: (i) materials strategies, i.e., providing protection or replacing the tubing, and (ii) design or operating strategies to try to control the local gaseous environment. 2. Specific actions should be chosen in conjunction with an analysis of remaining life and long-term plan for monitoring and periodic re-evaluation.

It may not be possible to remove completely the root cause for many fireside corrosion problems. Knowing how to minimize the wastage rate and the application of a remaining life assessment process, including periodic inspection and monitoring, will be the keys to economic handling of fireside corrosion problems. If an initial design condition is at the root cause of a fireside corrosion problem, there are three approaches that can be taken: correct the problem, accept the fault and the corresponding corrosion rates and expect an increased outage frequency, or accept the fault and seek a material solution. Figure 18-6 outlines most of the available strategies for waterwall fireside corrosion. As shown in that figure, there are two primary and not mutually exclusive routes that can be followed: (i) materials solutions and (ii) operating options. The circled numbers in Figure 18-6 are used to identify options for the discussion that follows and no ranking of the possible solutions is thus implied; however, boxes outlined in bold indicate those options which have been the most successful. Remaining life assessment (option 1, Figure 18-6) Utilities with fireside corrosion of waterwalls should initiate a systematic program of baseline measurement, monitoring rates of wastage, application and monitoring of control measures, and assessment of remaining life. Such an assessment will include establishing a baseline condition, predicting expected wastage rates, and periodic reinspection. It is required to relate the rate of corrosion wastage to the desired life of the unit, or the time available to implement a prevention strategy, for example, at the next

18-12

Waterwall Fireside Corrosion

scheduled outage. Therefore a remaining life assessment should be undertaken in parallel with any course of correction; periodic reassessment of condition will help avoid forced outages. Remaining life methods are discussed in Chapter 8, Volume 1. Wall thickness measurements should be made where a problem has been identified, usually with UT and preferably on grit-blasted surfaces. Monitoring of the flue gas, metal and steam temperatures, combustion conditions, and fuel composition should also be considered as these can determine corrosion rates while the unit is still on-line.9 Particularly important are step changes in key parameters. Coatings (option 2) A variety of processes have been suggested for applying metallized coatings to tubes in situ as a means of increasing corrosion resistance. The advantage of coatings is that very corrosion-resistant materials can be applied at specific susceptible sites, so it is not necessary to replace entire sections of tubing to overcome localized problems. Among the coating methods that have been tried are: surface nitriding, chromizing and aluminizing. Flame- or plasma spraying with and without subsequent heat treatment have also seen significant development work. The various coatings and application techniques have had variable and generally poor results in U.S. utilities. The primary problems have been getting the coatings to remain on the tubes, the reproducibility of coating techniques, and the need to have very well prepared surfaces. In general, coatings should be regarded as a quick fix which will require continued maintenance.

It is very important that a full assessment of the tube condition is made prior to coating. This should include knowledge of the minimum wall thickness and composition of any internal tube deposits. If the application of coatings results in a flow disruption at the internal tube surface, then it becomes a precursor to underdeposit corrosion mechanisms, such as hydrogen damage. Furthermore, if copper is present in internal tube deposits it can become

molten during the coating operation, leading to copper embrittlement of the grain boundaries. The former CEGB has tried a number of coatings for corrosion and erosion resistance.14, 15 As a result of many trials, alloys with compositions 50Cr-50Ni or 65Cr-35Ni have been found to have the best combination of corrosion resistance and resistance to spalling as caused by thermal cycling.15

Coatings were evaluated two years after their application in boiler trials. Corrosion was found to be less than the accuracy of the measurement technique and less than 1% of the coated surface area had spalled. Protection by coating was adopted for widespread use, implemented by mechanized spraying for most straight waterwall tubing and by a manual process for curved tubing or small isolated areas.15 Additional

Corrosion rate confirmed Extent determined

1

Remaining life assessment

Operating solutions

Materials solutions

Provide protection

2 Coatings

Replace component

8

Modify fuel

Modify combustion and/or flow conditions

Ð Change fuel Ð Blending Ð Washing More resistant material

6 Monolithic 7 Co-extruded

Same material

Gas side

9

Increased thickness 4 Same geometry 5 Different geometry Original thickness

3

11 Mills Burners 12 13

Water side Chemical clean deposits

Soot blowing 10

14 Check combustion

Air curtaining

Notes: a) Remaining life assessment (1) is almost mandatory to decide which option should be adopted b) Boxes outlined in bold indicate options that have been most successful c) Numbers refer to main text

Figure 18-6. Strategies for preventing repeat failures by waterwall fireside corrosion in coal-fired units.

Volume 2: Water-Touched Tubes

18-13

important features described included the following14:

• Importance of clearly specifying the areas to be coated, materials to be used and thickness required. Time available and other restrictions to site applications should also be clearly specified. • Importance of surface preparation to provide a contaminant-free, roughened surface. • Importance of taking samples throughout the application as a check that process parameters are producing consistent and acceptable quality. Trials of other coating types have included some refractory/ceramic coatings; in cyclone-fired units, refractories are used routinely in the furnace zone from around the location of the burners down to the slag hoppers Replacement with the same material of either the same thickness (option 3), increased thickness (option 4), or a changed geometry (option 5) If the corrosion rate is only slightly higher than that required to reach the desired life as determined from a remaining life assessment, tube replacement can be made in-kind.5 For a target lifetime of about 100,000 hours for waterwall tubes, the wastage rate should be much less than 40 nm/hr (~ 14 mils/year). If significantly higher corrosion rates than allowable are being experienced, then the options might be thicker tubes of the same alloy or a change of tube geometry. Asymmetrically-walled tubes, also called faceted, profiled, or omega tubes, have a thicker side oriented toward the flame where the most rapid wastage is experienced; they have been used in the petrochemical industry but generally not in the utility industry.4 The tubes are designed to fit into standard walltube spacing. The advantage is that more material is at the locations where corrosion is a problem. The

18-14

Waterwall Fireside Corrosion

disadvantages are such tubes are more expensive, more difficult to bend and install, and there is some reduction in heat transfer so that the external wall surface temperatures could be increased. Change to a more resistant ÒmonolithicÓ material (option 6) If it is necessary to go to a more resistant alloy, that is, where the corrosion rate is > 40 nm/hr (~ 14 mils/year), experience has shown that there is little difference in the resistance of ferritic materials with Cr levels up to 9%.4 Therefore, it has been recommended that a change in alloy be made to a material such as 9Cr-1Mo or higher. The principal disadvantage is cost, which can be several times the cost of the original material. Change to a more resistant composite material such as co-extruded tubing (option 7) It is possible to use tubes that are co-extruded so that an inner layer of load-bearing mild steel is metallurgically bonded with an outer layer of corrosion-resistant material, usually one that forms Cr2O3 in the furnace environment, as this restricts the formation of sulfides. The objective is to take advantage of the corrosionresistant outer material without the expense incurred by using it for the entire tube.

25%, and niobium to a minimum of 8X the carbon level. Note that sulfide scales have been found to form on co-extruded T310 which is a strong chromia former.7 The inner layer has typically been carbon steel. After 30,000 hours of operation in the first membrane wall trials carried out by the CEGB, a performance improvement of 2.5 over adjacent carbon steel tubes was measured. Tubes were 63.5 mm (2.5 in.) outside diameter with 3.4 mm (0.13 in.) Type 310 on 3.0 mm (0.12 in.) carbon steel. Tube bending characteristics of coextruded material have been found to be identical to monolithic materials.16 Weld procedures have generally matched weld metals to the base metal to maintain property levels. Conventional welding techniques and normal quality control have been found sufficient to ensure good weld quality. The CEGB experience base was more than 70,000 welds through 1984, without weld failure. No preheat or post-weld heat treatments have been required. Additional information about welding co-extruded tubes can be found in Chapter 11, Volume 1.

The former CEGB has used such tubes on a large-scale since 1976. By 1984 about 230 km of tubing had been installed in twelve CEGB units. Such a materials solution has been found to be the most economical approach to dealing with fireside corrosion problems which are caused by the aggressive nature of the coals burned in U.K. boilers.16

Fuel change, blending, washing (option 8) If the root cause of the problem is an aggressively corrosive coal, some change in fuel supply such as a change of source, cleaning or blending, can be beneficial. However, such procedures can often be economically prohibitive as transportation and/or handling costs increase. Further, there are usually overriding considerations not related to corrosion, such as to SO2 requirements.

The majority of furnace wall applications were in tangent wall construction through 1984. The primary material for the outer layer was Type 310 stainless (25 Cr 20 Ni) or a modified Type 310 which had increased silicon content to 0.75 1.5%, chromium to a minimum of

A method of investigating the potential effects of coal changes, blending, and washing is the application of the Coal Quality Impact Model (CQIM).17,18 Such a method will provide information on the total economic impact, as well as about the potential for fireside corrosion.

Change frequency, check effectiveness of sootblowing (option 9) Increased wall blowing can be tried. This is usually determined by slagging conditions and is not usually practiced for corrosion control. Combustion process changes to prevent local ÒreducingÓ conditions (option 10) Generally, changes in operating conditions are limited and consist of bringing the furnace back to design points. An exception would be where there has been a significant fuel change from the design coal. Nonetheless there are a number of operating checks that can be considered. One of the first courses of remedial action is to ensure present settings for burners and fuel nozzles match the design values unless there has been a change in coal type. Improvements to the stoichiometric balance of coal and air distribution to each burner can be tried, particularly to reflect changes in coal from design conditions. For example, in many front-wall fired boilers, secondary air registers can be used to control airflow and swirl. If adjustments are made, only one burner at a time should be adjusted with intermediate measurements of the CO levels in critical locations. A further approach to avoid substoichiometric regions might be to operate the unit at an increased excess air level. However, there have been no known cases where increases in air flow have been used as a corrosion prevention technique. Furthermore, this may come into

direct conflict with modifications specifically aimed at NOx reduction, and the increased mass flow might contribute to downstream erosion problems Checks can be made for oxygen at the economizer exit and carbon dioxide level after the ID fans to determine changes in the overall combustion process. However, lack of significant CO at the discharge to the ID fan does not preclude the existence of reducing conditions near furnace wall tubes. Mills (option 11) Equally loaded mills and equal pressure drop along the pulverized fuel lines to the burners will help ensure even fuel distribution. It is also possible that changes in coal fineness could be made to ensure complete combustion.19 This will help to avoid impingement of partially burnt coal particles. In practice, changes are generally not made except to restore original specifications. Pulverized fuel in each burner line should be sampled to ensure that the fineness specification is met (see Section 3.5.1). Inspection and adjustment of burners (option 12) As with general changes of the combustion process, there are only a few modifications of existing burners that can be made economically, and there are often overriding technical issues such as NOx control that will take precedence. However, some adjustments to burners might be indicated such as centering the fireball and ensuring that it is of proper size and

shape, adjustment or maintenance of the burners to account for changed air distribution and/or to avoid flame impingement, and checking and adjusting fuel nozzle tilt and aiming. Burners with separate air and coal ports may have separate adjustment problems. Air blanketing or air curtaining (option 13) One approach is to try to prevent substoichiometric zones by redistribution of air in the flame or arranging for some portion of the air to blanket the waterwalls in affected locations. A number of utilities have tried air blanketing or curtaining, which is included as standard practice for one boiler manufacturer, A difficulty is to verify that the desired level of air is present near the walls. One of the findings has been that a monitoring or diagnostic device which could provide an instantaneous reading of the local oxygen partial pressure would be useful.4 Chemical cleaning to remove waterside deposits (option 14) If the excessive temperatures in waterwall tubes are caused by excessive internal deposits, chemical cleaning should be conducted followed by a monitoring campaign to optimize the feedwater chemistry. Additional detail about the cycle chemistry targets can be found in Chapter 3, Volume 1, and about chemical cleaning in Chapter 4, Volume 1. This aspect is also discussed in Chapter 19 on supercritical waterwall cracking.

Volume 2: Water-Touched Tubes

18-15

7. Case study Waterwall Fireside Corrosion Case Study: Field Experience The most recent survey of experience with fireside corrosion and flyash erosion in the United States was in 1987.4 The survey included 21 utility telephone contacts and 10 plant visits. In total, information was obtained from 42 coal-fired stations. Boilers ranged in age from 2 to 34 years, and in size from 75 to 1300 MW. Five boiler manufacturers were represented in the data base and a variety of coal types were fired. Eleven of the 42 stations reported fireside corrosion of waterwalls as a serious problem and an additional 13 as a moderate problem. About half of the units reporting serious or moderate problems with fireside corrosion of waterwalls were cyclone-fired units. Table 18-2 summarizes the magnitude of the fireside corrosion problem at four of the ten utilities visited. As an example, at Utility T, significant problems with fireside corrosion were encountered. About 60 percent of the waterwall tube failures in one boiler were caused by this mechanism. The primary cause was a reducing atmosphere brought about by maladjusted air while maintaining load. At $25 per MWhr, forced outages caused by waterwall fireside corrosion in this unit still resulted in replacement power costs estimated to be about $1.7 million per year.

Units exhibiting serious problems were generally fired with Eastern coals with relatively high sulfur content. Problems appeared related not only to the composition of the coal, but also to operating conditions. The most common cause was presence of reducing conditions near the waterwalls or where there were short periods of overheating. Slagging problems were not well correlated with the incidence of fireside corrosion.The most common remedial actions reported were readjustment of combustion parameters, including coal fineness, burner positioning and nozzle adjustments, to provide uniform distribution of coal and air to each burner. The use of air blanketing had been tried in a few units with mixed opinion about the effectiveness. Of the material substitution alternatives, sprayed coatings had been tried at a number of utilities, thicker walled tubes at a few, and co-extruded tubing had generally been considered too expensive except for the lower furnace regions of cyclone boilers. Part of the problem in assessing the effectiveness of the various solutions is the interrelated nature of corrosion and other fireside problems. One factor cited was that the effect of recent coal source and firing procedural changes to meet environmental constraints

was still unknown. The survey concluded that: “Overall, as judged by the response of the utility personnel contacted to questions concerning reducing atmosphere corrosion, fireside corrosion of the waterwalls appears to be a problem of increasing importance.” Another example was utility P in the survey (see Table 18-2). Fireside corrosion was a continuing problem at one station that consisted of four 800 MW supercritical boilers constructed in the early 1970s. The units operated base load and burned Eastern coal that was not washed. All four units have been affected and replacement of as much as 2000 sq. ft. of waterwall per unit had occurred. About two forced outages per year per unit had occurred as a result of fireside corrosion of waterwall tubes in the early years of operation. Corrosion damage attributed to reducing atmosphere conditions was observed in moderately large areas of the waterwalls located at, and somewhat above, the burner elevation. Adjustment of burners, air supply, etc. was tried but had not been successful; installation of a curtain air system was at that time being considered. Best control of the problem had been obtained by metallizing the affected areas on a repetitive basis at most planned outages.

Table 18-2 Magnitude of Fireside Corrosion Problems (Annual) for Four Survey Utilities Utility

Days Lost to Forced Outages

Forced Outage Power Costs, MWhr

Number of Tube Leaks

T

4.9

68,800

—

Q

3.1

44,800

6.2

P

2.0

52,600

—

M

—

—

5 to 10

Source: D.N. Williams, et al.4 18-16

Waterwall Fireside Corrosion

8. References 1Culter,

A.J. B., T. Flatley and K.A. Hay, CEGB Research, October, 1978. 2Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 3Personal

Communication from D. French (David N. French, Inc.) to R.B. Dooley, February, 1995. 4Williams,

D.N., H.R. Hazard, H.H. Krause, L.J. Flanigan, R.E. Barrett, and I.G. Wright, Fireside Corrosion and Fly Ash Erosion in Boilers, Research Project 2711-1, Final Report CS-5071, Electric Power Research Institute, Palo Alto, CA, February, 1987. 5Holmes,

D.R. and D.B. Meadowcroft, “Fireside Corrosion and Problems of Tube Life Prediction”, Symposium on Thermal Utilities Boiler Reliability, McMaster University, Hamilton, Ontario, May, 1983. 6Personal

Communication from P. Daniel (Babcock & Wilcox) to R.B. Dooley, February 15, 1995. 7Personal

Communication from P. James and J. Davison (PowerGen, U.K.) to R.B. Dooley, February, 1995.

12American

Society for Testing and Materials, Standard E1131-86, “Standard Test Method for Compositional Analysis by Thermogravimetry, 1992 Annual Book of ASTM Standards: General Methods and Instrumentation, Volume 14.02, American Society for Testing and Materials, Philadelphia, PA, 1992. 13American

Society for Testing and Materials, Standard E794-85 (1989), “Standard Test Method for Melting Temperatures and Crystallization Temperatures by Thermal Analysis, 1992 Annual Book of ASTM Standards: General Methods and Instrumentation, Volume 14.02, American Society for Testing and Materials, Philadelphia, PA, 1992. 14Morgan-Warren,

E.J., “Thermal Spraying for Boiler Tube Protection”, Welding and Metal Fabrication, Jan/Feb, 1992, pp. 25-31. 15Bennett,

A.P. and M.B.C. Quigley, “The Spraying of Boiler Tubing in Power Plants”, Welding and Metal Fabrication, November, 1990, pp. 485-489. 16Flatley,

T. and T. Thursfield, “Review of Corrosion Resistant Co-Extruded Tube Development for Power Boilers”, in R.D. Sisson, Jr., ed., Coatings and Bimetallics for Aggressive Environments, American Society for Metals, Metals Park, OH, 1985.

8Latham,

E., D.B. Meadowcroft, and L. Pinder, “The Effects of Coal Chlorine on Fireside Corrosion”, Chlorine in Coal, J. Stringer and D.D. Banerjee, eds., Elsevier Science Publishers, Amsterdam, 1991, pp. 225-246.

17Davidson, P.G, et al., Development and Application of the Coal Quality Impact Model: CQIMTM, Research Project 2256-2, Final Report GS-6393, Electric Power Research Institute, Palo Alto, CA, April, 1990.

9Laxton,

18Pavlish,

J.W., D.B. Meadowcroft, F. Clarke, T. Flatley, C.W. King, and C.W. Morris, The Control of Fireside Corrosion in Power Station Boilers, Third edition, Central Electricity Generating Board, 1987. 10Brooks, S. and D.B. Meadowcroft, in Corrosion Resistant Materials for Coal Conversion Systems, D.B. Meadowcroft and M.I. Manning, eds., Applied Science, London, 1983. 11Personal

Communication S. Brooks and K.S. Gilroy to E. Latham, et al., cited in Reference 8, 1984.

J.H., P.R. Miller, N.C. Craig, and A.K. Mehta, “CQIM - An Analytical Tool Used to Evaluate Performance and Economic Issues”, Proceedings: Ninth Annual International Pittsburgh Coal Conference, October, 1992. 19Dooley, R.B. and H.J. Westwood, Analysis and Prevention of Boiler Tube Failures, Report 83/237G-31, Canadian Electrical Association, Montreal, Quebec, November, 1983. 20Sotter,

J.G., et al., Guidelines for Fireside Testing in Coal-Fired Power Plants, Research Project 1891-3, Final Report CS-5552, Electric Power Research Institute, Palo Alto, CA, March, 1988.

Volume 2: Water-Touched Tubes

18-17

ACTIONS for Waterwall Fireside Corrosion Two paths for the BTF team to take in the investigation of waterwall fireside corrosion damage begin here. The goal of these actions is to see if further investigation of fireside corrosion is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a waterwall BTF has occurred and fireside corrosion is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by waterwall fireside corrosion.

Action 1a: If a waterwall BTF has occurred and fireside corrosion is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of waterwall fireside corrosion: ➠ Review Figure 18-4 for typical boiler regions. ➠ Review main text, section 1.2 for description of susceptible locations such as: at a local “reducing” (substoichiometric) environment, usually around burners.

➠ Confirm that the macroscopic appearance of the failure includes such features as:

• Significant wall thinning across a number of tubes on the fireside. • Maximum attack at the crown of the tube facing the flame and extending around 120° of the circumference. • Grooving at tube surfaces with grooving consistent with “alligator hide” appearance. See Figure 18-3. • Hard, fired deposits at tube surface (inner layer) with loosely bonded ash on the outer layers.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

18-18

Waterwall Fireside Corrosion

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by fireside corrosion.

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit: • Any evidence of wall loss observed or measured. • Evidence of a locally substoichiometric environment. • Change to a more aggressive coal. • Evidence of carbon particle impingement or flame impingement. • Evidence that tubes may be overheating.

➠ These precursors can signal the potential for waterwall tube failures by fireside corrosion. If one or more has occurred, go to Action 3 which reviews root causes and outlines the steps to confirm the influence of each.

Action 2: Determine (confirm) that the mechanism is fireside corrosion. A waterwall tube failure has occurred which the BTF team has tentatively identified as being fireside corrosion damage (Action 1a). Action 2 should clearly identify fireside corrosion as the primary mechanism or point to another cause. The actions listed will be executed by confirming the macroscopic appearance of the failure, removing representative tube sample(s) followed by detailed visual and metallographic analysis. A primary objective is to identify constituents of the external tube deposits which are distinctive of fireside corrosion.

➠ Evaluate locations of failure. Are tubes subject to wastage in locations common to a “reducing” environment? See Section 1.2 main text for detailed list of typical locations.

➠ Characterize extent of damage. Is there significant wall thinning across a number of tubes on the fireside?

➠ Characterize nature of damage on a failed tube. Is maximum attack on crown of the tube facing the flame?

➠ Analyze external tube deposits. Does metallographic analysis of deposits detect the presence of (i) pyrosulfates, sulfides, or other evidence of attack by sulfidation, (ii) unburned carbon or other evidence of poor combustion or local reducing conditions, or (iii) chlorides in the layer next to the tube?

Underlying root cause is probably not related to a locally substoichiometric condition; however, mechanism may still be fireside corrosion. Continue through flow chart for alternative cause resolution.

If damage is localized, it may be flyash or coal particle erosion damage or overheating; however continue through flow chart, particularly deposit analysis.

May still be waterwall fireside corrosion caused by an off-center fireball.

Mechanism is probably not fireside corrosion. Review other possibilities including an erosive process such as flyash erosion (Chapter 14) or coal particle erosion (Chapter 28), or overheating (Chapter 23).

Probable mechanism is fireside corrosion.

➠ Go to Action 3: Root Cause Determination

Volume 2: Water-Touched Tubes

18-19

Action 3: Determine root cause of fireside corrosion A BTF failure has occurred and the mechanism has been confirmed as fireside corrosion (Action 2) or a precursor to fireside corrosion has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes of fireside corrosion, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

References to other sources of information:

➠ (a). Metallographic analysis to detect key contaminants in deposits to confirm which fireside corrosion mechanism is active.

• Main text (this chapter) reviews the mechanism and the distinctive nature of external tube deposits. • A summary of the typical steps in a metallurgical examination can be found in Chapter 6, Volume 1.

➠ (b). Monitor for levels of O2 (< 0.1%), H2S, and CO (> 1%). ➠ (c). Field testing to detect combustion conditions in susceptible areas with waterwall deposition probes to collect deposits or heat flux measurement in corrosion areas. ➠ (d). As in (b) and (c). ➠ (e). Visual examination to detect localized flame impingement. ➠ (f). Monitor for change in furnace slagging conditions. Use waterwall deposition probe, as needed. ➠ (g). Analytical techniques to identify melting points of compounds in deposits. ➠ (h). Metallographic analysis to detect key contaminants in deposits. ➠ (i). Measure pressure drop across waterwall circuits that would be indicative of increasing deposits on waterside. ➠ (j). Direct metal temperature measurements with thermocouples or heat-flux meters. ➠ (k). Selective sampling to determine internal deposit amount and composition. ➠ (l). Analysis of internal oxide and deposits. ➠ (m). Analysis of cycle chemistry monitoring devices.

➠ (n). Analysis of propensity for coal corrosivity via available index methods.

➠ (o). Visual and metallographic examination to detect key contaminants in deposits to confirm which fireside corrosion mechanism is active. ➠ (p). Periodic sampling from mills to ensure proper level of coal fineness.

18-20

Waterwall Fireside Corrosion

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Evaluation will be based on detecting wall thinning. Wastage rates in excess of 40 nm/hr (~14 mils/yr) are of concern.

➠ Identify all locations to be examined. Refer to Section 1.2 of main text and Figure 18-4 for typical locations. Damage may be widespread and missed locations are sites for future failures.

➠ Perform UT survey to measure extent of damage via wall thinning. A review of UT methods is provided in Chapter 9, Volume 1.

➠ Perform tube sampling to confirm results of NDE inspection and to determine the degree of damage.

➠ Use results interactively with Action 3.

➠ Go to Action 5: Implement Repairs, Immediate Solutions and Actions. Begin remaining life assessment.

Volume 2: Water-Touched Tubes

18-21

Action 5: Implement repairs, immediate solutions and actions The most important actions for the BTF team are to: (i) initiate a remaining life assessment based on the assessment of wastage rate derived from the NDE survey, (ii) choose a repair strategy based on remaining life assessment, and (iii) coordinate long-term strategy from options shown in Figure 18-6.

3.2 Influence of a substoichiometric environment

salts

3.2.1 Poor combustion conditions (general)

3.3.1 Overheated tubes

3.2.2 Where air distribution has been modified 3.2.3 Poorly adjusted or worn burners 3.4 Influences related to fuel factors 3.4.1 Change in fuel to coal with high

3.3 Influences of the deposition of alkali

18-22

Waterwall Fireside Corrosion

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will be based on clear identification of the underlying root cause and an economic evaluation to ensure that the optimal strategy has been chosen. It will also include an analysis of remaining life.

Major Root Cause Influences

➠ Long-Term Actions

corrosivity

➠Choose long-term strategy from those shown in Figure 18-6. ➠ Implement on-going program of remaining life assessment and monitoring. See Chapter 8, Volume 1 for additional detail.

3.5 Root causes of direct carbon deposition 3.5.1 Carbon particle impingement

➠ As above, plus ➠ Develop a fireside testing program. Guidance is provided in reference 20. ➠ As above. ➠ As above, plus ➠ Alternate burner designs, as needed. ➠ Choose long-term strategy from those shown in Figure 18-6. ➠ Implement on-going program of remaining life assessment and monitoring. See Chapter 8, Volume 1 for additional detail. ➠ As above, plus ➠ Develop optimum feedwater chemistry strategy so as to minimize ingress of feedwater corrosion products. ➠ Investigate the use of oxygenated treatment as a means to eliminate the internal boiler deposits. ➠ As above. ➠ As above, plus ➠ Develop a fireside testing program. Guidance is provided in reference 20. ➠ Investigate coal changes with Coal Quality Impact Model (CQIM) or equivalent, including economics evaluation. ➠ Choose repair strategy based on severity of corrosion rate. ➠ Implement long-term strategy from choices in Figure 18-6 in conjunction with on-going program of remaining life assessment and monitoring. ➠ As above, plus ➠ Develop a fireside testing program. Guidance is provided in reference 20.

Volume 2: Water-Touched Tubes

18-23

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components implied by the presence of fireside corrosion in the waterwalls or by its precursors.

18-24

Waterwall Fireside Corrosion

Waterwall Fireside Corrosion Aspect

Alert for Other Cycle Components

➠ Actions Indicated

Corrosive coal

• Potential for superheater/reheater fireside corrosion see Chapter 33, Volume 3. • Potential for back-end corrosion.

➠ Develop a fireside testing program such as provided in reference 20. ➠ Investigate coal changes with Coal Quality Impact Model (CQIM) or equivalent, including economics evaluation. ➠ Mitigate negative aspects of coal composition if possible by fuel switch, blending, or washing.

Poor combustion conditions

• Low unit efficiency. • Poor mill performance. • Combustion is delayed and occurring in the convective passes, which could lead to corrosion of SH/RH surfaces.

➠ Combustion adjustments to improve unit efficiency. ➠ Correct mill performance.

Tube overheating by thick internal deposits or ripple magnetite

• Overheating in tubes. • Alert of poor feedwater treatment or controls.

➠ Chemical clean unit if necessary. See guidance in Chapter 4, Volume 1. ➠ Implement program to clean up and ensure proper cycle chemistry. See overview of issues in Chapter 3, Volume 1.

Chapter 19 • Volume 2

Supercritical Waterwall Cracking

Introduction In its most common manifestation, this damage type appears as circumferential cracking in the waterwalls of coal-fired supercritical units. Although not as common, such damage has also been found in subcritical units. The underlying mechanism has been termed corrosion-enhanced thermal fatigue. A related damage type can also occur in oil-fired/gas-fired units where the higher heat flux leads to more rapid tube temperature increases and damage accumulates as oxidation-enhanced creep cracking; cracks are found in either circumferential or longitudinal orientations. Wastage, caused by fireside corrosion, has been found in some cases to be associated with this type of cracking. The problem of circumferential cracking in waterwalls has been fairly widespread in supercritical

units in the U.S. and in a number of other countries. Research over the last twenty years has helped to clarify the situation and successful prevention is in hand, although a complete explanation of the basic phenomenon is still evolving. Cracks are often not detected until a failure occurs unless care is taken to remove all of the scale from tube surfaces. They are usually not observed until a unit has been in service for a number of years, but once detected and repaired, seem to return within a matter of months. This type of cracking has been called by a number of names, including circumferential cracking, horizontal cracking, transverse cracking, craze cracking, elephant hiding, and alligator-skin cracking. Note that some care in interpreting these names is needed as many of them have also been applied to damage other than that addressed in this chapter.

Volume 2: Water-Touched Tubes

19-1

1. Features of Failure and Typical Locations Supercritical Waterwall Cracking: Identification Keys 1. Damage generally forms as regular, parallel cracking, typically oriented circumferentially. It is most commonly found on the fireside of waterwalls and membranes between tubes of coal-fired supercritical units. 2. In cross-section, the cracks are found to be sharp, vee- or “dagger”-shaped, generally transgranular and filled with oxide. The presence of a sulfur-containing central spline and sulfur in the oxide is typical. 3. Highest heat flux zones are the most susceptible.

1.1 Features of failure Damage typically consists of multiple, parallel cracks perpendicular to the direction of maximum tensile stress. Most often cracking is circumferential, however incidences of longitudinal cracking have also been observed, particularly in oil-fired and gas-fired units. Figure 19-1 shows the general nature of the cracking. This sample was removed from an 800 MW supercritical boiler and illustrates the multiple, parallel, veeshaped cracks. There is great variability in the cracking density from tube to adjacent tube, even within parallel waterwall tubes in a single pass. Cracks on a given tube are often uniformly spaced with a density that is typically 20-40 cracks per tube inch.

Although regularly spaced, adjacent cracks are usually of different lengths.1 Tubes containing these types of cracks may be found in conjunction with significant fireside wastage, up to 50% of the tube wall thickness in some units. However, there have also been may cases in which cracks occurred with little wastage.2 Chapter 18 discusses fireside wastage in water-touched tubes. Circumferentially-oriented cracks are sharp, vee-shaped or finger-like in cross section. Figures 19-2a and b show details of the typical appearance of cracking. Figure 19-2a shows typical sharp-pointed transgranular cracking. An oxide coating along the depth of the crack is shown. A central spline containing

Figure 19-1. Cracked waterwall tubes from an 800 MW supercritical boiler. Source: M.D. Kurre, F.B. Stulen and I.G. Wright1

19-2

Supercritical Waterwall Cracking

Figure 19-2a. Cross section showing the typical appearance of cracking: sharppointed features and the oxide (dark) and sulfide (light) corrosion products. Source: M.D. Kurre, F.B. Stulen, and I.G. Wright1

Figure 19-2b. Closeup of typical cracking indicating appearance at crack tip and presence of a minor amount of intergranular sulfidation at the tip. The right hand of the two photographs is of etched material. Source: M.D. Kurre, F.B. Stulen, and I.G. Wright1

oxide (dark) and sulfide (light) corrosion products is shown. Figure 19-2b shows a close-up of a typical crack tip showing the presence of intergranular sulfidation at the crack tip. The presence of spheroidization and graphitization near only the outside surface indicates that there was a significant thermal gradient at that location. However, this is not a consistent feature as other tubes removed from service with circumferential cracking were found to have no microstructural changes consistent with overheating.1

a)

b)

The formation of waterside magnetite with a “ripple” appearance is typical. Figures 19-3a-d show the macroscopic appearance and its details. In oil-fired/gas-fired units, cracking is typically oriented longitudinally and is found, upon examination, to be essentially creep; that is, such cracks have creep voids at crack tips and an appearance that is similar to long-term overheating failures.

c)

d)

Figure 19-3. Scanning electron micrographs of rippled magnetite surface. (a) View along the tube. (b) Details of the structure of a ripple. (c) Magnetite crystals at the crest of a ripple (area “A” in Figure 19-3b). (d) Magnetite crystals in the trough of the ripple (area “B” in Figure 19-3b). Source: M.D. Kurre, F.B. Stulen, and I.G. Wright1

Volume 2: Water-Touched Tubes

19-3

1.2 Locations of failure Failures are located primarily on the fireside of waterwall tubing and to some extent in the membranes between tubes. Usually cracks are the deepest at the fireside crown of the tube. Cracking tends to be very localized, it may be limited to a specific pass of multiple tube passes in the radiant section, and tends to be limited to a narrow range of elevations in the boiler, usually related to the maximum heat flux zone at the burner level. This is illustrated in Figure 19-4 which is a map of circumferential cracking made during a field survey. This type of cracking has also been observed in tube bends, particularly near burners. Cracking has also been seen on the hopper slope tubes.5 Fireside wastage, if it occurs, is also in the same areas.

Degree of Grooving Slight

Moderate

Firing circle

B u r n e r C o r n e r 1

2

Severe

Firing circle

3

4

50 100 150 200 250 50 100 150 200 250 Tube Numbers Tube Numbers Front Wall

Figure 19-4. Typical areas of supercritical waterwall cracking. Source: A.L. Plumley and W.R. Roczniak4

19-4

Supercritical Waterwall Cracking

2. Mechanism of Failure Supercritical Waterwall Cracking: Mechanism 1. The basic mechanism is corrosion-enhanced thermalfatigue. 2. A number of factors that tend to increase the likelihood of damage and its severity have been identified. However, it has been found that controlling the internal “ripple” magnetite deposits will eliminate the problem, leading to a focus on this as the primary contributing factor. 3. Once excessive amounts of internal deposits accumulate, the superposition of stress and temperature cycles from a variety of causes, notably the slagging/ deslagging process, in conjunction with a corrosive fireside environment, will lead to the rapid accumulation of damage.

The survey did provide the basis for lines of inquiry into causative factors. Identified factors that tended to promote cracking in the units surveyed were2: (i) heat flux into the waterwall tubes at levels above design, (ii) low levels of fireside excess oxygen, resulting in “reducing” conditions, (iii) excessive deposits on the internal tube surfaces, (iv) furnace-pressure cycles in balanced-draft units, (v) rapid startup procedures, and (vi) frequent load cycling. Section 2.2 below reviews the work conducted as a follow-on to this early study, including a refinement to some of these factors, and the addition of several more.

2.1 Introduction and brief review of the international experience base A brief overview of the international experience base will help set the stage for a discussion of the current thinking on the mechanism, and of the possible approaches to prevention outlined later. A 1986 survey conducted on 56 coal-fired supercritical units in the U.S. found about 45% had experienced cracking/tube failures by this mechanism. Twenty units (36 percent of the total) had experienced cracking considered to be “severe” by the operators. Analysis of the gathered information failed to uncover simple, direct relationships between the propensity for cracking and (i) the nature of the coal burned, (ii) unit design, (iii) operation, or (iv) maintenance parameters.

The experience base with circumferential cracking in the front and rear waterwalls of Russian supercritical units6 provides some excellent insight as to the nature of the problem. The cracking reported had

Pressure Drop (bar) 50

AVT

Boiler ÒAÓ

Oxygenated

40

✽

✽ Chemical clean

✽

30 50

Boiler ÒBÓ

AVT

Oxygenated

40

✽ 30 50

40

30 0

Boiler ÒCÓ

Oxygenated

2000

4000

6000 Time (hours)

8000

10,000

12,000

Figure 19-5. Typical boiler pressure drop recorded in three German supercritical units. Note: 1 bar = 14.504 psi. Source: A. Bursik9

Volume 2: Water-Touched Tubes

19-5

features similar to those described above and the cause was stated to be corrosion fatigue driven by stresses induced by thermal changes such as fluctuations in heat flux7 or from slagging and deslagging8. The widespread introduction of oxygenated feedwater treatment has drastically reduced the deposition in waterwall tubes and greatly decreased the incidence of circumferential cracking in these boilers.1 The case study later in this chapter presents some specific results from these units.

minimize the formation of ripple magnetite, and many use spiral-wound or serpentine waterwall construction which can provide more uniform heat flux distribution than vertical waterwall designs; these two factors are thought to be primarily responsible for the avoidance of this damage type. It should be noted that prior to changing to oxygenated treatment, both German and Japanese units had experienced problems with pressure drop losses associated with ripple magnetite. Figure 19-5 shows this experience. Pressure drop (left axis) is a measure of the amount of internal tube deposits, usually as ripple magnetite. Chemical cleaning to remove deposits, or changing to oxygenated treatment have been shown

This form of cracking has not been reported in German supercritical units. Those units typically employ oxygenated treatment, intended to

50

60 Density (r)

50

40 30

40 30

20

r, lb/ft3

4000 psia

Viscosity (m)

20

10

10

0

35

1400 4000 psia

30

1200 25 20

Enthalpy (H)

Thermal conductivity (k)

1000

15

H, Btu/lb

k, 10-2 Btu/h.ft. °F

m,10-6 lbm/sec. ft

70

800 10 600

5

Cp, Btu/lb °F

8 SH, 3000 psia SH, 4000 psia

6 4 2 0 600

650

700 750 800 Fluid Temperature, °F

850

900

Figure 19-6. Temperature dependence of some properties of supercritical steam. Data from ASME steam tables. Compiled by: I.G. Wright, et al.10

19-6

Supercritical Waterwall Cracking

to minimize pressure drop (tube internal deposition).

2.2 Contributing factors to circumferential cracking damage During the past five years, a significant research effort has continued to identify specific factors that may contribute to this form of damage, particularly in supercritical units. This section reviews those factors currently thought to be most important.1,10 The underlying process, as it occurs in coal-fired units, continues to be considered by many investigators to be corrosion-enhanced thermal fatigue. It has been suggested that corrosion effects include enhancement of crack initiation by corrosive attack along grain boundaries and cracks in the external oxide/scale, and enhancement of crack growth by corrosion-product wedge opening of cracks.2 When first placed into operation, or after a chemical clean, T11 tube material, typical of waterwall construction, develops an internal protective oxide of Fe3O4 as discussed in detail in Chapter 2, Volume 1. There is relatively little deposition on the inside tube surface, and normal tube metal temperatures are generally found to be in the range of 400425°C (~ 750-800°F). As with other damage types, the problems begin when conditions change from the nominally protective situation. The two major influences on the mechanism seem to be (i) increased tube surface temperature because of the formation of internal magnetite deposits and (ii) tube metal temperature excursions because of slagging/deslagging. Other potential influences that have been identified include (iii) the fireside conditions and (iv) cyclic stresses such as caused by unit load changes, flow instabilities, and local variations in heat flux. These variations are particularly important in supercritical units because at fluid temperatures above 390°C (~ 735°F), the ability of the fluid to absorb heat changes markedly, as illustrated in measures of specific heat, thermal conductivity, density and viscosity (Figure 19-6).

As illustrated in Figure 19-5, units that control the buildup of internal deposits, either by using oxygenated treatment or by periodic chemical cleaning, have been able to control boiler pressure drop losses and tube temperatures and are thus less prone to cracking. An analysis of the experience in Russian units indicated that thermal excursions from fireside slagging processes, discussed below, would by themselves not cause a problem, but that in conjunction with the increased metal temperature as a result of the buildup of internal deposits, would cause cracking.8 Flow instabilities can also result from the buildup of rippled deposits. Such instabilities can lead to additional tube-to-tube variations in flow, heat absorption, and, as a result, tube metal temperatures and propensity for cracking.10 Finally the observation that there tends to be more deposition of mag-

139

250 200

Fluid Temp. 700°F Pressure 3800 psi Porosity 0.65

150

Heat flux 300,000 Btu/ft2Ðhr

111

200,000

56

100 100,000

50 0

83

0

10

20 30 Deposit Weight (g/ft2)

28 0 40

Deposit Temperature Drop (°C)

Deposit Temperature Drop (°F)

2.2.1 Increased tube temperatures as result of internal magnetite deposits. As a result of the transport of feedwater corrosion products, there is a gradual buildup of internal tube deposits. In supercritical units, the first layer next to the tube surface is the compact protective magnetite; above this layer are porous deposits of magnetite, often in the form of “rippled” magnetite as illustrated in Figure 19-3. Such deposits have several effects. The temperature of the tube wall increases as a result of the insulation of the tube metal from the cooling effects of the fluid flow. Figure 19-7 shows that the temperature increase is a function of local heat flux: the higher the heat flux and the greater the deposit weight, the greater the increase in tube temperature. Tube metal temperatures can easily reach an average of 450°C (~ 840°F) with measured maximum tube temperatures as high as 475°C (~ 890°F).10

Figure 19-7. Illustrating increasing temperature drop across the internal deposit as a function of incident heat flux.

netite in regions of highest heat flux is consistent with the typical locations of most significant cracking. The case study in this chapter provides confirmatory evidence that the internal deposits are the primary factor in the development of damage by this mechanism. 2.2.2 Surface temperature excursions as a result of slagging/ deslagging. Complex reactions between the external tube surface and fireside environment are also occurring, as described in Chapter 2, Volume 1. The formation of fireside scale may include a mixture of iron oxide and sulfide. A layer of slag will develop which controls the fireside corrosive environment and insulates the tube. Significant thermal cycling of the tube occurs with the slagging and deslagging of tubes, either naturally or by sootblowing operations. Figure 19-8 illustrates the result schematically: a gradually increasing tube temperature caused by buildup of internal deposits of ripple magnetite is overlain with thermal excursions caused by the slagging/deslagging process.

It is a plot of the temperature of a clean tube with a nominal temperature around 400°C (~ 750°F). Significant spikes in temperature, up to 68°C (124° F), resulted from a large slag fall that occurred at 4.6 hours. The highest spikes were in those temperatures at 72° around the tube (fireside crown at 90°), and at the fireside web thermocouple. Interestingly, the fluid temperature increased only about 5.5°C (10°F) and then settled back to the steadystate level. Temperature spikes up to 140°C (~ 250°F) caused by deslagging, and occurring in less than 30 seconds, have been observed.10 As the ash layer reformed, the metal temperatures returned to normal. Measurements of “dirty” tubes, those with extensive internal deposit buildup, have shown that they can also be subject to significant thermal spikes caused by deslagging of up to 167°C (300°F). Improper sootblowing, in which water was inadvertently or intentionally blown on tubes has also led to thermal shock and thermal fatigue damage of tubes.11

Figure 19-9 shows measurements obtained from an instrumented unit.

Volume 2: Water-Touched Tubes

19-7

2.2.3 The contribution of other fireside conditions

Temperature, °F

2.2.3.1 Oxide crack formation. Stress analysis results indicate that compressive stresses develop on the fireside crown of a tube after slag fall. That stress results in axial creep strain which permanently shortens the tube near the crown of the tube. With slag cover, temperatures return to their normal levels, the tube surface is then placed in tension, and small, regularly-spaced cracks in oxide/scale layers can form, which allow attack of the external tube surface by corrosive processes.10

Temperature spikes due to deslagging; gradually increasing peak temperatures

1000 950 900 850 800

Gradually increasing base tube metal temperature due to internal deposits

750

Time

Chemical Clean

Figure 19-8. Schematic representation of increasing tube wall temperature caused by internal deposits and slag falls.

Temperature (°F) 900

800

72 45 22 FW, CW FL CC

700

72 45 22 FW FL CW CC

600 0

2

4 Time (hours)

6

8

Figure 19-9. Typical rapid changes of tube temperatures caused by a slag fall. (72, 45, and 22) are the thermocouples at those angles to the membrane; fireside crown would be 90°; (FW and CW) are the temperatures of the fireside and cold side of the membrane, respectively; (CC) is the crown on the cold side; (FL) is the temperature of the fluid. Source: I.G. Wright, et al.10

19-8

Supercritical Waterwall Cracking

Modeling and analysis of oxide layers has shown that damage is often manifested as a surprisingly regular array of cracks. This phenomenon and how it relates to corrosion fatigue damage is discussed in Chapter 13. It has been suggested that the cracking process results in stress relaxation in the oxide, with the highest relaxation adjacent to the crack. Therefore, the maximum remaining (unrelaxed) stress will be centered between cracks. With an increasing strain applied to the oxide layer, the next crack will then form at this center location between the existing cracks assuming a layer with uniform properties. Figure 1910 illustrates salient features of the model. The spacing between cracks is a function of the applied strain level. This model explains nicely the regular array of parallel cracks that forms. By this process, a number of equally spaced penetrations of the oxide are formed with “V”-shaped openings (Figure 19-11a). These cracks can then fill with oxide and be “wedged open” causing further crack propagation during thermal cycling (Figure 19-11b), and hence the formation of transgranular, straight cracks. The presence of a sulfide spline in the oxides filling the cracks suggests the repeated cracking of the oxide and the replenishment of the corrodent at the crack tip.10

d

b

w

Ds

Stress in film (s)

o

( d2 )

s¦

x

Figure 19-10. Schematic representation of the development of a regular array of evenly-spaced cracks in supercritical waterwall cracking. (sf ) is fracture stress; (Ds(x)) is a measure of stress relaxation. Source: A.G. Crouch and R.B. Dooley12

Fireside oxide Tube material

a) "V" shaped cracks

b)

Propagating crack with central spline

Figure 19-11. Development of a regular array of cracks on supercritical waterwalls. (a) The regularly spaced cracks in the oxide (see Figure 19-10) develop penetrations of oxide into the tube material. (b) The cracks propagate and develop a central sulfur "spline".

which occur by internal magnetite buildup and slagging/deslagging. 2.2.3.2 Effect of a substoichiometric environment. The fireside environment can influence the development of supercritical waterwall cracking in two distinct ways: (i) fireside corrosion caused by a substoichiometric environment, and (ii) the presence of low melting point ash species. Both of these effects are exacerbated by the increased surface temperatures

If a substoichiometric environment exists at the waterwall then it can be a direct cause of fireside corrosion by the mechanism discussed in Chapter 18. Under these local environmental conditions, and with elevated tube metal temperatures, there will be severe wall loss via the fireside corrosion mechanism. However, such wall loss has not been in evidence in all cases of supercritical waterwall cracking, so this aspect of the problem is not universal.

ash species. Molten ash deposits can form if tube metal temperatures are elevated into the regime of the lower melting point compounds. When liquid compounds form on the waterwalls, the bond between slag, or ash cover, and the tube metal surface may be weakened.13 Subsequently, slag shedding more readily occurs, increased tube metal temperatures result, and the consequent increased oxide strain can lead to crack formation. Such a sequence would allow for the formation of waterwall cracking without the presence of fireside corrosion. 2.2.4 Other possible contributors: the role of cyclic stresses and operation Work by Getsfrid, et al.8, indicated that damage from circumferential cracking of waterwalls was dominated by large thermal gradients. Large numbers of cycles with small temperature differences, such as might be expected from flame fluctuations, were negligible; however, large fluctuations in temperature resulted in a significant loss of tube life. A number of conditions that could cause thermal excursions of this extent can be postulated, including (i) slag shedding, (ii) unit load cycling, (iii) rapid unit startups, and (iv) a pressure imbalance between forced-draft and induceddraft fans that induces additional slag shedding because of waterwall vibrations, and also imposes direct bending stresses.10 The temperature of waterwall tubes, particularly in supercritical units, can also be subject to significant excursions above normal conditions because of the “sensitivity” of the circuit.10 This is the result of variation in heat absorption in waterwalls resulting from differences in (i) performance or aiming of burners, (ii) coal slag characteristics and deslagging tendency, and (iii) the pattern of operation of the sootblowers. This problem may be particularly acute in supercritical circuits where the tube-to-tube variations that result will be accentuated by the reduced ability of supercritical

2.2.3.3 Effect of low melting point

Volume 2: Water-Touched Tubes

19-9

fluid to accept heat.

2.3 Implications As noted above, the thermal cycles of slagging/deslagging alone are not thought to be sufficient to cause the development of cracks8; the necessity of increased baseline tube temperatures and associated boiler pressure drop losses because of internal deposits have been confirmed on a number of units in a number of countries.9,14 The conversion to oxygenated treatment in those units and the concurrent elimination of the problem as a result, also lends credence to the view that

19-10

Supercritical Waterwall Cracking

internal deposit buildup is the key contributor to the cracking problem, particularly as the fireside conditions, and slagging/deslagging occurrences, do not change with the conversion to oxygenated treatment. Further evidence is that in the U.S., units that have been chemically cleaned on a 11/2 to 2 year cycle rather than a 3 year cycle have drastically reduced the occurrence of failures.

2.4 A note on waterwall cracking in oil/gas-fired supercritical units The initial stage in the development of longitudinal cracking in oil-/gas-

fired units is similar to that described for coal-fired units; that is, internal deposits begin to accumulate thus raising the average tube metal temperature. However in the case of oil- and gas-fired units, the much greater heat flux accelerates the fluid side deposition rate and in a much shorter period of time tube temperatures reach a point where creep damage begins to accumulate. The deposits again are always rippled. Subsequent tube failures by long-term overheating (creep) are the result. Very little thermal cycling because of repeated slagging and deslagging occurs in these units

3. Possible Root Causes and Actions to Confirm Supercritical Waterwall Cracking: Root Causes 1. The primary root cause, and the one which is most preventable, is the buildup of excessive internal deposits in the tubes. 2. The second major root cause is slagging/deslagging that leads to the imposition of thermal or stress cycles; the role of fireside corrosion in the direct wastage of tubes is also influential in some units.

due to the fact that there is little slag on the tubes with gas- and oil-firing.

3.1 Introduction It is becoming clearer that the primary root cause influence, and the one which is the most preventable for this damage type, is the buildup of excessive internal deposits in the tubes. Once tube temperatures begin to rise, the superposition of stress cycles, particularly those which are induced thermally, along with the corrosive effects of the fireside environment, result in the accumulation of damage and possible fireside corrosion. There are some other factors that have been identified as also influencing the accumulation of damage. A review of the range of root cause influences follows; Table 19-1 summarizes the potential root cause influences, and actions to confirm, and corrective actions.

3.2 Influence of excessive internal deposits leading to increased tube metal temperatures The effect of increased waterwall deposits as a function of deposit weight and heat flux is shown in Figure 19-7. Research results have shown that the formation of such deposits and the resulting increase in tube metal temperature are a necessary and sufficient precondition to this type of damage.8 Such a gradual increase in deposits is the result of improper or non-optimum feedwater treatment and corrosion product control. It can also be regarded as the result of the expected normal operating condition in those units operating under deoxygenated (with N2H4) all-volatile treatment. Figure 19-5 shows the gradual increase in unit pressure drop measured in three units; the first two, boiler “A” and “B” show

typical pressure increases with operating hours. The marked decrease in pressure drop after a chemical clean is also clearly illustrated for boiler “A” at the 4,000 operating hours mark. The positive effect of changing to oxygenated treatment in these units is also evident. Actions to confirm this root cause: (a). Metallurgical analysis of the tube and internal deposits, specifically to determine the presence and extent of ripple magnetite. (b). Evaluate unit pressure drop, even if pressure drop is not a normal operating constraint. A plot of pressure drop with operating hours such as shown in Figures 19-5 and 19-12 can clearly identify a buildup of internal deposits. (c). Evaluate chemical cleaning frequency and records. The continual need to clean the unit is another sign that excessive deposition is occurring. Conversely, units which have been kept the cleanest, through the use of OT or by frequent chemical cleaning, appear to be less prone to circumferential cracking. (d). Evaluate chemical records, particularly for levels of feedwater oxygen and corrosion products. Monitoring results will indicate very low oxygen levels, perhaps 1 ppb or less at the economizer inlet, Fe levels between 5-10 ppb, and a very reducing environment, probably below - 300 mV measured with an ORP meter.

3.3 Influence of thermal cycling caused by slagging/deslagging Large, sudden increases in tube metal temperatures (up to 167°C (300°F)) have been measured during the deslagging process. Such extreme thermal cycles have been shown to cause crack development in laboratory tests. Field research has indicated that the cycles from

Volume 2: Water-Touched Tubes

19-11

Table 19-1 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

3.2 Excessive internal deposits (a). Metallurgical analysis of tube and inter- • Determine extent of leading to increased tube nal deposit, specifically to determine the damage, institute apprometal temperatures. presence and extent of ripple magnetite. priate repairs or replace(b). Evaluate boiler pressure drop even if it ments. See Chapter 11, has not been an operating constraint. Volume 1 for an Plot pressure drop measured versus overview of tube repairs. operating hours. See Figures 19-5 and • Perform chemical clean19-12 for examples. ing if indicated by the (c). Evaluate chemical cleaning frequency level of deposits that and records. have formed. Plot pres(d). Evaluate chemical records, particularly sure drop before and for levels of feedwater oxygen and after. corrosion products.

• Control formation of internal deposits, preferably by use of oxygenated treatment. If not possible, implement control steps such as periodic and frequent chemical cleaning, minimizing feedwater corrosion product production and transport. See additional information in Chapter 3, Volume 1.

3.3 Thermal cycling caused by slagging/deslagging.

(e). Analyze temperature transients (magni- • Periodic testing of soottude, frequency and timing) by the blowers to ensure proper installation of chordal thermocouples. operation such as func(f). Evaluate sootblower operation and tion of the water removal maintenance to determine whether systems. excessive conditions have occurred. See also separate writeup on sootblower erosion (Chapter 22).

• Optimize use of sootblowers through fireside testing. See additional detail in discussion of sootblower erosion (Chapter 22).

3.4 Fireside conditions.

(g). If significant fireside wastage is evident, review Chapter 18 for actions to pinpoint the most prominent causes. As a minimum, metallurgical analysis to determine extent and nature of fireside deposits is indicated.

• Review materials options if fireside corrosion is a significant contributor to the circumferential cracking damage. See the discussion in Chapter 18.

3.5 Large, cyclic stresses and other influences of operation.

(h). Review unit operating records for sources of cyclic stresses, number of starts, ramp rates, etc. (i). Install chordal thermocouples and review tube temperatures as in (e) above.

19-12

Supercritical Waterwall Cracking

Actions to confirm the influence of this factor are:

Boiler Pressure Drop Ð psi 600 (Chemical clean Nov. 92) 550

(g). If significant fireside wastage is evident in the tubes, for example at a rate > 25 nm/hr (8.6 mils/yr), review Chapter 18 for a series of actions to confirm the most prominent causes and to implement corrective actions. Metallurgical analysis of the damage and nature of the fireside deposits is required, as a minimum.

500 Hydrazine feed stopped

450

Oxygen injection started

400

3.5 Influence of large, cyclic stresses and operation

350 300 31 Jan

22 March

11 May

30 June 1993

19 Aug

8 Oct

27 Nov

Figure 19-12. Pressure drop with time in a U.S. boiler. The pressure drop can be reduced by eliminating the feedwater N2H4 and injecting oxygen. Source: R.B. Dooley, et al.15

slagging/deslagging operations alone will not cause the development of cracks without increased average tube metal temperatures from internal deposits.

excessive thermal shock to tubes. See also the discussion of sootblower operation and maintenance issues that is provided in Chapter 22.

Actions to confirm:

3.4 Influence of fireside conditions

(e). Analyze temperature transients by the installation of chordal thermocouples. The magnitude, frequency and timing of thermal transients can be confirmed for such operations as sootblower operation and for slag falls. (f). Evaluate sootblower operation and maintenance to determine whether excessive conditions or too frequent operation have occurred. For example, condensate introduced into the sootblower media can cause

A variety of factors can lead to an increased propensity for fireside corrosion of waterwalls; they are reviewed in Chapter 18. Such factors include the influence of a substoichiometric environment, the development and deposition of alkali sulfates, change to a fuel with a higher corrosivity, and direct carbon deposition. Chapter 18 should be reviewed for actions to confirm and solutions if it appears that fireside wastage is a significant contributor to the development of circumferential cracking.

The most significant damage seems to be done by a few, large temperature cycles. Thermal excursions and the corresponding cyclic stresses can influence the amount of tube life that is consumed by this damage type. Such cyclic stresses might arise from (i) significant unit load changes, (ii) excessive thermal excursion caused during rapid unit starts, or (iii) a pressure imbalance between forced-draft and induceddraft fans. In a supercritical unit there can also be significant tube-to-tube variation in temperature caused by such differences as (i) performance or aiming of burners, (ii) coal slag characteristics and deslagging tendency, and (iii) the pattern of operation of the sootblowers. The result will be additional tube temperature variation leading to an increase in damage accumulation in the hotter areas. To confirm: (h). Review of unit operating records for number of starts and ramp rates on start, to identify potential sources of excessive cyclic stresses.

Volume 2: Water-Touched Tubes

19-13

4. Determining the Extent of Damage (i). Installation of chordal thermocouples and review of tube temperatures as in (e) above. Once an initial failure has occurred,

a survey of susceptible areas should be instigated. Locations, of highest heat flux, (see Figure 19-4 for a typical pattern), as well as those listed in Section 1.2 should be included. Cracks are often not visible until oxide and scale have been removed

5. Background to Repairs, Immediate Solutions and Actions Supercritical Waterwall Cracking: Immediate Solutions and Actions

by grit-blasting. Wall loss can be measured by standard ultrasonic surveys; such methods are described in Chapter 9, Volume 1.

Immediate actions are to replace damaged tubes and, if appropriate, to remove excessive internal deposits by chemical cleaning. Most other activities can be performed as part of a longer term strategy.

In addition to determining the extent of damage, the first steps to be taken are the repair/replacement of affected tubes. Chapter 11, Volume 1 discusses repair and replacement methods in more detail.

19-14

Supercritical Waterwall Cracking

Control of the buildup of excessive

internal tube deposits is a primary consideration. Most of the options to do so are long-term solutions. However, if excessive deposits are found, such as in tube samples removed when repairs are made, a proper chemical cleaning is indicated. If other unit problems have been identified as root causes, such as incomplete water removal from sootblowers, improper water level control and drains in ash hoppers, etc.,

6. Background to Long-Term Actions and Prevention of Repeat Failures Supercritical Waterwall Cracking: Long-Term Actions 1. Controlling tube temperatures by limiting the formation of excessive internal tube deposits is the optimal longterm solution. 2. Oxygenated treatment can be employed for units with all-ferrous metallurgy in the feedwater train. 3. The use of more resistant materials, particularly chromizing of standard waterwall materials, has been shown to be partially effective in those cases where fireside corrosion is a significant contributor to the problem.

then these deficiencies should be corrected as soon as feasible and thus may fall into the category of immediate repairs.

6.1 Options for the control of internal deposits The most complete solution to the problem of controlling waterwall deposits in supercritical units is oxygenated treatment (OT). Additional detail on this topic can be found in Chapter 3, Volume 1; its use in preventing supercritical waterwall cracking is reviewed here. OT has been shown to reduce iron feedwater corrosion products to less than 1 ppb at the economizer inlet and the waterwall deposition rate to less than 0.5 mg/cm2/1000 hrs (see Figure 3-10, Volume 1). This means that deposition is minimized, and at these levels, it is likely that the unit will not require chemical cleaning over the expected service life. The case study that follows provides some data from the field experience indicating specific deposit reductions. If it is not possible to change the unit to full oxygenated treatment, then a series of control steps to prevent excessive deposition are indicated. These include:

• A periodic and frequent program of chemical cleaning. Chapter 4, Volume 1 discusses some of the key aspects of waterwall chemical cleaning. • Optimizing the feedwater treatment process to minimize the production and transport of corrosion products. This might include for example, eliminating N2H4 if the feedwater train is all ferrous or optimizing the use of O2 scavengers if a mixed metallurgy feedwater system is used. Figure 19-12 illustrates an example of eliminating N2H4 in a U.S. unit. Additional information can be found in the case study later in this chapter.

Volume 2: Water-Touched Tubes

19-15

6.2 Overview of materials options

economic appraisal which considers future repeat failures.

A materials solution may be indicated in those cases where fireside corrosion is a significant problem and its presence exacerbates the development and propagation of circumferential cracking. Such a solution will not address the underlying root cause and thus requires careful

A variety of materials have been evaluated for their ability to address the combined effects of this mechanism and particularly the fireside corrosion.4,16,17 Chromized panels

have shown good results. For example, in side-by-side testing, chromized tubing showed essentially no fireside wastage over a four year period, during which adjacent bare metal tubes had wall losses in the range 1.78 to 2.29 mm (0.070 to 0.090 inches).17 In one case, the annual number of tube failures

7. Case study Supercritical Waterwall Cracking: Field Experience dropped from about 45 to less than 10 and the lost availability at the unit dropped from 8% to less than 2%, which was attributed to the use of chromized panels. The chromizing process promotes diffusion of chromium into the surface of a boiler tube; the alloy layer that results is usually 0.20 to 0.38 mm (0.008 to 0.015 inches) thick and ranges in chromium content from almost 80% at the surface to 18% at the interface with the base material.17 Some problems with chromized tubes have been encountered including17: (i) degradation in butt welds made with low-alloy steel weld materials, a solution is to apply stainless steel caps on the butt welds by E308 or E309 filler, (ii) spalling of the chromized layer, the solution is control the quality of the chromized layer by ensuring uniform thickness and by minimizing voids, (iii) cracking of the chromized layer which can be controlled by ensuring a minimum layer thickness of 0.254 mm (0.010”) as thinner layers have cracked in field

service, (iv) decarburization can reduce material hardness and strength and thus make the tubes more susceptible to failures by short-term overheating. This can be controlled by normalizing the panels after chromizing, (v) removal of chromized layer by abrasives used in ash and scale removal processes prior to inspections, this can be prevented by the use of high pressure water only.

There is now a significant worldwide experience (estimated to be nearly 320 units) with the conversion of supercritical units to oxygenated treatment (OT) and of the ability of such conversion to eliminate the problem of waterwall cracking and wastage. This case study summarizes the international experience. In Russian coal-fired units, peak heat flux is on the order of 250-300 kW/m2. The experience of units operating on AVT indicated that chemical cleaning was required about every 20,000 hours at a deposit level in excess of 30 mg/cm2 (~ 28 g/ft2). At these levels of deposits, tube temperatures reached 525-530°C (~ 975 to 985°F), and a significant number of tube failures occurred by the mechanism described in this chapter. Change to OT typically resulted in significantly less deposition; for example, over 100,000 hours of operation has resulted in accumulated deposits in an amount less than 10 mg/cm2 (~ 9.3 g/ft2). Tube temperatures remained in the normal (clean tube) range and tube failures stopped.14 For oil-fired units, peak heat flux is on the order of 500-550 kW/m2. Units operating under AVT typically

19-16

Supercritical Waterwall Cracking

required cleaning after as few as 5,000 operating hours. Tube temperatures were reaching levels on the order of 600°C (~ 1110°F) and failures occurred by overheating (creep). Here, as with the coal-fired units, tube failures essentially stopped occurring by this mechanism following the change to OT, and chemical cleaning is no longer required. Figure 19-5 shows data from three German units. The buildup of deposits is indicated by the increasing pressure drop experienced in the waterwall circuits plotted versus operating hours. Fairly significant pressure drops are recorded over only a few thousand hours of operation on AVT. An immediate improvement can be seen for chemical cleaning (see Boiler “A” at the 4,000 hours mark), with a subsequent buildup of deposits following the clean. The decrease in deposits with units switched over to OT from AVT (Boilers “A” and “B”) and the constant level of deposits indicated for the unit on OT from the start (Boiler “C”) show the advantage of OT. As noted in Figure 19-7, there is a direct correlation between deposit weight and tube metal temperature, and as shown in field testing, between increased tube metal temperatures and the propensity for damage by circumferential cracking. The Japanese experience with deposits and changing from AVT to OT was consistent with that shown in Figure 19-5.14 With AVT, there was a typical increase in deposits and thus pressure drop in the unit. The change to OT resulted in a dramatic improvement and reversal of the deposit process. Finally, a U.S. unit showed the similar pattern upon the change to OT, as dramatically illustrated in Figure 19-12. A leveling off of the pressure drop occurred when N2H4 feed was stopped. The institution of oxygen treatment began to reverse the trend of increasing pressure drop. Although the unit had neither a problem with circumferential cracking nor excessive pressure drop, the favorable indicators of lower waterwall deposits as indicated by unit pressure drop, bode well for avoiding future problems and the need for periodic, frequent chemical cleans.

Cracking on the Waterwalls of Supercritical Boilers, Volume 1: Background to the Problem and Experimental Approach, Research Project 1890-8, Final Report TR104442, Electric Power Research Institute, Palo Alto, CA, December, 1995. 2Cialone,

H.J., and I.G. Wright, R.A. Wood, and C.M. Jackson, Circumferential Cracking of Supercritical Boiler Water-Wall Tubes, Research Project 1890-4, Final Report CS-4969, Electric Power Research Institute, Palo Alto, CA, December, 1986. 3Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 4Plumley,

A.L. and W.R. Roczniak, “Practical Prevention of Waterwall Distress: A Metallurgical Approach”, in B. Dooley, ed., Proceedings: International Conference on Boiler Tube Failures in Fossil Plants, held in San Diego, California November 5-7, 1991, Proceedings TR-100493, Electric Power Research Institute, Palo Alto, CA, April, 1992, pp. 3-45 through 3-81. 5Personal

Communication from D. French (David N. French, Inc.) to R.B. Dooley, February, 1995. 6Degtev,

O.N., et al., “Analyzing the Reasons for Corrosion-Fatigue Damage to the Waterwalls of P-57 and PK-39 Boilers”, Teploenergetika, Volume 35, No. 11, 1988, pp. 39-43. 7Shakhsuvarov,

K.-L.V., V.A. Chetverikov, and A.Y. Yalova, “Influence of Temperature Fluctuations on Lower Radiant Section Tubes on Their Service Life”, Teploenergetika, Volume 24, No. 6, 1977, pp. 25-28. 8Getsfrid,

W.I., M.A. Petrov, and A.V. Rudyka, “Evaluation of the Life of Tubes in the Lower Radiant Section of a P57 Boiler with Random Fluctuations in Temperature and an Increase in Internal Deposits”, Telpoenergetika, Volume 34, Number 3, 1987, pp. 150. 9Bursik,

8. References 1Kurre,

M.D., F.B. Stulen, and I.G. Wright, Circumferential

A., “Eight Years of Modified AVT with Elevated Oxygen Level for Once-Through Steam Generators”, Proceedings of the International Water Conference, Volume 47, 1986, pp. 227-231. 10Wright,

I.G., D. Anson, H.J. Cialone, G.O. Davis, M.D.

Volume 2: Water-Touched Tubes

19-17

ACTIONS for Supercritical Waterwall Cracking Two paths for the BTF team to take in the investigation of supercritical waterwall cracking begin here. The goal of these actions is to see if further investigation is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred and supercritical waterwall cracking is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by supercritical waterwall cracking.

Kurre, and F.B. Stulen, “Root Cause of Circumferential Cracking in the Waterwalls of Supercritical Boilers: A Report on Work in Progress”, in op. cit. reference 4, pp. 3-1 through 337. 11Ellery,

A.R., T.R. Johnson, and J.D. Newton, “Investigation into the Likelihood of Thermal Fatigue Damage to Furnace and Superheater Tubes Caused by OnLine Water Deslagging”, Transactions of the ASME, April, 1974, pp. 138-144. 12Crouch,

A.G. and R.B. Dooley, “The Mechanical Integrity and Protective Performance of Silica Coatings”, Corrosion Science, Volume 16, 1976, pp. 341-347. 13French,

D.N., “Circumferential Cracking and Thermal Fatigue in Fossil-Fired Boilers”, Paper No. 133, presented at the NACE Conference/Corrosion 88, St. Louis, Mo, 1988. 14Bursik,

A., B. Dooley, and B. Larkin, Guidelines for Oxygenated Treatment for Fossil Plants, Research Project 1403-45, Final Report TR102285, Electric Power Research Institute, Palo Alto, CA, December, 1994. 15Dooley,

R.B., J. Mathews, R. Pate, and J. Taylor, “Optimum Chemistry for ‘All-Ferrous’ Feedwater Systems: Why Use an Oxygen Scavenger?”, Proceedings of the 55th International Water Conference, Pittsburgh, PA, October 31-November 2, 1994. 16Bakker,

W.T., E.C. Lewis, and A. Plumley, “The Use of Diffusion Coatings and Claddings for Fireside Corrosion Prevention” in B. Dooley

19-18

Supercritical Waterwall Cracking

and D. Broske, eds., Boiler Tube Failures in Fossil Power Plants: Conference Proceedings, Conference held in Atlanta, Georgia, November 10-12, 1987, CS-5500-SR, Electric Power Research Institute, Palo Alto, CA, 1988, pp. 2-85 through 2-100. 17Bonnington,

A.J. and T.M Cullen, “Mitigation of Circumferential Cracking and Fire-Side Corrosion in Supercritical Boilers by the Installation of Chromized Waterwall Panels”, Proceedings of the EPRI Conference on Welding and Repair Technology, held in Williamsburg, Va, March 20-25, 1994, TR-104588, Electric Power Research Institute, Palo Alto, CA, in publication.

Action 2: Determine (confirm) that the mechanism is supercritical waterwall cracking. A tube failure has occurred which the BTF team has tentatively identified as being supercritical waterwall cracking (Action 1a). Action 2 should clearly identify supercritical waterwall cracking as the primary mechanism or point to another cause. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Analyze internal scale.

Lack of internal deposits, unless removed by the failure event itself, indicates that the mechanism is probably not waterwall cracking.

➠ Analyze appearance of cracking.

May still be one of the varieties of supercritical waterwall cracking, particularly in oil-/gas-fired units, or longitudinal cracks which may evidence creep voids at or near crack tips.

Is there evidence of a buildup of “ripple” magnetite?

Are cracks sharp, “vee”-shaped or “dagger”-shaped? Are they oxide filled?

➠ Evaluate the extent of fireside corrosion. Is there evidence of extensive fireside corrosion?

Circumferential cracking may or may not be associated with fireside corrosion; see also fireside corrosion in Chapter 18.

➠ Analyze oxide in cracks. Is there evidence of sulfide in the oxide? In particular, is there a central spline of sulfide present?

➠ Analyze the microstructure of the tube metal, particularly near the fireside surface. Is there significant evidence of overheating such as spheroidization or graphitization?

Circumferential waterwall cracking may or may not be accompanied by signs of significant overheating. Longitudinal waterwall cracking will generally have associated material degradation and overheating.

Probable mechanism is supercritical waterwall cracking.

➠ Go to Action 3: Root Cause Determination

Action 1a: If a waterwall BTF has occurred and supercritical waterwall cracking is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of supercritical waterwall cracking:

Volume 2: Water-Touched Tubes

19-19

Action 3: Determine root cause of the supercritical waterwall cracking A BTF failure has occurred and the mechanism has been confirmed as supercritical waterwall cracking (Action 2) or a precursor has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences • On the fireside of waterwall tube, particularly at the crown, or in the membranes between tubes. • In the highest heat flux locations, such as in a narrow range of elevations around the burners.

➠ Confirm that the macroscopic appearance of the failure includes such features as:

• Multiple, parallel cracks oriented circumferentially; may be longitudinal in oil-fired/gas-fired units. • Variability of crack density from tube to adjacent tube. • May be found with significant external wastage, for example, up to 50% of the tube wall thickness.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

19-20

Supercritical Waterwall Cracking

➠ Actions to Confirm ➠ (a). Metallurgical analysis of tube and internal deposit, specifically to determine the presence and extent of ripple magnetite. ➠ (b). Evaluate boiler pressure drop even if it has not been an operating constraint. Plot pressure drop measured versus operating hours. See Figures 19-5 and 19-12 for examples. ➠ (c). Evaluate chemical cleaning frequency and records. ➠ (d). Evaluate chemical records, particularly for levels of feedwater oxygen and corrosion products. ➠ (e). Analyze temperature transients (magnitude, frequency and timing) by the installation of chordal thermocouples. ➠ (f). Evaluate sootblower operation and maintenance to determine whether excessive conditions have occurred. See also separate writeup on sootblower erosion (Chapter 22). ➠ (g). If significant fireside wastage is evident, review Chapter 18 for actions to pinpoint the most prominent causes. As a minimum, metallurgical analysis to determine extent and nature of fireside deposits is indicated. ➠ (h). Review unit operating records for sources of cyclic stresses, number of starts, ramp rates, etc. ➠ (i). Install chordal thermocouples and review tube temperatures, as in (e) above.

Action 4: Determine the extent of damage or affected areas Damage is generally easy to detect by visual examination once external oxide and scale is removed by grit-blasting. If wastage is occurring, use of ultrasonic testing to detect the extent of damage in typical survey mode is indicated.

Action 5: Implement repairs, immediate solutions and actions The primary immediate action is to execute the appropriate repairs and replacements. Most other actions can be considered long-term. Any obvious problems, such as condensate in the sootblower media, etc., should be corrected immediately.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by supercritical waterwall cracking.

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit: • Problems controlling feedwater corrosion product levels, evidence of erosion/corrosion in the feedwater system, and/or evidence of fouling on other feed-

Volume 2: Water-Touched Tubes

19-21

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF Team. The proper choice of long-term actions will be based on clear identification of the underlying root cause (Action 3) and an economic evaluation to ensure that the optimum strategy has been chosen.

Major Root Cause Influences water or boiler parts such as the BFP or orifices, for example. • Evidence of excessive deposit buildup on inside surface of tubes especially of ripple Fe3O4. Linked with the need to clean on a frequent (² 2 years) basis.

➠ These precursors can signal the potential for future tube failures by supercritical waterwall cracking. If one or more has occurred, go to Action 3 which reviews root causes and outlines the steps to confirm the influence of each.

➠ Long-Term Actions ➠ Control formation of internal deposits, preferably by use of oxygenated treatment. If not possible, implement control steps such as periodic and frequent chemical cleaning, minimizing feedwater corrosion product production and transport. See additional information in Chapter 3, Volume 1. ➠ Optimize use of sootblowers through fireside testing. See additional detail in discussion of sootblower erosion (Chapter 22). ➠ Review materials options if fireside corrosion is a significant contributor to the circumferential cracking damage. See the discussion in (Chapter 18).

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components implied by the presence of supercritical waterwall cracking, or its precursors.

19-22

Supercritical Waterwall Cracking

Supercritical Waterwall Cracking Aspect

Alert for Other Cycle Components

➠ Actions Indicated

High levels of feedwater corrosion products.

• Erosion and/or corrosion may be occurring in the feedwater system: in the feedwater heaters, deaerators, piping, or at the economizer inlet. • Corrosion products have probably deposited in other locations such as the boiler feed pump and at boiler orifices. The latter could lead to BTF by overheating.

➠ Need to develop an optimized cycle chemistry control program, preferably by instituting oxygenated treatment, but consisting at a minimum of periodic chemical cleaning and optimizing feedwater treatment.

Improper feedwater treatment process such as use of N2H4 (all ferrous) or excessive use of O2 scavengers (mixed metallurgy).

• Various problems throughout cycle, such as condenser tube problems caused by ammonia grooving.

➠ As above.

Chapter 20 • Volume 2

Thermal Fatigue Economizer Inlet Header Tubes Introduction Waterside-initiated cracking in carbon steel economizer inlet headers is a relatively recently recognized boiler tube damage mechanism. The generic mechanism is induced primarily by cyclic or transient thermal loading; the damage has the characteristics of thermal fatigue and the cracking has a morphology similar to corrosion fatigue. It is manifested by multiple cracks perpendicular to the principal direction of stress, with one crack usually becoming dominant and causing wall penetration. The stub tube failures that result can cause costly forced outages, and damage to the header is a serious safety issue which could keep a unit unavailable until a new header is fabricated and installed. Effective solutions include (i) improved operating conditions to limit through-wall thermal gradients,

along with (ii) improved header design details that have served to reduce inherent stress concentration factors. This is one of three mechanisms that affects the same general location on the economizer header inlet tubes. The other two are flexibility-driven fatigue (initiated from the OD) and an erosion-corrosion damage mechanism (on the ID). All three are compared in Chapter 7, Volume 1. Some comments are provided in this Chapter on distinguishing the damage caused primarily by thermal cycling from the other two damage types. The erosion-corrosion mechanism is the subject of Chapter 21. This damage mechanism has arisen in conjunction with the cycling of traditionally base-loaded units. Its root cause has been established and corrective actions have been fully characterized.

Volume 2: Water-Touched Tubes

20-1

1. Features of Failure and Typical Locations Thermal Fatigue in Economizer Inlet Header Tubes: Identification Keys 1. Thermal fatigue damage in economizer inlet header tubes will generally occur at the toe of header-to-stub tube attachment welds and be first noticed as a pin-hole leak. 2. Further examination will generally reveal numerous longitudinal cracks with morphology typical of thermally-induced fatigue with possible enhancement by corrosion, i.e., straight, transgranular cracks filled with oxide. They will be located on the inside surface of the tube stub, in the header borehole and then extending across ligaments between boreholes on the header inside surface. 3. Careful consideration of the damage mechanism is required to distinguish between (i) thermal fatigue induced damage (this chapter), (ii) flexibility-induced fatigue, and (iii) erosion-corrosion. All affect similar areas of the economizer header inlet tubes, but will require different corrective actions.

1.1 Features of failure Cracks can begin to form at any location along the header where the cyclic thermal stress is sufficiently high. Damage is initiated at stress concentrations associated with the bore hole and the tube attachment to the header (Locations A and B on Figure 20-1). The first appearance of thermally-induced damage is often a pin-hole leak in the toe of the weld at the header-to-stub tube attachment weld of economizer inlet headers. As shown in Figure 20-2, the inside surface of the damaged tube stub will manifest numerous longitudinal cracks. The cracks may be distributed completely around the bore of the tube, or may appear only in the higher stress locations. Cracks that begin to form in the header near tube penetrations can progress across the ligament between two tubes (Figure 20-1). Extensive cracking will invariably

lead to replacement of the header because of the large number of boreholes, difficulty of access for repair, and questions about safety and reliability of the header. Microscopic examination will show the cracking to consist of straight transgranular cracks, filled with oxide; i.e., damage that is typical of a thermal fatigue mechanism with some corrosion signs. Figure 20-3 shows the typical crack appearance.

1.2 Location of failures Service experience indicates that the worst damage is usually found in tubes closest to the feedwater inlet. However, damage can also occur away from the inlet tee in areas affected by service loading, the condition of the water, the header geometry, or header composition. Stub tube failures have generally occurred at the toe of the fillet weld on the tube side.

Typical longitudinal cracks Tube leak location

Economizer inlet header stub tubes

Weld

Tube thickness

A ID ligament spacing

B

CL tube

Header thickness

Figure 20-1. Cross section through economizer inlet header and tubes showing stub tube leak location and typical longitudinal pattern of cracking in the tube and header bore. Source: R.B. Dooley1

20-2

Thermal Fatigue Economizer Inlet Header Tubes

Figure 20-2. Damage developed from a tube penetration in an economizer inlet header.

Figure 20-3. Typical thermal fatigue cracking morphology. Note regular spacing of cracks and that they become thinner and straighter with propagation.

Volume 2: Water-Touched Tubes

20-3

2. Mechanism of Failure Thermal Fatigue in Economizer Inlet Header Tubes: Mechanism The mechanism is one of thermally-induced fatigue, with possible corrosion assistance, primarily caused when relatively cold feedwater is introduced into the hot header during transient operations such as startup, restart and shutdown, and off-line drum top-up periods.

Figure 20-4 shows that large header temperature differentials are associated with rapid changes in the feedwater flow typical of a unit hot start. Tests have shown that ÆT values in excess of 80°C (~ 175°F) are possible.3 Note also that there are several peaks per startup which are all associated with large spikes of feedwater for drum top-up as shown in the simplified schematic of Figure 20-5. Similar temperature spikes are experienced during shutdown. Evidence of enhancement of the damage by corrosion, such as the presence of heavy oxide thickness associated with the thermal fatigue cracks, has also been observed. Pitting is very often found in this region attributable to poor shutdown practices. Criteria to determine whether headers are likely to have developed damage or to be susceptible for developing future cracks are: • Header has accumulated a large number of operating cycles. • Unit is being converted to cyclic operation. • Ligament spacing between bore holes is small (on the order of 3.5 cm (1-3/8 inch) or less). • Thickness of the header is well above minimum code thickness. • Header-to-stub tube joint is a partial fillet weld (as in Figures 20-1 and 20-2), as compared to a full penetration weld (Figure 20-9). • The header has experienced stub tube leaks.

20-4

Thermal Fatigue Economizer Inlet Header Tubes

• The design of the header has a history of problems at other power plants. • The header experiences large through-wall thermal gradients. • Feedwater has been introduced intermittently at high flow rates during start-ups and off-line drum top-ups. The mechanism may be misdiagnosed as a flexibility problem. Stub tube leaks can develop because of flexibility problems related to header support, rotation, tube routing, etc. In that case, damage will appear generally above the fillet weld and cracks will propagate in the circumferential direction. Tube leaks caused by flexibility problems are generally restricted to the ends of the headers. The key difference is that these cracks initiate from the outside surface at the stress concentration, e.g., at the toe of the attachment weld to the tube. The thermal fatigue mechanism may also be confused with erosioncorrosion which can occur in this region. Erosion-corrosion failures also occur from the inside surface. A primary means of distinguishing the two will be visual inspection of the tube ID. Erosion-corrosion failures will typically have an “orange peel” appearance. Additional detail about this mechanism can be found in Chapter 21. Additional discussion of the distinctions between all three mechanisms can be found in Chapter 7, Volume 1.

Econ. header DT (difference between inner and outer temperature) Outer temp. > inner temp. = positive D T

60

Hot start

1. Feedwater flow (kg/s) 2. Unit Load (MW) 3. Drum level (+20 in -20 in)

30 0

2

-30

3

20

3

300

1 -60

150

0

-90 0

1

2

-20 3 Time (hours)

0 4

5

Load and FW Flow

Metal Temperature DT (F°)

90

6

Figure 20-4. Header temperature gradients during a unit hot start before operating changes were made to the economizer inlet flow rate. Source: G.G. Stephenson2

0

Economizer header DT

0 Drum level

0

0

Feedwater flow

Turbine speed Start firing Time

Figure 20-5. Simplified schematic of an economizer inlet header ÆT during shutdown/startup. Note: Temperature spikes which correspond to drum top-up during shutdown

Volume 2: Water-Touched Tubes

20-5

3. Possible Root Causes and Actions to Confirm Thermal Fatigue in Economizer Inlet Header Tubes: Root Causes 1. The most common root cause is operating conditions that produce large through-wall thermal gradients in the header. This condition usually accompanies increased cyclic operation of the unit. 2. Header design and construction provide locations of stress concentration that can exacerbate the basic problem.

3.1 Introduction Table 20-1 summarizes potential root causes, actions to confirm, and corrective actions for this damage mechanism. To confirm that this is the active damage mechanism: (a). Metallurgical analysis of removed tube samples is the principal action to distinguish thermallyinduced fatigue from flexibilityinduced cracking or erosioncorrosion.

3.2 Influence of large throughwall temperature gradients The most common root cause of this damage is a large temperature differential induced by the introduction of relatively cold feedwater into the hot header during transient operations such as startup, restart and shutdown. Off-line drum top-up with cold water is a common and necessary occurrence. Actions to confirm will consist of: (b). Measure thermal gradients during all operation periods. Thermocouples need to be attached to the surface of the header and installed in a drilled hole within one centimeter of the inside surface. The pre-

20-6

Thermal Fatigue Economizer Inlet Header Tubes

ferred location is adjacent to the feedwater inlet, but other locations can be monitored as well.2 A typical set of thermocouple locations is illustrated in Figure 20-6. Thermal gradients should be monitored during all typical operating procedures and shutdown.

3.3 Influence of stress concentrations Header design and construction details will introduce stress concentrators that may exacerbate the problem. Actions to confirm will include: (c). Analysis of inspection information to highlight the locations of maximum damage and determine if they are associated with known stress concentrations. For example, if the bole hole surfaces on the ID have been radiused, the stresses developed will be lower than if there is no radius. Full penetration welds will have lower stress concentrations than “paste-on” or socket welds. A primary use of this information will be as input to the redesign of headers if replacement is indicated.

Table 20-1 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Immediate Actions and Solutions

Actions to Confirm

Long-Term Actions and Prevention of Repeat Failures

3.1 For all root cause influences

(a). Metallurgical analysis of removed tube sample to confirm orientation, initiation sites and extent of cracking. Ensure that damage is in fact thermally-induced and not either flexibility-induced or caused by erosion-corrosion.

• Analysis with tools such as fracture mechanics and fatigue analysis to assess the safety of continued operation, for example leak-before-break. • Repair, replace, run decision required.

3.2 Cyclic operation that introduces large ÆT excursions through the wall of the header.

(b). Measure through-wall thermal gradients • Confirm mechanism. during all operating periods, including • Inspect to determine the feedwater flow, drum top-up, and during extent of damage. • For minor damage, shutdown. repair, modify operating procedures and institute long-term monitoring. • For major damage, replace header, modify operating procedures, institute long-term monitoring.

• Long-term monitoring and alarm of through-wall temperatures, particularly for ÆT. • Introduction of trickle feed system to prevent spikes of cold feedwater and to minimize ÆT. See main text for discussion of this and additional operating options. • Set re-inspection intervals to confirm efficacy of modifications, and to monitor damage accumulation. • Set re-evaluation period and execute periodic life assessment.

3.3 Stress concentrations

(c). Evaluate inspection data indicating locations of damage.

South End of Header, S

S3 S4

Feed Line, F

1st platen (S-N) S1

S3 S4

36th platen (S-N) Wall position

R3 R4

F4

R2

R5

Top row of tubes

F5

F1 F3

• Possible header redesign to lower stress concentrations, and stress levels caused by temperature differentials (when replacing header). See Figure 20-9 for typical modifications.

Recirculation Line, R

F3 F2 F4

S2

• See long-term strategies.

R6

F5

R1

10th platen (N-S)

R3

Wall position

R4

R5 R6

Valve S2, F2, R2, surface OD thermocouples between adjacent longitudinal stub tubes

Figure 20-6. Schematic of typical thermocouple locations on economizer inlet header. Thermocouple locations are designated by a letter (S, F, R) followed by an identifying number. Source: G.G. Stephenson2

Volume 2: Water-Touched Tubes

20-7

a

4. Determining the Extent of Damage

Determining the extent of damage may be difficult because of access restrictions. Figure 20-1 indicates locations where damage has been typically found. Since the worst damage is expected around the feedwater inlet, tubes removed should be from this area. The stub tube should be examined metallurgically to assess the degree of damage.

Visual inspection of the borehole and across the inside ligament will indicate if the header is cracked. If cracks are detected, random checks across the header can be used to indicate the extent of damage. This can be done by videoprobe.Other areas to be checked might include (i) header boreholes, (ii) the ligament region between boreholes on the inside of the header, (iii) the external stub tube to header welds that have not been removed, and (iv) the ID of the stub tubes that have not been removed.

Visual examination, dye penetrant, magnetic particle or ultrasonic testing (UT) may be used depending on the location to be examined and access, as shown in Figure 20-7. Surface preparation will depend on the technique to be used. None is required for visual examination of the ID, scale removal is required for dye penetration or magnetic particle inspection; grinding is necessary to prepare the outside surface for UT. Sizing of cracks may be difficult. Grinding or UT techniques may be used to determine crack depth if access and geometry permit. Procedures are available for sizing locations in ligament, girth weld and tee locations. An example of UT sizing with tandem probes is given in reference 4. Alternative methods of current injection and eddy-current inspection for shallow cracks are also possible.3

MP

MP

a MP

Visual/DP/MP Visual

Figure 20-7. Inspection methods and areas to be inspection. (MP) - magnetic particle inspection, (DP) - dye penetrant inspection. Source: G.G. Stephenson2

20-8

Thermal Fatigue Economizer Inlet Header Tubes

5. Background to Repairs, Immediate Solutions and Actions Thermal Fatigue in Economizer Inlet Header Tubes: Immediate Solutions and Actions 1. Apply step-by-step method to assess header condition; repair damage or replace header as indicated by the condition assessment. 2. Modify operating procedures if possible to minimize damaging conditions. 3. Include periodic inspection of susceptible areas as a longterm action.

5.1 Assessment methodology Once damage has been identified, a condition assessment methodology should be applied to analyze its severity. In fact, a utility should initiate the use of the methodology when (i) existing tube or header damage is found, (ii) cyclic unit operation is begun or contemplated, or (iii) an inspection is required for life assessment studies. The basic methodology consists of confirming the root cause, assessing condition of the tubes and of the header body, evaluating the effect of observed temperature excursions, determining the significance of cracking and damage found, performing a residual life assessment of the header, performing the indicated repairs, modifying operation to minimize the conditions that promote cracking, and monitoring for effectiveness of the imposed solutions and for continued crack growth. Damaged headers should be monitored for temperature differentials, which are the primary driving force behind the accumulation of damage via this mechanism. As noted above, thermocouples should be installed at the surface and in drilled holes within 1 cm (0.39 in.) of the inside surface. These deep wall thermocouples or thermowells5 are recommended because the analysis of surface temperatures alone does not appear to be sufficient to monitor rapid thermal through-wall gradients.3

Operating parameters such as load, feedwater flow, and pressure should be compiled along with the header thermal gradient. This information is then used in a stress analysis and fracture mechanics analysis to determine whether the stresses induced by thermal excursions are sufficient to propagate existing cracks. Note that the analysis of condition, by these methods is not a simple process. Complexities include, but are not limited to: the need to use accurate material properties for weld metal, base metal, and heat-affected zones, realistic models of geometry, accurate evaluations of stresses, and realistic flaw sizes and shapes.2,3,6 The methods starting with the existing crack depth, will evaluate the expected incremental crack growth as a function of measured levels of ÆT-induced stresses and other stress components, and determine the expected number of cycles to failure depending on the final allowable crack size. The analysis should also indicate the maximum thermal gradients that can be tolerated without crack growth. It should be emphasized that headers with significant cracking, e.g., cracks extending across the ligament, require analysis to determine whether they are safe to operate in that condition or need immediate replacement. The stress and fracture analysis should provide the information needed to make this decision.

Monitoring across the full range of operating procedures is required, including shutdown periods.

Volume 2: Water-Touched Tubes

20-9

5.3 Repairs/Replacement Econ. header DT (difference between inner and outer temperature) Outer temp. > inner temp. = positive DT

60

1. Feedwater flow (kg/s) 2. Unit Load (MW) 3. Drum level (+20 in -20 in)

30

Hot start

0 300

20

-30

12 3 -60

150

0

-90 0

1

2

-20 3 Time (hours)

0 4

5

Load and FW Flow

Metal Temperature DT (F°)

90

6

Figure 20-8. Header temperature gradients during a unit hot start following operating modifications. Source: G.G. Stephenson2

5.2 Modifications Once the monitoring and assessment has been performed, then a decision can be made about changes to operating procedures to minimize future damage. It has been clearly established that such modifications can reduce the thermal transients that are the responsible driving mechanism. For example, an attempt should be made to introduce the feedwater into the header on a more even basis. If this is not

20-10

possible, then the solutions should be sought among options such as (i) the use of a trickle feed system, (ii) intermittent feedwater admission by valving or (iii) increasing the feedwater temperature. Figure 20-8 shows temperature gradients during a unit startup on the same unit detailed in Figure 20-4 after procedure changes were introduced. Specifically, feedwater is now trickled into the header on a continuous basis and as a result the ÆT is much lower.

Thermal Fatigue Economizer Inlet Header Tubes

The repair of cracks on the inside surfaces of economizer inlet headers cannot normally be economically justified because of the large number of tube bore holes and the problem of access. Therefore, unless it can be shown by fracture mechanics and fatigue analysis that a cracked header can be operated safely and reliably by minimizing and monitoring crack propagation, the header will require replacement. It is important to recognize that any repair program which does not address the base cause of the problem will result in repeat failures. If the analysis, including monitoring and fracture mechanics, specifies a maximum ÆT that will not cause cracks to initiate and grow, and if this is easily achievable with modified operating procedures such as suggested above, then the current header with minor repairs can be used. If cracking was found to be too severe then a new, replacement header will be needed. If cracking was found to be serious and steps to reduce thermal excursions to acceptable levels would be an operating restraint, then a new header design is indicated. Improved design features are mostly targeted at reducing stress concentrations inherent in past designs. A

redesigned header will include such modifications as: • Decreased header thickness within design Code allowables. • Full penetration welds for the tube attachments. Full penetration tube attachment welds

• A radius on the hole lips on the inside surface of the header. • Designing the pitch on the tube holes to maximize all ligament areas on the header. • Installing headers outside the gas path.

Contoured inner hole radii

Figure 20-9 shows a schematic of some of these design improvements.

Increased ligament length

Decreased header thickness Figure 20-9. Schematic of modified header design. Source: G.G. Stephenson2

Options must be reviewed in conjunction with geometric and operating characteristics. Decreased wall thicknesses will tend to reduce through-wall temperature gradients. The use of materials with higher fatigue resistance, such as 1 CrMo steel may also be considered. Economic evaluations are required to judge which options are cost effective, and will include such considerations as number of stub tube failures, condition of the economizer, the extent of header cracking, expected type of operation, and the desired life of the header and unit.

6. Background to Long-Term Actions and Prevention of Repeat failures The long term strategy to prevent failures will extend those activities outlined above. Whether the existing header is used or new (and/or redesigned), it will be necessary to include a program of monitoring with appropriate alarms in the control room on the ÆT levels in the header. An understanding of how those thermal gradients will affect

header tube cracking should be in hand from the stress and fracture mechanics analysis. These actions should be performed in conjunction with periodic reinspection and confirmation that damage has been controlled to within acceptable levels.

Volume 2: Water-Touched Tubes

20-11

7. Case study Thermal Fatigue in Economizer Inlet Header Tubes: Case Study: Field Experience A survey of thirteen North American utilities was conducted during 1989. All had found some level of damage to economizer inlet header tubes. Table 20-2 illustrates that some common features were evident among those utilities. The following conclusions were reached: 1. For most utilities the first signs of damage were stub tube failures with leaks occurring at the toe of the fillet weld on the tube side or just above it. 2. Crack orientation was longitudinal. 3. Highest incidence of failure was at the feedwater inlet, although there was some cracking away

from the inlet region that was probably related to additional system stresses. 4. There was a strong correlation between unit cycling and tube and header cracking, found by comparing cracked and uncracked headers. 5. Design and fabrication features that were directly related to cracking included (i) use of socket welds instead of full penetration welds to join the stub tube to the header and (ii) lack of radii on inside edges of holes versus radiusing the hole lip such as required in German TRD design codes.

Source: G.G. Stephenson2

20-12

Thermal Fatigue Economizer Inlet Header Tubes

6. All U.S. boiler manufacturers’ headers have experienced cracking. 7. At the time of the survey, only a few utilities had instituted modified operating procedures to reduce the level of ÆT in the headers, typically by trickle feeding or raising the feedwater temperature before it was admitted to the header. A continuous slow feed of feedwater was also practiced; this was accomplished with either a low flow control valve or a low-capacity, variable-speed, boiler feed pump.

Table 20-2 Economizer Inlet Header Tube Cracking: Case Study Summary of Root Cause Similarities in Field Failures

Utility

Stub Tube Leaks?

Number of Headers Replaced

Cycling Operation?

ID Ligament Spacing (inches)

ÆT Measured?

Throughwall T/C?

T Limits? (°F)

A

Yes

0

Yes

1

Yes

Yes

70

B

Yes

0

Yes

13/8

No

No

None

C

No

1

Yes

78

/

No

No

None

D

Yes

9

Yes

12

/

No

No

None

E

Yes

8

Yes

38

/

Yes

Yes

55

F

Yes

1

Yes

-

No

No

None

G

No

0

Yes

-

No

No

None

H

No

1

Yes

12

/

No

No

None

I

Yes

1

Yes

11/2

No

No

50-100

J

Yes

0

Yes

-

Yes

Yes

43

K

Yes

0

Yes

34

/

Yes

Yes

None

L

Yes

2

Yes

-

No

No

None

M

Yes

3

Yes

13/8

No

No

None

8. References 1Dooley,

R.B., “Status of Economizer Inlet Header Cracking in Ontario Hydro Boilers”, Ontario Hydro Report TG31030, October 1, 1981.

4Moles,

2Stephenson, G.G., Guidelines for the Prevention of Economizer Inlet Header Cracking in Fossil Boilers, Research Project 1890-6, Final Report GS-5949, Electric Power Research Institute, Palo Alto, CA, November, 1989.

5Dunn,

3Parker, J.D., et al., Condition Assessment Guidelines for Fossil Fuel Power Plant Components, Research Project 2596-10, Topical Report GS-6724, Electric Power Research Institute, Palo Alto, CA, March, 1990.

M.D.C. and A.L. Allen, “Tandem Probe Ultrasonic Measurement of Cracks in Economizer Inlet Header Sections”, Materials Evaluation, May, 1984. K.M., J.R. Scheibel, and E. Schwarz, “Monitoring for Life Extension”, Combustion Engineering Report No. TIJ-PSG-85-001/. 6Mukherjee,

B., M.L. Vanderglas, and D.M. McCluskey, “Toughness Measurement and Structural Integrity Considerations of a Pressure Vessel”, Fifth International Conference on Pressure Vessel Technology, Volume 2, San Francisco. September, 1984.

Volume 2: Water-Touched Tubes

20-13

ACTIONS for Thermal Fatigue in Economizer Inlet Header Tubes: Two paths for the BTF team to take in the investigation of thermal fatigue in the economizer inlet header begin here. The goal of these actions is to see if further investigation of this mechanism is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred and thermal fatigue is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by thermal fatigue in the economizer inlet.

Action 1a: If a BTF has occurred and thermal fatigue is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of thermal fatigue, such as shown in Figure 20-1.

➠ Confirm that the macroscopic appearance of the failure includes such features as: • Pin-hole leaks or cracks at the toe of header-to-stub tube welds. See Figure 20-1. • Cracks extending across ligaments. • Damage to the inside surface of tube stubs (Figure 20-2).

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF in the economizer inlet header by thermal fatigue.

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit: • Has the unit recently been converted to cycling duty? • Has the header accumulated a large number of operating cycles? • Has the header experienced large thermal gradients? • Is the spacing of the ligament holes in the header small ( < 3.5 cm (1-3/8 inch)? • Is the thickness of the header well above Code minimum? • Are header to stub tube joints made with partial fillet welds? • Does this header design have a history of problems at other power plants? • Has feedwater been introduced intermittently at high flow rates during start-ups and off-line top-ups?

➠ These precursors can be root cause influences on thermal fatigue. If one or more has occurred, go to Action 3 which outlines the steps to confirm the influence of each.

20-14

Thermal Fatigue Economizer Inlet Header Tubes

Action 2: Determine (confirm) that the mechanism is thermal fatigue A failure has occurred which the BTF team has tentatively identified as being thermal fatigue (Action 1a). Action 2 should clearly identify thermal fatigue as the primary mechanism or point to another cause. The actions listed will be executed by (i) visual inspection of tubes and bore holes, and (ii) metallographic analysis of removed tube(s).

➠ Evaluate location along header. Is damage located primarily near to the feedwater inlet?

➠ Evaluate location on tube. Is damage initiation on the tube ID? See Figures 20-1 and 20-2.

➠ Evaluate location relative to weld. Is damage associated with the toe of a weld, particularly with a partial fillet weld (as opposed to a full penetration weld)?

➠ Evaluate damage appearance. Does sample of cracked/ damaged material show an appearance like an orange peel?

Problem may be flexibility induced cracking.

Problem is more likely to be flexibility induced cracking.

Damage is more likely to have been caused by erosion-corrosion, however, continue with flowchart to confirm.

Mechanism is more likely to be erosion-corrosion.

Probable mechanism is thermal fatigue. Steps to confirm will include:

➠ Sampling to confirm evidence of ID, bore hole damage.

➠ Go to Action 3: Root Cause Determination

Volume 2: Water-Touched Tubes

20-15

Action 3: Determine root cause of thermal fatigue damage A BTF failure has occurred and the mechanism has been confirmed as thermal fatigue (Action 2) or a precursor to thermal fatigue is evident (Action 1b). The goal for this Action 3 is for the BTF team to review the root causes of thermal fatigue at the economizer inlet header and take those steps needed to confirm that they are operative. This step must be taken so that the condition assessment, which is central to correction, can be performed. Execute, in parallel, Action 4 to determine the extent of damage.

20-16

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.1 For all root cause influences

➠ (a). Metallurgical analysis of removed tube sample to confirm orientation, initiation sites and extent of cracking. Ensure that damage is in fact thermally-induced and not either flexibility-induced or caused by erosion-corrosion.

3.2 Cyclic operation that introduces large ÆT excursions through the wall of the header.

➠ (b). Measure through-wall thermal gradients during all operating periods, including feedwater flow, drum top-up, and during shutdown.

3.3 Stress concentrations

➠ (c). Evaluate inspection data indicating locations of damage.

Thermal Fatigue Economizer Inlet Header Tubes

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis) the BTF Team should determine the extent of damage. Access is a key concern; for accessible locations, standard NDE inspection methods are usable.

➠ Determine the areas and extent of the inspection from review of header design, operating history, and stub tube failures. Refer to Figure 20-1 and main text for typical locations to inspect.

➠ Pre-inspection activities. Install scaffolding and remove insulation. Remove a handhole cap or cut a stub tube to provide access since worst damage is expected around the feedwater inlet.

➠ Take tube samples to assess the degree of damage.

➠ Surface preparation if any. None required for ID visual examination, scale removal for dye penetrant or magnetic particle inspection, grinding to prepare outside surface for UT. See Figure 20-7 for sample locations.

➠ Perform NDE. Perform visual, dye penetrant, magnetic particle, and/or UT examination as required. If cracks are detected, determine their depth by grinding or UT if access permits; alternative methods include current injection and eddy-current inspection for shallow cracks.

➠ Use results in conjunction with condition assessment.

➠ Go to Action 5: Implement Repairs, Immediate Solutions and Actions.

Volume 2: Water-Touched Tubes

20-17

Action 5: Implement repairs, immediate solutions and actions The BTF Team must ensure that repairs and immediate solutions are directly tied to the underlying cause. The following flowchart indicates importance of the condition assessment in choosing the correct immediate actions.

Do results of header inspection indicate that there is severe cracking?

➠ Install thermocouples and monNo

Yes

Does an analysis of the damage, including stress analysis and fracture mechanics, indicate that header can continue to be operated?

Yes

itor ÆT levels for the full range of operation including feedwater flow and drum top-up, as well as shutdown periods.

➠ Perform stress analysis and fracture mechanics analysis of the severity of damage, including results of inspection for extent of damage, and results of thermocouples for ÆT.

No

Can existing header, with or without repairs, continue in service? No

Is existing design sufficient to achieve desired life?

➠ Replace header in-kind. Yes

No

➠ Redesign header, replace. See Figure 20-9 for typical design modifications.

➠ Evaluate whether changes in operating procedures are needed to lower ÆT and thus achieve desired life.

➠ See Action 6: Long-Term Solutions to Prevent Repeat Failures

20-18

Thermal Fatigue Economizer Inlet Header Tubes

Yes

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will be based on clear identification of underlying root cause (Action 3) and an economic evaluation to ensure that the optimal strategy has been chosen.

Major Root Cause Influences

➠ Long-Term Actions

For all root cause influences

➠ Analysis with tools such as fracture mechanics and fatigue analysis to assess the safety of continued operation, for example leak-before-break. ➠ Repair, replace, run decision required.

Cyclic operation that introduces large ÆT excursions through the wall of the header.

➠ Long-term monitoring and alarm of through-wall temperatures, particularly for ÆT. ➠ Introduction of trickle feed system to prevent spikes of cold feedwater and to minimize ÆT. See main text for discussion of this and additional operating options. ➠ Set re-inspection intervals to confirm efficacy of modifications, and to monitor damage accumulation. ➠ Set re-evaluation period and execute periodic life assessment.

Stress concentrations

➠ Possible header redesign to lower stress concentrations, and stress levels caused by temperature differentials (when replacing header). See Figure 20-9 for typical modifications.

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the ramifications to other cycle components that are implied by thermal fatigue of the economizer inlet header. For this BTF, if poor shutdown conditions contributed to pitting in the economizer inlet tubes, similar pitting may occur in other economizer regions.

Volume 2: Water-Touched Tubes

20-19

20-20

Thermal Fatigue Economizer Inlet Header Tubes

Chapter 21 • Volume 2

Erosion/Corrosion in Economizer Inlet Headers Introduction This is one of three boiler tube failure mechanisms for stub tubes in economizer inlet headers. The other two are thermal fatigue, covered in Chapter 20, and flexibility-driven fatigue, covered in Chapter 26. As

all three can occur in the same general area, there is a possibility of misdiagnosing the active mechanism even through they are distinctly different. Distinguishing characteristics of the three are compared in Chapter 7, Volume 1.

1. Features of Failure and Typical Locations 1.1 Features of failure Failures by erosion/corrosion will be manifested as tube wastage on the inside surface. The surface appearance is that of “orange peel”. Progressive wall thinning leads eventually to failure by ductile overload. Figure 21-1 shows the typical appearance. A cross section through this failed tube is shown in Figure 21-2. Note the absence of protective magnetite on the inside tube surface. Some care is required to distinguish failures by erosion-corrosion from the other two mechanisms which can occur in the same location. Primary distinguishing features will be (i) the damage manifestation, (ii)

its location of origin, and (iii) the orientation. Manifestation: damage caused by thermal fatigue and flexibility-induced fatigue will be manifest as cracks, erosion-corrosion as wastage. Location: erosion-corrosion and thermal fatigue are ID-initiated; flexibility-induced cracking, OD-initiated. Orientation: thermal-fatigue will typically be oriented longitudinally (parallel) to the stub tube axis, flexibility-induced cracking is generally circumferential around the toe of the weld, and erosion-corrosion damage will depend on local flow characteristics, appearing finally as a ductile overload in the middle of the largest gouge on the inside surface. Additional detail can be found in Chapter 7, Volume 1.

Volume 2: Water-Touched Tubes

21-1

1.2 Typical locations The most common location is in economizer inlet header stub tubes nearest to the point of the feedwater inlet. Wastage by erosion/corrosion is found on the tube inside surface and generally extends to a length of approximately 10 to 12 cm (~ 4 to 5”) from the inside surface of the header. Figure 21-3 shows the typical location.

Figure 21-1. Appearance of an economizer inlet tube that failed by erosion-corrosion. The weld is the header nipple weld about 2 inches from the header. The "orange-peel" appearance of erosion-corrosion is clearly visible.

Figure 21-2. Cross-section of the failed tube shown in Figure 21-1. Note the almost complete absence of protective magnetite on the inside surface.

21-2

Erosion/Corrosion in Economizer Inlet Headers

a These tubes have lost tube wall thickness

a Tube failure location

Economizer inlet header stub tubes

Weld

Tube thickness

ID ligament spacing

CL tube

Header thickness

Figure 21-3. Cross section through the economizer inlet header and tubes showing locations of erosion-corrosion in the tubes. The tube bore shows the "orange-peel" appearance. This erosion-corrosion peaks after a distance of about 1-2 inches into the tube.

2. Mechanism of Failure

The mechanism is a combination of flow-induced corrosion (erosion/corrosion) caused by the local flow of water turning into the tubes from the inlet header, and corrosion caused by reducing feedwater conditions. Under normal operating conditions there is a balance between the growth of Fe3O4 (discussed at length in Chapter 2, Volume 1) and its flow-induced removal. Attack by an erosion/corrosion process results when the removal (dissolution) of the oxide occurs faster than its growth and thus leads to attack of the tube surface.

Ironically, this boiler tube failure mechanism occurs predominantly when the feedwater has very low O2 levels ( > 20 ppb). Under these conditions the feedwater becomes very reducing (typically 5 ppb to high purity water (cation conductivity < 0.15 mS/cm), provides a substantial reduction of erosion/corrosion transported feedwater corrosion products.

3. Possible Root Causes and Actions to Confirm The mechanism is caused by flowinduced corrosion under conditions of low oxygen activity, thus the primary actions to confirm the mechanism will involve reviewing chemistry records and monitoring systems to determine whether (i) reducing conditions exist ( 20 ppb N2H4 in the feedwater are present. High Fe levels at the economizer inlet (>> 5 ppb) may also indicate a feedwater chemistry problem.

4. Determining the Extent of Damage Thinning can be detected by ultrasonic testing of the stub tubes nearest to the location where the feedwater enters the economizer inlet header.

5. Background to Repairs, Immediate Solutions and Actions About the only immediate action is replacement of damaged tubes. However, the decision about whether to replace in-kind with carbon steel or upgrade to a material containing Cr such as T11 will depend upon (i) the desired unit life,

(ii) an estimate of the time-to-failure and (iii) perhaps most importantly, with larger considerations involved with other components in the feedwater train, as discussed in the next section.

6. Background to Long-Term Actions and Prevention of Repeat Failures Erosion-corrosion in the feedwater system because of the chemical environment indicated above can occur in a number of locations including: (i) economizer inlet header tubes, (ii) HP heater inlet tubes and tube sheet, (iii) the deaerator shell, and (iv) connecting pipework. Thus, while the immediate problem for the first can be overcome by using T11 tubes, the optimum approach for all areas is to address the root cause: the feedwater chemistry. The steps to be taken will depend on whether the feedwater train is allferrous or mixed metallurgy. For either case, a series of steps can be implemented over time as follows:

• For all-ferrous feedwater systems. The O2 scavenger (or N2H4) can be eliminated. As shown in Figure 21-4, the removal of N2H4 (or the O2 scavenger) will increase the oxidation-reduction potential (ORP) to between 0 and +100 mV, i.e., into the oxidizing range. The feedwater at the economizer inlet should be monitored for Fe, O2, cation conductivity, and ORP to ensure that the elimination of N2H4 produces the desired effect.

The preferred approach is to convert the unit to oxygenated treatment (OT). In this case, the O2 level at the economizer inlet depends on whether the unit is once-through or drum. For either of these approaches, the use of carbon steel for replacement tubing should suffice, as the oxidizing environment will eliminate the erosion-corrosion. This is illustrated in the tube cross section shown in Figure 21-5. This tube, from the same unit as manifested the failed tube shown in Figures 21-1 and 21-2, shows that after conversion to OT, a protective magnetite layer has been reestablished on the inside tube surface. • If the feedwater train has mixed metallurgy it is necessary to keep the reducing environment to protect the Cu-based heater tubes. In this case, the monitoring campaign is specifically to optimize the feedwater chemistry, i.e., to balance the feed of O2 scavengers with the Fe, Cu levels at the economizer inlet. In this case, it may be necessary to replace the tubes with a Cr-containing alloy such as T11.

Volume 2: Water-Touched Tubes

21-5

N2H 4

ORP mV 200

ppb

50

Fe ppb 16 14

100

40 ORP

12

0 30 -100

10 8

Fe 20 -200 N 2H 4

6 4

10 -300 2 -400

1 10

20

30 40 50 60 Time (days)

70 80 90

0

0

Figure 21-4. Change in the oxidizing-reducing potential (ORP) and total iron with the reduction in hydrazine in the feedwater. Note that ORP increases into the oxidizing regime and the iron levels decrease markedly, indicating lower erosion-corrosion rates. Source: D. Platt and D.A. Vinnicombe3

Figure 21-5. Economizer inlet tube removed from a unit which had previously experienced failure by erosion-corrosion (see Figures 21-1 and 21-2). This drum unit was converted to oxygenated treatment (OT); as a result, as can be seen here, the protective magnetite, previously lacking, has been restored. Importantly no further loss of wall thickness occurred after operating on OT for a year.

7. Case Study None for this mechanism.

8. References 1Dooley,

R.B., J. Mathews, R. Pate, and J. Taylor, “Optimum Chemistry for ‘All-Ferrous’ Feedwater Systems: Why Use an Oxygen Scavenger?”, Proceedings of the 55th International Water Conference, Pittsburgh, PA, October 31-November 2, 1994.

2Bates,

A.J., G.J. Bignold, K. Garbett, W.R. Middleton, D. Penfold, K. Tittle, and I.S. Woolsey, “The Central Electricity Generating Board Single-Phase ErosionCorrosion Research Programme”, Nuclear Energy, No. 6, December, 1986, pp. 361-370. 3Platt,

D. and D.A. Vinnicombe, “Operating of a Drum Boiler Without Hydrazine”, ESKOM, Johannesburg, South Africa, June, 1994.

21-6

Erosion/Corrosion in Economizer Inlet Headers

ACTIONS for Erosion/Corrosion in Economizer Inlet Header Tubes Two paths for the BTF team to take in the investigation of erosion/corrosion begin here. The goal of these actions is to see if further investigation of erosion/corrosion is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred in economizer inlet header stub tubes and erosion/corrosion is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit which indicates that there could be a future BTF by erosion/corrosion.

Action 1a: If a BTF has occurred in economizer inlet header stub tubes and erosion/corrosion is the likely mechanism.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by erosion/ corrosion:

➠ Determine whether the failure has

• Erosion-corrosion seen during

occurred in typical locations i.e. in those stub tubes nearest to the point of feedwater inlet, on the tube inside surface, and over the first 10 to 12 cm (4 to 5”) of tube length from the header. See Figure 21-3.

➠ Confirm that the macroscopic appearance of the failure includes the features shown in Figure 21-1, including: • Wall thinning • Orange peel surface appearance • Ductile final fracture

➠ Determine whether the protective magnetite is missing from the inside surface of tube, such as shown in Figure 21-2.

maintenance inspection in the HP feedwater heater inlet tubes or tube sheet, on the deaerator shell, or in connecting pipework. • Fouling of orifices or BFP impellers are other indications that excessive erosion-corrosion is taking place in feedwater systems. • Evidence of persistent reducing feedwater conditions ( 20 ppb N2H4 in the feedwater.

➠ If these indicators have been found, go to Action 3 which outlines the steps needed to confirm the influence of each.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

Volume 2: Water-Touched Tubes

21-7

Action 2: Determine (confirm) that the mechanism is erosion/corrosion. A failure has occurred which the BTF team has tentatively identified as being erosion/corrosion (Action 1a). Action 2 should clearly identify erosion/corrosion as the primary mechanism or point to another cause. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Analyze the macroscopic damage I. Does damage have features including: inside surface damage, wall thinning, orange peel appearance, thin-edged failure surfaces or pin-hole leaks?

➠ Analyze the macroscopic damage II. Is there longitudinal cracking on the inside tube surface, and particularly in the bore of the header or between bore holes?

Suspect that damage may be thermal fatigue (Chapter 20) or if damage is OD-initiated it is more likely to be flexibility-induced fatigue (Chapter 26).

Suspect thermal fatigue (Chapter 20).

➠ Probable mechanism is erosion/corrosion.

➠ Go to Action 3: Root Cause Determination.

Action 3: Determine root cause(s) of the erosion/corrosion A BTF failure has occurred and the mechanism has been confirmed as erosion/corrosion (Action 2), or a precursor has occurred (Action 1b). Although the underlying mechanism has contributors from both flowinduced corrosion and chemistry-induced corrosion, the controllable aspect is the latter. Therefore, the goal of this Action 3 is for the BTF Team to review the feedwater conditions to determine whether reducing conditions exist ( 20 ppb N2H4 in the feedwater. High Fe levels at the economizer inlet (>> 5 ppb) may also indicate a feedwater chemistry problem.

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Subject to access constraints, detection of erosion/corrosion is possible through ultrasonic examination for wall thinning. Chapter 9, Volume 1 provides some summary information.

21-8

Erosion/Corrosion in Economizer Inlet Headers

Action 5: Implement repairs, immediate solutions and actions Routine repairs and replacement practices are generally sufficient to deal with the immediate failure. A decision needs to be made whether to replace in-kind with carbon steel which will eventually result in a repeat failure, or to upgrade to a Cr-Mo steel such as T11. The decision will be based, in part, on the desired remaining life of the unit, the expected time to failure for carbon steel material, and on the choice of long-term actions discussed in (Action 6).

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. Long-term options will depend on whether the feedwater train is all-ferrous or mixed metallurgy. In the case of all-ferrous feedwater systems actions will consist of (i) eliminating oxygen scavengers such as N2H4 to raise the ORP into the oxidizing range, (ii) monitor Fe, O2, cation conductivity, and ORP to ensure desired effect, or (iii) convert the unit to oxygenated treatment. If these steps are taken then carbon steel will be sufficient as replacement materials. For mixed metallurgies, a reducing environment must be maintained to protect Cu-based heater tubes. A monitoring campaign should be used to optimize feedwater chemistry, i.e., balance the feed of oxygen scavengers with the levels of Fe and Cu measured at the economizer inlet. Replacing carbon steel tubes with a Cr-containing material such as T11 may be required.

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components implied by the presence of erosion/corrosion damage or its precursors. A primary consideration is the potential for development of erosion/corrosion problems elsewhere in the feedwater train, i.e., deaerators, high-pressure carbon steel feedwater heater tubes and tube sheets, and feedwater connecting pipework. This is very important because in each known instance of erosion-corrosion in economizer inlet tubes, a subsequent inspection of the deaerator has found erosioncorrosion on the shell. Erosion-corrosion of the pipework could be a safety problem. Actions such as outlined in Action 6 above will be useful for control in these components as well.

Volume 2: Water-Touched Tubes

21-9

21-10

Erosion/Corrosion in Economizer Inlet Headers

Chapter 22 • Volume 2

Sootblower Erosion (Water-Touched Tubes)

Introduction This chapter covers aspects of sootblower erosion that are particular to water-touched tubes. The main discussion, and thus additional detail on sootblower erosion can be found in the description of that mechanism in SH/RH tubes (Chapter 38, Volume 3).

1. Features of Failure and Typical Locations 1.1 Features of failure Macroscopic features of failure from sootblower-induced erosion will be those common to other erosive processes: (i) wall thinning caused by external tube surface wastage, and (ii) little or no ash deposits on the tube. Thermal fatigue cracking may also be present if there is water in the first steam flow from a sootblower. Since waterwall tubes are exposed across only one half of their circumference, the appearance of wall blower erosion will be somewhat different from the symmetric wastage flats seen for sootblower erosion in SH/RH tube. As a result, the erosion pattern will be angled to the tubes from the direction of the blow. Waterwall tubes subjected to sootblower erosion will have little or no

ash on the tube surface and, in common with other erosive processes, a distinguishing feature is the formation of fresh rust on tubes only a few hours after boiler washing which indicates that the protective scale has been removed. As erosion becomes more severe, tubes begin to thin, flattened areas develop, and eventually internal pressure leads to tube rupture. If the erosion is rapid, the failure may be thin-edged, a pin-hole shape or a long, “thin” blowout.

1.2. Locations of failure Typical failure locations for sootblower erosion in water-touched tubes are in a circular pattern around wall blowers. The corner effects are important.

Volume 2: Water-Touched Tubes

22-1

2. Mechanism of Failure Sootblower-induced erosion causes accelerated tube wastage by direct material removal, and removal of the fireside oxide also increases the fireside oxidation rate. The rate and extent of all of the erosive processes are affected by impacting particle or fluid velocity, angle of impact, particle composition and shape, and erosive resistance including compositional and temperature variations of the tube surface. The separate writeup on flyash erosion in Chapter 14 provides a detailed discussion of erosion in general. Sootblower erosion can be caused by either solid particles, essentially an acceleration of existing ash particles caused by the sootblowing operation, or by liquid “jets” when

condensed water gets into the sootblowing media. Either projectile results in characteristic wear patterns on the tube. For example, condensate in the blowing media will generally originate at the sootblower valve and continue along the blower path until the moisture is cleared from the blower. There can also be the appearance of gouges on the external tube surface where eddying of the steam occurs between adjacent tubes.1 Final failure generally occurs as a result of stress rupture when the thinned tube wall can no longer support the internal pressure; the final failure appearance is ductile and thin walled.

3. Possible Root Causes and Actions to Confirm Primarily, sootblower erosion is caused by improper operation and maintenance, such as incorrectly setting the blowing temperature, excessive use of the sootblowers, or malfunction of sootblowers. Condensation in steam or air supply lines can result from a variety of causes including2: (i) improper drainage, (ii) temperature changes, (iii) insufficient steam superheat temperature, and (iv) inadequate aftercooling of sootblower air compressors. Table 22-1 summarizes the possible root causes, actions to confirm that this is indeed the operative root

22-2

Sootblower Erosion (Water-Touched Tubes)

cause, immediate and long-term actions appropriate to prevent recurrence of failures by sootblower erosion. Actions to pinpoint the particular shortcoming that is causing the boiler tube wear include: (a). Visual examination can detect misalignment, etc. (b). Calibration and testing to measure key parameters such as blowing temperature and pressure or operation of moisture traps.

Table 22-1 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences Improper maintenance or operation of sootblowers: - Incorrect setting of blowing temperature (insufficient superheat) - Condensate in blowing media - Improper operation of moisture traps - Excessive sootblowing pressures - Improper location of sootblower - Misalignment of sootblower - Malfunction of sootblower - Excessive sootblowing

Actions to Confirm (a). Visual examination to determine obvious maintenance shortcomings or blower problems. (b). Calibration and testing to measure key parameters: - blowing temperature and pressure - operation of moisture traps - checking travel and sequence times.

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

• Evaluate the extent of wall thinning and erosion damage to determine whether repairs/ replacements are required. • Effect applicable repairs/replacements. See Chapter 11, Volume 1 for an overview of the applicable methods. • Avoid the use of temporary measures such as pad welding, shielding and/or coatings unless they are absolutely required to get unit to next scheduled outage. • Repair sootblower inadequacies and/or modify operation to prevent repeat failures.

• Determine the optimal period of sootblowing. It should not be simply a matter of once/shift or once/day. Fireside testing with probes to determine the rate of buildup of ash on tubes is useful. • Success has been achieved by having a sootblower maintenance team so that maintenance is performed on a regular basis and not on an as-needed basis. • Institute periodic visual examination and a program of calibration and testing of sootblower operation to prevent future failures. • Make needed modifications to hardware or operating procedures to prevent condensate from forming in blowing media.

4. Determining the Extent of Damage As with other erosion processes, visual examination may identify a serious sootblower erosion problem where significant wastage has occurred, or it may uncover indirect signs of a problem, such as rusted tube locations within a few hours of a boiler wash, indicating the removal of protective surface oxides.

blowers. An ultrasonic testing (UT) survey to detect wall thinning can determine the degree of damage that has occurred, a necessary precursor to rational repair/replace decisions. Chapter 9, Volume 1 provides an overview of the use of UT to detect wall thinning.

Damage should be localized to a circular pattern around the wall

Volume 2: Water-Touched Tubes

22-3

5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures Table 22-1 lists the available immediate actions and solutions. Where the loss of wall thickness is not significant, the damaged tubes should be evaluated as to whether they can remain in service, or whether they will require repair or replacement. Chapter 8, Volume 1, addresses remaining life calculations; procedures for the repair and replacement of tubes are discussed in Chapter 11, Volume 1. Palliative solutions such as the use of pad welds and spray coatings to increase the wear resistance of the tube, or shielding to protect local eroded areas, should only be used as emergency repairs to get the unit back on-line. They should be replaced at the next scheduled outage as these weld measures can themselves introduce further problems such as copper embrittlement or introduction of flow disruption if the weld bead penetrates to the inside surface. The pitfalls of these repair methods are covered in Chapter 11, Volume 1. It is also important to recognize that if the underlying sootblower problem is not addressed, the result will be ongoing repeat failures. Tube failures by sootblower erosion are preventable through improved maintenance or operation of sootblowers, and setting of optimized sootblowing parameters and frequency. Since a modern boiler can have over 100 sootblowers and they are subject to a harsh environment, they present a constant maintenance chore, to the point where some utilities keep dedicated crews

22-4

Sootblower Erosion (Water-Touched Tubes)

to perform sootblower maintenance on a regular basis.1 This is considered to be the optimum approach. Periodic checking of steam traps and drains are essential to keep water out of the system, as is regular blowdown of the sootblowing air lines for compressed air systems. Thermocouples have been used in steam trap lines to detect the presence of water. If condensate is forming in the blowing media then appropriate actions may include3: (i) allowing for steam warmup, (ii) completely draining the supply piping through thermal drain valves or impulse condensate drain valves, or (iii) the use of air dryers. Excessive blowing pressures will increase the likelihood for erosion to occur since the rate of erosion is a function of velocity to an exponent that ranges from 2 to 4, depending upon a number of factors. Long-term, the prevention of failures will also be found in a program that confirms, by visual inspection and calibration of the sootblowers and components, that the optimum operation continues to be achieved. Fireside testing with probes in the area to determine the rate of ash buildup on tubes will allow the optimum intervals between sootblowing to be set. Sootblowing should not be simply performed once/shift or once/day, but on an as-needed basis. Alignment problems, which can develop as a result of inadvertent forces during operation, can be found by inspection of tubes adjacent to sootblowers during planned maintenance outages.

7. Case Study None for this mechanism.

8. References 1Dooley,

R.B. and H.J. Westwood, Analysis and Prevention of Boiler Tube Failures, Report 83/237G-31, Canadian Electrical Association, Montreal, Quebec, November, 1983.

3Lamping,

G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985.

2Pack, R.W. and P.J. Resetar, State-of-the-Art Maintenance and Repair Technology for Fossil Boilers and Related Auxiliaries, Research Project 2504-1, Final Report CS-4840, Electric Power Research Institute, Palo Alto, CA, March, 1987.

Volume 2: Water-Touched Tubes

22-5

ACTIONS for Sootblower Erosion (Water-Touched) Two paths for the BTF team to take in the investigation of sootblower erosion begin here. The goal of these actions is to see if further investigation of sootblower erosion is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred and sootblower erosion is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by sootblower erosion.

Action 1a: If a BTF has occurred and sootblower erosion is the likely mechanism.

➠ Determine whether the failure has occurred near to a wall blower and shows wastage pattern consistent with erosion.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by sootblower erosion.

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit:

➠ If the BTF does not seem to have

• Burnishing or polishing of the tube in the blower path. This is typical of the early stages of an erosion problem.

features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

• Flat spots, ovality and formation of edges on straight tube sections typical of erosion in an advanced stage. • Fresh rust found on tubes following a boiler washing.

➠ Determine whether inspection or testing of sootblowers indicates such problems as (i) misalignment of sootblowers or sootblower stuck in one position, (ii) incorrect blowing temperature, (iii) presence of condensed water in blowing medium, or (iv) improper drainage.

➠ If one or more of these indicators have been found, go to Action 3 which outlines the steps to confirm the influence of each.

22-6

Sootblower Erosion (Water-Touched Tubes)

Action 2: Determine (confirm) that the mechanism is sootblower erosion. A failure has occurred which the BTF team has tentatively identified as being sootblower erosion damage (Action 1a). Action 2 should clearly identify sootblower erosion as the primary mechanism or point to another cause. The actions listed will be executed by visual examination of affected areas and removal of representative tube sample(s) for analysis.

➠ Evaluate the extent of damage. Is damage localized in a circular pattern around blower?

➠ Eliminate other erosive processes as candidates. Is the damage found in the blower path and/or obviously associated with sootblowing?

➠ Evaluate appearance of damage. Is wear manifested as wastage of the tube angled back toward the blower?

If spread over a wide area, problem may be generalized corrosion or generalized erosion; however continue with balance of flowchart to eliminate sootblower erosion as the cause.

Possibility that another erosive mechanism - flyash (Chapter 14), coal particle (Chapter 28), or falling slag (Chapter 29) is responsible for the damage. Should be able to distinguish by tube location; see discussion of these alternative erosion mechanisms if there is uncertainty in diagnosis

Damage may be flyash erosion if evidenced by smooth, polished wastage of the tube, particularly on the side facing into the gas flow.

Probable failure mechanism is sootblower erosion.

➠ Go to Action 3: Root Cause Determination

Volume 2: Water-Touched Tubes

22-7

Action 3: Determine root cause(s) of sootblower erosion A BTF failure has occurred and the mechanism has been confirmed as sootblower erosion (Action 2) or a precursor has been identified (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes of sootblower erosion, identify probable ones, and take those actions that are needed to confirm. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4: to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

Improper maintenance or operation of sootblowers: • Incorrect setting of blowing temperature (insufficient superheat) • Condensate in blowing media • Improper operation of moisture traps • Excessive sootblowing pressures • Improper location of sootblower • Misalignment of sootblower • Malfunction of sootblower • Excessive sootblowing

➠ (a). Visual examination to determine obvious maintenance shortcomings or blower problems. ➠ (b). Calibration and testing to measure key parameters: • blowing temperature and pressure • operation of moisture traps • checking travel and sequence times

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Evaluation will be based on detecting obvious signs of erosion and wall thinning. Ultrasonic testing can be used to measure the amount of wall thinning.

Action 5: Implement repairs, immediate solutions and actions The most important actions for the BTF team are to (i) make the tube repairs necessary to get the unit on-line and (ii) fix the underlying sootblower problem.

➠ Implement repairs or replacement of affected tubes as identified from the NDE Survey (Action 4). ➠ See Chapter 11, Volume 1 for summary of applicable tube repair techniques. ➠ Temporary pad welds, spray coating, or shielding may be used, but are not long-term solutions, as they will most likely lead to continual repairs. Plan to remove at the next outage.

➠ Perform sootblower maintenance as required.

22-8

Sootblower Erosion (Water-Touched Tubes)

➠ Modify sootblower operation as required. Most important is to develop an understanding of when sootblowing is needed and thus move away from an automatic sootblowing cycle. Other modifications to sootblowing operation may include changing sootblowing parameters, regular blowdown of sootblowing air lines to avoid condensation, and installation of thermocouples in steam trap lines to detect water.

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. Optimized longterm actions will include periodic inspection and calibration of the sootblower and its components. Modifications to prevent the development of condensate in the blower media may be indicated. As always, the required steps will be based on the clear identification of the underlying root cause (from Action 3).

Major Root Cause Influences

➠ Long-Term Actions

Improper maintenance or operation of sootblowers: • Incorrect setting of blowing temperature (insufficient superheat) • Condensate in blowing media • Improper operation of moisture traps. • Excessive sootblowing pressures • Improper location of sootblower • Misalignment of sootblower • Malfunction of sootblower • Excessive sootblowing

➠ Determine the optimal period of sootblowing. It should not be simply a matter of once/shift or once/day. Fireside testing with probes to determine the rate of buildup of ash on tubes is useful. ➠ Success has been achieved by having a sootblower maintenance team so that maintenance is performed on a regular basis and not on an as-needed basis. ➠ Institute periodic visual examination and a program of calibration and testing of sootblower operation to prevent future failures. ➠ Make needed modifications to hardware or operating procedures to prevent condensate from forming in blowing media.

Action 7: Determine possible ramifications/ancillary problems None for this mechanism.

Volume 2: Water-Touched Tubes

22-9

22-10

Sootblower Erosion (Water-Touched Tubes)

Chapter 23 • Volume 2

Short-Term Overheating in Waterwall Tubing Introduction Short-term overheating in watercooled tubes occurs because of abnormal coolant flow or excessive combustion gas temperature. As a result, the tube is subjected to excessively high temperature, often hundreds of degrees above design, which results in rapid failure.

Long-term overheating (creep) failures can occur in water-touched tubing but are much less prominent than in steam-touched tubes. As a result, only brief mention is made of long-term overheating in this chapter. A review of the more common case of creep in SH/RH tubing can be found in Chapter 32, Volume 3; short-term overheating of superheater/reheater tubing is covered in Chapter 36, Volume 3.

Volume 2: Water-Touched Tubes

23-1

1. Features of Failure and Typical Locations 1.1 Features of failure Damage by short-term overheating will display distinctive characteristics depending upon the temperatures experienced by the tube. The most general features for short-term overheating include (i) a considerable increase (> 5%) in the inside or outside diameter of affected tubes,

a) Delta iron

3000 (1649)

Delta iron + liquid

2800 (1538)

Liquid

2600 (1427) Austenite + liquid

2200 (1204)

1600 (871)

Ferrite + Austenite Austenite + Cementite

Pearlite + Ferrite

1000 (538)

Irons

Pearlite + Cementite 2.0% C

1400 (760)

Steels

32 (0)

0

0.5

Austenite, Eutectic + Cementite

Cementite, Pearlite + transformed Eutectic

4.3% C Eutectic (Ledeburite)

1800 (982)

1200 (649)

b)

Cementite Ledeburite

Austenite

2000 (1093)

0.80 % Eutectoid (Pearlite)

Temperature, °F (°C)

Cementite + liquid

Delta iron + Austenite

2400 (1316)

Cast irons

1 2 Carbon, %

3

4

5

1800 (982) Upper-critical shor t-term overheating 1600 (871) Inter-critical shor tter m overheating

1400 (760) A1 1200 (649)

Subcritical shor t-term overheating

0.80% C Eutectoid

Temperature, °F (°C)

A3

1000 (538) 826 (441)

Long-ter m overheating Nor mal tube design allowable

Steels 32 (0)

0.5

1

Figure 23-1 (a.) Equilibrium diagram for iron-iron carbide. (b.) Detail of equilibrium diagram, showing short-term overheating and long-term overheating regimes, along with the normal tube design allowable.

23-2

Short-Term Overheating in Waterwall Tubing

which, depending upon the underlying cause, may be over only a small circumferential arc of tube metal, and (ii) a ductile final failure showing a thin-edged fracture surface and “fish-mouth” appearance. Thickedged failure surfaces are also possible under two distinct overheating conditions, as discussed below. Pronounced microstructural changes are evident in this type of failure; those changes can be used as a diagnostic to estimate the tube temperature reached at burst.1 Three levels of “short-term” overheating have been classified depending on whether the temperature at burst was (i) below the lower critical temperature, A1, (termed subcritical shortterm overheating), (ii) between A1 and A3 - the upper critical temperature (termed intercritical short-term overheating), or (iii) above A3 (termed upper critical short-term overheating). Figures 23-1a and b show the equilibrium diagram for iron-iron carbide; they show the regions where these temperature ranges are operative. The defining characteristics of each of the three degrees of overheating are summarized in Table 23-1. The typical base metal has a microstructure consisting of ferrite and pearlite; the normal limit on its operating temperature is about 440°C (~ 825°F). As shown in Table 23-2, at this temperature, a minimum time to rupture of about 33 years is estimated for a tube operating at 55 MPa (8 ksi).2 As the temperature experienced by the tube increases, the time to failure decreases and the resulting microstructure upon cooling will change. At temperatures just above the design allowable for the material, the tube metal is in the long-term overheating regime where the dominant mechanism will be creep. If the temperature experienced by the tube is significantly higher, then the mechanism will become short-term overheating. For example, at a tube metal temperature of 510°C (950°F), the predicted life is only 70 days.

Table 23-1 Distinguishing Features of the Three Levels of Short-Term Overheating for Waterwall Materials Type of Overheating

Temperature Range

Fracture Surface

Extent of Tube Swelling

Fracture Mechanism

Microstructure (for ferritic tubing)

Hardness Characteristics

Subcritical shortterm overheating

> Design < Lower critical temperature, A1

Thin-lipped, “fish-mouth”

Considerable

Transgranular void formation by power law creep.

Ferrite and spheroidized pearlite or bainite.

Near that of original hardness.

Intercritical short-term overheating

Between the lower critical temperature, A1and the upper critical temperature, A3

Thin-lipped, “fish-mouth”

Considerable

Transgranular or mixed inter- and transgranular void formation by power law creep.

Ferrite, transformational products (pearlite, bainite, and/or martensite). Some spheroidized pearlite or bainite may also be present.

Variable, with hardness near transformation products being higher than the original.

Upper critical short-term overheating

> Upper critical temperature, A3

Thick-lipped, “fish-mouth”

Little

Inter- or transgranular creep fracture.

Near rupture, transformational products (pearlite, bainite, and/or martensite). Some ferrite may also be present.

Above original.

Table 23-2 Sample Minimum Rupture Times as a Function of Tube Temperature Temperature, °C (°F)

Minimum Time to Rupture at 55 MPa (8 ksi)

Design Allowable 441°C (826°F)

288,000 hrs. (33 yrs.)

455°C (~ 850°F)

100,000 hrs. (11 yrs.)

510°C (~ 950°F)

1686 hrs. (70 days)

565°C (~ 1050°F)

49 hrs. (2 days)

620°C (~ 1150°F)

2 hrs.

675°C (~ 1250°F)

0.14 hrs (8.5 minutes)

Lower critical temperature 737°C (1358°F)

0.01 hrs. (36 seconds)

> 737°C (1358°F)

< 30 seconds

The effect of even higher temperatures on expected life is shown in Table 23-2. The explanation for why the microstructural differences occur for different levels of overheating can be seen in that portion of the iron-iron carbide phase diagram pertinent to waterwall materials, Figures 23-1a and b. If the temperature before burst exceeds the A1 temperature, the pearlite will be transformed to austenite. If the A3 temperature is exceeded the original material will all be transformed to austenite and because of the quenching effects of tube rupture, upon examination the microstructure will consist of martensite and bainite. Thus the maximum temperature reached can be determined by the relative amounts of ferrite, bainite, and martensite in samples of the failed tubing.

For SA-210 Grade A-1. Source: S.R. Paterson, et al.2

Volume 2: Water-Touched Tubes

23-3

In the case of overheating to levels above the A3 temperature, the final fracture will generally be thickedged as indicated in Table 23-1. It will also show microstructural evidence of the complete transformation of the ferrite structure to austenite, and subsequent transformation to martensite or bainite on cooling. Thick-edged failures can occur as a result of overheating under a second set of conditions in waterwall and water-touched tubing: when there has been only slight overheating of the material and the failure is by long-term creep. Under these conditions, the thick-edged fracture will be found in conjunction with microstructural changes that include secondary cracking and intergranular cavitation. Care should be taken to distinguish such thick-edged failures, indicative of lower temperatures from those of the more typical failures by short-term overheating in waterwall tubes. See also lowtemperature creep of water-touched tubing in Chapter 24.

Table 23-3 Typical Locations for Short-Term Overheating Failures usually do not occur where the interruption of tube flow occurs, but in the higher (or highest) heat flux zone above. Failure locations might therefore include: Above places where flow has been partially or completed blocked by prior maintenance activities such as: • In tubes where weld repairs have been performed and weld spatter has been left in the tube, or • Where tools or repair materials have been left in tubing Above those orifices in lower waterwalls where blockage or restricted flow results from deposition of feedwater corrosion products across the orifice. Locations, such as horizontal tubing, which are affected when a “slug” of steam comes down the downcomer from the steam drum.

1.2 Locations of failure Short-term overheating in waterwall tubing results from the partial or full interruption of flow of water in tubes. Such flow interruptions are distinct from the local flow disruption that can be an integral part of underdeposit corrosion mechanisms. The most common locations of failures by short-term overheating are shown in Table 23-3. As noted in that table, the failure usually does not occur where the interruption of flow takes place but rather in the higher (or highest) heat flux zone. Regions of the waterwalls where overheating failures have occurred are highlighted in Figure 23-2.

23-4

Figure 23-2. Typical boiler locations where short-term overheating in waterwalls can occur.

Short-Term Overheating in Waterwall Tubing

2. Mechanism of Failure Once flow conditions are interrupted as described above, the normal cooling effects of the water no longer occur and the tube metal temperature rises rapidly. Pronounced local bulging occurs because of the decrease in strength of the tube metal. Failure can occur

within a matter of minutes as indicated in Table 23-2. The level of overheating (subcritical, intercritical or upper critical) will depend on the temperature reached by the tube and the length of time of the transient or operating feature that underlies the problem.

3. Possible Root Causes and Actions to Confirm 3.1 Introduction The primary causes of short-term overheating in waterwalls are related to sources of flow interruption. Three primary causes are (i) partial blockages due to maintenance activities, (ii) feedwater corrosion product deposits on orifices, and (iii) poor drum level control. One other secondary cause, the loss of coolant because of an upstream failure, is also discussed below. Table 23-4 summarizes the potential root causes, actions to confirm, immediate and long-term actions.

3.2 Partial blockage caused by maintenance activities This is a common cause of shortterm overheating failures in waterwall tubing. Typical problems stem from (i) tools left in tubes after a repair or (ii) improperly executed weld repairs that leave materials in the tube such as tube shavings from the weld preparation, tube preparation materials, or weld spatter. Such repair debris drops to the lowest point in the tube and causes partial or complete blockage. Failures by this root cause usually occur relatively soon after a unit overhaul or tube repair. Actions to confirm: (a). Check flows through tubes and/or for signs of obvious blockage such as tools remaining in tubes. (b). Review repair records to see whether the tube circuit was recently repaired.

3.3 Plugging of orifices by feedwater corrosion products Orifices are used in controlled circulation units to distribute flow evenly around the waterwalls, and in other units where flow control is needed. Feedwater corrosion products flow into the economizer inlet and can deposit at the orifices; the deposition mechanism is not completely understood, however, it appears to be controlled by zeta potential or streaming potential effects.3 The deposits accumulate with time, leading to a progressive decrease in flow in that particular waterwall tube until there is insufficient cooling water flow and a BTF then occurs by short-term overheating. The deposits are those typical of feedwater corrosion products, containing Fe and Cu and are very soft: they can usually be blown off. Usually if a tube fails by short-term overheating due to this cause, there will not be any deposits left on the orifice; however, it is necessary to check the conditions on all of the remaining orifices of the unit for evidence of deposition. The actions to confirm this cause are: (c). Inspect orifices in other lower waterwall areas for evidence of blockage. (d). Check records of pressure drop across boiler circulation pumps.

Volume 2: Water-Touched Tubes

23-5

Table 23-4 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

3.2 Partial blockage caused by maintenance activities: - tools left in tubes - poor maintenance practices, particularly improperly executed weld repairs such as where weld spatter is allowed to fall into a tube

• Institute repair and (a). Check flows through tubes, and/or for replacement as required. signs of obvious blockage. (b). Review repair records to see whether the tube circuit was recently repaired.

• Develop maintenance procedures and welding practices for optimized repair of tubing. See Chapter 11, Volume 1 for a summary of methods. • Institute additional quality control steps to prevent welding process errors.

3.3 Plugging of waterwall orifices by feedwater corrosion products

(c). Inspect orifices in other lower waterwall • Clean orifices. • Institute repair and areas for evidence of blockage. replacement as required. (d). Check records of pressure drop across boiler circulation pumps.

• Chemically clean. • Keep deposits to acceptable level. See guidance in Chapter 4, Volume 1. • Minimize feedwater corrosion products through control of chemistry particularly pH and O2 scavenger additions; Fe < 5 ppb and Cu < 2 ppb at economizer inlet. See also Chapter 3, Volume 1. • Monitor for effectiveness of chemistry control. • Monitor pressure drop across pumps on a continuous basis.

3.4 Poor control of drum level.

(e). Review operating records including drum level control.

• Institute repair and replacement as required. • Check drum internals and operation.

• Modify operating procedures or improve monitoring instrumentation to prevent recurrence of low flow and low drum water levels.

3.5 Loss of coolant because of upstream tube failure.

(f). Review of past BTF locations in relation • Institute repair and to current problem. replacement as required.

• Develop and institute a boiler tube failure repair philosophy that checks the whole waterwall tube circuit.

3.4 Poor control of drum level A large steam bubble can descend the downcomer and enter the waterwall tubes, particularly on startup, if the drum level is too low and a swell occurs. Carryover of steam down the downcomers can also be caused by low hydraulic pressure. These events will cause an instantaneous temperature increase and tube blow out. Monitoring of tube temperatures with thermocouples will not provide sufficient warning for operators to prevent the problem

23-6

because the effects are large and instantaneous. Horizontal tubes are most affected. Failures by this means generally manifest upper critical short-term overheating features because of the rapid heating and high temperatures reached. (e). A review of operating records including drum level and control will provide a confirmation of this root cause of short-term overheating.

Short-Term Overheating in Waterwall Tubing

3.5 Loss of coolant because of an upstream tube failure Action to confirm: (f) Compare the locations and timing of tube failures, which should indicate if there is a cause and effect relationship. The failure by short-term overheating either occurs at the same time as the primary BTF or when the unit is brought back online after repair. Bulging upstream of obvious tube ruptures should not be overlooked.

4. Determining the Extent of Damage areas need to be examined. As short-term overheating failures can occur over a very short period, it is important that inspection be sufficient to ensure that there are no additional tubes affected. That is, if the root cause is likely to be widespread, such as header debris that blocks a number of tubes, then extensive inspection may be required. If recent, improper weld repairs caused an isolated problem, then only those potentially affected

It is important that the entire tube length be checked for signs of bulging, particularly above the shortterm overheating failure. NDE methods to detect wall thinning, tube blockage, excessive internal tube deposits or tube swelling are appropriate. Chapter 9, Volume 1, provides an overview of the available methods and their use.

5. Background to Repairs, Immediate Solutions and Actions It is important for this mechanism that the problem is immediately identified. Both the specific root cause and the extent of the problem should be addressed before the unit goes back on line in order to avoid the predictable forced outage that will occur otherwise. Removal of all tube blockages and repair/replacement of all affected tube bends should be performed prior to the unit being returned to service. This is particularly applicable if the root cause is poor maintenance practices associated with whole waterwall panel replacements. If the failure was a result of the blockage of a waterwall orifice, all orifices should be inspected. If

additional deposition is found, the likely case, two options are possible: (i) the deposits can be blown off, which is very time consuming, or (ii) the unit can be chemically cleaned. It should be noted that the timing of periodic chemical cleans is often tied to a specified drop in pressure across the circulation pumps. If waterwall orifices are blocked, pressure readings may seem to indicate a dirty boiler overall and thus cause the utility to perform the chemical clean too early. If poor drum level control is at the root of the problem, procedures need to be established that allow the level to be maintained and ensure that swell is not large enough to allow the entry of steam bubbles into the downcomers.

Volume 2: Water-Touched Tubes

23-7

6. Background to Long-Term Actions and Prevention of Repeat Failures Longer term corrective actions involve instituting measures (i) to ensure that blockages of tubes do not occur by developing optimized maintenance and welding/repair procedures4, (ii) to ensure control of drum water levels via appropriate operating procedures, and (iii) to ensure coolant circulation. Keeping the boiler clean and deposits to an acceptably low level is a major preventive step for this as well as other failure mechanisms. Overall optimization of feedwater treatment is the key to prevention.

Chapter 3, Volume 1 summarizes key requirements, including recommended steps to reduce the levels of Fe and Cu at the economizer inlet by optimizing the treatments chosen. Guidance about chemical cleaning of waterwalls, including when it is needed and major steps in the process, can be found in Chapter 4, Volume 1. Redesign of tubing may be necessary, or relocation of inclined or horizontal tubing may be required, to prevent future failures.

7. Case Study None for this mechanism.

8. References 1French,

D.N., Metallurgical Failures in Fossil-Fired Boilers, John Wiley & Sons, Wiley-Interscience Publications, New York, 1993. 2Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993.

23-8

Short-Term Overheating in Waterwall Tubing

3Warwood,

B.K., “Fundamental Mechanisms of Deposition in Power Plants”, in R.B. Dooley and R. Pate, eds., Proceedings of the 4th International Conference on Fossil Plant Cycle Chemistry, held in Atlanta, Ga, September, 1994, Final Report TR-104502, Electric Power Research Institute, Palo Alto, CA, January, 1995, pp. 30-1. 4Stephenson,

G.G. and J.W. Prince, Guidelines on Fossil Boiler Field Welding, Research Project 2504-02, Final Report TR-101699, Electric Power Research Institute, Palo Alto, CA, January, 1993.

ACTIONS for Short-Term Overheating In Waterwall Tubing Two paths for the BTF team to take in the investigation of shortterm overheating in waterwalls begin here. The goal of these actions is to see if further investigation of short-term overheating is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a waterwall BTF has occurred and short-term overheating is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by short-term overheating.

Action 1a: If a waterwall BTF has occurred and short-term overheating is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of short-term overheating:

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by short-term overheating.

➠ Determine whether one or more of

➠ Review Figure 23-2 for typical boiler regions.

the following precursors has been found or is likely to have occurred in the unit:

➠ Review main text, section 1.2 and Table 23-3 for description of susceptible locations

• Pressure drop across circulation pumps indicating that the waterwall orifices are plugging.

➠ Confirm that the macroscopic appearance of the failure includes such features as: • Considerable increase in tube diameter (“swelling”) • Ductile failure with “fish-mouth” appearance.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

• Indications of high levels of feedwater corrosion products. • Any upstream tube leaks. • Indication that previous maintenance activities may have been poor. • Extensive waterwall/panel replacement. • Any outbreak of blocked tubes.

➠ These precursors can signal the potential for waterwall tube failures by a short-term overheating mechanism. If one or more has occurred, go to Action 3 which reviews root causes and outlines the steps to confirm the influence of each.

Volume 2: Water-Touched Tubes

23-9

Action 2: Determine (confirm) that the mechanism is short-term overheating. A waterwall tube failure has occurred which the BTF team has tentatively identified as being short-term overheating (Action 1a). Action 2 should clearly identify short-term overheating as the primary mechanism or point to another cause. The primary identifiers will be the ductile appearance of the failure and pronounced microstructural changes due to the overheated material.

➠ Confirm ductile nature of failure. Is fracture thin-edged? Does it have a “fish-mouth” appearance? Is it accompanied by significant swelling of the tube?

➠ Evaluate microstructure of the tube. Are microstructural transformation products consistent with one of the three levels of shortterm overheating listed in Table 23-1?

➠ Check hardness of material near fracture. Is material harder near fracture surface than in base material removed from the failure point?

An overheating failure manifesting signs of brittle failure may be caused by long-term overheating (creep).

Lack of transformation products may indicate another mechanism is operative, causing wall thinning. The subsequent ductile failure may not be because of overheating, but caused by low temperature creep (Chapter 24), or from wall loss by flyash erosion (Chapter 14), sootblower erosion (Chapter 22), etc.

If material is softened near point of failure, suspect a long-term overheating mechanism.

Probable mechanism is short-term overheating.

➠ Go to Action 3: Root Cause Determination

References to other sources of detailed information:

• Main text (this chapter) provides the background to mechanism and its development.

• Summary of the steps and methods of metallurgical investigation of boiler tube failures can be found in Chapter 6, Volume 1.

23-10

Short-Term Overheating in Waterwall Tubing

Action 3: Determine root cause of the short-term overheating A BTF failure has occurred and the mechanism has been confirmed as short-term overheating (Action 2) or a precursor to shortterm overheating has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.2 Partial blockage caused by maintenance activities: • tools left in tubes • poor maintenance practices, particularly improperly executed weld repairs such as where weld spatter is allowed to fall into a tube

➠ (a). Check flows through tubes, and/or for signs of obvious blockage. ➠ (b). Review repair records to see whether the tube circuit was recently repaired.

3.3 Plugging of waterwall orifices by feedwater corrosion products

➠ (c). Inspect orifices in other lower waterwall areas for evidence of blockage. ➠ (d). Check records of pressure drop across circulation pumps.

3.4 Poor control of drum level.

➠ (e). Review operating records including drum level control.

3.5 Loss of coolant because of upstream tube failure.

➠ (f). Review of past BTF locations in relation to current problem.

Volume 2: Water-Touched Tubes

23-11

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage or affected areas. Damage accumulation by this mechanism is not progressive, that is, it is unlikely that “a little” short-term overheating can be identified and then monitored over a period of time. Detection will be indirect, looking for evidence of a precursor to shortterm overheating - poor weld repairs, excessive waterside deposits, developing tube blockages, and the like.

➠ Identify all locations to be examined. Refer to Section 1.2 of main text and Figure 23-2 for typical locations. Because failure is very rapid, missed precursors will cause failure soon after unit re-start. Also, missed damage, farther up the tube, will fail on restart/ repressurization.

➠ Perform NDE survey to (i) measure tubes for indications of swelling, (ii) detect tube blockages (may require radiographic or UT methods), (iii) measure waterside deposits, if they are suspected of being a potential problem. A review of the basics of these NDE methods is provided in Chapter 9, Volume 1.

➠ Perform tube sampling to confirm results of NDE inspection and to determine the degree of damage by analysis of the transformation products in the microstructure. See additional detail on metallographic methods in Chapter 6, Volume 1.

➠ Perform visual inspection of remaining waterwall tube circuit for bulging and other signs of obvious tube distress.

➠ Use results interactively with Action 3 to confirm root cause.

➠ Go to Action 5: Implement Repairs, Immediate Solutions and Actions.

23-12

Short-Term Overheating in Waterwall Tubing

Action 5: Implement repairs, immediate solutions and actions The most important immediate actions for the BTF team are to (i) correct all sources of tube overheating before the unit goes back on line, and (ii) effect required tube repairs or replacements.

➠ Identify sources of tube blockages and remove, if applicable.

➠ Implement repairs or replacement of affected tubes as identified from the NDE Survey (Action 4). ➠ Ensure that the full extent of damage is removed as indicated by the results of the NDE examination. ➠ See Chapter 11, Volume 1 for summary of applicable tube repair techniques.

References to other sources of detailed information:

• Main text (this chapter) and Table 23-4 provide additional detail on repairs, immediate solutions and actions and relates them to underlying root causes.

• Guidance on chemical cleaning can be found in Chapter 4, Volume 1.

➠ Address blocked orifices by chemically cleaning waterwalls or by local removal of deposits (very time consuming).

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will be based on the clear identification of the underlying root cause (Action 3).

Major Root Cause Influences

➠ Long-Term Actions

Partial blockage caused by maintenance activities: • tools left in unit • poor maintenance practices, particularly improperly executed weld repairs.

➠ Develop maintenance procedures and welding practices for optimized repair of tubing. See Chapter 11, Volume 1 for a summary of methods; reference 4 for detailed discussion. ➠ Institute additional quality control steps to prevent welding process errors.

Plugging of waterwall orifices by feedwater corrosion products

➠ Chemically clean. ➠ Keep deposits to an acceptable level. See guidance in Chapter 4, Volume 1. ➠ Minimize feedwater corrosion products through control of chemistry particularly pH and O2 scavenger additions; Fe < 5 ppb and Cu < 2 ppb at economizer inlet. See also Chapter 3, Volume 1. ➠ Monitor for effectiveness of chemistry control ➠ Monitor pressure drop across pumps on a continuous basis.

Poor control of drum level.

➠ Modify operating procedures or improve monitoring instrumentation to prevent recurrence of low flow and low drum water levels.

Loss of coolant because of upstream tube failure.

➠ Develop and institute a boiler tube failure repair philosophy that checks the whole waterwall tube circuit.

Volume 2: Water-Touched Tubes

23-13

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components that might be implied by the presence of short-term overheating in waterwalls, or by its precursors.

23-14

Waterwall Short-term Overheating Aspect

Alert for Other Cycle Components

➠ Actions Indicated

Orifice deposits may indicate high levels of feedwater corrosion products

• Poor feedwater chemistry control (probably iron levels at the economizer inlet are > 10 ppb and Cu levels are > 5 ppb). • High Cu levels in deposits might indicate Cu deposition in HP turbine.

➠ Implement stricter cycle chemistry control program and instrumentation. See Chapter 3, Volume 1. ➠ Develop monitoring program to optimize feedwater chemistry and use of O2 scavengers. See Chapter 3, Volume 1.

Excessive deposits

Potential BTF by overheating and creep

➠ Sampling to determine nature and extent of deposit problem. See Chapter 6, Volume 1 for metallographic methods overview; Chapter 9 for sampling methods. ➠ Apply guidelines for chemical cleaning. See Chapter 4, Volume 1.

Short-Term Overheating in Waterwall Tubing

Chapter 24 • Volume 2

Low-Temperature Creep Cracking

Introduction Creep damage accumulates over time through the synergistic effects of stress and temperature. Creep damage is most commonly associated with high temperatures and as such is not typically a consideration in water-touched tubing. It is, however, one of the most common damage forms in high temperature SH/RH tubing as discussed in Chapter 32, Volume 3, and is life limiting for SH/RH tubes.

In this chapter a second manifestation of creep is discussed. It is primarily stress-driven and occurs in relatively low temperature tubing. It can be found in either water-touched or steam-touched tubing as it occurs over the temperature range 300420°C (~ 570-790°F). The tubing involved is typically located in the low temperature reheater, primary superheater, or economizer. This damage type has also been observed in cold bent steam piping.

Volume 2: Water-Touched Tubes

24-1

1. Features of Failure and Typical Locations 1.1 Features of failure Cracking typically initiates in high stress locations, notably the outside surface of tube bends. The direction of propagation will depend on the applied and residual stress fields; circumferentially oriented cracks are common, but longitudinal cracks on economizer bends have also been observed. The appearance of typical cracking is seen in Figure 24-1 that shows a reheater tube constructed with SA210-A1. The final failure is thick-edged.

Microscopically, cracking is found to be predominantly intergranular with significant branching and associated secondary cracking. Cracking may also be transgranular, particularly for higher stresses and lower hardnesses.2 A low-temperature creep failure may appear to be superficially similar to stress corrosion cracking or fatigue cracking. Low-temperature creep cracking will generally display evidence of grain boundary creep cavitation and the formation of creep

Figure 24-1. Low-temperature creep in the 135° bend of a reheater tube. Source: J. Hickey (ESB Ireland)1

24-2

Low-Temperature Creep Cracking

voids as shown in Figure 24-2; these features will not accompany stress corrosion cracking. Fatigue damage also will not show manifestation of microscopic creep damage. Observing a fatigue crack surface under scanning electron microscopy may also indicate the presence of either beach marks or ratchet marks typical of fatigue. Damage may accumulate over a long period. If it does, evidence of fracture surface oxide, extensive secondary cracking and creep cavitation will generally be seen. The crack shown in Figure 24-2 contains approximately 16 mm of oxide after approximately 40,000 hours of service.1

1.2 Typical locations By definition, failures are associated with high stress locations, most commonly in tube bends, where high residual stresses remain from fabrication, in conjunction with high service stresses. Failures will be found in higher temperature regions of economizer tubes, lower temperature regions of the reheater, and primary superheater. This mechanism does not occur in very low or very high temperature regions.

Figure 24-2. Micrograph of section through cracking. Indicative of low-temperature creep damage are the intergranular fracture, associated secondary cracking, grain boundary creep cavitation, and creep voids in the tube material. Source: J. Hickey (ESB Ireland)1.

Volume 2: Water-Touched Tubes

24-3

2. Mechanism of Failure The failure mechanism is creep cracking where the dominant factor is high stress. Failures by this mechanism have been preceded by a period of stable crack growth, which can be either transgranular or intergranular. Failures by this mechanism are typically in carbon-manganese or low-carbon steels, such as SA210 and occur in conjunction with sources of high stresses such as: • High residual stresses from the cold forming process, which may be particularly important in initiating cracks. • Enhanced membrane stresses caused by pipe ovality at bends. Figure 24-3 shows an ovality > 8% in a reheater tube which failed by low-temperature creep cracking. • High service stresses which may be most important in propagating the initial cracks. It has also been suggested that there is a correlation between failure probability and hardness of the tube material with harder bends being the more likely to fail. Figure 24-4 shows the results of hardness tests on two failed reheater tube bends, an intact tube bend and an intact straight section of adjacent tubing. The intact bend and straight tubing show hardnesses below about 200 HV whereas the failed bends have hardnesses above 220 HV.

24-4

Low-Temperature Creep Cracking

As a rule of thumb, bends with ovality greater than 8% or hardness greater than 220-240 HV are considered to be at the greatest risk. There is some evidence that high initial stresses, primarily residual stresses from forming, may be sufficient to initiate cracking. Although these stresses will relax with operating time, service stresses, if high enough can continue to drive the cracking to fail in later years.2 Pre-straining, such as induced by cold forming has two effects on creep crack propagation. The first effect is an increase in immobile dislocation density which will increase crack propagation rates. It will retard the relaxation of crack tip stresses and thus favor cavitation or microcracking. This effect will also inhibit the relaxation of stress concentrations at microstructural features such as carbides, inclusions, and grain boundaries. The second effect of pre-straining is the generation of internal stress fields due to the formation of dislocation pile-ups. Both of these effects of pre-straining can result in transgranular creep failure which is often observed during the crack initiation phase. Initiation has also been reported to often be associated with surface defects such as pipe-making laps or indentations from hammer blows.

Figure 24-3. Cross-section through a failed reheater tubes showing ovality in excess of 8%. Source: J. Hickey (ESB Ireland)1

Fracture Failed bend Intact bend Straight tube Failed bend

Hardness Hv (20kg) 240 220 200 180 160 140 120 100

0

50

250 150 100 200 Circumferential Position (Degrees)

300

350

Figure 24-4. Plot of tube hardness as a function of circumferential position. The failed tubes demonstrated maximum hardnesses that exceeded 220 HV; unfailed tubes had maximum hardnesses that were typically less than 200 HV. Source: J. Hickey (ESB Ireland)1

Volume 2: Water-Touched Tubes

24-5

3. Possible Root Causes and Actions to Confirm Unanticipated sources of high stress are at the root of BTF by this mechanism. A combination of high residual and/or service stresses, or high hardness are root causes. Actions to help pinpoint the most dominant underlying root cause include:

(b). Measure distortion (ovality) in susceptible locations. (c). Measure residual stresses. This action may not be definitive as relaxation during service or removal of the tube from the boiler may have lowered initial stresses.

(a). Perform an in-situ hardness test.

4. Determining the Extent of Damage Cracks will be surface-connected so that methods such as visual examination and magnetic particle examination may be used. However, access is generally difficult, and if the problem is likely to be widespread i.e., affecting a number of bends, the time to implement a tube-by-tube inspection will be pro-

hibitive. In these cases, a hydrotest has been suggested as the best means to determine whether there are leaking tubes. Note that this will not necessarily find all cracked tubes, only those with damage which has progressed to the leaking stage, or induced to be so by the hydrotest itself.

5. Background to Repairs, Immediate Solutions and Actions All affected tube bends should be replaced. Access will probably be difficult.

24-6

Low-Temperature Creep Cracking

6. Background to Long-Term Actions and Prevention of Repeat Failures Over the long-term, the key objective is to determine how widespread the problem is. Is it limited to a few bends? Is it limited to only one material? It is possible that the tubes subject to the lowest temperature will be out of the creep range altogether and the higher temperature tubes hot enough to permit thermally activated recovery to restore creep ductility?

A fracture mechanics analysis including an analysis of crack growth rates, if possible, in conjunction with stress analysis may enable a prediction of times to failure. Sizing the existing defects and determining the distribution of damage will be subject to the problems outlined in Section 4 above.

7. Case Study None for this mechanism.

8. References 1Personal

Communication from J. Hickey (ESB Ireland) to R.B. Dooley, February 16, 1995.

2Gooch,

D.J., “Creep Crack Growth in Cold Formed Carbon Manganese Steels at 320-380°C”, TPRD/L/2529/R83, Central Electricity Generating Board, November, 1983.

Volume 2: Water-Touched Tubes

24-7

ACTIONS for Low-Temperature Creep cracking Two paths for the BTF team to take in the investigation of lowtemperature creep cracking begin here. The goal of these actions is to see if further investigation of low-temperature creep cracking is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred and low-temperature creep cracking is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit which indicates that there could be a future BTF by lowtemperature creep cracking.

Action 1a: If a BTF has occurred and low-temperature creep cracking is the likely mechanism.

➠ Determine whether the failure has occurred in typical locations, i.e. does it appear to be associated with a weld, tight hair-pin bend, or other cold bend?

➠ Confirm that the macroscopic appearance of the failure includes such features as: • OD-initiated • Thick-edged failure

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) or for steam-touched tubing (Table 31-1) to pick a more likely candidate.

24-8

Low-Temperature Creep Cracking

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by low-temperature creep cracking. • Routine inspection and/or tube sampling finds an indication of significant hardness or ovality, particularly associated with tube bends.

➠ If this indicator has been found, go to Action 3 which outlines the steps to confirm its influence.

Action 2: Determine (confirm) that the mechanism is low-temperature creep cracking. A failure has occurred which the BTF team has tentatively identified as being low-temperature creep cracking (Action 1a). Action 2 should clearly identify low-temperature creep cracking as the primary mechanism or point to another cause. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Analyze the macroscopic damage. Does damage have features including: thick-edged failure surface, obvious OD-initiation, and association with a tight tubing bend?

➠ Analyze microscopic appearance of cracking. Do signs of creep damage accompany the cracking? Such evidence may include creep cavitation and creep void formation in the tube material. Is cracking intergranular?

Damage may not be low-temperature creep cracking. If final failure is thin-edged suspect an external wastage mechanism as caused by flyash erosion (Chapter 14). If damage is ID-initiated, suspect corrosion fatigue (Chapter 13).

Cracking is more likely to be either (i) a corrosion-assisted mechanism, such as corrosion fatigue in waterwalls (Chapter 13) or stress corrosion cracking (Chapter 37, Volume 3) or (ii) caused by fatigue (see Chapter 26, Volume 2 or Chapter 39, Volume 3).

Probable mechanism is lowtemperature creep cracking.

➠ Go to Action 3: Root Cause Determination

References to other sources of detailed information:

• Summary of the steps and methods of metallurgical investigation of boiler tube failures can be found in Chapter 6, Volume 1.

Volume 2: Water-Touched Tubes

24-9

Action 3: Determine root cause(s) of the low-temperature creep cracking A BTF failure has occurred and the mechanism has been confirmed as low-temperature creep cracking (Action 2) or a precursor has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes of low-temperature creep cracking, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

Unanticipated sources of high residual or service stress and/or high hardness material.

➠ (a). Perform an in-situ hardness test. (Hardness is over 220 HV). ➠ (b). Use a visual examination or dimensional measurement to detect tube ovality (> 8%). ➠ (c). Measure residual stresses with rosette strain gauges prior to removing the tube bend from the boiler.

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Subject to access constraints, detection of low-temperature creep cracking is possible through routine NDE methods that detect surface cracking such as magnetic particle and visual examination. Chapter 9, Volume 1 provides additional information. Hydrotesting may be required, if damage is widespread, to ensure that all leaking tubes have been detected.

24-10

Low-Temperature Creep Cracking

Action 5: Implement repairs, immediate solutions and actions Tube replacement is needed for failures associated with bends. Pad welding/grinding should not be employed. Thereafter, most actions should be considered for the longer-term (Action 6).

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will be based on the clear identification of the underlying root cause (Action 3), a knowledge of the extent of affected material (Action 4), and an economic evaluation to ensure that the optimum strategy has been chosen.

Major Root Cause Influences

➠ Long-Term Actions

Unanticipated sources of high residual or service stress and/or high hardness material.

➠ Perform fracture mechanics analysis to determine probable replacement needs.

Action 7: Determine possible ramifications/ancillary problems None for this mechanism.

Volume 2: Water-Touched Tubes

24-11

24-12

Low-Temperature Creep Cracking

Chapter 25 • Volume 2

Chemical Cleaning Damage: Waterwalls

Introduction Two types of improper chemical cleaning of waterwalls can result in an increased incidence of boiler tube failures. The first, is where excessive deposits are left as a result of incomplete cleaning or flushing. Such residual deposits can lead to failures by underdeposit corrosion mechanisms via (i) hydrogen damage given a source of acid contamination, (ii) acid phosphate corrosion in the presence of mono- and/or and excessive di-sodium phosphate, or (iii) caustic gouging with a concentration of sodium hydroxide.

A second form of boiler tube damage is generalized corrosion caused by one or more improper operations during the cleaning process. This chapter provides some general comments about the effects on tubes of these types of chemical cleaning errors. A brief review of the chemical cleaning process for waterwalls and economizer tubing is included in Chapter 4, Volume 1. Here the focus is on damage done to tubes as a direct result of the chemical cleaning process.

Volume 2: Water-Touched Tubes

25-1

1. Features of Failure and Typical Locations The primary problem is chemical attack, manifested as generalized corrosion of affected tube surfaces, Figure 25-1. Depending upon the severity of the problem, the appearance of the affected tubes will be as jagged, rough, straight-sided or undercut pits (see Figures 25-2a

and b) or as generalized wall thinning that can occur around the entire tube circumference. If the damage is found relatively soon after chemical cleaning then pitting will be found to be relatively free of oxides and deposits.1

Figure 25-1. Internal surface of failed tube exhibiting a rough pitted appearance typical of acid cleaning corrosion (MAG:1.2X) Source: S.R. Paterson, et al.1

25-2

Chemical Cleaning Damage: Waterwalls

Figure 25-2. Cross sections of the pitted region revealing straightsided and undercut pit morphologies associated with acid cleaning corrosion. Note also the absence of deposits within the pits, also characteristic of acid cleaning corrosion. Source: S.R. Paterson, et al.1

Volume 2: Water-Touched Tubes

25-3

2. Mechanism of Failure The failure is corrosion of the base tube metal by the solvent used in, or remaining after, chemical cleaning, particularly for hydrochloric acid cleans.

3. Possible Root Causes and Actions to Confirm The root causes of this form of damage are one or more improper operations in the chemical cleaning process including: • Use of an inappropriate cleaning agent • Excessively strong acid concentration • Excessively long cleaning times • Too high a temperature • Failure to neutralize, drain and rinse after cleaning • Breakdown of inhibitors as a result of temperature excursions. Inhibitors are used to prevent corrosion of the base tube material after the scale and deposits have been removed. Most inhibitors have a maximum temperature above which they will decompose or lose effectiveness.2 Monitoring of the chemical cleaning process by means of a side loop, particularly for Fe in the cleaning solution, can access corrosion by this mechanism. If, at some later

4. Determining the Extent of Damage The primary means to assess the extent of damage will be measurements at suspect locations for evidence of wall thinning. Ultrasonic test methods for doing so are discussed in detail in Chapter 9, Volume 1.

25-4

Chemical Cleaning Damage: Waterwalls

date, tube failures are suspected to have been caused by this mechanism, actions to confirm consist of: (a). Review of chemical cleaning procedures, chemical pumping systems, and chemical control logs. Items of particular concern are those listed above which would lead to significant damage. (b). Review of cycle chemistry monitoring records to detect a pH depression on start-up of the unit after cleaning, indicating improper rinsing of acid from the unit. (c). Sampling of affected tubes to allow examination of the inside surface for evidence of generalized corrosion. Part of an optimized procedure for chemical cleaning will be sampling of selected tubes to confirm the efficacy of the cleaning process. These samples can be used to determine whether excessive damage has accumulated. Wall thickness measurements can provide a quick screening as to whether excessive tube corrosion has occurred.

5. Background to Repairs, Immediate Solutions and Actions Immediate solutions will consist of repairing or replacing the damaged tubes (see Chapter 11, Volume 1 for a discussion of these methods) and immediate chemical clean followed by proper neutralizing and rinsing.

6. Background to Long-Term Actions and Prevention of Repeat Failures If procedures were inadequate to control the chemical cleaning process, including sampling for deposit extent and composition, solvent choice, planning, cleaning procedures, post-cleaning inspection, etc., then proper procedures must be established as a means to prevent recurrence of the problem. See summary of guidance in Chapter 4, Volume 1 or other available reference.3

7. Case Study None for this mechanism.

8. References 1Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993.

3Bartholomew,

R.D., W.E. Chesney, R.D. Hopkins, J.S. Poole, J.W. Siegmund, J.P. Williams, and S. Yorgiadis, Guidelines for Chemical Cleaning of Fossil-Fueled SteamGenerating Equipment, Research Project 2712-06, Final Report TR-102401, Electric Power Research Institute, Palo Alto, CA, June, 1993.

2Lamping,

G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985.

Volume 2: Water-Touched Tubes

25-5

ACTIONS for Waterwall Chemical Cleaning Damage Two paths for the BTF team to take in the investigation of chemical cleaning damage begin here. The goal of these actions is to see if further investigation is warranted or whether another BTF mechanism should be investigated.

Action 1a: If a BTF has occurred and chemical cleaning damage is the likely mechanism.

➠ Determine whether damage appears to be generalized corrosion over a large number of affected surfaces.

➠ Confirm that the macroscopic appearance of the failure includes extensive pitting (jagged, rough, straight-sided or undercut) or is manifested as generalized wall thinning. If found relatively soon after chemical cleaning, pits might be relatively free of deposits of oxides.

➠ Follow Action 1a: If a BTF has occurred and chemical cleaning damage is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF because of chemical cleaning damage.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by chemical cleaning damage.

➠ Determine whether there may have been one or more of the following during a recent chemical cleaning operation: • Use of an inappropriate cleaning agent • Excessively strong acid concentration • Excessively long cleaning times • Too high a temperature • Failure to neutralize, drain and rinse after cleaning • Breakdown of inhibitors.

➠ Determine whether monitoring during the chemical cleaning process indicated that the level of Fe in the cleaning solution continued to increase instead of leveling out to indicate that the chemical clean had finished.

➠ Determine whether one of more of the following precursors has been found or is likely to have occurred in the unit: • Wall thinning found during a routine inspection • Excessive waterside deposits found relatively soon after a chemical clean.

➠ These precursors can signal the potential for BTF because of an improper chemical clean. If one or more has occurred, go to Action 3 which reviews root causes and outlines the steps to confirm the influence of each.

25-6

Chemical Cleaning Damage: Waterwalls

Action 2: Determine (confirm) that the mechanism is chemical cleaning damage. A failure has occurred which the BTF team has tentatively identified as being caused by chemical cleaning (Action 1a). Action 2 should clearly identify this as the primary mechanism or point to another cause. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Review extent of damage (in conjunction with Action 4). Is damage manifested as generalized attack over affected tube surfaces?

➠ Characterize nature of damage. Does damage have the appearance of jagged, rough, straightedged or undercut pits?

Very localized attack may be via a pitting mechanism, also review underdeposit corrosion mechanisms for “gouging”-like appearance. See discussion of hydrogen damage, acid phosphate corrosion and/or caustic gouging in Chapters 15-17 respectively..

Compare to pitting damage (Chapter 27).

Probable mechanism is chemical cleaning damage.

➠ Go to Action 3: Root Cause Determination

Action 3: Determine root cause of chemical cleaning damage A BTF failure has occurred and the mechanism has been confirmed as chemical cleaning damage (Action 2) or a precursor has occurred (Action 1b). The goal for this Action 3 is to pinpoint the particular source of damage and in conjunction with Action 4, to determine the extent of the affected area so that appropriate repairs can be implemented.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

One or more improper operations during the chemical cleaning process

➠ (a). Review chemical cleaning procedures, chemical pumping systems, and chemical control logs for evidence of (i) inappropriate cleaning agent, (ii) excessively strong solvent concentration, (iii) excessively long cleaning times, (iv) failure to properly neutralize, drain and rinse after cleaning, (v) levels of Fe that continue to increase in cleaning solution, (vi) breakdown of inhibitors, (vii) too high a temperature. ➠ (b). Review cycle chemistry monitoring records for pH depression on startup, indicating a hideout and return of acid. ➠ (c). Sample affected tubes for evidence of generalized corrosion.

In parallel, determine the extent of affected area, Action 4. Volume 2: Water-Touched Tubes

25-7

Action 4: Determine the extent of damage or affected areas In parallel with Action 3, the BTF team should determine the extent of the affected area.

➠ Determine the areas and extent to be inspected from a review of the chemical cleaning operation and likely problems.

➠ Perform NDE. Ultrasonic examination of suspect locations to detect wall thinning extent. See Chapter 9, Volume 1 for additional background on methods.

➠ Selective tube sampling to confirm results of NDE survey.

➠ See Action 5: Implement Repairs, Immediate Solutions and Actions.

Action 5: Implement repairs, immediate solutions and actions ➠ Implement repairs or replacement The primary immediate objective for the BTF team is to ensure that a proper cleaning occurs and that tube damage which has compromised minimum wall requirements is repaired or the affected tubes replaced.

of affected tubes as identified from the NDE survey (Action 4). Immediate repair will be required for those tubes which fall below minimum wall. Over the longer term it will be necessary to replace other roughened surfaces as they provide sites of flow disruption that can lead to underdeposit corrosion failures. ➠ See Chapter 11, Volume 1 for summary of applicable tube repair techniques.

25-8

Chemical Cleaning Damage: Waterwalls

Action 6: Implement long-term actions to prevent repeat failures Tubes which still have wall thickness in excess of minimum wall requirements, but which have been damaged, need careful attention over the long term. Specifically, the roughened or pitted tube internal surface provides a site for flow disruption, which in conjunction with a source of concentration, can lead to BTF by an underdeposit corrosion mechanism (see separate discussions of hydrogen damage, acid phosphate corrosion, and/or caustic gouging in Chapters 15-17). Those sites can also act as points of initiation for corrosion fatigue (Chapter

13). Careful monitoring of cycle chemistry is thus also indicated. A tube replacement program to address the worst damage will probably be required. If procedures or controls for the chemical cleaning processes were inadequate they should be modified to include: sampling for deposit extent and composition, solvent choice, planning, cleaning procedures, post-cleaning inspection. See the summary provided in Chapter 4, Volume 1 or other available guidance.3

Action 7: Determine possible ramifications/ancillary problems There is a concern here with volatile carryover, especially if the root cause of the problem was an excessive temperature. The chemical “carries over” and inadvertently “cleans” the superheater. Since these steam circuits are not rinsed, material removed by the chemical clean remains in the tubes, causing blockages and subsequent failures by short-term overheating. This type of damage is discussed in more detail in Chapter 36, Volume 3. Reference may be made to the separate Chapter 43, Volume 3, that

Volume 2: Water-Touched Tubes

25-9

25-10

Chemical Cleaning Damage: Waterwalls

Chapter 26 • Volume 2

Fatigue in WaterTouched Tubes

Introduction This chapter discusses boiler tube failures caused essentially by fatigue without regard to the source of the cyclic stresses e.g., thermal, mechanical, vibration etc. All the failures in this category are manifested by cracking that is OD-initiated, are predominantly transgranular, and are not primarily attributable to the effect of either the fireside or water environment. That is, for the failures described here, the effect of environment is so small that even if all environmental effects were removed, the failure would still occur. Cyclic stresses are obviously also an important contributor to cracking mechanisms that do have a major contribution from the environment, such as corrosion fatigue (Chapter 13), and stress corrosion cracking (Chapter 37, Volume 3), and creepfatigue interactions. This chapter provides some means to distinguish among these effects; however, it

concentrates on fatigue as the dominant failure mechanism. Two specific BTF that are caused predominantly by thermal fatigue, and are thus similar to those described in a more generic way in this chapter, are covered in separate chapters: (i) failures at the tube stub in economizer inlet headers (Chapter 20), and (ii) circumferential cracking of waterwall tubes in supercritical units (Chapter 19). These two mechanisms merit an independent look because of (i) their relatively widespread occurrence in numerous units during the past few years, (ii) the specific locations affected, and (iii) the specific long term solutions. Two general classifications of BTF caused by fatigue are reviewed in this chapter: tubing-related and header-related. A separate chapter addresses fatigue failures in steam-touched tubing (Chapter 39, Volume 3).

Volume 2: Water-Touched Tubes

26-1

1. Features of Failure and Typical Locations 1.1 Features of failure Fatigue failures, either tubing- or header-related, generally result in thick-edged failures. Figure 26-1 shows the fatigue failure of a wall tube from a once-through boiler. A micrograph taken through the fatigue crack, Figure 26-2 illustrates the predominant characteristic that the cracking tends to be straight and transgranular. The appearance of beach marks or ratchet marks is typical, although they may be totally obliterated by oxidation. Cracking

Figure 26-1. Fatigue failure of a wall tube from a once-through boiler. The tube is a finned tube and forms part of a man-hole door opening. The failure initiated on the O.D. of the tube, at the toe of the fin/tube weld. The tube is shown here with the fin removed by mechanical grinding to allow removal from the boiler. Source: J. Hickey, Irish Electricity Supply Board

Figure 26-2. Micrograph taken through the crack showing transgranular cracking typical of fatigue. Source: J. Hickey, Irish Electricity Supply Board

26-2

Fatigue in Water-Touched Tubes

can be intergranular when occurring in creep damaged materials. Depending upon the service conditions, the cracks may be filled with oxide. Fatigue failures originate on the outside surfaces of tubes in contrast to corrosion fatigue failures which are ID-initiated. Table 26-1 lists several characteristics that distinguish corrosion fatigue in particular from thermal- or mechanical-fatigue in watertouched tubing.

1.2 Typical locations Tubing-related failures are generally found in locations associated with (i) attachments, particularly solid attachments or jammed sliding attachments, or (ii) bends in tubing. At attachments, failures by fatigue are often associated with weldments, particularly in cases where the weld or condition of the attachment does not allow for thermal expansion. Examples are (i) the end of the membrane of waterwall tubing, either in the lower slope region near the ash hopper or at the top of the rear wall at the entrance to the rear gas passage, or (ii) at tie bars, K bars or beams. Figure 26-3 shows typical spacers and sliding attachment details where fatigue in watertouched tubes can occur. The force of the failure may cause the attachment to be ripped from adjacent tubes. Also, if the attachment is large, then it can be corroded (fireside) indicating that it reached severe temperatures. At bends in the tubing, particularly “U”-bends, failures can be initiated at the intrados, extrados or neutral axis as shown in Figure 26-4. Header-related fatigue failures occur most frequently at the end of the header and are related to flexibility of the header and/or thermal expansion problems. A common location is at the nipple or stub tube welds as shown in Figure 26-5. In watertouched tubes a primary location of concern is the economizer inlet header where flexibility-induced stresses are at the root of the problem. Failures have also been experienced near the welded connection to the lower waterwall inlet header.1 Failures from gas-flow-induced vibration fatigue can occur in welded tie-type spacers between vertical waterwalls and horizontal economizer tubing (Figure 26-3), or other water touched tubes. Poor weld geometry or execution such as overfill, poor fillet weld profile, or undercut, may also lie at the root of a fatigue problem.

Table 26-1 Distinguishing Corrosion Fatigue from Mechanical Fatigue of Water-Touched Tubes Characteristic

Corrosion Fatigue

Mechanical Fatigue

Initiation Location

• Inside surface

• Outside surface

Proximity to stress riser at outside surface, such as the toe of a weld

• Possible, especially if the weld is the attachment weld of the tube to a non-pressure part. Cracking may be slightly removed from the exact toe position.

• Typical condition exactly at the toe of the weld.

Appearance when grinding out defect

•Damage appears to increase as grinding proceeds from outside surface. Flaw increases in size towards its initiating site at inside surface.

• Damage decreases as grinding progresses from outside surface inward.

Timing of appearance

• Usually later in life of boiler

• Usually earlier in life of boiler

a) C and T ÒSlidingÓ Spacer

b) Large Rigid Spacer or Beamer

c) Sliding or Rigid Support

Cracks Cracks Cracks Economizer circuit

Vertical waterwall

Figure 26-3. Typical spacers or sliding supports where fatigue in water-touched tubing can occur.

Volume 2: Water-Touched Tubes

26-3

Neutral axis

Extrados Intrados Figure 26-4. Three possible locations for tubing-related fatigue failures in tight 180° bends.

a) Cold

b) Hot

Outlet header Outlet tubes

Cracks at weld

...Water wall...

c) Close-Up of Tube Attachment to Header and Crack Location

Crack

Figure 26-5. Schematic illustrating failures caused by inflexibility to the movement between header and waterwall.

26-4

Fatigue in Water-Touched Tubes

2. Mechanism of Failure Fatigue is the phenomenon of damage accumulation caused by cyclic or fluctuating stresses. It is manifested as the initiation and stable propagation of a crack. Final failure ensues when a critical crack size is reached and failure occurs by fracture or overload. Fatigue is dependent upon the frequency and magnitude of the stress cycles and is generally independent of stress duration. With the exception of flowinduced vibration failures, these fatigue failures are low-cycle and usually related to the operation or cycling of the unit. Obviously, at high temperatures where creep fatigue becomes the dominant mechanism, there is dependence on stress duration. Stress cycling may be induced mechanically or thermally.

Fatigue in boiler tubes is generally driven by excessive local stress levels, typically in welded connections. Such connections may be tube attachments and supports or at header nipple and stub welds, as described above. Stresses may be induced by (i) excessive (unanticipated) mechanical loads, (ii) restrained thermal expansion, (iii) vibration such as produced by the flow of combustion gases, or (iv) as a result of a poor weld geometry, such that, although the global stresses are within the design, they are locally excessive because of the geometry of the joint. In many cases, the excessive stresses, particularly thermally-induced, are produced by unit cycling or two-shift operation, leading to conditions that did not arise during baseload operation.

Volume 2: Water-Touched Tubes

26-5

3. Possible Root Causes and Actions to Confirm 3.1 Introduction Unanticipated or excessive cyclic stresses (strains) are the primary cause of fatigue failures in boiler tubes. Three basic sources of excessive strains were discussed above: those caused by constraint of thermal expansion, excess mechanical loads, or flow-induced vibration. The first of these is responsible for the bulk of tube failures experienced by this mechanism. Excessive stresses can also arise in welded connections if the expansion of the welded joint is inadequate. These four causes are discussed below. A summary of the appropriate actions to confirm, immediate and longer term actions for each are summarized in Table 26-2.

3.2 Excessive strains caused by constraint of thermal expansion The typical locations for the occurrence of this problem were discussed above. Actions to confirm the influence of this root cause include: (a). Visual examination to detect distortion or bending of adjacent tubes. Tubes at headers may also be physically distorted or even pulled away. (b). Strain gauging of susceptible locations to determine the level of strains experienced during thermal excursions of the unit, such as cycling operation. (c). LVDT measurements to monitor the relative movement of the header/ tube during transients.

26-6

Fatigue in Water-Touched Tubes

3.3 Poor design and/or manufacture giving rise to excessive mechanical stresses Actions to confirm this root cause include: (d). Strain gauging of susceptible locations to indicate unexpected mechanical loads in excess of design assumptions. (e). as in (c) above. (f). For tight, hair-pin bends, may need to determine if residual stresses are high.

3.4 Vibration (flue gasinduced) by direct flow or vortex shedding Actions to confirm include: (g). Metallurgical examination and location usually provide confirmation. Also if vibration is occurring, it is extremely noisy and thus obvious.

3.5 Poor welding, particularly poor geometry of final joint Actions to confirm this root cause include: (h). Analysis of weld profile and quality to determine whether poor welding was a likely contributor to the failure.

Table 26-2 Major Root Cause Influence, Confirmation and Corrective Actions Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and and Prevention of Repeat Failures

(a). Visual examination for distortion or bending in adjacent tubes. (b). Strain gauging of suspect locations to evaluate strains during unit starts and cycling operation. (c). LVDT measurements to monitor the relative movement of the header/tube during transients.

• Identify similar damaged locations. • Repair/replace affected tubes. See Chapter 11, Volume 1 for an overview of methods.

• Evaluate modifications to attachment design or to header/tube connection to reduce stress levels. • Institute periodic inspection program, particularly of susceptible header locations in units that are being, or to be, cycled. • Improve header/tube flexibility and confirm with LVDT.

3.3 Poor design and/or manufacture giving rise to excessive mechanical stresses.

(d). Strain gauging to measure actual strains experienced at the local area during operation. (e). As in (c) above. (f). For tight, hair-pin bends, determine whether residual stresses are high.

• As above.

• Evaluate modifications to attachment design or to header/tube connection to reduce stress levels. • Institute periodic inspection program, particularly of susceptible header locations in units that are being, or to be, cycled.

3.4 Vibration (flue gasinduced) by direct flow or vortex shedding.

(g). Metallurgical examination to determine high cycle fatigue. (h). Visual and microscopic examination of weld quality.

• As above.

• Evaluate and install modifications such as snubbers or vibration restraints to reduce stresses induced by vibration.

• As above.

• Institute program of weld quality control based on guidelines such as provided in reference 2.

Major Root Cause Influence

3.2 Excessive strains caused by constraint of thermal expansion.

3.5 Poor welding, particularly poor geometry of final joint.

4. Determining the Extent of Damage Detection of fatigue damage once a crack has initiated is well established by a number of methods including liquid penetrant, magnetic particle, eddy current, ultrasonic testing, and radiography. Since fatigue cracks in boiler tubes are OD-initiated, identification of cracked locations is generally

straightforward, within the constraints of access. The analysis of fatigue is likewise well established and sufficient materials data exist for all common constructional materials to allow complete analysis of the expected life of components subject to fatigue.

Volume 2: Water-Touched Tubes

26-7

5. Background to Repairs, Immediate Solutions and Actions Cracks can usually be ground out and rewelded as a temporary measure. Tight “U” bends should be replaced. Most of the rest of the actions can be taken over the longer term.

6. Background to Long-Term Actions and Prevention of Repeat Failures Longer-term actions will consist of modifications to prevent a recurrence of the problem. For example, if the problem was at an attachment and strains were caused by differential thermal expansion, modifications to the attachment design should be made to reduce the stress level. It is important to establish clearly that the stresses in the new design are in fact lower than those of the original design.

In the case of vibration-induced strains, solutions are typically found in the installation of vibration baffles to break up the critical vibration modes for gas flow, or by using snubbers and vibration restraints. A periodic inspection program is indicated for susceptible locations, particularly in the header-related locations and if the unit is cycled.

7. Case Study None for this mechanism.

8. References 1Lamping,

G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985.

26-8

Fatigue in Water-Touched Tubes

2Stephenson,

G.G. and J.W. Prince, Guidelines on Fossil Boiler Field Welding, Research Project 2504-02, Final Report TR-101699, Electric Power Research Institute, Palo Alto, CA, January, 1993.

ACTIONS for Fatigue in Water-touched Tubes Two paths for the BTF team to take in the investigation of fatigue begin here. The goal of these actions is to see if further investigation of fatigue is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred and fatigue is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit which indicates that there could be a future BTF by fatigue.

Action 1a: If a BTF has occurred and fatigue is the likely mechanism.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by fatigue.

➠ Determine whether the failure has

• Routine inspection and/or tube

occurred in typical locations, i.e. does it appear to be associated with a weld, tight hair-pin bend, and either tubing-related or header-related?

➠ Confirm that the macroscopic appearance of the failure includes such features as: •

OD-initiated

•

Thick-edged failure

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

sampling finds an indication of cracking, broken or bent tubing or other macroscopic evidence of fatigue.

• If unit has recently been converted to cycling or two-shifting duty, the potential for fatiguerelated tube failures will increase. Check the flexibility of watertouched headers and connecting tubes.

➠ If these indicators have been found or are in effect, go to Action 3 which outlines the steps to confirm the influence of each.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

Volume 2: Water-Touched Tubes

26-9

Action 2: Determine (confirm) that the mechanism is fatigue. A failure has occurred which the BTF team has tentatively identified as being fatigue (Action 1a). Action 2 should clearly identify fatigue as the primary mechanism or point to another cause. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Analyze the macroscopic damage. Does damage have features including: thick-edged failure surface or pin-hole leak, obvious OD-initiation, association with a weld or tight tubing bend? Also there should be no material loss on the OD.

➠ Analyze microscopic appearance of cracking. Is cracking transgranular? Is cracking from the OD?

Damage may not be fatigue. If final failure is thin-edged suspect an internal wastage mechanism as caused by either acid phosphate corrosion (Chapter 16) or caustic gouging (Chapter 17). If internal wastage accompanies a thickedged failure, suspect hydrogen damage (Chapter 15). If damage is ID-initiated, suspect corrosion fatigue (Chapter 13).

Intergranular cracking may still be primarily fatigue-related; however it may also indicate that a corrosionassisted mechanism, such as corrosion fatigue (in waterwalls) or stress corrosion cracking, is in effect, or that the damage is a combination of creep and fatigue. Confirm the latter with a determination of whether creep damage is seen such as void formation, etc. See also Chapter 24 on low-temperature creep cracking.

➠ Probable mechanism is fatigue.

➠ Go to Action 3: Root Cause Determination

References to other sources of detailed information:

• Summary of the steps and methods of metallurgical investigation of boiler tube failures can be found in Chapter 6, Volume 1.

26-10

Fatigue in Water-Touched Tubes

Action 3: Determine root cause(s) of the fatigue A BTF failure has occurred and the mechanism has been confirmed as fatigue (Action 2) or a precursor has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes of fatigue, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.2 Excessive strains caused by constraint of thermal expansion.

➠ (a). Visual examination for distortion or bending in adjacent tubes. ➠ (b). Strain gauging of suspect locations to evaluate strains during unit starts and cycling operation. ➠ (c). LVDT measurements to monitor the relative movement of the header/tube during transients.

3.3 Poor design and/or manufacture giving rise to excessive mechanical stresses.

➠ (d). Strain gauging to measure actual strains experienced at the local area during operation. ➠ (e). As in (c) above. ➠ (f). For tight, hair-pin bends, determine whether residual stresses are high.

3.4 Vibration (flue gas-induced) by direct flow or vortex shedding.

➠ (g). Metallurgical examination to determine high cycle fatigue.

3.5 Poor welding, particularly poor geometry of final joint.

➠ (h). Visual and microscopic examination of weld quality.

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Subject to access constraints, detection of fatigue is possible through a variety of routine NDE methods. Chapter 9, Volume 1 provides additional information. Results of the survey for damage will be used interactively with Action 3 to determine root cause and with Actions 5 and 6 to develop a rational prevention strategy.

Action 5: Implement repairs, immediate solutions and actions Routine weld repairs are generally sufficient to deal with the immediate attachment or header-related failures. Tube replacement is needed for failures associated with bends. Pad welding/grinding are not satisfactory. Thereafter, most actions should be considered for the longer-term (Action 6).

Volume 2: Water-Touched Tubes

26-11

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will be based on the clear identification of the underlying root cause (Action 3) and an economic evaluation to ensure that the optimum strategy has been chosen.

Major Root Cause Influences

➠ Long-Term Actions

Excessive strains caused by constraint of thermal expansion.

➠ Evaluate modifications to attachment design or to header/tube connection to reduce stress levels. ➠ Institute periodic inspection program, particularly of susceptible header locations in units that are being, or to be, cycled. ➠ Improve header/tube flexibility and confirm with LVDT.

Poor design and/or manufacture giving rise to excessive mechanical stresses.

➠ Evaluate modifications to attachment design or to header/tube connection to reduce stress levels. ➠ Institute periodic inspection program, particularly of susceptible header locations in units that are being, or to be, cycled.

Vibration (flue gas-induced) by direct flow or vortex shedding.

➠ Evaluate and install modifications such as snubbers or vibration restraints to reduce stresses induced by vibration.

Poor welding, particularly poor geometry of final joint.

➠ Institute program of weld quality control based on guidelines such as provided in reference 2.

Action 7: Determine possible ramifications/ancillary problems None for this mechanism.

26-12

Fatigue in Water-Touched Tubes

Chapter 27 • Volume 2

Pitting in WaterTouched Tubes

Introduction Pitting is a form of damage that can occur throughout the boiler. In water-touched tubing, especially economizers, it is primarily a result of poor shutdown practices with oxygen-saturated, stagnant water. It may also be caused by direct acidic attack as a result of poor chemical cleaning practice; this topic is discussed separately in Chapter 25.

Pitting in steam-touched tubes as a result of shutdown practices or carryover of Na2SO4 which combines with moisture from condensate to form an aggressive agent, is also discussed separately (Chapter 41, Volume 3). Damage caused by improper chemical cleaning in SH/RH sections is discussed in Chapter 43, Volume 3.

Volume 2: Water-Touched Tubes

27-1

1. Features of Failure and Typical Locations 1.1 Features of failure Pitting is localized part- or throughwall dissolution of tube metal. Pits associated with low pH conditions tend to be numerous and closely spaced. The affected metal is usually clean and free of deposits (if found soon after the acid attack); the surface is jagged and rough. Pits are generally sharp-walled and free of oxide or corrosion products if fresh. Damage from excessively aggressive chemical cleaning in water-touched tubes is of this type; reference should be made to Chapter 25 for additional detail about its appearance. The second type of pitting results from stagnant, oxygen-saturated water formed during shutdown. That damage, particularly in the economizer, is the subject of this chapter. Such pits can either be numerous and closely spaced, or isolated. They are often covered with caps of corrosion products as shown in

27-2

Pitting in Water-Touched Tubes

Figure 27-1. Such pitting is generally found to be filled with corrosion products, particularly red hematite iron oxide (Fe2O3) which is the thermodynamically favored reaction of iron with oxygen under conditions of high oxygen activity (see Figure 2-3, Volume 1). This form of pitting may undercut the surface. Figure 27-2 shows a cross section through an oxygen pit showing it to be filled with corrosion products and topped by a distinctive corrosion cap.

1.2 Locations of failure Pitting typically appears in those locations where boiler water stagnates in the tubes during unit shutdown and/or layup periods. This will occur as a result of improper venting, draining, shutdown and storage procedures.

Figure 27-1. Oxygen pitting in a carbon steel economizer tube. Pits are covered with caps of corrosion products (arrow) (MAG:1.6X) Source: S.R. Paterson, et al.1

Figure 27-2. Cross section through oxygen pit showing corrosion product cap and corrosion products in the pit. Source: S.R. Paterson, et al.1

Volume 2: Water-Touched Tubes

27-3

2. Mechanism of Failure 2.1 Development of pitting corrosion Pitting is a form of corrosion that is distinguished by the aspect ratio of the damage: it tends to be deep through-wall relative to the defect length. It is an insidious form of damage because (i) a relatively small amount of metal loss can lead to through-wall failure with catastrophic results, (ii) the small size coupled with the fact that pits are often filled with corrosion products makes them hard to detect, (iii) pits often undercut the surface, which can complicate their detection, (iv) laboratory simulation of field pitting is difficult, and (v) perhaps most interesting, pitting is autocatalytic; that is, conditions within the pit stimulate the continued activity of the corrodant.2 A breakdown in the passivity of a metal surface initiates the pitting process. An electrolytic cell is formed, the anode is a small area of active metal, the cathode a large area of passive metal.3 A large potential difference exists (about 0.5 V for 300-series stainless steels) which results in considerable current flow and rapid corrosion. Figure 27-3 shows the process for the growth of a pit in a metal M caused by a concentrated solution of aerated NaCl. Dissolution of metal in the pit forms metal ions (M+) which results in the inward migration of Cl- from solution to maintain charge neutrality. The metal chloride (M+Cl-) formed combines with water to form hydroxide and free acid: (M+Cl-)

+ H2O ® MOH +

H+Cl(27-1)

Although the pH of the bulk solution remains neutral, the concentration of acid in the pit lowers the pH values (to the range 1.5 to 1.0).3 This accelerates the dissolution of the metal. At the same time, a cathodic reaction such as oxygen reduction, is occurring on nearby surfaces which suppresses corrosion on these adjacent areas. The suppression of cor-

27-4

Pitting in Water-Touched Tubes

rosion around the edges of the pit and the acceleration of corrosion within the pit result in increasing rates of damage accumulation. This model for pitting initiation and growth provides an explanation for other observations about pitting. For example, pitting generally displays long initiation periods followed by what can be quite rapid propagation. Why? Initiation is thought to require a momentarily high concentration of metal dissolution to start the process.4 Such an initial concentration might be caused by a surface scratch, dislodged defect or random solution variation.2 However, just as it forms, the initial concentration could be swept away by the flow along the surface. Thus initiation might nearly start and stop many times before a “permanent’ concentration forms that is sufficient to initiate a pit. In tubing, other possible initiators are deposits of feedwater corrosion products, which might contain a number of different cations (such as Fe, Cu, Ni, Zn, etc.) and anions (Cl, SO4, etc.). A number of theories have been proposed to explain pitting initiation. Two general classifications are: (i) kinetic theories that explain breakdown of passivity based on competitive absorption between chloride ions in solution and oxygen on the metal surface, and (ii) thermodynamic theories that consider that the critical pitting potential to be the potential where chloride ions are in equilibrium with the oxide film.3,5,6,7 Pit growth is influenced by gravity, a fact that is also explained by the mechanistic process described above. The formation of a dense, concentrated solution and its retention favors the propagation of the pit, a process that is facilitated if the pit is oriented to the pull of gravity. Similarly, it is now clear why pitting is most prominent in stagnant conditions. The fact that the fluid is static allows the initiation of pitting to occur unimpeded and stagnant conditions allow for the concentrations

Na+ O2

ClO2

Na+

Na+ O2 Na+ Cl-

O2

Na+

O2

Cl-

Cl-

Na+ O2

Cl-

O2

Cl-

Cl-

ClO2

O2

O2 Cl-

O2

O2

O2

OH-

OH-

M+ OH-

OH-

M+

OH-

M+

ClM+

H+ ClCl-

Cl-

M+

Cl-

M+

Cl-

H+ M+

H+ Cl-

M+

e e e

M+

M+

M+

H+

Cl-

Cl-

H+ M+

M+

M+

Cl-

Cl-

M+

M+

ClM+ Cl-

H+

e

e

Figure 27-3. Autocatalytic processes occurring in a corrosion pit. From: M.G. Fontana and N.D. Greene, Corrosion Engineering, 1967, McGraw-Hill, New York, NY. Reproduced with permission of the McGraw-Hill Companies.

needed to accelerate the attack. Once initiation has begun, such as during shutdown of the unit, the continued growth of the pit under the more turbulent conditions of operation can continue unabated. Pitting remains a topic of active research interest. Additional information about pitting mechanisms and activities can be found in references 5, 8 and 9.

2.2 Development of pitting from oxygen-saturated, stagnant condensate The most common cause of pitting in economizer tubing is the presence of oxygen-saturated, stagnant water during unit shutdown periods caused if improper blanketing and/or protection procedures were employed to shut down the unit.

It should be noted that this effect of oxygen in stagnant water as a cause of pitting should not be confused with the effects of oxygenated treatment (OT) as a cycle chemistry option. OT when properly applied, does not cause pitting. The levels of oxygen under OT are on the order of 30-150 ppb as opposed to the ppm levels under stagnant shutdown conditions. In OT the levels are carefully controlled to encourage the growth of protective oxides throughout the feedwater train. Very little of the economizer tubing is affected (passivated) by the OT and, of course, the oxygen injection should be turned off at least an hour prior to shutdown and when the cation conductivity of the feedwater exceeds 0.3 mS/cm. Additional background on this important topic can be found in Chapter 3, Volume 1 and in references 10 and 11.

Volume 2: Water-Touched Tubes

27-5

3. Possible Root Causes and Actions to Confirm 3.1 Introduction Table 27-1 summarizes the major root causes of pitting, actions to confirm each, and the corresponding immediate and long-term actions. Pitting will result in watertouching tubing as a result of improper acid cleaning which is covered in Chapter 25, or by poor shutdown practices.

3.2 Influence of poor shutdown practice (presence of stagnant, oxygen-saturated water) As discussed above, pitting can form if shutdown procedures allow accumulations of stagnant, oxygenated water. Actions to confirm this root cause include: (a). Analysis of corrosion products present in and around the pitting will help to pinpoint the source of the problem. In the case of oxygenated attack, hematite (Fe2O3) should typically be found.

(b). Selective tube sampling can be used to evaluate whether localized corrosion cells have been formed. (c). A critical evaluation of shutdown procedures and unit conditions during shutdown should be performed to ensure that they are optimal for the unit. Such procedures are provided in Chapter 4 of Volume 1. A review of the chemistry logs of analyses performed during the shutdown will provide clues. A partial list of questions to be answered might include the following: (i) If the boiler pH was increased to around 10 with ammonia and around 200 ppm of hydrazine, were these monitored during shutdown/layup periods? (ii) Were nitrogen blanketing and caps properly applied and checked? (iii) Was sufficient N2H4 added,? (iv) Have sources of air inleakage been detected?

Table 27-1 Major Root Cause Influence, Confirmation and Corrective Actions Major Root Cause Influence

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

3.2 Influence of poor shutdown practice (presence of stagnant, oxygenated water)

(a). Analyze corrosion products in and around pits; specifically looking for presence of hematite. Methods of metallurgical examination are reviewed in Chapter 6, Volume 1. (b). Selective tube sampling to evaluate for localized corrosion cells. (c). Perform critical evaluation of shutdown procedures and of unit condition during shutdown.

• Identify damaged locations. • Replace affected tubes. See Chapter 11, Volume 1 for an overview of methods. • Initiate implementation of long-term options.

• Confirm or establish unit shutdown and layup procedures that will prevent pitting precursors. See additional discussion in Chapter 4, Volume 1.

27-6

Pitting in Water-Touched Tubes

4. Determining the Extent of Damage Pitting can be difficult to detect unless it is extensive. If there has been significant loss of wall thickness, the standard method of NDE is ultrasonic testing. Typical problems with access and with surface preparation will be present. Chapter 9,

Volume 1 provides an overview of UT methods as applied to wall thickness measurement. Selective tube sampling and metallographic analysis can be used to detect localized corrosion cells.

5. Background to Repairs, Immediate Solutions and Actions The primary immediate actions are to (i) identify damaged locations, (ii) replace the affected tubes, and (iii) repair obvious contributing conditions such as sources of air inleakage, etc., and (iv) revise shutdown/layup procedures. Procedures

for tube replacement are summarized in Chapter 11, Volume 1. The balance of actions can be conducted as a part of long-term strategies.

6. Background to Long-Term Actions and Prevention of Repeat Failures Long-term actions will consist of establishing procedures to prevent precursor conditions from occurring. A detailed discussion of appropriate shutdown and layup procedures is presented in Chapter 4, Volume 1. A brief overview of key considerations is provided here. There are five general approaches to protecting components during layup:

approaches. For short-term layups, keeping high quality water and protective blanketing with nitrogen, including a 5 psig overpressure, are the most common procedures. For longer term, layups either raising the pH of the water with ammonia and adding significant hydrazine with a nitrogen cap or a dry layup with nitrogen overpressure are the optimum choices.

• Protect surfaces with vapor-phase corrosion inhibitors.

Although lay-up with clean, dry, dehumidified air is seldom used in U.S. units, it can offer significant benefit, particularly where pitting or generalized corrosion during layups has been a problem. Even if reasonable draining and venting procedures for unit shutdown are used to reduce the amount of moisture that condenses in susceptible tubes, the presence of high humidity (greater than 60%) can significantly increase the potential for corrosion of unprotected tubes.

Of particular usefulness to prevent pitting are the first and third general

All layup conditions should be monitored to ensure that layup water or air quality is being maintained.

• Exclude impurities such as oxygen and carbon dioxide by nitrogen blanketing or by filling with specially treated layup water, and maintaining a positive pressure. • Maintain an alkaline pH in the layup water. • Keep surfaces dry and clean. • Maintain the highest purity boiler water with no additives.

Volume 2: Water-Touched Tubes

27-7

7. Case Study None for this mechanism.

8. References 1Paterson,

S.R., T.A. Kuntz, R.S.Moser, H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 2Fontana, M.G. and N.D. Greene, Corrosion Engineering, McGraw-Hill, New York, 1967. 3Asphahani,

A.I. and W.L. Silence, “Pitting Corrosion” in Metals Handbook Volume 13: Corrosion, ASM International, Metals Park, OH, 1987. 4Evans,

U.R., Corrosion, Volume 7, Number 238, 1951.

5Shreir,

L.L, R.A. Jarman, and G.T. Burstein, eds., Corrosion Volume 1: Metal/Environment Reactions, 3rd Edition, Butterworth-Heinemann, Oxford, 1994. 6Uhlig,

H.H. and J. Gilman, Corrosion, Volume 19, 1963, p. 261t.

27-8

Pitting in Water-Touched Tubes

7Vermilyea,

D., Journal of the Electrochemistry Society, Volume 118, 1971, p. 529. 8Isaacs,

H., U. Bertocci, J. Kruger, S. Smialowska, Advances in Localized Corrosion, NACE-9, National Association of Corrosion Engineers, p. 221. 9Evans,

U.R., The Corrosion and Oxidation of Metals, Arnold, London, 1961. 10Dooley,

R.B., J. Mathews, R. Pate, and J. Taylor, “Oxygenated Treatment for Fossil Plants”, Paper IWC-9216, Proceedings of the 53rd International Water Conference, Pittsburgh, PA, October, 1992. 11Bursik,

A., B. Dooley, and B. Larkin, Guidelines for Oxygenated Treatment for Fossil Plants, Research Project 1403-45, Final Report TR-102285, Electric Power Research Institute, Palo Alto, CA, December, 1994.

ACTIONS for Pitting in Water-touched Tubes Two paths for the BTF team to take in the investigation of pitting begin here. The goal of these actions is to see if further investigation of pitting is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred and pitting is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by pitting.

Action 1a: If a BTF has occurred and pitting is the likely mechanism.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by pitting:

➠ Determine whether the failure has

➠ Determine whether there is evi-

occurred in typical locations.

➠ Confirm that the macroscopic appearance of the failure includes such features as: • Obvious corrosion pits on tube ID.

dence of a shortcoming during a unit shutdown/layup such as: • Uncertainty about the water and/or air quality maintained during the shutdown or layup period.

• Corrosion product “caps”. See Figures 27-1 and 27-2.

• Inadequate nitrogen blanketing.

• Pits are deep, relative to their length.

• Evidence of air inleakage.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

• Insufficient N2H4. • Indication that stagnant, oxygenated water may have remained in tubes.

➠ Determine whether routine inspection and/or tube sampling finds indication of localized corrosion cells, measurable wall thinning or evidence of pitting.

➠ If one or more has occurred, go to Action 3 which outlines the steps to confirm the influence of each.

Volume 2: Water-Touched Tubes

27-9

Action 2: Determine (confirm) that the mechanism is pitting. A failure has occurred which the BTF team has tentatively identified as being pitting (Action 1a). Action 2 should clearly identify pitting as the primary mechanism or point to another cause. If the cause is pitting, the source of the problem (low pH or stagnant oxygenated water) should be identified through reliance on analysis of any deposits found in and around the pits. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Analyze the macroscopic damage. Does damage consist of pits that are deep relative to their length, with corrosion “caps”, and evidence of deposits?

Probable mechanism is pitting.

➠ Analyze deposits in and around pits to help determine root cause.

• Deposits primarily of hematite provide evidence that pitting most likely is a result of oxygenated, stagnant condensate during shutdown. • If deposits are missing, as is surface oxide, suspect chemical cleaning damage. See Chapter 25.

➠ Go to Action 3: Root Cause Determination

References to other sources of detailed information:

• Main text (this chapter) provides the background to mechanism and the development of pitting.

• Summary of the steps and methods of metallurgical investigation of boiler tube failures can be found in Chapter 6, Volume 1.

27-10

Pitting in Water-Touched Tubes

Action 3: Determine root cause(s) of the pitting A BTF failure has occurred and the mechanism has been confirmed as pitting (Action 2) or a precursor has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes of pitting, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.2 Influence of poor shutdown practice (presence of stagnant, oxygenated water)

➠ (a). Analyze corrosion products in and around pits; specifically looking for presence of hematite. Methods of metallurgical examination are reviewed in Chapter 6, Volume 1. ➠ (b). Selective tube sampling to evaluate for localized corrosion cells. ➠ (c). Perform critical evaluation of shutdown procedures and of unit condition during shutdown. Check records or logs of the chemistry during shutdown periods.

Volume 2: Water-Touched Tubes

27-11

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Detection of extensive pitting may be possible with ultrasonic testing (UT), otherwise sampling for localized damage in suspect locations (see main text) will be required.

Sampling may also be required to confirm the root cause. Results of the survey for damage will be used interactively with Action 3 to determine root cause and with Actions 5 and 6 to develop a strategy to prevent repeat failures.

Action 5: Implement repairs, immediate solutions and actions The BTF Team must ensure that repairs and immediate solutions are directly tied to the underlying cause. Most actions can be considered for the longer-term (Action 6) but several underlying problems can be dealt with in the short-term.

➠ Implement repairs or replacement of affected tubes ➠ Develop a plan based on results of NDE survey (Action 4) to replace affected tubing, including an assessment of the anticipated future failure rate. ➠ See Chapter 11, Volume 1 for summary of applicable tube repair techniques. ➠ Repair any obvious mechanical problems that are contributing to the problem, for example, by allowing the ingress of oxygen during shutdown.

27-12

Pitting in Water-Touched Tubes

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will be based on the clear identification of the underlying root cause (Action 3) and an economic evaluation to ensure that the optimum strategy has been chosen.

Major Root Cause Influences

➠ Long-Term Actions

Influence of poor shutdown practice (presence of stagnant, oxygenated water)

➠ Confirm or establish unit shutdown and layup procedures that will prevent pitting precursors. See additional discussion in main text this chapter and Chapter 4, Volume 1.

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components implied by the presence of pitting damage or its precursors. In particular, improper shutdown/ layup procedures can also lead to problems in other areas such as feedwater heaters, the condenser and turbine.

Volume 2: Water-Touched Tubes

27-13

27-14

Pitting in Water-Touched Tubes

Chapter 28 • Volume 2

Coal Particle Erosion

Description of Coal Particle Erosion and its Manifestation Coal-particle erosion is one of five fireside erosion mechanisms discussed in this book. The others are flyash erosion, falling slag erosion, sootblower erosion, and the erosion of in-bed tubes of bubbling fluidized bed units. Coal particles can cause erosion in some designs when protective devices no longer perform their function because of wear or damage.1 Two primary problems have been observed. The first, in cyclone burners, is the wear of replaceable wear liners located near the end of the burner and of refractories covering waterwall tubes in the furnace. High velocity combustion air is used to impart a whirling motion to coal particles. The impact of these particles can wear out resistant liners and refractory coatings that are intended to protect tube surfaces; thereafter direct erosion of the tube occurs. The second kind of problem occurs in burners, particularly front- or rearfired, where the direct impact of the coal stream, before ignition, erodes tubes in the throat or quarl region. The problem is particularly acute if refractory protection is missing or deficient.

Features of failed tubes by a coal particle erosion mechanism include: wall thinning, external wastage flats, little or no surface ash, a shallow layer of surface hardening caused by the particle impact, and in some cases, grooving of the tube surface by abrasion. As with other erosion processes, final failure of the tube occurs when the thinned wall of the tube is no longer able to contain service stresses and a ductile failure occurs. Visual examination of refractory coatings and wear-resistant liners can verify the mechanism. Ultrasonic testing (UT) can also detect the extent of wall thinning. Observation of flow patterns established in the burner by the introduction of secondary and tertiary air may reveal locations where significant wear will occur.2 Actions to minimize future failures will consist of (i) a periodic program of inspection and replacement of wear-resistant liners and refractory coatings as needed to protect tube surfaces, and (ii) adjustment of secondary and tertiary air inlet dampers, as needed, to change flow patterns and control erosion rates on liners and coatings.2

Volume 2: Water-Touched Tubes

28-1

References 1Lamping,

G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985.

28-2

Coal Particle Erosion

2Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993.

ACTIONS for Coal-Particle Erosion Two paths for the BTF team to take in the investigation of coal particle erosion begin here.

➠ Follow Action 1a: If a BTF has occurred and coal particle erosion is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by coal particle erosion.

Action 1a: If a BTF has occurred and coal particle erosion is the likely mechanism.

➠ Determine if there is tube damage in (i) a cyclone burner and whether there is evidence of extensive wear to protective devices (wear liners and/or refractories), or (ii) in the throat or quarl of conventional front/rear wall burners.

➠ Confirm that the macroscopic appearance of the failure includes such features as: • Tube wall thinning or wastage flats on the tube surface

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by coal particle erosion.

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit: • Extensively eroded wear liners or refractories. • Deteriorating refractory in throat/quarl of burner.

➠ These can be precursors to coal particle erosion. If one or more has occurred, go to Actions 2-4 for further steps.

• Thin-edged, ductile failures, or the formation of pin-holes, or a long “thin” blowout

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

Volume 2: Water-Touched Tubes

28-3

Action 2: Determine (confirm) that the mechanism is coal particle erosion. Action 3: Determine root cause(s) of coal particle erosion. Action 4: Determine the extent of damage or affected areas A failure has occurred which the BTF team has identified as being coal particle erosion damage (Action 1a). Actions 2, 3 and 4 should clearly identify coal particle erosion as the primary mechanism, confirm the root cause and determine the extent of damage. The actions listed will be executed by visual examination and an ultrasonic survey of wall thinning, along with the removal of representative tube sample(s) for analysis if needed.

➠ Confirm the condition of wear liners and refractories (cyclone units) and throat/quarl of conventional front/rear wall burners. Are there signs that either wear or damage has occurred in these locations?

➠ Determine the extent of damage via visual examination and UT (where needed) to quantify wall thinning.

➠ Evaluate appearance of damage. Is tube damage consistent with an erosive process characterized by: wall thinning, external wastage, thin-edged ductile failure, and/or external grooving consistent with abrasion?

Failure mechanism is probably by coal particle erosion.

➠ Go to Action 5: Implement Repairs, Immediate Solutions and Actions

28-4

Coal Particle Erosion

Damage is probably not coal particle erosion. Check for possibility that damage is caused by fireside corrosion (Chapter 18).

Damage is probably not coal particle erosion. Check for possibility that damage is caused by fireside corrosion (Chapter 18).

Action 5: Implement repairs, immediate solutions and actions The most important actions for the BTF team are to (i) make the tube repairs necessary to get the unit on-line, and (ii) fix the underlying problem with wear liners or refractories.

Action 6: Implement long-term actions to prevent repeat failures Long-term actions will include (i) periodic inspection and replacement of wear-resistant liners and refractory coatings as needed to protect tube surfaces, and (ii) adjustments, if needed, to secondary and tertiary air inlet dampers to change the flow patterns and control erosion rates on the liners and coatings.

Action 7: Determine possible ramifications/ancillary problems None for this mechanism.

Volume 2: Water-Touched Tubes

28-5

28-6

Coal Particle Erosion

aa Chapter 29 • Volume 2

Falling Slag Damage

aa Description of Falling Slag Damage and its Manifestation

Falling slag damage results from impacts by fused coal ash deposits or resolidified molten material (slag) that detach from furnace walls and superheater pendants. The slag is directed toward the bottom ash hopper by the lower furnace sloping wall as illustrated schematically in Figure 29-1. Damage from either erosion or mechanical impact results to waterwall tubing in either sloping

wall tubes and/or the ash hopper. As shown in Figure 29-1, the ash does not fall evenly along the length of the furnace bottom opening; it is concentrated in the first 0.9 to 1.2 m (3 to 4 ft) along each end. As a result, these are the locations of the heaviest erosion damage. As with other erosion processes, tube wastage occurs on a progressive basis leading eventually to a final failure of the tube when the thinned wall is no longer able to

Side wall

Rear wall

Front wall

1Ð2% of total ash/ft of furnace hopper bottom

20Ð25% of total ash at each end

Figure 29-1. Distribution of falling ash along furnace hopper opening. The higher concentrations of falling ash through the first 3-4 feet at each end of the bottom opening result in significant fireside wall thinning. Source: Combustion Engineering, Inc.

Volume 2: Water-Touched Tubes

29-1

contain service stresses. The fracture will be thin-edged with ductile characteristics. Additional detail on the contributing factors to erosion failures can be found in Chapter 14 on flyash erosion. Mechanical damage from impacts can cause (i) short-term overheating failures higher in the furnace as a result of the flow restriction, or (ii) breakage of tubes and attachments.1

a Panel tube (furnace side)

Web

1) Increased Wall Thickness 7¤16" hex bar Weld

a Slagging propensity is determined by the boiler design and by the fuel.2 A larger furnace plan area will have a larger wall and lower gas velocities, leading to more falling ash. The slagging potential of the coal can be determined from an ASTM Standard3 used to evaluate the potential for this type of damage. Coal properties used to estimate slagging potential include: (i) ashfusibility temperatures and temperature range, (ii) the base/acid ratio, (iii) the iron/calcium ratio, (iv) the silica to alumina ratio, (v) the iron to calcium ratio, (vi) the dolomite percentage, (vii) the ferric percentage, and (viii) the silica percentage.2 In general, high fusion temperatures result in low slagging potential.2 Tube failures will be part of the cost associated with burning highly slagging coal. Some additional time between tube failures can be purchased by using tubes of increased wall thickness or by installing sacrificial material such as wear bars shown schematically in Figure 29-2.

Panel tube (furnace side)

Web

2) 7¤16" Hexagonal Wear Bar

Panel tube (furnace side)

Web

3) Weld-Deposit Wear Bar

Figure 29-2. Options available for furnace bottom slope tube protection from falling slag. Source: Combustion Engineering, Inc.

References 1Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 2Lamping, G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985.

29-2

Falling Slag Damage

3American

Society for Testing and Materials, Standard D1857-87 (1994), “Standard Test Method for Fusibility of Coal and Coke Ash”, 1994 Annual Book of ASTM Standards: Gaseous Fuels; Coal and Coke, Volume 05.05, American Society for Testing and Materials, Philadelphia, PA, 1994.

ACTIONS for Falling Slag Damage Two paths for the BTF team to take in the investigation of falling slag damage begin here.

➠ Follow Action 1a: If a BTF has occurred and falling slag damage is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by falling slag damage.

Action 1a: If a BTF has occurred and falling slag damage is the likely mechanism.

➠ Determine if the tube damage in waterwall tubing, particularly sloping wall tubes is concentrated in the first 0.9 to 1.2 m (3 to 4 ft) along each end of the furnace bottom opening, or in the ash hopper.

➠ Confirm that the macroscopic appearance of the failure includes features consistent with either erosion damage or mechanical impact damage:

• Tube wall thinning or wastage flats on the tube surface; thinedged, ductile failures (erosion damage characteristics) • Impact damage visible on tubes, or tube failures where mechanical impacts are located remote from the failure site such as might occur if the secondary mechanism was overheating failures or breaking of attachments.

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by falling slag damage.

➠ Determine whether a high slagging coal is being used, or whether its use is anticipated in the future.

➠ Determine whether one of more of the following precursors has been found or is likely to have occurred in the unit:

• Broken sloping tube attachment points.

➠ Determine whether there have been overheating failures in waterwalls that seem to lack an identifiable root cause.

➠ These can be precursors to falling slag damage. If one or more has occurred, go to Actions 2-4 for further steps.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

Volume 2: Water-Touched Tubes

29-3

Action 2: Determine (confirm) that the mechanism is falling slag damage. Action 3: Determine root cause(s) of falling slag damage Action 4: Determine the extent of damage or affected areas A failure has occurred which the BTF team has identified as being falling slag damage (Action 1a). Actions 2, 3 and 4 should clearly identify falling slag damage as the primary mechanism, confirm the root cause and determine the extent of damage. The actions listed will be executed by visual examination and an ultrasonic survey of wall thinning, along with the removal of representative tube sample(s) for analysis if needed.

➠ Determine the extent of damage via visual examination and UT (where needed) to quantify wall thinning. Is damage primarily concentrated in typically susceptible locations such as sloping wall tubes within 0.9 to 1.2 m (3 to 4 ft) of the furnace bottom opening and ash hoppers?

➠ Evaluate appearance of damage. Is tube damage consistent with either (i) an erosive process characterized by wall thinning, external wastage, thin-edged ductile failures, and/or external grooving consistent with abrasion, or (ii) mechanical impact damage?

Failure mechanism is probably falling slag damage.

➠ Go to Action 5: Implement Repairs, Immediate Solutions and Actions

29-4

Falling Slag Damage

Damage is probably not falling slag damage.

Damage is probably not falling slag damage.

Action 5: Implement repairs, immediate solutions and actions The most important actions for the BTF team are to (i) make the tube repairs necessary to get the unit on-line, (ii) understand the implications of burning a high slagging coal and evaluate the costs (in future boiler tube failures) to do so, and (iii) gather the needed information to determine the long-term options.

Action 6: Implement long-term actions to prevent repeat failures Long-term actions will include choices among: (i) accepting tube failures, (ii) burning coals with lower slagging potential, (iii) replacing existing at-risk tubes with thicker walled substitutes, (iv) designing and installing sacrificial materials such as wear bars.

Action 7: Determine possible ramifications/ancillary problems None for this mechanism.

Volume 2: Water-Touched Tubes

29-5

29-6

Falling Slag Damage

Chapter 30 • Volume 2

Acid Dewpoint Corrosion (Economizer) Introduction This damage mechanism, which is variously termed “low temperature corrosion”, “cold end corrosion”, “acid dew point corrosion”, or simply, “dew point corrosion”, is not generally a problem for boiler tubes; however, it can be of considerable concern in the “back-end” of a unit (in air heaters, duct work, flue gas cleaning equipment, etc). If operating temperatures in the economizer were to fall below the acid dew-

point temperature, corrosion could occur as a result of the condensation of sulfuric acid from the flue gas. The mechanism can also occur in the colder regions of heat recovery steam generators (HRSG) which are becoming more popular in association with combined cycles. This chapter provides some brief comments about the mechanism in the unlikely event of its occurrence.

Volume 2: Water-Touched Tubes

30-1

1. Features of Failure and Typical Locations 1.1 Features of failure

1.2 Locations of failure

Manifestations of this mechanism in economizer tubing will be those of a general corrosion process:

Low temperature corrosion will appear in locations where (i) the boiler tube metal temperatures are below the acid dew point so that condensate forms on the metal surface, or (ii) where flue gas temperatures are below the acid dew point, so that the condensate will form on the fly ash particle.1

• The normal fireside oxide scale will probably be thin or absent in affected regions; • The corroded surface of the tube, after removing fireside deposits, if any, will have a gouged or “orange peel” appearance; • The final failure will be overpressurization caused by wall thinning and therefore the fracture will appear thin-edged, transgranular and ductile; • Because the attack is by sulfuric acid, the presence of sulfur in ash deposits remaining on the tube is likely. A white layer of iron sulfate may be present at the tube deposit interface.

30-2

Acid Dewpoint Corrosion (Economizer)

Oil-fired, stoker-fired and cycloneburner-fired (coal) boilers are more likely to experience dew point corrosion as these types of units produce less fly ash which acts to neutralize any acid formed.1

2. Mechanism of Failure Acid Dew Point Corrosion (Economizer): Mechanism 1. Acid dewpoint corrosion is primarily a result of the oxidation of SO2 to SO3 which combines with moisture to form sulfuric acid; it subsequently condenses onto surfaces which are below the condensation temperature, resulting in corrosion. 2. There are a variety of factors that influence the acid dew point temperature, although control is usually not by lowering the dewpoint, but by ensuring that the metal temperature is high enough to avoid condensation.

2.1 Introduction The oxidation of SO2 to SO3 in combination with moisture forms sulfuric acid. When temperatures are lowered to the dewpoint, the sulfuric acid condenses, leading to corrosion and exacerbating the fouling of the affected surface. The temperature at which condensate first appears depends upon a number of factors; however, for a 10 ppm SO3 concentration, dew point is about 140°C (~285°F). Since feedwater temperatures for most utility boilers are above this temperature and SO3 levels are usually below 10 ppm, low temperature corrosion of boiler tubing is seldom encountered. The mechanism is of considerable concern for the operation of air heaters and other “back-end” components. A number of parameters can affect the rate of acid dewpoint corrosion in these components where the normal operating temperatures can be around the acid dewpoint. With the understanding that condensation is seldom at issue for temperatures typical of economizer operation, a brief review of factors that can affect the acid dewpoint is included here: effect of fuel type, excess oxygen level, fuel firing, moisture level, surface temperature and air in-leakage.

2.2 Effect of fuel type Fouling and corrosion potential is a function of fuel type(s) utilized, including firing of single or multiple fuels and fuel switching.

Oil-firing does have a greater potential for corrosion and at a higher temperature.2 The back-end components of oil-fired boilers can be susceptible to low-temperature corrosion because of the presence of vanadium in the fuel ash which is a good catalyst for converting SO2 to SO3 and because of the small quantity of ash, which can serve as a buffer, in the flue gases. Sodium in the fuel oil can also contribute to high temperature conversion of SO2 to SO3. Acid dew points with residual oils generally range from 132148°C (~ 270-300°F) depending on sulfur content.3 Distillate oil combustion usually produces flue-gas acid dew points in the range 115-135°C (~ 240-275°F). There is a peak in the rate of acid deposition depending upon temperature, which various investigations have found to be within the range 22-55°C (~ 40-100°F) below the dew point.4-7 Coal-firing can also present an acid dewpoint corrosion problem. The most significant difference between oil and coal firing is the lower rate of conversion of sulfur to SO3. The ash in coal also acts as a physical barrier between iron-oxide coated surfaces and SO2 in the gas stream. Coal contains about 100 times more ash than residual oil, much of it as alkaline compounds, which theoretically are present in sufficient quantities to neutralize all of the SO3/H2SO4 produced during combustion. Acid dew points range from 120-140°C (~ 250-285°F) which correspond to SO3 concentrations of 2-20 ppm, respectively.8

Gas firing, because there is little or no sulfur and ash in the fuel, generally has little potential for corrosion.

Volume 2: Water-Touched Tubes

30-3

2.3 Effect of excess oxygen The higher the level of excess oxygen in the combustion process, the more SO2 that will be converted to SO3 and the higher the acid dew point. SO3 levels are extremely sensitive to residual oxygen levels below 2.0%. By controlling excess oxygen at a maximum of 1-2%, it is possible to avoid the formation of fully oxidized vanadium and sulfur compounds and thereby reduce boiler fouling and corrosion problems. Unfortunately, the amount of solids deposited and combustibles (carbon carryover) will increase at low O2 levels and may become a problem.9

2.4 Effect of fuel firing In addition to excess oxygen, other fireside conditions such as furnace design, furnace temperature, firing conditions and burner performance will affect the production of SO3. Furnace design. Cyclone-fired boilers depend on the use of coals that produce fusible slags of moderate viscosity at their operating temperatures. High combustion temperatures, on the order of 1650°C (~ 3000°F), produce high NOX and SO3 values, and volatize some ash components such as silica. SO3 levels encountered in these units are higher than those on pulverized coal equipment.10

30-4

Furnace temperature. Oxidizable sulfur compounds in the fuel are converted predominantly to SO2 during high temperature combustion. Very low levels of SO3 can be formed by direct oxidation from SO2 because equilibrium favors SO2 at flame temperatures. Theoretical investigations have shown that the maximum rate of SO3 formation occurs at a temperature of about 1100°C (~2010°F), which is three times the rate in effect at 900°C (~ 1650°F) and ten times the rate at 800°C (~ 1470°F). About 80% of the SO3 that forms is formed in the range 930-1330°C (~ 1705-2425°F). Firing conditions. Staged combustion, used to control NOX formation, can increase excess air and the amount of SO3 produced. The additional amount of SO3 may necessitate a change in additive control strategy.10 Burner performance. Poor atomization of liquid fuels will contribute to deposit formation. If combustion is poor or incomplete, the various constituents, separately or in combination, can form deposits. Carbon, while not corrosive itself, can carry free sulfuric acid and increase the severity of the corrosion attack if it is present in the flue gas.

Acid Dewpoint Corrosion (Economizer)

Pulverized coal fineness. Inadequately pulverized coal, i.e. excessive amounts retained on the 50 mesh screen, can also contribute to high carbon carryover and deposit buildup.

2.5 Effect of moisture level The higher the moisture level, the more SO3 that will be converted to sulfuric acid.

2.6 Effect of surface temperature The first temperature at which sulfuric acid condenses depends on the partial pressures of SO3 and water vapor and is usually around 120150°C (~ 250 to 300°F). The HCl dewpoint is below this point. The peak corrosion rate occurs at about 40°C (~ 70°F) below the H2SO4 dewpoint.

2.7 Effect of air in-leakage Air in-leakage providing oxygen and cooling effects can increase the corrosion rate. In balanced draft boilers where the back end of the boiler is under negative pressure, air ingress can occur through poorly sealed inspection doors, expansion joints, and corroded duct work.6

3. Possible Root Causes and Actions to Confirm Acid Dew Point Corrosion (Economizer): Root Causes The root cause of boiler tube failures by acid dew point corrosion can be that either the temperatures are below the dew point in the economizer, or that a high dew point is caused by a variety of unit factors.

3.1 Introduction Acid dewpoint corrosion in the economizer can occur when economizer temperatures are below the condensation point. As discussed above, this is generally a major concern for air heaters but not for economizers.

3.2 Economizer tube temperatures below the acid dew point This can occur, for example, when a number of feedwater heaters are out of service, or during unit shutdown or startup. Startup is of less consequence because of the short period of time involved.

3.3 High acid dewpoint caused by fuel or operating choices The second root cause of a problem with acid dewpoint corrosion is if fuel and operating practices are such that the dewpoint is raised. A variety of possible control options are available if this is the case as outlined below. Actions to confirm: (b). Perform an evaluation of dewpoint, or measurement with deposition probes, and determine the sensitivity to key operating and fuel parameters (fuel composition, additive choices, excess air levels, etc.)

Actions to confirm: (a). Measure economizer temperatures and compare to calculated or measured acid dewpoint. That calculation will be based on fuel type, fuel composition, firing conditions, excess air and air ingress, and moisture level. Dew point and deposition rates have been measured with dew point meters and gold disc deposition probes, although their application is mostly for setting operating conditions relative to air heaters. For pulverized coal-fired boilers, corrosion occurs only well below the calculated dewpoint. If corrosion seems to be occurring during shutdown periods, deposition rates can be monitored with deposition probes.

3.4 Local air inleakage Locally low gas temperatures caused by air inleakage can create a problem.11 High oxygen levels in the flue gas at the boiler outlet, coinciding with high carbon in the ash can, apart from mill or burner problems, indicate a leakage of air after the combustion chamber. While the operator is controlling outlet air he is really only reducing furnace air with often dire consequences.11 High air inleakage at the air heater can lead to poor burner air distribution and poor combustion. Actions to confirm: (c). Examine for localized wastage patterns such as downstream from door openings.

Volume 2: Water-Touched Tubes

30-5

4. Determine the Extent of Damage Inspection for damage will include visual examination for signs of corrosion and the use of ultrasonic testing (UT) to survey for wall thinning. Chapter 9, Volume 1 discusses this use of UT in more detail. Constraints

on accessibility may limit how much UT examination is possible. Deposition probes can be used to determine the corrosion potential for various fuel/additive/excess air combinations, if required.

5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long Term Actions and Prevention of Repeat Failures Acid Dew Point Corrosion (Economizer): Actions 1. Immediate actions will include confirming the extent of damage and making the necessary tube repairs or replacement. Some minor operating changes may also be feasible but most steps are longerterm. 2. Long-term actions will involve either (i) raising the gas and metal temperatures or (ii) lowering the acid dew point through a variety of operating options. 3. Eliminate sources of air ingress, e.g., around inspection doors. 4. For serious conditions, monitoring with deposition probes or dew point meters is indicated.

30-6

The necessary repairs should be effected once the extent of damage is determined. Chapter 11, Volume 1 reviews boiler tube repair procedures. Over the longer term, solutions will involve raising the gas and metal temperatures above the acid dew point or lowering the acid dew point; the latter has far more options available. If acid dewpoint corrosion is a problem, it can be mitigated by a variety of steps to reduce the level of sulfuric acid produced. This in turn implies reducing the SO3 in the flue gas. Although a complete description of options is beyond the scope of this book, some possible actions might include (i) using a lower sulfur fuel, (ii) periodic removal of ash deposits, (iii) reducing the amount of excess air in the furnace, or (iv) injecting an additive into the fuel. A discussion of additives used in oilfired units can be found in the writeup on SH/RH fireside corrosion in Chapter 34, Volume 3. Those of

Acid Dewpoint Corrosion (Economizer)

benefit for controlling that form of damage, such as MgO, have also been shown to be beneficial for limiting acid dew point corrosion in economizers. It is likely that such actions will be triggered by considerations other than the low temperature corrosion of economizer tubes, such as air heater fouling or corrosion, or desire to decrease SOX emission levels. Such positive steps, would, of course, also further protect against economizer corrosion. One control challenge when the economizer is expected to be at a temperature below the dewpoint, is during unit shutdown. The application of a corrosion inhibitor to tube surfaces following cleaning of external deposits and drying can be considered if a problem develops.12 Humidity control during shutdown has also been used to limit corrosion damage.

7. Case Study None for this mechanism.

8. References 1Lamping, G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985.

7Sirois,

2Macduff,

8Sotter,

E.J. and N.D. Clark, “Ljungstron Air Preheater Design and Operation - Part II: Corrosion and Fouling”, Combustion, March, 1976. 3Bennett,

R. and B. Handelman, “Solving Cold End Boiler Problems Through Innovative Chemical Technology”, Combustion, January, 1977. 4Clark,

N.D., et al., “Boiler Flue Gas Measurements Using a Dewpoint Meter”, ASME Paper 63-WA-108. 5Frisch,

N.W., “Analysis of Air Heater-Fly Ash-Sulfuric Acid Vapor Interactions”, in F.A. Ayer, compiler, Proceedings: Fifth Symposium on the Transfer and Utilization of Particulate Control Technology, Volume 2, held in Kansas City, MO, August 27-30, 1984, Research Project 1835-6, Proceedings CS-4404, Electric Power Research Institute, Palo Alto, CA, February, 1986, p. 28-1. 6Hernandez,

J.J., “Dewpoint Corrosion in Oil-Fired Boilers in Cycling Service”, in Proceedings: Fossil Plant Retrofits for Improved Heat Rate and Availability, held in San Diego, CA, December 1-3, 1987, Proceedings GS-6725, Electric Power Research Institute, Palo Alto, CA, December, 1989, pp. 31-1 through 31-10.

R.H. and G. Furman, “Acid Dewpoint: New Concern for an Old Problem in Fossil Plant Design”, presented at the 1986 American Power Conference, Chicago, IL, April 14-16, 1986. J.G., J.A. Arnot, and T.M. Brown, Guidelines for Fireside Testing in Coal-Fired Power Plants, Research Project 1891-3, Final Report CS-5552, Electric Power Research Institute, Palo Alto, CA, March, 1988. 9Stein,

F., et al., “Effect of Excess Air on Acid Deposition in a Regenerative Air Preheater”, presented in Proceedings: 1986 EPRI Power Plant Performance Monitoring and System Dispatch Improvement Workshop, Washington, D.C., 1986. 10Radway, J.E. and M.S. Hoffman, Operations Guide for the Use of Combustion Additives in Utility Boilers, Research Project 1839-3, Final Report CS-5527, Electric Power Research Institute, Palo Alto, CA, December, 1987. 11Personal

Communication from T. Healy (ESB Ireland) to R.B. Dooley, February, 1995.

12Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993.

Volume 2: Water-Touched Tubes

30-7

ACTIONS for Acid Dewpoint Corrosion Two paths for the BTF team to take in the investigation of acid dewpoint corrosion begin here. The goal of these actions is to see if further investigation of this mechanism is warranted or whether another BTF mechanism should be investigated.

Action 1a: If a BTF has occurred and acid dewpoint corrosion is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of acid dewpoint corrosion, such as a low-temperature section of the economizer, during shutdown periods, and preferentially in oil-fired, stoker-fired or cyclone-burner coal-fired units.

➠ Follow Action 1a: If a BTF has occurred and acid dewpoint corrosion is the likely mechanism.

➠ Confirm that the macroscopic appearance of the failure includes such features as:

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by acid dewpoint corrosion.

• Wall thinning • Fireside scale that is thinned or missing entirely • External surface with corroded or “orange peel” appearance after removing deposits, if any • Final failure with ductile characteristics such as thin-edged fracture surface.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for watertouched tubing (Table 12-1) to pick a more likely candidate.

30-8

Acid Dewpoint Corrosion (Economizer)

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by acid dewpoint corrosion.

➠ Determine whether one or more of the following precursors has been found, or is likely to have occurred in the unit: • Evidence of particularly corrosive combustion products in back-end portions of the boiler. • Routine inspections of the economizer that find evidence of corrosion damage - such as wall thinning or lack of fireside oxide or deposits on tubes. • Operation over an extended period with a number of feedwater heaters out of service.

➠ These precursors can be indicative of attack by an acid dewpoint corrosion mechanism. If one or more has occurred, go to Action 3 which outlines the steps to confirm the root cause.

Action 2: Determine (confirm) that the mechanism is acid dewpoint corrosion. A failure has occurred which the BTF team has tentatively identified as being acid dewpoint corrosion damage (Action 1a). Action 2 should clearly identify acid dewpoint corrosion as the primary mechanism or point to another cause. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Analyze in detail the failure surface. Is the failure surface a thinedged crack, and/or ductile in appearance?

➠ Analyze deposits. If deposits are present, do they contain sulfur?

➠ Evaluate ash. Is pH of ash in deionized water acidic?

➠ Determine extent of external corrosion. Is there evidence of external corrosion (“orange peel” appearance) of tube after removing deposits?

Smooth external surface and material removal may be caused by sootblower erosion, see Chapter 22.

Probable mechanism is acid dewpoint corrosion.

➠ Go to Action 3: Root Cause Determination

Volume 2: Water-Touched Tubes

30-9

Action 3: Determine root cause of acid dewpoint corrosion A BTF failure has occurred and the mechanism has been confirmed as acid dewpoint corrosion (Action 2) or a precursor to acid dewpoint corrosion has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

30-10

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.2 Economizer tube temperatures are below the acid dewpoint during operation, such as with a number of feedwater heaters out of service, or during shutdown

➠ (a). Measure economizer temperatures and compare to calculated or measured acid dewpoint. If corrosion seems to be occurring during shutdown periods, deposition rates can be monitored with deposition probes.

3.3 Excessively high dewpoint caused by fuel or operating conditions.

➠ (b). An evaluation of dewpoint or measurement with deposition probes. The sensitivity to key operating and fuel parameters such as fuel composition, additive choices, excess air levels, etc. should be determined. See main text discussion of mechanism for key variables.

3.4 Locally low gas temperatures caused by local air ingress.

➠ (c). Examine for localized wastage patterns, such as downstream from door openings.

Acid Dewpoint Corrosion (Economizer)

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Evaluation will be based on detecting (i) wastage and wall thinning, or (ii) obvious (visual) signs of corrosion.

➠ Identify all locations to be examined. Missed locations are sites for future failures.

➠ Perform UT survey and visual inspection to: (i) measure extent of damage via wall thinning or (ii) detect signs of corrosion. A review of UT methods is provided in Chapter 9, Volume 1.

➠ Perform tube sampling to confirm results of NDE inspection, as needed.

➠ Install deposition probes or dew point meters to monitor rates, as needed.

➠ Use results interactively with Action 3.

➠ Go to Action 5: Implement repairs, immediate solutions and actions.

Volume 2: Water-Touched Tubes

30-11

Action 5: Implement repairs, immediate solutions and actions Action 6: Implement long-term actions to prevent repeat failures ➠ Implement repairs or replacement

Immediate actions (Action 5) will be primarily to repair and replace damaged tubes as required. Most actions can be considered for the longer-term (Action 6) although several underlying problems can be dealt with in the short-term.

of affected tubes based on NDE survey (Action 4) ➠ Develop a plan to replace affected tubing, including an economic assessment of the future possible failure rate and the resulting optimal extent of new tubing. ➠ See Chapter 11, Volume 1 for summary of applicable tube repair techniques. ➠ Place the unit back in service.

➠ Monitor economizer temperatures, dew point temperature or deposition rates, as needed, to track development of problem.

➠ Evaluate economics of available options for controlling problem including: • Fuel effects. • Additive options. • Combustion options such as changes in excess oxygen

Action 7: Determine possible ramifications/ancillary problems level. • Unit shutdown frequency • Expected rates of wastage under various alternatives for control and replacement cost of expected number of tube failures based on those rates.

30-12

References to other sources of detailed information:

• Main text (this chapter) pro-

Acid Dewpoint Corrosion (Economizer)

vides additional detail on the effects on mechanism of various fuel and operating options.

Index

Acid contamination, 15-10 through 15-12 Acid deposition, 30-2, 30-3 Acid dewpoint corrosion (economizer), 30-1 through 30-12 actions, 30-8 through 30-12 determining the extent of damage, 30-6, 30-11 features of failure, 30-2, 30-8 locations of failure, 30-2 long-term actions and the prevention of repeat failures, 30-6, 30-12 mechanism, 30-3, 30-4, 30-9 precursors, 30-8 ramifications/ancillary problems, 30-12 repairs and immediate solutions/ actions, 30-6, 30-12 root causes and actions to confirm, 30-5, 30-10 Acid phosphate corrosion, 16-1 through 16-28 actions, 16-22 through 16-28 case study, 16-16 through 16-20 deposit characteristics, 7-1 through 7-4, 16-2, 16-4, 16-6, 16-19, 16-20 determining the extent of damage, 16-13, 16-25 distinguishing from hydrogen damage or caustic gouging, 7-1, 7-2, 16-3 features of failure, 7-1, 7-2, 16-2 through 16-4, 16-18 through 16-20, 16-22 locations of failure, 16-4, 16-5 long-term actions and the prevention of repeat failures, 16-14, 16-15, 16-27 mechanism, 16-6 through 16-8, 16-23 precursors, 16-22 ramifications/ancillary problems, 16-28 repairs and immediate solutions/ actions, 16-13, 16-26 root causes and actions to confirm, 7-4, 16-9 through 16-12, 16-24 Additives, oil-fired units, 34-10, 34-11, 34-18, 34-19 Air inleakage, 13-24, 27-6, 27-7, 27-9, 30-4, 30-5, 41-6

Alkali iron trisulfates, 33-2, 33-7, 33-8 Alkali salts, 33-2, 33-7, 33-8 All-volatile treatment (see also Feedwater treatment), 1-18, 3-9, 3-13 “Alligator hide”, 32-2, 33-3, 33-4, 34-5 American Society of Mechanical Engineers (ASME) Codes design, 2-2 through 2-6 non-destructive examination, 11-3 welding 11-3, 11-4 Ammonia, 3-8, 3-9 Ash analysis, 33-12 Austenitic welds (in dissimilar metal welds), 11-7, 35-2 through 35-9, 35-15 Availability losses and improvement, 1-20 Backing rings, 2-14, 11-4 Baffles (erosion), 14-12 Bell-shaped corrosion curve, 33-7, 33-8 Black boiler water samples, 16-11 Boiler pressure drop losses, 19-5, 19-6 Boiler Tube Failure (BTF) Reduction Program, 1-20, 5-1 through 5-3 corporate directives for BTF reduction, 5-2 goals, 1-20, 1-21, 5-2 multidisciplinary teams for BTF reduction, 5-2 Boiler tube failures formalizing programs for reduction of, 1-20, 5-1 through 5-6 historical developments in identification, correction and prevention, 1-16, 1-18 importance, 1-1 importance of operation and maintenance procedures in preventing, 4-1 influence of cycle chemistry, 1-18, 3-1 through 3-2 influence of fuel options, 1-18 influence of operating conditions, 1-18 influence of unit lay-up, 4-9 influence of unit transients, 4-8, 4-9 influencing or influenced by chemical cleaning, 4-2

largest availability losses, 1-1, 1-2 precursors to, 1-4, 1-10 through 1-15, 1-16, 12-7 through 12-12, 31-7 through 31-13 repeat failures, 1-20, 1-21 reporting and report form, 5-3 through 5-5 resulting from breakdown of protective magnetite in water-touched tubing, 2-11 resulting from breakdown of protective oxide in steamtouched tubing, 2-15 resulting from fireside conditions, 2-21 screening table, steam-touched tubes, 1-8, 1-9, 31-4, 31-5 screening table, water-touched tubes, 1-6, 1-7, 12-4, 12-5 steps in generic investigation 1-4, 1-5, 1-16, 12-2, 12-3, 31-2, 31-3 with significant microstructural changes, 10-2 worldwide statistics, 1-1 Boiler tubes (see also Superheater/ reheater tubes and Waterwalls and economizer tubes) design considerations, 2-2 through 2-6 materials and alloys, 2-2, 2-3, 2-6 maximum design and oxidation temperatures, 2-4, 23-2, 23-3 Boiler water treatment, 3-1 through 3-8 all-volatile treatment (see also Feedwater treatment), 1-18, 3-9, 3-13 caustic treatment, and caustic gouging, 17-5, 17-6, 17-10 guidelines for, 3-5, 3-6, 3-13 historical development of, 1-18, 3-5, 17-5 success factors for use of, 3-5 comparison of options, 3-6 effect on boiler tube failures 3-1, 3-2 factors during unit transients, 4-8 optimization of, 3-6 through 3-8

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-1

phosphate treatments, 3-3 through 3-5, 16-6 and acid phosphate corrosion, 16-6 through 16-8, 16-11, 16-12,16-16, 16-17 effect of chemical additions on operating regimes, 3-4 guidelines for, 3-5, 3-13 historical development of, 1-18, 3-3 Borio index, 33-9 Bubbling-bed FBC units, 47-1 through 47-12 chromized tubes, 47-10 plasma coatings, 47-10 tube armoring, 47-10 Burner misalignment, 15-10, 16-11, 17-10 Carbides, 10-5, 10-6 Carryover, 37-5, 37-6 of Na2SO4, 41-2, 41-5, 41-6 Caustic gouging, 17-1 through 17-22 actions, 17-16 through 17-22 case study, 17-14 deposit characteristics, 7-1 through 7-3, 17-2, 17-3, 17-6, 17-7 determining the extent of damage, 17-11, 17-19 distinguishing from hydrogen damage or acid phosphate corrosion, 7-1, 7-2, 17-2 electrochemical corrosion cell, 17-6, 17-7 features of failure, 7-1, 7-2, 17-2, 17-3, 17-16 locations of failure, 17-2 through 17-4 long-term actions and the prevention of repeat failures, 17-12, 17-13, 17-21 mechanism, 2-11, 2-14, 17-5 through 17-7, 17-17 precursors, 17-16 ramifications/ancillary problems, 17-22 repairs and immediate solutions/ actions, 17-11, 17-20 root causes and actions to confirm, 7-4, 17-8 through 17-10, 17-18 Caustic treatment (see also Boiler water treatment),1-18, 3-5, 3-6, 3-13, 17-5,17-6, 17-10

Chemical cleaning (see also Chemical cleaning damage in super heater/reheater tubes and Chemical cleaning damage: waterwalls) as indicator of non-optimized feedwater chemistry, 3-2 boiler tube failures influenced by, 4-2, 36-6, 36-8 effect of changing to oxygenated treatment, 3-11, 3-12 FBC units, 4-8 superheaters/reheaters, 4-5 through 4-7, 32-21, 33-21, 34-19, 37-5 through 37-10 importance of sampling, 4-6 locations to clean, 4-6 monitoring, 4-7 process optimization, 4-6, 4-7 reasons to perform, 4-5 solvent choice, 4-6 typical operations for, 4-7 when to clean, 4-6 waterwalls, 4-1 through 4-5 assessing cleanliness and deposit levels, 4-2, 4-3 guidelines for, 4-1 importance, 4-1 inhibitor breakdown, 25-4 monitoring Fe levels to determine finish, 4-5 possible problems that could lead to damage, 25-4 solvent choice, 4-3, 4-4 typical operations for, 4-5 when to clean, 4-2 Chemical cleaning damage in superheater/reheater tubes, 43-1 through 43-8 actions, 43-5 through 43-8 determining the extent of damage, 43-3, 43-7 features of failure, 43-2, 43-5 locations of failure, 43-2 long-term actions and the prevention of repeat failures, 43-4, 43-8 mechanism, 43-2, 43-6 precursors, 43-5 ramifications/ancillary problems, 43-8 repairs and immediate solutions/actions, 43-4, 43-8 root causes and actions to confirm, 43-3, 43-7

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-2

Chemical cleaning damage: water walls, 25-1 through 25-9 actions, 25-6 through 25-9 determining the extent of damage, 25-4, 25-8 features of failure, 25-2, 25-3, 25-6 long-term actions and the prevention of repeat failures, 25-5, 25-9 mechanism, 25-4, 25-7 precursors, 25-6 ramification/ancillary problems, 25-9 repairs and immediate solutions/ actions, 25-5, 25-9 root causes and actions to confirm, 25-4, 25-7 Chlorine in coal, 18-5, 18-6, 33-10, 33-11, 47-6 Chordal thermocouples, 9-8, 9-9 Chromizing waterwalls, 19-15, 19-16 Circulating-bed FBC units, 48-1 through 48-4 coatings, 48-2 erosion/abrasion, 48-2 through 48-4 underdeposit corrosion, 48-2 Coal composition (see also Combustion process and/or Fireside scale/ash) and corrosiveness, 18-4 through 18-6, 33-8 through 33-11 and erosiveness, 14-5 through 14-7 effect of chlorine content on fire side corrosion in steamtouched tubes, 33-10, 33-11 effect of chlorine content on fireside corrosion in watertouched tubes 18-5, 18-6 effect of sulfur level on fireside corrosion in water-touched tubes, 18-4 Coal particle erosion, 28-1 through 28-5 actions, 28-3 through 28-5 description and manifestation, 28-1 Coal Quality Impact Model (CQIM) 2-22, 33-14, 33-20 Coatings, 22-4, 48-2 for fireside corrosion in steamtouched tubing, 33-18, 34-16, 34-17 for fireside corrosion in watertouched tubing, 18-12 through 18-14 for sootblower erosion, 38-5

Cold air velocity test (CAVT) (see also Flyash erosion), 14-12 through 14-18 Co-extruded tubing for fireside corrosion in steamtouched tubes, 33-20, 34-17 for fireside corrosion in watertouched tubes, 18-14 welding, 11-7 Cold bent tubes and lowtemperature creep, 24-4 Cold end corrosion, 30-1 Combustion process, ash formation, erosiveness, and deposition, 2-22 through 2-24 formation of gaseous species, 2-22, 18-4 Commissioning of units, activities to prevent future boiler tube failures, 4-9, 4-11 Concentration in deposits, 2-13, 2-14, 15-4 through 15-6, 15-8, 15-10, 16-5 through 16-7, 17-4, 17-6, 17-7 Condenser leaks, 15-10, 15-11, 37-6 Congruent phosphate treatment (see also Boiler water treatment), 3-4, 16-6 Coordinated phosphate treatment (see also Boiler water treatment), 3-3, 3-4 Core monitoring parameters for cycle chemistry, 3-14 Corporate commitment needed to solve boiler tube failures, 5-1, 5-2 Corporate directives for BTF reduction, 5-2 Corrosion indices, 18-5, 18-6, 33-8 through 33-11 rates as a determinant of repair choices, 18-11 Corrosion fatigue, 13-1 through 13-41 actions, 13-35 through 13-41 analysis of field experience, 13-13 through 13-15 breakdown of magnetite, 13-10 through 13-12 case study, 13-30 through 13-32 determining the extent of damage, 13-26, 13-38 distinguishing from OD-initiated fatigue, 7-6, 7-7 environmental effects on initiation and propagation, 13-16 through 13-20

features of failure, 13-2 through 13-5, 13-35 Influence Diagram for the analysis of corrosion fatigue, 13-24 through 13-26, 13-30 through 13-32 locations of failure, 13-6 through 13-9 long-term actions and the prevention of repeat failures, 13-28, 13-29, 13-40 mechanism, 2-11, 13-10 through 13-20, 13-36 oxygenated treatment, effect on corrosion fatigue, 13-20 phosphate treatment, effect on corrosion fatigue, 13-18, 13-20 precursors, 13-35 ramifications/ancillary problems, 13-41 repairs and immediate solutions/actions, 13-27, 13-39 root causes and actions to con firm, 13-21 through 13-26, 13-37, 13-38 stress effects on initiation and propagation, 13-15, 13-16 Corrosion products, 1-17, 3-1, 3-2 Creep (see also Long-term overheating and Low-temperature creep cracking), 6-8, 7-6, 7-8, 24-1 through 24-11, 32-1 through 32-32 Creep cavitation, 10-6 Creep damage assessment techniques, 10-5 through 10-8 Larson-Miller Parameter (LMP), 10-2 through 10-6 Cycle chemistry (see also Boiler water treatment and Feedwater treatment), core monitoring parameters, 3-14 developing unit-specific guidelines, 3-12 through 3-13 diagnostic parameters, 3-14 goals for improvement program, 3-1, 3-2 guidelines documents for, 3-13 instrumentation and monitoring, 3-14 setting action levels, 3-12, 3-13 Cycling of units, 4-8, 4-9, 13-24, 20-4, 26-5, 35-12, 39-5 effect on boiler tube failures, 4-8, 4-9 effect on thermal fatigue in economizer inlet header tubes, 20-2

Departure from nucleate boiling (DNB), 2-12, 2-13 Deposit density, 4-2 Deposit weight, 4-2, 4-3 Deposits (see Waterside deposits, Feedwater corrosion products, Concentration in deposits, Oxides internal in steam-touched tubes, Oxides internal in water-touched tubes, Fireside scale/ash), 15-2, 15-4, 16-2 through 16-5, 17-2 through 17-4, 19-5 Diffusion screens (erosion), 14-14, 14-16 through 14-18 Dissolved oxygen, 13-11, 13-12, 13-16 through 13-20, 21-3, 21-4 Dissimilar metal welds, 35-1 through 35-25 actions, 35-19 through 35-25 case study, 35-17 determining the extent of damage, 35-12, 35-13, 35-22 features of failure, 35-2 through 35-5, 35-19 influence of welding variables, 35-7 through 35-9 locations of failure, 35-3 long-term actions and the prevention of repeat failures, 35-15, 35-16, 35-24 mechanism, 35-6 through 35-9, 35-20 microstructural changes in service, 35-6, 35-7 precursors, 35-19 ramifications/ancillary problems, 35-25 repairs and immediate solutions/ actions, 35-14, 35-23 root causes and actions to confirm, 35-10 through 35-12, 35-21 Distorted or misaligned tubes, 14-3, 14-4, 14-11, 33-6, 33-15, 33-21, 34-5, 34-15, 34-19, 35-10, 39-5, 40-1, 40-3 Distribution screens (erosion), 14-14, 14-16 through 14-18 DMW LIFE code, 35-16 Drum boiler water treatment, 3-3 through 3-8 Drum level control, 23-6, 37-6, 41-6 “Dutchman” repair, 11-7, 35-23 Economizer inlet header tube failures (see Erosion-corrosion of economizer inlet header tubes and/or Thermal fatigue in economizer inlet header tubes)

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-3

Electric resistance flash welding, 45-1 through 45-3 Equilibrium phosphate treatment (see also Boiler water treatment), 3-3, 3-5, 16-14 Erosion (see also Flyash erosion, Coal particle erosion, Falling slag damage, Sootblower erosion in superheater/reheater tubes, Sootblower erosion in water-touched tubing, Fluidized-bed combustion (FBC) units) abrasion index, 14-6, 14-7 basics of damage mechanism, 14-5 erosiveness of ash constituents, 2-23, 2-24, 14-5, 14-6 wear propensity calculation, 14-6, 14-7 Erosion-corrosion, general 3-9, 21-3, 21-4 Erosion-corrosion of economizer inlet header tubes, 21-1 through 21-9 actions, 21-7 through 21-9 determining the extent of damage, 21-5, 21-8 distinguishing from thermal fatigue and flexibility-induced cracking, 7-6, 7-7 features of failure, 21-1, 21-2, 21-7 locations of failure, 21-2, 21-3 long-term actions and the prevention of repeat failures, 21-5, 21-9 mechanism, 21-3, 21-8 precursors, 21-7 ramifications/ancillary problems, 21-9 repairs and immediate solutions/ actions, 21-5, 21-9 root causes and actions to confirm, 21-4, 21-8 Excess oxygen, high excess air in oil-fired units, 34-14, 34-19 low excess air, 18-1, 18-7, 34-14 Exfoliation of SH/RH steamside oxide, 2-17 through 2-21, 36-5, 36-7 effect of unit chemistry on, 2-21 effects, 2-17, 2-18, 2-21 failure criterion, 2-18, 2-20 rating severity of, 2-18, 2-19 susceptible materials, 2-18, 2-20, 2-21

Failure mechanisms fluidized-bed units, Chapters 47 and 48 list, 1-3 steam-touched tubes, Volume 3 waste-to-energy units, Chapter 49, Volume 3 water-touched tubes, Volume 2 Falling slag damage, 29-1 through 29-6 actions, 29-3 through 29-6 description and manifestation, 29-1, 29-2 Fatigue in superheater/reheater tubes, 39-1 through 39-12 actions, 39-9 through 39-12 determining the extent of damage, 39-7, 39-11 features of failure, 39-2, 39-9 locations of failure, 39-3, 39-4 long-term actions and the prevention of repeat failures, 39-7, 39-12 mechanism, 39-5, 39-10 precursors, 39-9 repairs and immediate solutions/actions, 39-7, 39-12 root causes and actions to confirm, 39-5, 39-6, 39-11 Fatigue in water-touched tubes, 26-1 through 26-12 actions, 26-9 through 26-12 determining the extent of damage, 26-7, 26-11 distinguishing from corrosion fatigue, 7-6, 7-7, 26-3 features of failure, 26-2, 26-9 locations of failure, 26-3, 26-4 long-term actions and the prevention of repeat failures, 26-8, 26-12 mechanism, 26-5, 26-10 precursors, 26-9 repairs and immediate solutions/ actions, 26-8, 26-11 root causes and actions to confirm, 26-6, 26-7, 26-11 Feedwater corrosion products, 1-17, 3-1, 3-2, 15-4, 15-14, 16-4, 16-5, 16-14, 17-2 through 17-4, 17-12, 23-5 Feedwater treatment, 3-8 through 3-12 all-volatile treatment (AVT), 3-9 guidelines for, 3-13 historical development of, 1-18 comparing AVT and oxygenated treatment, 3-9, 3-11, 3-12

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-4

factors during unit transients, 4-8, 4-9 importance of proper choice of, 3-8 optimizing for all-ferrous feedwater trains, 3-9 through 3-11, 21-5, 21-6 optimizing for mixed metallurgy feedwater trains, 3-11, 21-5, 21-6 oxygen scavenger use, 3-8 through 3-10, 21-3, 21-6 oxygenated treatment (OT), 3-9 through 3-11 effect on corrosion fatigue, 13-20 effect on oxide growth and exfoliation, 2-21 guidelines for in once-through and drum units, 3-13 historical development of, 1-18 to reduce deposition in waterwalls, 19-5, 19-6, 19-11, 19-13, 19-15 problems with erosion-corrosion throughout unit, 3-9 Fe-Fe carbide equilibrium diagram, 7-5, 23-2 FeO, 2-7, 32-9 Fe2O3, 2-7, 2-16, 2-17, 2-20, 32-9 Fe3O4, 2-7, 2-16, 2-17, 2-20, 32-9 Ferric oxide hydrate (FeOOH), 3-10 Film boiling, 2-12, 2-13 Finite element analysis for analyzing corrosion fatigue, 13-29 Fireside corrosion in SH/RH tubes (coal-fired units), 33-1 through 33-30 actions, 33-24 through 33-30 case study, 33-22 determining the extent of damage, 33-15, 33-27 distinguishing from long-term overheating, 6-8, 7-6 through 7-8, 33-4, 33-5 features of failure, 7-6, 7-8, 33-2 through 33-5, 33-24 locations of failure, 33-6 long-term actions and the prevention of repeat failures, 33-17 through 33-21, 33-28, 33-29 mechanism, 33-7 through 33-11, 33-25 precursors, 33-24 ramifications/ancillary problems, 33-30 repairs and immediate solutions/ actions, 33-16, 33-28

root causes and actions to confirm, 33-12 through 33-15, 33-26, 33-27 use of indices to predict likelihood of, 33-8 through 33-10, 33-15 Fireside corrosion in SH/RH tubes (oil-fired units), 34-1 through 34-26 actions, 34-21 through 34-26 determining the extent of damage, 34-14, 34-24 distinguishing from long-term overheating, 6-8, 7-6 through 7-8, 7-9, 34-5, 34-6 features of failure, 7-6, 7-8, 34-2 through 34-5, 34-21 locations of failure, 34-5 long-term actions and the prevention of repeat failures, 34-16 through 34-19, 34-25 mechanism, 34-7 through 34-10, 34-22 precursors, 34-21 ramifications/ancillary problems, 34-26 repairs and immediate solutions/ actions, 34-15, 34-24 root causes and actions to confirm, 34-11 through 34-14, 34-23 Fireside corrosion in water-touched tubes, 18-1 through 18-24 actions, 18-18 through 18-24 case study, 18-16 determining the extent of damage, 18-11, 18-21 effect of coal chlorine content on, 18-5, 18-6 features of failure, 18-2, 18-3, 18-18 locations of failure, 18-2, 18-3 long-term actions and the prevention of repeat failures, 18-12 through 18-15, 18-23 mechanism, 18-4 through 18-6, 18-19 precursors, 18-18 ramifications/ancillary problems, 18-24 repairs and immediate solutions/ actions, 18-11, 18-22 root causes and actions to confirm, 18-7 through 18-10 summary of field experience, 18-16

Fireside scale/ash, compositional analysis of, 33-12, 33-15 development on SH/RH tubing, 32-10 metallurgical analysis of, 6-9 Flame impingement, 15-10, 16-11, 17-10 Fluidized-bed combustion (FBC) units boiler tube failures in bubblingbed units, 47-1 through 47-12 boiler tube failures in circulatingbed units, 48-1 through 48-4 chemical cleaning of, 4-8 Fluxdome, 9-9 Flux meter, 9-9 Flyash erosion, 14-1 through 14-29 actions, 14-23 through 14-29 case studies, 14-19 through 14-21 cold air velocity test (CAVT), 14-12 through 14-18 determining the extent of damage, 14-11, 14-26 distinguishing from sootblower erosion in SH/RH tubes, 7-9 estimating solids loading, 14-16 features of failure, 14-2, 14-3, 14-23 locations of failure, 14-3, 14-4 long-term actions and the prevention of repeat failures, 14-12 through 14-18, 14-28, 14-29 mechanism, 14-5 through 14-7, 14-24 precursors, 14-23 protection options, 14-16 through 14-18 ramifications/ancillary problems, 14-29 repairs and immediate solutions/ actions, 14-11, 14-27 root causes and actions to confirm, 14-8 through 14-10, 14-25 Forging laps, 45-1, 45-2, 45-3 Fossil-fuel power plants, primary components, 1-16 Fretting, 40-1 through 40-5 Fuel changing, blending, washing, 14-10, 18-10, 18-14, 30-3, 30-5, 30-6, 33-15, 33-20, 34-7 Gas-touched length (GTL), 32-8, 34-5 plotting as a diagnostic tool, 32-15, 33-12, 33-15

Gas tungsten arc welding (GTAW), 11-4, 11-6, 11-7 Gouging of tubes, 15-2, 15-3, 16-2, 16-3, 17-2, 17-3 Graphitization, 42-1 through 42-11 actions, 42-9 through 42-11 determining the extent of damage, 42-6, 42-11 distinguishing from dissimilar metal weld failures, 42-3 distinguishing from long-term overheating (creep), 7-9, 42-3 features of failure, 42-2, 42-3, 42-9 kinetics of growth, 42-4, 42-5 locations of failure, 42-2 long-term actions and the prevention of repeat failures, 42-8, 42-11 mechanism, 42-4, 42-5, 42-10 repairs and immediate solutions/ actions, 42-7, 42-11 root causes and actions to confirm, 42-6, 42-11 Hardness assessing changes in, 10-4, 10-5 metallurgical analysis, 6-7 Header flexibility, 39-4 Heat flux effects of high levels, 15-10, 16-11, 17-10, 18-9 measuring with Fluxdome, 9-9 measuring with a flux meter, 9-9 monitoring, 9-9 Heat recovery steam generators (HRSG), 30-1 Hideout of phosphate, 3-4, 3-8, 16-6, 16-11, 16-12 Hydrazine, 3-8, 3-9, 21-4, 27-7 Hydrogen damage, 15-1 through 15-30 actions, 15-21 through 15-30 case studies, 15-16 through 15-19 deposit characteristics, 7-1 through 7-4, 15-3 determining the extent of damage, 9-1, 9-6, 9-7, 15-13, 15-25 distinguishing from caustic gouging or acid phosphate corrosion, 7-1, 7-2, 15-3 electrochemical corrosion cell, 17-6 features of failure, 7-1, 7-2, 15-2, 15-3, 15-7, 15-21 locations of failure, 15-4

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-5

long-term actions and the prevention of repeat failures, 15-14, 15-15, 15-27 mechanism, 2-11, 2-14, 15-5 through 15-7, 15-22, 15-23, 17-6 microstructural changes, 7-3, 15-2, 15-3, 15-7 precursors, 15-21 ramifications/ancillary problems, 15-29 repairs and immediate solutions/ actions, 15-14, 15-26 root causes and actions to confirm, 7-4, 15-8 through 15-12, 15-24 Hydrogen sulfide, 18-4 Hydrostatic testing, 9-10 In-bed wastage (in bubbling-bed FBC units), 47-5 through 47-8, 47-11, 47-12 Induction pressure welds (in dissimilar metal welds), 35-2, 46-2 Influence diagram, 13-24 through 13-26, 13-30 through 13-32 Inhibitor breakdown during chemical cleaning, 25-4 Inspection, 9-1 through 9-12 acoustic monitoring, 9-8 codes and standards, 9-3 eddy current testing, 9-1, 9-2, 9-7, 9-8 importance, 9-1 liquid penetrant testing, 9-1, 9-2, 9-7 magnetic particle testing, 9-1, 9-2, 9-7 NDE for different weld types, 46-4 radiographic testing, 9-1, 9-2, 9-7 ultrasonic testing, 9-1 through 9-7 detecting microstructural changes, 9-6, 9-7, 15-13 measuring steamside oxide thickness, 9-4 through 9-6 measuring wall thickness, 9-4 measuring waterside deposits, 9-6 Union Electric technique for dissimilar metal welds, 35-12, 35-13 visual examination, 9-7 Instrumentation for cycle chemistry monitoring, 3-14 Intergranular stress corrosion cracking (see also Stress corrosion cracking), 37-1 through 37-3

Lack of fusion weld defect (see also Welding/repair defects), 45-1 through 45-3 Laning of gas passages, 14-8, 32-16, 33-15, 34-14 Larson-Miller Parameter (LMP), 10-2 through 10-5 Lay-up, 4-9 through 4-11, 27-7, 41-5, 41-6, 41-8 Lifetime, tubes (see also Remaining life of tubes and Boiler tubes, design considerations), 2-2 through 2-6, 4-5, 8-1 through 8-8, 18-12, 23-2, 23-3, 32-18, 32-19 Long-term overheating (creep), 32-1 through 32-32 actions, 32-24 through 32-32 case study, 32-22 determining the extent of damage, 32-16, 32-29 distinguishing from fireside corrosion, 6-8, 7-6, 7-8, 7-9, 32-2 through 32-6 distinguishing from graphitization, 7-9 distinguishing from short-term overheating, 32-5 features of failure, 7-6, 7-8, 32-2 through 32-6, 32-24 locations of failure, 32-6 through 32-7 long-term actions and the prevention of repeat failures, 32-18 through 32-21, 32-31, 32-32 mechanism, 32-8 through 32-10, 32-25, 32-26 precursors, 32-24 ramifications/ancillary problems, 32-32 repairs and immediate solutions/ actions, 32-17, 32-30 root causes and actions to confirm, 7-6, 7-8, 32-11 through 32-16, 32-27, 32-28 Low excess air for Nox control, 18-1, 18-7 Low melting point ashes (see Melting points of fireside ashes) Low-temperature corrosion, 30-1 Low-temperature creep cracking, 24-1 through 24-11 actions, 24-8 through 24-11 determining the extent of damage, 24-6, 24-10 features of failure, 24-1, 24-2, 24-5, 24-8 locations of failure, 24-3

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-6

long-term actions and the prevention of repeat failures, 24-7, 24-11 mechanism, 24-4, 24-5, 24-9 precursors, 24-8 repairs and immediate solutions/ actions, 24-6, 24-11 root causes and actions to confirm, 24-6, 24-10 Magnetite, strain tolerance, 1-18, 2-18, 2-20, 13-10 Maintenance, effects on boiler tube failures, 4-1 through 4-12 Maintenance damage, 44-1 through 44-6 actions, 44-3 through 44-6 as a possible cause of short-term overheating in waterwall tubing, 23-5 description of the mechanism and its manifestation, 44-1, 44-2 Maricite, 16-2 through 16-4, 16-7, 16-8 Material flaws, 45-1 through 45-6 actions, 45-4 through 45-6 description of the mechanism and its manifestation, 45-1 through 45-3 Melting points of fireside ashes coal-fired, 33-7, 33-8 oil-fired, 34-2, 34-3, 34-7 through 34-10 waste-to-energy units, 49-3 through 49-5 Membrane fins, failures associated with, 45-1 through 45-3 Metallurgical analysis, 6-1 through 6-10 fireside scale/ash analysis, 6-9 flowchart of steps for, 6-2 importance of, 6-1 metallographic samples, 6-6, 6-7 oxide scale thickness and morphology, 6-7, 6-8 required background information, 6-4 ring sampling for dimensional checks, 6-6 sample evaluation form, 6-5 sample removal and shipping, 6-4 waterside deposits/scale analysis, 6-8, 6-9 MgO - V2O5 phase diagram, 34-10

Microstructure assessing changes in austenitic stainless steels, 10-4 through 10-5 assessing changes in ferritic steels, 10-1 through 10-4 Microvoids, 10-6 through 10-8 Misaligned or distorted tubes, 14-3, 14-4, 14-11, 33-6, 33-15, 33-21, 34-5, 34-15, 34-19, 35-10, 39-5, 40-1, 40-3 Molten deposits, 2-22 through 2-24 Molten salt attack, 32-10, 33-7 Monitoring displacements and strains, 9-10 heat flux, 9-9 temperatures, 8-6, 9-8, 9-9 Multidisciplinary teams for BTF reduction, 5-2 Multilaminated oxides, 2-16, 2-17 Municipal solid waste (MSW) units, BTF issues in, 49-1 through 49-7 Nickel-based welds (in dissimilar metal welds), 11-7, 35-2, 35-3, 35-5 through 35-9, 35-15 Nitrogen blanketing (see Layup) Nucleate boiling, 2-12, 2-13 Oil composition and corrosiveness, 34-7, 34-8 effect of additives on corrosiveness, 34-9, 34-10, 34-15, 34-18, 34-19 Oil-fired boilers fireside corrosion in, 34-1 through 34-26 maintenance damage while washing, 44-1 Operation and maintenance, effects on boiler tube failures, 4-1 through 4-12 Orifice plugging, 23-5 Ovality of tubes, 24-4, 24-5 Over-fire air, 18-1, 18-7 Oxide notch, 35-3, 35-4, 35-6 Oxide thickness (see also Oxides, internal in steam-touched tubes), 2-14 through 2-21, 4-5, 4-6, 6-7, 6-8, 8-2 through 8-6, 9-4 through 9-6, 10-2, 32-9 Oxides internal in steam-touched tubes, development and breakdown, 2-14 through 2-21, 10-2, 32-2, 32-9 exfoliation, 2-17 through 2-21, 36-5 through 36-7 failure criterion, 2-18, 2-20

growth on austenitic materials, 2-17, 8-4, 8-5 growth on ferritic materials, 2-16, 2-17, 8-4, 8-5, 10-2 influence on tube metal temperatures, 4-6, 8-4, 8-5, 9-4, 9-5, 32-2 life assessment analysis of, 8-2 through 8-4 life improvement by chemical cleaning of, 4-5 measuring by ultrasonic testing, 9-4 through 9-6 metallurgical analysis of, 6-7, 6-8 spalling, 2-17 through 2-21, 36-5, 36-6 Oxides, internal in water-touched tubes, comparing most common forms, 2-7 formation, 2-6 through 2-12, 19-7 general nature of, 1-18 model explaining regular array of cracking, 13-10, 13-11 Pourbaix diagram, 13-11, 13-12 protective magnetite breakdown and resulting boiler tube failures, 1-18, 2-10, 2-11, 13-10 through 13-13 protective magnetite growth, 2-8 strain tolerance of magnetite, 2-11, 13-10 Oxygen (see also Dissolved oxygen) effect on corrosion fatigue, 13-16 through 13-20 Oxygen scavengers 3-8, 3-9, 3-10, 21-3 through 21-6 Oxygenated treatment (see also Feedwater treatment), 1-18, 3-9 through 3-11, 3-13, 19-5, 19-6, 19-11, 19-13, 19-15 effect on corrosion fatigue, 13-20 effect on growth and exfoliation, 2-21, 19-5, 19-6, 19-11, 19-13, 19-15 Pad-type thermocouples, 9-8, 9-9 Pad welding (see also Repair and replacement of boiler tubes), 11-5, 11-6, 13-27, 15-15, 16-13, 16-14, 17-11, 17-12, 22-4, 38-6, 46-2, 46-3 Personnel, importance of training, 5-2 pH depression, 13-16 through 13-20, 13-23, 13-24, 15-10 through 15-12, 15-14, 15-15 pH elevation, 17-5 Phosphate control, 3-3, 3-4, 16-6 through 16-8 Phosphate control diagrams, 3-3, 3-4, 16-7

Phosphate hideout, 3-4, 3-8, 16-6, 16-11, 16-12 Phosphate treatment (see also Boiler water treatment), 1-18, 3-3 through 3-5, 3-13, 16-6 through 16-8, 16-12, 16-14, 16-16, 16-17 effect on corrosion fatigue, 13-18, 13-20 Pitting in superheater/reheater tubes, 41-1 through 41-14 actions, 41-10 through 41-14 determining the extent of damage, 41-8, 41-12 features of failure, 41-2, 41-3, 41-10 locations of failure, 41-2 long-term actions and the prevention of repeat failures, 41-8, 41-13 mechanism, 41-4, 41-11 precursors, 41-10 ramifications/ancillary problems, 41-13 repairs and immediate solutions/ actions, 41-8, 41-12 root causes and actions to confirm, 41-6, 41-7, 41-12 Pitting in water-touched tubes (see also Chemical cleaning damage: waterwalls), 27-1 through 27-13 actions, 27-9 through 27-13 determining the extent of damage, 27-7, 27-12 features of failure, 27-2, 27-3, 27-9 initiation, 27-4 locations of failure, 27-2 long-term actions and the prevention of repeat failures, 27-7, 27-13 mechanism, 27-4, 27-5, 27-10 precursors, 27-9 ramifications/ancillary problems, 27-13 repairs and immediate solutions/ actions, 27-7, 27-12 root causes and actions to con firm, 27-6, 27-11 Plasma coating (see Coatings) PODIS (Prediction of Damage in Service) code, 35-15, 35-16 Polythionic acid, 37-5, 37-6 Post-exposure testing of tubes 10-6, 10-8 Pourbaix diagram, iron, high temperature, 13-11, 13-12 Pressure drop across circulation pumps (orifices plugging), 23-4, 23-5

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-7

Pressure drop losses in boiler, 19-5, 19-6 Protective oxide, 1-18, 2-6 through 2-21 Pyrites (effect on erosion), 2-23, 2-24, 14-5 through 14-7 Quartz (effect on erosion), 2-23, 2-24, 14-5 through 14-7, 47-6 Reducing fireside conditions, 18-1 through 18-5, 18-7 Reducing feedwater conditions, 21-3, 21-4 Refuse-derived fuel (RDF) units (see also Waste-to-energy units) 49-1 through 49-7 Remaining life computer codes, 8-3 through 8-6 NOTIS, 8-3 TUBECALC, 8-3 TUBELIFE, 8-3 through 8-6, 10-8 TUBEPRO, 8-3 Remaining life of tubes, accelerated creep rupture testing, 8-5, 8-6 assessment, 8-1 through 8-8, 32-18, 32-19 assessment methods for SH/RH tubes, 8-1 through 8-7, 32-18, 32-19, 33-17, 33-18, 34-16 assessment methods for waterwalls and economizer tubes, 8-7 assessment to optimize actions for fireside corrosion, 18-12 computer codes, 8-3 through 8-6 for graphitization in SH/RH tubes, 42-4, 42-5 improvement by chemical cleaning of SH/RH tubes, 4-5 roadmap for analysis of, 8-3 statistical analysis, 8-6, 8-7 Repair and replacement of boiler tubes (see also Welding/repair defects), 11-1 through 11-8 boiler tube buildup, 11-6 codes for weld repair, 11-3 dissimilar metal welds 11-7 general requirements, 11-4 pad welding, 11-5, 11-6, 13-27, 15-15, 16-13, 16-14, 17-11, 17-12, 22-4, 38-5, 46-2, 46-3 repair strategies, 11-1, 11-2 roadmap for weld repair, 11-2 tube section replacement, 11-4, 11-5 welding co-extruded tubes, 11-7 welding problems that can lead to boiler tube failures, 46-2

window welding (canoe piece repairs), 11-6, 11-7, 15-15, 16-13, 17-11 Residual oils, 34-7 high vanadium, 34-7 low vanadium, 34-8 Mexican, 34-8 Rifled tubes, 2-13, 15-15, 16-14, 17-12 “Ripple” magnetite, 2-10, 19-3 Root passes in welding repairs, 11-4, 11-5 Rubbing/fretting failures, 40-1 through 40-5 actions, 40-3 through 40-5 description of the mechanism and its manifestation, 40-1, 40-2 Rupture times, 23-2, 23-3 Rust on tubes following washing, 14-2, 22-1, 38-2 Sampling, 9-10 Secondary tube failures, identifying, 7-10, 7-11 Shielded metal arc welding (SMAW), 11-4, 11-6, 11-7 Shields for corrosion resistance, 33-18, 33-19, 34-16 for erosion resistance, 14-12, 22-4 Short-term overheating in superheater/reheater tubes, 36-1 through 36-16 actions, 36-12 through 36-16 case study, 36-10 determining the extent of damage, 36-9, 36-15 distinguishing from long-term overheating, 36-2, 36-3 features of failure, 36-2 through 36-4, 36-12 locations of failure, 36-3, 36-4 long-term actions and the prevention of repeat failures, 36-10, 36-16 mechanism, 36-4, 36-13 precursors, 36-12 ramifications/ancillary problems, 36-16 repairs and immediate solutions/ actions, 36-9, 36-15 root causes and actions to confirm, 36-5 through 36-8, 36-14 Short-term overheating in waterwall tubing, 23-1 through 23-14 actions, 23-9 through 23-14 determining the extent of damage, 23-7, 23-12

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-8

distinguishing among the three levels of, 7-5, 7-6, 23-2 through 23-4 features of failure, 23-2 through 23-4, 23-9 locations of failure, 23-4 long-term actions and the prevention of repeat failures, 23-8, 23-13 mechanism, 23-5, 23-10 precursors, 23-9 ramifications/ancillary problems, 23-14 repairs and immediate solutions/ actions, 23-7, 23-13 root causes and actions to confirm, 23-5, 23-6, 23-11 Shutdown of units, 4-8, 4-9, 27-6, 37-10, 41-4 through 41-6 Sigma phase of austenitic stainless steels, 10-4, 10-5, 32-3 Slagging, 2-22, 2-23, 19-6 through 19-8, 29-2, 36-12 Slagging propensity, 29-2 Solid particle erosion in the turbine, 32-32, 36-5, 36-12, 36-16 Solvent choice for chemical cleaning, 4-3, 4-4, 4-6 Sootblower erosion in superheater/ reheater tubes, 38-1 through 38-10 actions, 38-7 through 38-10 determining the extent of damage, 38-5, 38-9 distinguishing from flyash erosion, 7-9, 38-3 features of failure, 38-2, 38-3, 38-7 locations of failure, 38-3 long-term actions and the prevention of repeat failures, 38-5, 38-6, 38-10 mechanism, 38-4, 38-8 precursors, 38-7 repairs and immediate solutions/ actions, 38-5, 38-10 root causes and actions to confirm, 38-4, 38-9 Sootblower erosion in water-touched tubing, 22-1 through 22-9 actions, 22-6 through 22-9 determining the extent of damage, 22-3, 22-8 features of failure, 22-1, 22-6 locations of failure, 22-1 long-term actions and the prevention of repeat failures, 22-4, 22-9

mechanism, 22-2, 22-7 precursors, 22-6 repairs and immediate solutions/ actions, 22-4, 22-8 root causes and actions to confirm, 22-2, 22-3, 22-8 Sootblower operation and maintenance practices (see also Sootblower erosion in superheater/reheater tubes and Sootblower erosion in watertouched tubes), 22-2 Spacers, 26-3 Spalling of SH/RH steamside oxide, 2-17 through 2-21, 36-5, 36-6 Spray coatings (see Coatings) Spheroidization, 10-3, 32-3, 42-4 Stagnant water, 27-1, 27-4, 27-5, 41-2, 41-4, 41-5 Startup of units, 4-8, 4-9, 27-6 Steam blanketing, 2-12, 2-13, 15-5, 15-6, 16-6 through 16-8, 17-5 through 17-7 Steam flow redistribution, 32-19 through 32-21, 33-21, 34-19 Steam impingement, importance of identifying, 7-10, 7-11 Steam monitoring, 3-14, 41-8 Steamside oxide (see Oxides, internal in steam-touched tubes) Strain age embrittlement, 45-1 Strains, monitoring, 9-10 Stress analysis for analyzing corrosion fatigue, 13-29 Stress corrosion cracking, 37-1 through 37-16 actions, 37-12 through 37-16 case study, 37-10 determining the extent of damage, 37-8, 37-15 distinguishing from stress corrosion cracking and intergranular corrosion, 7-10 features of failure, 37-2, 37-3, 37-12 locations of failure, 37-3 long-term actions and the prevention of repeat failures, 37-9, 37-16 mechanism, 37-4, 37-5, 37-13 precursors, 37-12 ramifications/ancillary problems, 37-16 repairs and immediate solutions/ actions, 37-8, 37-15 root causes and actions to confirm, 37-6 through 37-8, 37-14

Substoichiometric fireside conditions, 18-1 through 18-5, 18-7 Sulfidation, 18-4, 33-7, 33-8 Supercritical steam properties, 19-6 Supercritical waterwall cracking, 19-1 through 19-22 actions, 19-19 through 19-22 case study, 19-16 chromizing waterwalls, 19-15, 19-16 determining the extent of damage, 19-14, 19-21 features of failure, 19-2, 19-3, 19-18 in oil-/gas-fired units, 19-10 international experience base, 19-5, 19-6, 19-16 locations of failure, 19-4 long-term actions and the prevention of repeat failures, 19-15, 19-16, 19-22 mechanism, 19-5 through 19-10, 19-19 precursors, 19-18 ramifications/ancillary problems, 19-22 repairs and immediate solutions/ actions, 19-14, 19-21 root causes and actions to confirm, 19-11 through 19-13, 19-20 Superheater/reheater chemical cleaning (see also Chemical cleaning), 4-5 through 4-7 solvent choice, 4-6 Superheater/reheater tubes, basics, 2-5, 2-6, 32-8 failure mechanisms screening table, 1-8, 1-9, 31-4, 31-5 maximum metal temperatures, 32-8, 32-9 temperature distribution in, 32-11, 32-14, 32-15 Supports, 26-3, 35-10, 35-11, 39-3, 39-4 Temperature measurements, in economizer inlet headers, 20-6, 20-7, 20-10 in SH/RH tubes, 32-11, 32-14 10 o’clock - 2 o’clock flats, 32-2, 32-10, 33-2, 33-3, Thermal-hydraulic regimes in boiler tubes, 2-12 through 2-14 conditions that lead to deposit formation, 2-13, 2-14 global, 2-12, 2-13 local, 2-13, 2-14

Thermal fatigue in economizer inlet header tubes, 20-1 through 20-19 actions, 20-14 through 20-19 assessment methodology, 20-9 case study, 20-12, 20-13 determining the extent of damage, 20-8, 20-17 distinguishing from erosioncorrosion and flexibilityinduced cracking, 7-6, 7-7, 20-4 features of failure, 20-2, 20-3, 20-14 locations of failure, 20-2 long-term actions and the prevention of repeat failures, 20-11, 20-19 mechanism, 20-4, 20-5, 20-15 precursors, 20-14 ramifications/ancillary problems, 20-19 repairs and immediate solutions/ actions, 20-9 through 20-11, 20-18 root causes and actions to confirm, 20-6, 20-7, 20-16 Thermocouples, 8-6, 9-8, 9-9 chordal thermocouples, 9-8, 9-9 pad-type thermocouples, 9-8, 9-9 Thermogravimetry analysis, 33-12, 33-15 Transgranular stress corrosion cracking (see also Stress corrosion cracking), 37-1 through 37-3 TUBELIFE, 8-3 through 8-6, 10-8 Tube blockage, 23-5, 36-5 Tube build-up, 11-6 Tube manufacturing laps, 45-1, 45-2, 45-3 Tube ovality, 24-4, 24-5 Tube temperatures increased by increasing oxide thickness, 4-5, 8-3, 8-4, 9-4 measuring via thermocouples, 8-6, 9-8, 9-9 predicted by oxide growth laws compared to thermocouple measurements, 8-4 through 8-6 Two phase flow, 2-12, 2-13 U-bends in tubes as fatigue site, 26-3, 26-4, 39-3 Ultrasonic measurement of oxide thickness, 4-6, 9-4 through 9-6, 32-11

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-9

Underdeposit corrosion, acid phosphate corrosion, 16-1 through 16-28 caustic gouging, 17-1 through 17-22 distinguishing among the types, 7-1 through 7-5 hydrogen damage, 15-1 through 15-30 in bubbling-bed FBC units, 47-3, through 47-5, 47-9 in circulating-bed FBC units, 48-2 Unit lay-up, as a cause of boiler tube failures, 4-9, 27-7, 41-1, 41-6, 41-8 options, 4-9 through 4-11, 27-7, 41-8 Unit startup and shutdown, effect on boiler tube failures, 4-8, 4-9 effect on pitting in water-touched tubes, 27-6 effect on pitting in SH/RH tubes, 41-6 effect on stress corrosion cracking in SH/RH tubes, 37-10

V2O5 - MgO phase diagram, 34-10 V2O5 - Na2O phase diagram, 34-3 Vanadates, 32-2, 32-3 Vibration in tubes as cause of fatigue, 26-6, 39-6, 39-11 Vortex shedding, 26-6, 39-6, 39-11 Waste-to-energy units, BTF issues in, 49-1 through 49-7 additives, 49-5 erosion, 49-2, 49-3, 49-7 fireside corrosion of SH/RH, 49-3 through 49-6 fireside corrosion of waterwalls, 49-3 through 49-6 high chlorides, 49-2 Water chemistry (see Boiler water treatment and/or Feedwater treatment) Waterside fireside corrosion (see Fireside corrosion in water-touched tubes) Water-steam cycle ingress, corrosion and deposition in drum units, 1-17 ingress, corrosion and deposition in once-through units, 1-17 introduction to 1-16

Chapters 1-11 can be found in Volume 1; 12-30 in Volume 2; 31-49 in Volume 3

I-10

Waterwall deposits, effect on tube metal temperatures, 19-7, 19-8 local tube conditions that can cause, 2-13, 2-14, 15-4, 15-5, 15-6, 15-8, 15-10, 16-5, 17-4 measuring by ultrasonic testing, 9-6 metallurgical analysis of, 6-8, 15-2, 15-3, 16-2, 16-3, 16-18 through 16-20, 17-2, 17-3 rate of accumulation, 2-9 Waterwalls and economizer tubes, basics, 2-4, 2-5 failure mechanisms screening table, 1-6, 1-7, 12-4, 12-5 Weld build-up, 11-6 Welding/repair defects, 45-1, 46-1 through 46-7 actions, 46-5 through 46-7 description of the mechanism and its manifestation, 46-1 through 46-4 Welding repairs (see also Repair and replacement of boiler tubes), 11-1 through 11-8 Wick boiling, 2-13, 2-14 Window welds (canoe piece repairs), 11-6, 11-7, 15-15, 16-13, 17-11

Boiler Tube Failures: Theory and Practice Volume 3: Steam-Touched Tubes

R. B. Dooley Electric Power Research Institute and W. P. McNaughton Cornice Engineering, Inc.

i

About EPRI Electricty is increasingly recognized as a key to societal progress throughout the world, driving economic prosperity and improving the quality of life. The Electric Power Research Institute delivers the science and technology to make the generation, delivery, and use of electricity affordable, efficient, and environmentally sound. Created by the nation’s electric utilities in 1973, EPRI is one of America’s oldest and largest research consortia, with some 700 members and an annual budget of about $500 million. Linked to a global network of technical specialists, EPRI scientists and engineers develop innovative solutions to the world’s toughest energy problems while expanding opportunities for a dynamic industry. EPRI . POWERING PROGRESS

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THIS BOOK WAS PREPARED BY THE ORGANIZATIONS NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, THE ORGANIZATIONS NAMED BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS BOOK, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY’S INTELLECTUAL PROPERTY, OR (III) THAT THIS BOOK IS SUITABLE TO ANY PARTICULAR USER’S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS BOOK OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS BOOK. ORGANIZATIONS THAT PREPARED THIS BOOK: ELECTRIC POWER RESEARCH INSTITUTE CORNICE ENGINEERING, INC.

This book is EPRI Licensed Material and contains a single-user, shrink-wrapped license.

ISBN 0-8033-5060-0

ORDERING INFORMATION Requests for copies of this book should be directed to the EPRI Distribution Center, 207 Coggins Drive, P.O. Box 23205, Pleasant Hill, CA 94523, (510) 934-4212. Electric Power Research Institute and EPRI are registered service marks of Electric Power Research Institute, Inc. Copyright © 1996 Electric Power Research Institute, Inc. All rights reserved.

ii

Table of Contents Volume 3: Steam-Touched Tubes

Chapter

Page

31 31.1 31.2 31.3 31.4

Introduction to Volume 3 Subject Matter and Objectives for This Volume Organization of Volume 3 Optimizing the Use of this Volume For BTF Mechanisms Not Covered by This Book

31-1 31-1 31-1 31-2 31-2

32

Long-Term Overheating/Creep Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS

32-1 32-1 32-2 32-8 32-11 32-16 32-17

33

34

SH/RH Fireside Corrosion/Coal-Fired Units Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS SH/RH Fireside Corrosion/Oil-Fired Units Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

32-18 32-22 32-23 32-24 33-1 33-1 33-2 33-7 33-12 33-15 33-16 33-17 33-22 33-22 33-24 34-1 34-1 34-2 34-7 34-11 34-14 34-15 34-16 34-20 34-21

iii

Table of Contents Volume 3: Steam-Touched Tubes (continued)

Chapter

Page

35

35-1 35-1 35-2 35-6 35-10 35-12 35-14

36

37

38

iv

Dissimilar Metal Weld Failures Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS Short-Term Overheating in SH/RH Tubing Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS Stress Corrosion Cracking Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 7. Case Study 8. References ACTIONS SH/RH Sootblower Erosion Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

35-15 35-17 35-18 35-19 36-1 36-1 36-2 36-4 36-5 36-9 36-9 36-10 36-10 36-11 36-12 37-1 37-1 37-2 37-4 37-6 37-8 37-8 37-9 37-10 37-11 37-12 38-1 38-1 38-2 38-4 38-4 38-5 38-5 38-5 38-6 38-7

Table of Contents Volume 3: Steam-Touched Tubes (continued)

Chapter

Page

39

39-1 39-1 39-2 39-5 39-5 39-7 39-7

Fatigue in Steam-Touched Tubes Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

39-7 39-8 39-9

40

Rubbing/Fretting Steam-Touched Tubes Description of Boiler Tube Failure and Its Manifestation References ACTIONS

40-1 40-1 40-2 40-3

41

Pitting in Steam-Touched Tubes Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

41-1 41-1 41-2 41-4 41-6 41-8 41-8

42

43

Graphitization Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS Chemical Cleaning Damage: SH/RH Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References ACTIONS

41-8 41-9 41-10 42-1 42-1 42-2 42-4 42-6 42-6 42-7 42-8 42-8 42-9 43-1 43-1 43-2 43-2 43-3 43-3 43-4 43-4 43-4 43-5 v

Table of Contents Volume 3: Steam-Touched Tubes (continued)

Chapter

Page

44

Maintenance Damage Description of Boiler Tube Failure and Its Manifestation References ACTIONS

44-1 44-1 44-2 44-3

45

Rubbing/Fretting Steam-Touched Tubes Description of Boiler Tube Failure and Its Manifestation References ACTIONS

45-1 45-1 45-3 45-4

46

Welding/Repair Defects Description of Boiler Tube Failure and Its Manifestation References ACTIONS

46-1 46-1 46-4 46-5

47

BTF Issues in Bubbling Bed FBCs Introduction 1. Features of Failure and Typical Locations 2. Mechanism of Failure 3. Possible Root Causes and Actions to Confirm 4. Determining the Extent of Damage 5. Background to Repairs, Immediate Solutions and Actions 6. Background to Long-Term Actions and Prevention of Repeat Failures 8. References

47-1 47-1 47-3 47-5 47-7 47-9 47-9 47-10 47-12

48

BTF Issues in Circulating Bed FBCs BTF Issues in Circulating Bed FBCs Erosion-Abrasion of Waterwall Tubes Potential for Waterside Corrosion in Horizontal Tubing References

48-1 48-1 48-2 48-4 48-4

49

Issues in BTF of Waste-to-Energy Units Introduction Types of Boilers Combusting MSW Overview of the Demands Placed on Boiler Tubes in WTE Units Mechanism: Waterwall Thinning Along the Grate Line Mechanism: Corrosion by Deposits Containing Low-Melting Point Chlorides and Sulfates Mechanism: Corrosion by Combustion Gases Mechanism: Erosion-Assisted Corrosion References

49-1 49-1 49-1 49-2 49-3

Index

vi

49-3 49-6 49-7 49-7 I-1

A:

B:

C:

BTF Mechanism Unknown

BTF Known Mechanism (Table 31-2)

Anticipating Future BTF (Table 31-3)

Compare Macroscopic Appearance to Table 12-1, Volume 2 (Water-touched) or Table 31-1, (Steam-touched) Tubes to identify candidate(s)

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes. Follow actions

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes.

Action 1a: Perform Screening Analysis: Is it possible that this boiler tube failure is caused by this mechanism?

Action 1b: Screening Analysis: ¥ Review precursor list in mechanism chapter ¥ Remove tube sample to determine extent of damage

No

Yes Action 2: Determine (confirm) mechanism

Yes

Are BTF likely to occur in the future by this mechanism?

Chapter 31 • Volume 3

Introduction and Use of Volume 3

No Action 3: Determine root cause

Action 4: Determine extent of damage or affected areas

Action 5: Implement repairs, immediate solutions and actions

Action 6: Implement long-term solutions to prevent repeat failures Action 7: Determine possible ramifications/ancillary unit problems

31.1 Subject Matter and Objectives for This Volume The primary objective of this volume is to provide the most recent knowledge about boiler tube failures (BTF) in steam-touched tubing of conventional fossil-fueled power plants. Constituent objectives are:

• To provide sufficient theory and background information so that the reader can: (i) identify boiler tube failure mechanisms, (ii) determine their root cause, and (iii) apply immediate solutions to correct the problem, and (iv) implement longer-term strategies to prevent their reoccurrence. • To provide direct, easy-to-follow actions to be taken in the event that a boiler tube failure or precursor has occurred.

31.2 Organization of Volume 3 Each chapter deals with a specific boiler tube failure mechanism. With only a few exceptions each chapter consists of two parts. The first half covers the Theory and Background about the mechanism; the second half addresses Actions to be taken.

31.2.1 Theory and Background Ð the first half of each chapter The Theory and Background matter generally includes the following topics: 1.0

Features of Failure and Typical Locations

2.0

Mechanisms of Failure

3.0

Possible Root Causes and Actions to Confirm

4.0

Determining the Extent of Damage

5.0

Background to Repairs, Immediate Solutions and Actions

6.0

Background to Long-Term Actions and the Prevention of Repeat Failures

7.0

Case Studies

8.0

References

A key part of each Theory and Background section is a Table that ties together the possible root causes, actions to confirm, immediate actions/solutions and long-term actions. It is crucial that the root cause of the damage be clearly identified so that the correct actions (short- and longterm) can be properly chosen. To fail in either identification or correction is to open the door to repeat failures.

Volume 3: Steam-Touched Tubes

31-1

31.2.2 Actions - the second half of each chapter The second half of each chapter contains Actions to be followed by the investigator or BTF team if (i) a boiler tube failure has occurred and a particular mechanism is suspected, or (ii) if a unit precursor has occurred that might lead to a future BTF by this mechanism. Note that throughout the three volumes, actions are generally marked with a special symbol, “➠”. The Actions are numbered in a manner consistent with the Theory and Background section. That is, Action 2 corresponds to Section 2.0 of the Theory and Background section; the former details specific actions to be taken to confirm the mechanism, the latter provides additional information about the mechanism, why these specified actions are to be taken and how the mechanism develops.

31.3 Optimizing the Use of this Volume Figure 31-1 shows that three avenues are open to the investigator or BTF team depending upon the status of the BTF event:

• A: BTF with mechanism unknown. If a BTF in steamtouched tubing has occurred and the mechanism is not known, then Table 31-1 should be consulted. The aim of this table is to provide a starting point for the investigation based on the macroscopic appearance of the failure and a description of typical locations. From it, a preliminary choice of mechanism can be made, then the Actions for that mechanism followed to confirm that the choice was correct. Note that as shown in Table 31-1, two BTF mechanisms (those caused by low temperature creep cracking and flyash erosion), common to both steam-touched and watertouched tubing, are covered in Volume 2.

31-2

Introduction and Use of Volume 3

• B: BTF with known mechanism. If the BTF Team has knowledge from past failures that a particular mechanism is the likely cause, then Table 31-2, an index to Volumes 2 and 3, can be used to go directly to the appropriate chapter.

• C: Anticipating future BTF. The BTF Team should continually anticipate possible failures by reviewing key unit/boiler operating events, that can lead to future BTF. Table 31-3 is a tool that can help to anticipate BTF. It is organized as a series of “unit precursors”. These are events or conditions that experience has shown should be cause for detailed evaluation of the potential for future BTF, even though no BTF has yet occurred. The process is not unlike routine inspection of components; it may take only one identification of an incipient failure to justify the cost-effectiveness of the practice to even the most cost-conscious management. Table 31-1 is organized in five sections: (1.0) inspection results or appearance, (2.0) cycle chemistry events, (3.0) maintenance-related, (4.0) operation-related, and (5.0) specific equipment events. The BTF Team or investigator may find that the best way to implement this table is to work through each precursor and ask: “Has this precursor occurred in our utility/unit?”, or “Have we taken this action recently?” If the answer to either is “yes”, then a review of the mechanism(s) indicated in the final column may be indicated. Note that this table includes both water-touched and steam-touched tubing. In compiling this table, an attempt has been made to limit the “precursor” list to those which (i) can be easily identified, (ii) are important observations and will be useful for indicating a potential BTF problem, (iii) are not direct indications of boiler tube damage (an inspection

that finds cracks at the toe of a tube/attachment weld would be a direct indicator of a BTF), and (iv) are reasonably likely to lead to a BTF based on past evidence. Clearly, it is not possible to put every possible precursor in Table 31-3, but it is hoped that two objectives are achieved. First, that forced outages by BTF are reduced by anticipating the pre-conditions to the most common mechanisms. Second, that a first step will be taken to improve the understanding of the complex, interconnected nature of cycle chemistry, operating practice, combustion processes, and maintenance effects on BTF. As a final note, the list should not pre-empt good engineering judgment. If a precursor is found that you think should be an alert of a future problem, follow it up, even if it is not in this particular list.

31.4 For BTF Mechanisms Not Covered in this Book If, having gone through the above procedure, it appears that the BTF experienced is not covered in this book, or if multiple mechanisms appear to be operative, then the generic investigation procedure shown in Figure 31-1 is still applicable. Specifically, it is important that the following sequence be followed: Understand the mechanism ¯ Determine the root cause ¯ Apply proper long-term solution Removal of a tube sample and use of metallurgical techniques should enable an understanding of the underlying damage processes (erosion, corrosion, overheating, creep, fatigue, environmentally-assisted cracking, etc.) and may facilitate assignment of the BTF to one of the categories discussed here, which will then provide additional guidance to the investigator.

A:

B:

C:

BTF Mechanism Unknown

BTF Known Mechanism (Table 31-2)

Anticipating Future BTF (Table 31-3)

Compare Macroscopic Appearance to Table 12-1, Volume 2 (Water-touched) or Table 31-1, (Steam-touched) Tubes to identify candidate(s)

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes. Follow actions

Tentative identification of mechanism(s). Go to Volume 2 (Water-touched) or Volume 3 (Steamtouched) Tubes.

Action 1a: Perform Screening Analysis: Is it possible that this boiler tube failure is caused by this mechanism?

Action 1b: Screening Analysis: ¥ Review precursor list in mechanism chapter ¥ Remove tube sample to determine extent of damage

No

Yes Action 2: Determine (confirm) mechanism

Yes

Are BTF likely to occur in the future by this mechanism?

No Action 3: Determine root cause

Action 4: Determine extent of damage or affected areas

Action 5: Implement repairs, immediate solutions and actions

Action 6: Implement long-term solutions to prevent repeat failures Action 7: Determine possible ramifications/ancillary unit problems

Figure 31-1. Flowchart of actions for identifying, evaluating, and anticipating boiler tube failures.

Volume 3: Steam-Touched Tubes

31-3

Table 31-1 Screening Table for Steam-Touched Boiler Tube Failures Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 3 (or 2 as noted)

Thick-Edged Fracture Surface Chap. 24 Volume 2

Thick-edged

Outside surface initiated, intergranular crack growth with evidence of grain boundary creep cavitation and creep voids.

Predominant in lower temperature regions in tube bends, particularly at intrados on outside surface, and other locations subject to high residual, forming, or service stresses.

Low-Temperature Creep Cracking

Thick-edged

Internal thick scales, may be accompanied by external wastage at 10 o’clock and 2 o’clock positions; generally longitudinal (axial) orientation; damage on heated side of tube.

Highest temperature locations: near material transitions, where there is a variation in gas-touched length, in or just beyond cavities, in the final leg of tubing just prior to the outlet header.

Long-Term Overheating (Creep)

32

Thick-edged, leak

Usually fusion line cracking on low alloy side of weld, circumferential orientation.

At dissimilar metal welds.

Dissimilar Metal Weld Failure

35

Thick-edged (may manifest as a pin-hole)

Cracking is transgranular or intergranular usually with significant branching; initiation can be at ID (most common) or on OD, circumferential or longitudinal orientation.

Bends and straight tubing with low spots; high stress locations are particularly susceptible at bends, welds, tube attachments, supports or spacers.

Stress Corrosion Cracking

37

Thick-edged

Transgranular cracking, OD-initiated and associated with tubing (at tube bends or attachments) or headers (particularly at the ends)

Tubing-related failures associated with attachments or bends in tubing; header-related generally at ends of header.

Fatigue

39

Thick-edged, leak

Most commonly in HAZ of C or C-Mo steel tubes; key is microstructure appearance of graphite particles or nodules

Adjacent to weld fusion line at heat affected zone most common.

Graphitization

42

Most prominent in backpass regions; bends near to walls.

Flyash Erosion

Chap. 14 Volume 2

Thin-Edged Fracture Surface Thin-edged (unless creep-assisted)

31-4

External polishing of tube surface; very localized damage

Introduction and Use of Volume 3

Table 31-1 Screening Table for Steam-Touched Boiler Tube Failures (continued) Typical Fracture Surface Appearance

Other Likely Macroscopic and Metallographic Features

Typical Locations

Possible Mechanism

Chapter in Volume 3 (or 2 as noted)

Thin-Edged Fracture Surface (continued) Thin-edged

External damage; wastage at 10 and 2 o’clock (flue gas at 12 o’clock); longitudinal cracking; perhaps “alligator hide” appearance; real key to identification will be the presence of low-melting point ash in external deposits

Highest temperature tubes: leading tubes, near transitions, tubes out of alignment, tubes around radiant cavities.

Fireside Corrosion (coal-fired units and oil-fired units)

33 (Coal-fired units)

Thin-edged

Often shows signs of tube bulging or “fish-mouth” appearance, longitudinal orientation.

Most commonly near bottom bends in vertical loops of SH/RH; outlet legs, and near material transitions.

Short-Term Overheating

36

Thin-edged, pin-hole or “thin” longitudinal blowout

External wastage flats at 45° around tube from sootblower direction, little or no ash.

First tubes in from wall entrance of retractable blowers; tubes in direct path of retractable blowers.

Sootblower Erosion

38

Thin-edged

External damage; obvious metal-tometal contact on tube surface

Rubbing/ Fretting

40

Chemical Cleaning Damage or Pitting

41 or 43

34 (Oil-fired units)

Pinhole Damage Pitting

Internal tube surface damage.

For pitting: Tubes where condensate can form and remain during shutdown: bottoms of pendant loops on either SH or RH, low points in sagging horizontal tubes.

Various Other Damage Types Depends on the underlying cause

Usually obvious from type of damage and correspondence to past maintenance activity

Depends on defect Usually thick-edged or pinholes

Care required to separate weld defects from another problem located at a weld

Maintenance Damage

44

Materials Flaws

45

Welding Flaws

46

Note: This table is based on simple, macroscopic features of failure and should be used as a guide to a particular chapter for further analysis. The more detailed discussions, starting with Actions can then be used for identification and confirmation of the actual mechanism.

Volume 3: Steam-Touched Tubes

31-5

Table 31-2 Index to BTF Mechanisms Chapter In Volume 3

Water-Touched Tubes

Chapter In Volume 2

Chemical cleaning damage in SH/RH tubes

43

Acid dewpoint corrosion

30

Dissimilar metal weld failures

35

Acid phosphate corrosion

16

Fatigue in steam-touched tubes

39

Caustic gouging

17

Fireside corrosion in SH/RH tubes (coal-fired units)

33

Chemical cleaning damage

25

Fireside corrosion in SH/RH tubes (oil-fired units)

34

Coal particle erosion

28

Corrosion fatigue

13

Steam-Touched Tubes

Flyash erosion

14 (Volume 2)

Graphitization

42

Erosion-corrosion (economizer inlet headers)

21

Long-term overheating/creep

32

Fatigue in water-cooled tubes

26

Falling slag erosion

29

Low-temperature creep cracking

24 (Volume 2)

Maintenance damage

44

Fireside corrosion (coal-fired units)

18

Material flaws

45

Flyash erosion

14

Pitting in steam-touched tubes

41

Hydrogen damage

15

Short-term overheating

36

Low-temperature creep

24

Sootblower erosion in SH/RH tubes

38

Maintenance damage

44 (Volume 3)

Stress corrosion cracking

37

Material flaws

45 (Volume 3)

Rubbing tubes/fretting

40

Pitting in water-touched tubes

27

Welding flaws

46

Short-term overheating

23

Sootblower erosion in water-cooled tubes

22

Supercritical waterwall cracking

19

Thermal fatigue of economizer inlet headers

20

Welding flaws

31-6

Introduction and Use of Volume 3

46 (Volume 3)

Table 31-3 Unit Precursors and Potential Future BTF 1.0 2.0 3.0 4.0 5.0

Inspection/Appearance Cycle Chemistry Maintenance Related Operation Related Specific Equipment

Category

Precursor

1.1 Water-touched Excessive waterside deposits ( >> 30 mg/cm2) for hightubes (waterside) pressure boilers.

1.2 Water-touched tubes (fireside)

Mechanism(s) of Concern (Chapter, Volume)

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), short-term overheating (23,V2)

Excessive waterside deposits, such as ripple Fe3O4 in oncethrough (O/T) and supercritical units.

Supercritical waterwall cracking (19,V2)

Boiler water samples that appear black (high suspended solids).

Acid phosphate corrosion (16,V2)

Corrosion/erosion in feedwater system; fouling in boiler feed pump or orifices.

• For supercritical or O/T units: supercritical waterwall cracking (19,V2) • For subcritical or non-O/T units - hydrogen damage (15,V2), acid phosphate corrosion (16,V2), or caustic gouging (17,V2) • Erosion-corrosion of economizer inlet header (21,V2)

Pressure drop across circulation pumps (orifices are plugging).

Short-term overheating in waterwall tubing (23,V2)

Flame impingement due to burner change or misalignment, leading to excessive tube deposits.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), fireside corrosion (18,V2)

Excessive furnace slagging that could lead to overheating in convective passes (or fuel change).

Short-term overheating in SH/RH tubing (36,V3)

Fresh rust found on tubes after unit washing, external flat spots, burnishing or polishing.

Flyash erosion (14,V2), sootblower erosion - waterwalls (22,V2), coal particle erosion (28,V2)

Failed tubes, any upstream tube leaks, as a warning to scout for the potential short-term overheating.

Short-term overheating in waterwall tubing (23,V2)

Significant hardness or ovality, particularly associated with tube bends, found during routine inspection.

Low-temperature creep cracking (24, V2)

1.3 Steam-touched Excessive steamside oxide (detected by UT measure of oxide tubes (steamside) thickness, or analysis of removed tube samples, evidence of excessive exfoliation like solid particle erosion in turbine). Steamside deposits in RH tubing - particularly of sodium sulfate, or high Na or SO4 levels in steam.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33&34,V3), dissimilar metal weld failures (35,V3), short-term overheating (36,V3) Pitting and failure in steam-touched tubes (41,V3)

Volume 3: Steam-Touched Tubes

31-7

Table 31-3 Unit Precursors and Potential Future BTF (continued) 1.0 2.0 3.0 4.0 5.0

Inspection/Appearance (continued) Cycle Chemistry Maintenance Related Operation Related Specific Equipment

Category

Precursor

1.4 Steam-touched tubes (fireside)

31-8

Mechanism(s) of Concern (Chapter, Volume)

Excessive flue gas temperature, displaced fireball, delayed combustion, periodic overfiring or uneven firing of burners.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33&34,V3)

High levels of excess oxygen.

SH/RH fireside corrosion: oil-fired units (34,V3)

Blockage or laning of boiler gas passages observed during boiler inspection.

Flyash erosion (14,V2), long-term overheating/creep (32,V3), SH/RH fireside corrosion: coal/oil units (33&34,V3)

Excessive temperatures measured by thermocouples in vestibule or header area.

Long-term overheating/creep (32,V3), dissimilar metal weld failures (35,V3)

Evidence of “alligator hide” appearance on external tube surface, observed during boiler inspection, associated with wall loss or thinning.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33&34,V3)

Fresh rust found on tubes after unit washing, external flat spots, burnishing or polishing.

Flyash erosion (14,V2), sootblower erosion in SH/RH (38,V3)

Significant hardness or ovality, particularly associated with tube bends, found during routine inspection.

Low-temperature creep cracking (24, V2)

Distortion or misaligned tube rows found during routine inspection.

Flyash erosion (14,V2), SH/RH fireside corrosion (33&34,V3), dissimilar metal weld failures (35,V3), fatigue of steam-touched tubing (39,V3), rubbing/fretting (40,V3),

Failed tube supports and lugs, location of dissimilar metal welds close to fixed supports.

Fatigue of steam-touched tubing (39,V3), dissimilar metal weld failures (35,V3)

Introduction and Use of Volume 3

Table 31-3 Unit Precursors and Potential Future BTF (continued) 1.0 2.0 3.0 4.0 5.0

Inspection/Appearance Cycle Chemistry Maintenance Related Operation Related Specific Equipment

Category

2.1 All units

2.2 Units on Phosphate Treatments

2.3 Units on AVT

2.4 Units on Caustic Treatment

Precursor

Mechanism(s) of Concern (Chapter, Volume)

Problem with high levels of feedwater corrosion products; operating ranges for pH, cation conductivity or dissolved oxygen consistently outside recommended ranges, including persistent reducing conditions or excessive use of oxygen scavengers.

Corrosion fatigue (13,V2), hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), waterwall fireside corrosion (18,V2), supercritical waterwall cracking (19,V2), erosion/corrosion in economizer inlet header (21,V2), short-term overheating in waterwall tubing (23,V2),

Carryover of volatile chemicals from boiler, such as NaOH for units on caustic treatment, or excess of Na, SO4, and/or chloride; steam limits exceeded.

Stress corrosion cracking (37,V3), pitting in steam-touched tubes (41,V3)

Major acid contamination event (pH < 8) when unit is at full load; condenser leak, or breakdown of makeup or condensate polisher regeneration chemical.

Hydrogen damage (15,V2)

Evidence of a persistent problem with phosphate hideout, particularly where mono-sodium and/or an excess of di-sodium phosphate has been added to the boiler.

Acid phosphate corrosion (16,V2)

Persistent phosphate hideout with phosphate return causing a pH depression (7-8).

Corrosion fatigue (13,V2)

Caustic level in excess of that necessary for optimal control (>> 2 ppm).

Caustic gouging (17,V2)

Caustic, used in excess of that necessary for optimal control of contaminant ingress (to counteract pH depressions on startup).

Caustic gouging (17,V2)

pH depression during shutdown and early startup (pH around 7-8). Hideout/return of sulfate.

Corrosion fatigue (13,V2)

Caustic, used in excess of that necessary for optimal control (>> 2 ppm).

Caustic gouging (17,V2)

Volume 3: Steam-Touched Tubes

31-9

Table 31-3 Unit Precursors and Potential Future BTF (continued) 1.0 2.0 3.0 4.0 5.0

Inspection/Appearance Cycle Chemistry Maintenance Related Operation Related Specific Equipment

Category

Precursor

3.1 Chemical cleaning

3.2 Repairs

31-10

Mechanism(s) of Concern (Chapter, Volume)

Evidence of shortcoming in chemical cleaning process such as inappropriate cleaning agent, excessively strong concentration or long cleaning time, too high a temperature, failure to neutralize, breakdown of inhibitor, inadequate rinse.

Chemical cleaning damage in waterwalls (25,V2) or SH/RH (43,V3), short-term overheating (23,V2 & 36,V3).

Shortcoming in SH/RH cleaning process such as inadequate rinse, improper flow verification.

Short-term overheating in SH/RH tubing (36,V3)

Evidence that level of Fe in cleaning solution continued to increase instead of leveling out when cleaning process was ended.

Chemical cleaning damage in waterwalls (25,V2) or SH/RH (43,V3)

Need for excessive cleaning in supercritical units (interval < 2 years).

Supercritical waterwall cracking (19,V2)

Contamination in SH/RH (particularly by chlorides) during chemical clean of SH/RH (breakdown of inhibitors or improper flushing of solvents) or waterwalls (caused by poor backfill procedures that failed to protect SH circuits).

Stress corrosion cracking (37,V3)

In water-touched tubes: use of backing rings, pad welds, canoe pieces, weld overlay that penetrates to inside surface as a source of flow disruption and excessive deposits.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2)

Application of shielding, baffles, palliative coatings to mitigate flyash erosion without the use of a cold-air velocity test.

Flyash erosion (14,V2)

In water-touched tubes, Cu in water-side deposits.

Hydrogen damage (15,V2), welding defects (46,V3)

Introduction and Use of Volume 3

Table 31-3 Unit Precursors and Potential Future BTF (continued) 1.0 2.0 3.0 4.0 5.0

Inspection/Appearance Cycle Chemistry Maintenance Related Operation Related Specific Equipment

Category

4.1 Startup Procedures

4.2 Combustion conditions

4.3 Fuel choices and changes

4.4 Cycling

Precursor

Mechanism(s) of Concern (Chapter, Volume)

Feedwater introduced intermittently into economizer inlet at high flow rates during startups and particularly during off-line top-ups.

Economizer inlet header thermal fatigue (20,V2)

Rapid unit startups that cause the reheater to reach temperature before full flow starts (no furnace exit gas temperature control).

SH/RH fireside corrosion (33&34,V3)

Heat flux change caused by change to higher BTU-value coal, dual firing with gas, changeover to oil- or gas-firing leading to excessive tube deposits in waterwalls; new burners causing impingement.

Hydrogen damage (15,V2), acid phosphate corrosion (16,V2), caustic gouging (17,V2), fireside corrosion (18,V2)

Implementing low excess air strategies for NOx control and the potential for waterwall fireside corrosion (note that unlike the other precursors in this Table, this is a possibility based on understanding the mechanism; to date no failures have been directly attributed to this cause).

Waterwall fireside corrosion (18,V2)

Operation with high levels of excess oxygen in oil-fired units (> 1%).

SH/RH fireside corrosion in oil-fired units (34,V3)

Change to a fuel that either contains more ash or contains elements which are more erosive such as quartz.

Flyash erosion (14,V2)

Change to a more corrosively-aggressive coal, particularly one high in chlorine, Na, K, or S contents.

Waterwall fireside corrosion (18,V2), acid dewpoint corrosion (30,V2), SH/RH fireside corrosion (33&34,V3)

Use of Mg-based additives (oil-fired units) leading to coating of waterwalls, reflecting heat into convection passes.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion in oil-fired units (34,V3)

Conversion of the unit to cycling operation or an increase in the number of cycles.

Corrosion fatigue (13,V2), economizer inlet header thermal fatigue (20,V2), fatigue in water-touched (26,V2) or steamtouched tubing (39,V3), dissimilar metal weld failures (35,V3)

Volume 3: Steam-Touched Tubes

31-11

Table 31-3 Unit Precursors and Potential Future BTF (continued) 1.0 2.0 3.0 4.0 5.0

Inspection/Appearance Cycle Chemistry Maintenance Related Operation Related (continued) Specific Equipment

Category

Precursor

4.5 Shutdown or layup

4.6 Other

31-12

Mechanism(s) of Concern (Chapter, Volume)

Evidence of a shortcoming during unit shutdown/layup such as uncertainty about water and/or air quality during period, insufficient nitrogen blanketing, insufficient N2H4, evidence of air inleakage.

Pitting in water-touched (27,V2) or steam-touched tubes (41,V3), and maybe corrosion fatigue (13,V2)

Indication that stagnant, oxygenated water may have rested in tubes during shutdown or layup particularly in economizer and RH.

Pitting in water-touched (27,V2) or steam-touched tubes (41,V3)

Evidence that condensate is forming in SH/RH bends during unit shutdown, exacerbated if steam purity is not good (as determined by elevated levels of SO4).

Short-term overheating in SH/RH tubes (36,V3), pitting in steam-touched tubes (41,V3)

Operation above the maximum continuous design rating, with excess air flow settings above design, with unbalanced fans or air heaters - leading to nonuniform gas flows.

Flyash erosion (14,V2)

Low drum level.

Short-term overheating (23,V2)

Long-Term Overheating/Creep

Table 31-3 Unit Precursors and Potential Future BTF (continued) 1.0 2.0 3.0 4.0 5.0

Inspection/Appearance Cycle Chemistry Maintenance Related Operation Related Specific Equipment

Category

5.1 Condensers

Precursor

Mechanism(s) of Concern (Chapter, Volume)

Major condenser leaks or minor leaks that have occurred over a long period of time.

Hydrogen damage (15,V2)

Condenser leak leading to condenser cooling water constituents in attemperator spray water.

Stress corrosion cracking (37,V3)

Upset in water treatment plant or condensate polisher regeneration chemicals leading to low pH condition in boiler (pH < 8).

Hydrogen damage (15,V2)

Upset in water treatment plant or condensate polisher regeneration chemicals leading to high pH condition.

Caustic gouging (17,V2)

Carryover test indicates high mechanical carryover.

Stress corrosion cracking (37,V3), pitting in steam-touched tubing (41,V3)

Operating with high drum level allowing excessive carryover into steam.

Pitting in steam-touched tubing (41,V3)

5.4 Sootblowers

Poor sootblower maintenance.

Sootblower erosion in waterwalls (22,V2), SH/RH sootblower erosion (38,V3)

5.5 Low temperature headers

Header has large number of operating hours, has experienced large thermal gradients, spacing of ligament holes is small (< 3.5 cm), header thickness is well above Code minimum, header-to-stub tube joints made with partial penetration welds.

Economizer inlet header thermal fatigue (20,V2)

5.2 Water treatment plant/condensate polisher

5.3 Drum

5.6 High temperature Excessive relative movement of header/ tube during unit headers transients, restricted movement, header is not allowed to expand freely (maybe ash-related), unit change to cycling.

Fatigue in steam-touched tubing (39,V3).

5.7 Turbine

A problem with solid particle erosion (SPE) in the turbine.

Short-term overheating SH/RH tubing (36,V3), long-term overheating /creep (32,V3)

5.8 SH/RH Circuit (redesign)

Redesign of the SH/RH circuit may change the absorption patterns through other SH/RH sections and increase tube temperatures.

Long-term overheating/creep (32,V3), SH/RH fireside corrosion (33 & 34,V3), dissimilar metal weld failures (35,V3)

5.9 Supports/ Attachments (redesign)

Addition of supports without consideration of their impact on the stresses of dissimilar metal welds.

Dissimilar metal weld failures (35,V3)

Redesign of waterwall tube attachments to increase flexibility without analysis to determine whether solution is actually beneficial.

Corrosion fatigue (13,V2)

Volume 3: Steam-Touched Tubes

31-13

31-14

Long-Term Overheating/Creep

Chapter 32 • Volume 3

Large fracture opening

Long-Term Overheating/Creep

Introduction Superheater (SH) and reheater (RH) tubes operate in a regime where creep is significant and oxidation resistance is important. As discussed in Chapter 2, Volume 1, SH/RH tube design involves a choice of wall thickness and alloy type to withstand the expected pressures and temperatures and still provide at least a minimum specified level of life. Creep damage is strongly dependent on stress level and on tube metal temperature. Therefore conditions which are only a slight departure from design levels leading to increased stress (such as wall thinning by oxidation, corrosion or erosion) or increased temperature (such as caused by the buildup of internal oxide scale) will result in operating tube lives that are significantly shorter than expected. In a typical SH/RH tube, each 0.001 inch (0.0254 mm) of internal oxide buildup increases the tube metal

temperature by approximately 3°F (1.7°C).1 Thus, for a 0.020 inch (0.51 mm)-thick internal scale, tube metal temperatures increase by about 60°F (~ 33°C), which corresponds to a five-fold increase in accumulation of damage by creep. Higher than expected temperatures in SH/RH tubes can be a contributor to a number of damage types. The reader is specifically directed to discussions of other mechanisms which might be confused with longterm overheating such as fireside corrosion (Chapter 33 for coal-fired units and Chapter 34 for oil-fired units), and short-term overheating (Chapter 36). Failure by high temperature creep remains a significant cause of boiler tube failures in the United States. The proper assessment of the remaining life of SH/RH tubes is one of the primary challenges of an integrated program for boiler tube failure prediction and control.

Volume 3: Steam-Touched Tubes

32-1

1. Features of Failure and Typical Locations Long-Term Overheating/Creep: Identification Keys 1. Final failure by creep will manifest features of low ductility, usually longitudinal (axial) in orientation. In superheater tubing, failures will be thickedged, they will be somewhat thinner in reheater tubing because of the thinner walls. 2. Thick, internal oxide scales, often longitudinally cracked, are indicative of overheating in ferritic materials. 3. Microstructural features indicative of overheating and creep damage will be present. In ferritic materials such features include spheroidization and decreased fireside surface hardness. In austenitic stainless steels, the presence of sigma phase microstructure, sensitization and grain boundary cavities will be characteristic. 4. Care should be taken to distinguish failures by a longterm overheating mechanism from those caused primarily by fireside corrosion. The presence of low melting point ash components on the external tube surface and a higher ratio of wall thinning to steamside oxide scale buildup will be characteristic of fireside corrosion.

Precursors to final failure may include wastage flats on the tube at the 10 o’clock and 2 o’clock positions (flue gas at 12 o’clock) with the maximum amount of “alligator hide” in the middle of the flats. Note that the flats can be present at other tube locations depending on tube alignment in the bundle.

1.1 Features of failure Final failures by long-term overheating/creep are generally longitudinal (axial to the tube), and located on the heated side of the tube. Failures are generally thick-edged, corresponding to low ductility. The range of final failure appearance in ferritic materials is shown schematically in Figure 32-1. A typical appearance is shown in the failed tube pictured in Figure 32-2. In reheater tubes, because of the thinner-walled materials, final failures tend to look more ductile than in superheater tubes.

Microscopically, there will be evidence of damage to the material microstructure by overheating and intergranular or transgranular creep; distinguishing features will be material and stress level dependent.

Primary evidence for the overheating of SH/RH tubes is thickened external scales, often with Y-shaped grooves which give the appearance of an alligator hide. Often there are also thick, internal oxide scales, cracked longitudinally.

Ferritic materials will demonstrate: • A critical thickness of steamside oxide, which may often be exfoliated. A discussion of the growth and exfoliation of such oxide layers is presented in Chapter 2, Volume 1.

BLISTER OVERHEATING Localized Axially

A

FISHMOUTH OVERHEATING General Along Tube Axis

A

A

A Large fracture opening

Little or no creep swelling away from blistered area

Small fracture opening

Measurable creep swelling usually greater than 10% in very rapid overheating. In longer term failures less creep swelling is present, but there is always some measurable creep, even if only 1%, in very long-term stress rupture.

Little or no wall loss in non-blistered area "A-A"

"A-A"

Some creep reduction in wall away from fracture in rapid overheating

Considerable wall reduction (knife edge) in very rapid failure. Fracture edge in longer term overheating shows less of a reduction, with extremely long-term failures showing little creep wall loss.

Figure 32-1. Types of high temperature creep failure in ferritic tubing. Source: G.A. Lamping and R.M Arrowood, Jr.2

32-2

Long-Term Overheating/Creep

• A spheroidized microstructure as illustrated in Figure 32-3. Chapter 10, Volume 1 discusses levels of spheroidization and their interpretation; additional information can be found in reference 1. • A reduction in material hardness, usually with a decarburized layer at the fireside interface. This effect is illustrated in Figure 32-4. Austenitic stainless steels will manifest damage with sigma phase microstructure and grain boundary cavities. Note that improper etching may etch out sigma phase and give the appearance of creep cavities.4 Graphitization may be present in carbon or carbon molybdenum steels. The nucleation, growth and interlinking of creep voids or cavities, termed creep cavitation, will occur at the crack tip, as shown in Figure 32-5. However, such voids will not be found in the bulk of the tube away from the crack.

Figure 32-2. Long-term creep failure in a 2 1/4 Cr- 1 Mo reheater tube.

It is important that the root cause of the failure be identified accurately. In particular, the fireside wastage mechanisms, such as fireside corrosion or erosion, can lead to wall thinning which will then result in the accumulation of creep deformation. Before settling on long-term overheating as the primary damage mechanism, it is important to evaluate whether these other mechanisms are also contributors. Table 32-1 lists some key distinguishing differences among longterm overheating, short-term overheating, and fireside corrosion in SH/RH tubing. This topic is also discussed in Chapter 7, Volume 1.

Figure 32-3. Enlarged view of the secondary cracking and spheroidization associated with long-term creep damage. Source: S.R. Paterson, et al.1

Volume 3: Steam-Touched Tubes

32-3

With a Knoop indentor and a 300 gram load 80 HRB

Knoop Hardness

160 75 HRB

150

70 HRB

140

65 HRB

130

Cold side away from rupture

Rockwell Hardness Conversion

170

120 0

0.1

0.2 0.3 0.4 0.5 0.6 0.7 Distance from Rupture Line (inch)

0.8

0.9

Figure 32-4. Microhardness traverse to illustrate the loss in hardness in the vicinity of a long-term overheat failure. Source: S.R. Paterson, et al.1

Figure 32-5. Typical grain boundary creep cavitation/microcracking at and adjacent to a crack. Source: J. Hickey, Irish Electricity Supply Board

32-4

Long-Term Overheating/Creep

Table 32-1 Comparison of Characteristics of Long-Term Overheating/Creep, Short-Term Overheating, and Fireside Corrosion (Coal-Fired Units) In Superheater/Reheater Tubing Characteristic

Long-Term Overheating

Short-Term Overheating

Fireside Corrosion

Fracture Surface and Appearance of Failure

• Generally thick-edged, brittle final failure. • Generally accompanied by external tube wastage at the 10 o’clock and 2 o’clock positions.

• Usually thin-edged, ductile final failures. • Swelling of tubes without ovalization. • “Fish-mouth” appearance of tube rupture.

• Tube wastage, particularly at the 10 and 2 o’clock positions. • Longitudinal cracking, final failure can be (but not necessarily) by overheating.

Internal Scale?

Yes, generally extensive, multilaminated and exfoliating.

Not necessarily thick. Depends on age of tube at failure.

Yes, particularly if tube metal overheating was an influencing factor.

External Scaling?

•Yes, thick, laminated and often longitudinally cracked. • Usually two layers - (i) a hard, porous outer layer with composition typically that of flyash, and (ii) a black glossy inner layer mostly oxide but may contain some sulfates and sulfides of iron.

Not necessarily thick.

Yes, with multi-layers: (i) a hard, porous layer - composition typically of flyash, (ii) an intermediate layer containing complex alkali sulfates, and (iii) a black, glossy inner layer mostly of oxides, sulfates, and sulfides of iron.

Outside surface appearance after removal of scale/deposits

Characteristic longitudinal grooving and pitting (“alligator hide”).

Swelling, stretch marks on tube metal.

Characteristic longitudinal grooving and pitting (“alligator hide”). Sometimes the corroded area is smooth and featureless. Sometimes “orange peel” appearance at extremities of severe corrosion.

Composition of External Scales/ Deposits

Does not contain low melting point ash compounds such as alkali iron sulfates

Not relevant.

Does contain low melting point compounds such as alkali-iron sulfates (coal-fired units).

Wall Thinning?

Typically wastage flats at 10 o’clock and 2 o’clock positions caused by accelerated oxidation. Can be at other locations depending on tube position. There is always a layer of oxide adjacent to the tube.

Only because of bulging of tube material.

Primary feature of failure, may be worse at the 10 and 2 o’clock positions. Can be at other locations depending on tube position. Depending upon the rate of corrosion, a protective oxide layer may remain on the tube or may have been fluxed off.

Ratio of wall loss to steamside oxide thickness

Typically less than 3:1

Not relevant.

Typically greater than 3:1; for ratios greater than 5:1 fireside corrosion or erosion is the dominant mechanism.

Tube Material Degradation

Yes, generally extensive signs of overheating and/or of creep damage, particularly near crack tips. Creep voids will not be found removed from crack tip.

Depends on the material and the maximum temperature reached.

If overheating has been a problem, yes; otherwise, no. Fireside corrosion can occur in a tube at design temperatures.

Change in material hardness

Localized softening near the rupture is typical.

Localized hardening near the rupture is likely.

Hardness change is not necessary; if there has been no overheating, there will be no change in hardness.

Volume 3: Steam-Touched Tubes

32-5

Note that long-term overheating, where the damage accumulates because of temperatures that are somewhat over the design level, should also not be confused with short-term overheating where a pressurized tube is heated to well above its design temperature and failure occurs in a much shorter time. SH/RH tube failures by shortterm overheating are the subject of Chapter 36.

1.2 Locations of failure Figures 32-7 a and b show the typical locations of damage by longterm overheating/creep. The following locations are susceptible5:

• The most common location is near material changes, such as in the middle of a tube circuit just

32-6

Long-Term Overheating/Creep

Reheater Tube With Long-term Overheating/Creep 0.2

Wall Loss (Inches)

In evaluating the contributing mechanisms, distinguishing between fireside corrosion and long-term overheating is perhaps the most common cause of confusion. A key difference is that if corrosive attack has occurred, external deposits will contain low melting-point ash components. Figure 32-6 illustrates a second, although less definitive, means of distinguishing between the two damage types: the ratio of loss of wall thickness to internal oxide thickness. If this ratio is large, certainly greater than five and perhaps as low as three or more, then a wastage mechanism such as fireside corrosion is operative. If the ratio is small, i.e., relatively little wall thinning relative to the build-up of steamside oxide, then overheating is probably the predominant problem.

Y = 5X

Flue gas flow Probable mechanism is fireside corrosion

0.15 A B

F

Y = 3X

0.1 C

E

A B

D

0.05

D E C

0 0

Probable mechanism F is long-term overheating/creep

0.01 0.02 0.03 Steamside Oxide Scale Thickness (Inches)

0.04

Figure 32-6. Schematic representation of steamside oxide scale thickness versus tube wastage (wall loss). Such a plot can be used to distinguish between overheating/long-term creep and fireside corrosion mechanisms. Source: S.R. Paterson, et al.1

before the change to a higher grade of material. This location represents the severest condition for the lower grade material as it is subjected, even under normal conditions, to the highest temperatures. • Near changes in wall thickness in the same alloy. • Where there is a variation in the gas-touched length among tubes of the same material, leading to the earlier failure of those subjected to higher temperatures.

• On the lowest tube in a horizontal platen or the leading tube in a pendant section. • Either in a cavity or just beyond a cavity, where radiant effects can lead to higher tube temperatures. • In the final leg of tubing just before the outlet header, where steam temperatures are the highest.

(a) VerticalTube Circuits 321H

E F D B G C A H N M

Outlet header P

T-22

T-11

Hanger

Hanger

Hanger

(b) Horizontal Tube Circuits

Front water wall

D C E G B F H A

Inlet header

Figure 32-7. Typical boiler locations where long-term overheating/creep failures can occur in vertical platen elements (a) and in horizontal platens (b). Solid circles represent typical failure locations; letters in circles indicate locations of interest described in the main text in Section 3.2.

Volume 3: Steam-Touched Tubes

32-7

2. Mechanism of Failure Long-Term Overheating/Creep: Mechanism 1. Superheater/reheater tubing is designed for finite, but long (> 100,000 hours) lifetimes by allowing for a prescribed amount of creep damage. 2. A number of factors can lead to elevated tube temperatures; only slight elevation (about 1015°F) of the temperature above that anticipated in the design can lead to a significant reduction in tube life. 3. The growth of internal scale is a major contributor to increased tube temperatures; careful diagnosis of the extent of scaling can provide useful information about past service temperatures as well as about expected remaining tube life. 4. Final failure occurs by creep attributable to degraded material, reduced wall thickness, increasing hoop stresses, reduced wall thicknesses and cracking of the internal and external oxides.

2.1 Introduction Chapter 2, Volume 1 discusses the fundamental constraints associated with the design of and material choices for SH/RH tubing. A few brief remarks are made here to set the stage for the discussion of the creep damage mechanism. In the absence of other boiler tube failure mechanisms, it is not unexpected that creep will be the life-limiting damage type for SH/RH tubing. In fact, such tubing is designed for a finite life based on a conservative creep criterion. Design temperatures can range from 400 to 600°C (752 to 1112°F) depending upon location. Increasingly higher tube metal temperatures demand either increased wall thickness and/or material changes. Carbon steel can therefore be used in primary stages, whereas low-alloy steels which exhibit increased creep and oxidation resistance are used for most of the SH/RH, except for the finishing stages where austenitic stainless tubes are normally used. On the basis of laboratory oxidation experiments each manufacturer specifies a maximum operating temperature for each material. Representative values are shown in Table 32-2. There can be major differences (100-150°F) between the operating temperature and the temperature expected for the design. This can lead to significant shortening of actual tube life. Figure 32-8 shows the effects of stress and temperature on the length of time to failure for a typical boiler tube material.

32-8

Long-Term Overheating/Creep

A number of other factors, discussed in the root cause section below, can result in higher than expected temperatures. In addition to poor initial design (such as tubes with gas-touched length longer than the design estimate), the main factors are (i) the growth of internal oxide which insulates the tube from the cooling effects of the steam flow, (ii) the loss of wall thickness caused by accelerated oxidation in the case of overheating, but may also be caused by fireside corrosion or erosion, or (iii) excessive operating conditions such as high flue gas temperature, displaced fireball, periodic overfiring, or uneven firing of fuel burners, etc.

2.2 The accumulation of creep damage in superheater/reheater tubes The effects on material microstructure that occur as a result of exposure to temperatures consistent with creep damage are discussed in Chapter 10, Volume 1. Here the manifestation of creep in SH/RH tubes is reviewed. The accumulation of creep damage can occur through an elevation of tube metal temperature or stress as noted above. Tube stresses increase primarily as a result of localized wall thinning and correspondingly higher hoop stresses. Processes which occur on both the steamside and fireside of the tube can contribute to premature tube failures by creep.

Table 32-2 Maximum Tube Metal Temperatures ASME Max.1 °F (°C)

B&W Max.2 °F (°C)

C-E Max.3 °F (°C)

Riley Max.4 °F (°C)

SA-178 C

1000 (538)5, 6

950 (510)

850 (454)

850 (454)

SA-192

1000 (538)5, 6

950 (510)

850 (454)

850 (454)

SA-210 Al

1000 (538)5, 6

950 (510)

850 (454)

850 (454)

SA-209 T1

1000 (538)7

—

900 (482)

900 (482)

SA-209 T1a

1000 (538)7

975 (524)

—

—

SA-213 T11

1200 (649)

1050 (566)

1025 (552)

1025 (552)

SA-213 T22

1200 (649)

1115 (602)

1075 (580)

1075 (580)

SA-213 321H

1500 (816)

1400 (760)

—

1500 (816)

SA-213 347H

1500 (816)

—

1300 (704)

—

SA-213 304H

1500 (816)

1400 (760)

1300 (704)

—

Tube Steel Type

ASME Spec. No.

Carbon steel

Carbon Moly

Chrome Moly

Stainless

Notes: 1. From reference 6, Table PG-23.1. This is the highest metal temperature for which maximum allowable stress values are given. 2. From reference 7, 1978 edition, p. 29-11, Table 3. 3. From reference 8, 1981 edition, p. 6-43, Table IV. 4. From reference 9, 1983 edition, p. 263, Table VI. 5. Upon prolonged exposure to temperatures above about 800°F (427°C), the carbide phase of carbon steel may be converted to graphite. 6. Only killed steels shall be used above 850°F( 454°C). 7. Upon prolonged exposure to temperatures above about 875°F (468°C), the carbide phase of carbonmolybdenum steel may be converted to graphite. Direct comparison of maximum metal temperatures is not meaningful without information on design heat transfer analysis and actual material properties. Compilation source: G.A. Lamping and R.M Arrowood, Jr.2

2.2.1 Steamside processes A protective oxide forms on carbon and low-alloy steels when they are exposed to steam. The initial formation is rapid and “parabolic” in nature in the absence of heat flux. The oxide formed is stable and tenacious. In addition to a protective function, oxide layers on the steamside of carbon and low-alloy steel tubing are important because they provide useful information about the service temperatures to which the tube has been exposed. An excess of steamside scale can lead to an increased tube temperature which is a contributor to damage in superheater/reheater tubes by long-term and short-term overheating, fireside corrosion, and graphitization mechanisms. Three solid phase iron oxides can form: wustite (FeO), magnetite (Fe3O4), and hematite (Fe2O3), depending on oxygen activity and temperature. The various oxides form in layers in the order listed from the tube metal outward, e.g., wustite, if it forms will be between the tube metal and magnetite; hematite is normally found in the outer layer of steamside oxide. The oxides of steam-touched materials normally form in two distinct layers; the inner layer forming by the inward diffusion of oxygen ions (oxidizing the base metal) and the outer layer growing by the outward diffusion of iron.10 The buildup of steamside scale insulates the tubes from cooling steam which is reflected in an increase in metal temperature. This also accelerates the accumulation of creep damage and will eventually lead to failure by creep, even in the absence of all the external tube effects described above. As the metal temperature increases, the rate of oxide growth also increases, leading to even faster failure times. A detailed discussion of the mechanism of steamside oxide growth and the various effects of exfoliate of that oxide can be found in Chapter 2, Volume 1.

Volume 3: Steam-Touched Tubes

32-9

2.2.2 Fireside effects In the normal (design) condition, a protective oxide scale forms on the exterior of SH/RH tubing. Ash accumulates on the side of the tube facing into the flow of flue gas as shown in Figure 32-9a. On a tube that operates above the design condition, there can be a reduced thickness of ash at the 10 and 2 o’clock positions (flue gas flow at 12 o’clock) due to aerodynamic factors. Several consequences result: (i) the thinner layer of ash at the 10 and 2 o’clock positions allows for higher heat transfer rates and more radiant effects which raise the temperature of the tube, (ii) the easier access to the tube metal allows for increased oxidation rate leading to thinner tubes and thus higher stress levels, and (iii) an increased corrosion rate occurs (oxidation in a sulfur-containing environment) which also leads to material removal and thus higher stress levels. The latter two consequences also change the kinetics of oxide growth from parabolic to linear and, as a result, it becomes nonprotective (Figure 32-9b).

Stress, 1000 psi 50 1.0 hr. to rupture 40

10

30

100 20 1000 10

10,000 100,000

0 1000

1100 1050 1150 Temperature, °F

1200

1000 psi = 6.9 MPa

Figure 32-8. Time-to-rupture curves for 2 1/4 - 1 Mo SA-213 T22 steel.

Operating Above Design Temperature

Normal

Wastage flats

Steam

Ash

Flue gas

Corrosion rate

(b) Linear (locally)

Steam (a) Ash

Parabolic Time

(a)

Thick insulating ash layer

Flue gas

Thin ash layer

(b)

Figure 32-9. Schematic of the development of long-term overheating failures. (a) Represents the normal situation. (b) Shows the development of "flats" at 10 and 2 o'clock. The flats can be present at other tube locations, such as the tube crown, depending on tube alignment in the bundle.

32-10

Long-Term Overheating/Creep

The increasing stress level and overtemperature in the tube accelerate the creep process leading to tube failure. The process can be further accelerated if tubes are also subject to attack by other corrosive processes, such as “molten salt” attack, the topic of Chapters 33 (for coal-fired units) and 34 (for oil-fired units). Tubing attack by erosive processes, such as flyash erosion (Chapter 14, Volume 2) can further exacerbate the tubing degradation. In summary, creep damage will accumulate at a rate higher than the design estimates if there is (i) a relatively continuous period of slight overheating such as caused by shortcomings in the initial design, (ii) a slowly increasing level of temperature such as by increasing internal oxide thickness, or stress caused by decreasing wall thickness, or (iii) the accumulation of periods of excessive overheating due to operating factors.

3. Possible Root Causes and Actions to Confirm Long-Term Overheating/Creep: Root Causes 1. Overheating of the tube metal by a variety of causes, and/or increased tube stress level primarily because of wall thinning, is at the root of tube failures by a creep mechanism. 2. Important causes relate to inadequate initial design and material choices. These are exacerbated by the normal increase of oxide thickness.

3.1 Introduction Table 32-3 summarizes the major root cause influences, actions to confirm each, and corrective actions. Verification of long-term overheating/creep, independent of cause, can be direct or indirect. Direct verification and an indicator of the current temperatures can be obtained by thermocouples. Indirect verification and an indication of the history of tube metal temperatures can be obtained from analysis of the steamside oxide scale. Independent of the source of overheating the following four actions are used to confirm that the condition exists. Additional actions pertinent to specific root causes are provided below. (a). Direct measurement by thermocouples. Thermocouples can be used to measure tube temperatures directly and also to determine furnace gas temperatures; in the latter case, they verify abnormal flow patterns. The placement of thermocouples and interpretation of results is complicated by the variations in temperature around an individual tube (Figure 32-10) and across a superheater or reheater (Figure 32-11). General guidance can be obtained from the temperature profiles across the unit that are recorded in the

vestibule or header area on the permanently installed operating thermocouples. More detail about thermocoupling of SH/RH tubes can be found in Chapter 9, Volume 1. (b). Perform metallographic analysis of tube samples. Ferritic steels form an oxide scale in high temperature steam. The thickness of the layer formed is a function of tube metal temperature and time. Given one of the variables, usually time, the other can be determined. The thickness of the oxide scale can be measured nondestructively or directly via tube sampling. Tube sampling can also be used to check for blockages and flow restrictive deposits. Metallographic analysis is used to determine the degradation of the tube material. (c). Measure steamside scale buildup non-destructively. Ultrasonic and eddy current methods have been applied to measure steamside scale non-destructively. An estimate of the oxide thickness, in conjunction with analysis as described in Chapter 8, Volume 1 can then estimate tube metal temperatures. (d). Visual examination to look for evidence of slag buildup, laning, bowed or misaligned tubes acting as leading tubes.

Volume 3: Steam-Touched Tubes

32-11

Table 32-3 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

3.1 All causes of overheating • Local repairs as appro(a). Direct measurement of temperatures priate. by thermocouples. • Perform selective (b). Metallurgical analysis of tube sampling and/or ultrastructure and oxide thickness and sonic measurement to morphology of selected tube samples. determine extent of (c). Ultrasonic testing for direct measureproblem. ment of steamside oxide thickness. • Perform remaining life (d). Visual examination for evidence of estimate of affected slag buildup, laning, bowed or mistubes. aligned tubes acting as leading tubes. • See long-term actions.

• Determine remaining life of affected tubes based on actual temperatures, stress levels and materials properties. See Chapter 8, Volume 1 for additional detail. • Make the change to a higher grade of steel (next higher chromium level). • Tube/circuit realignment; consider steam flow redistribution. See main text for summary of methodology. • Major SH/RH redesign and replacement. • Redesign tube bank. • Retube with same material depending on expected and desired life of the SH/RH.

3.2 Influences of initial design and/or material choice • As above. • Original alloy inadequate for (e).Review temperature data from actual operating temperatures. thermocouples installed in vestibule or across the header. • Inadequate heat treatment (f). Review SH/RH circuit material of original alloy. diagrams, calculate and plot GTL as • Tubes at failure location a function of steam and metal have gas-touched lengths temperatures, plot positions of failures. longer than design estimate and/or row-to-row variation in gas-touched length. • Side-to-side or local gas temperature differences. • Radiant cavity heating effects. • Lead tube/wrapper tube material not resistant enough to temperature.

• As above: steam flow redistribution will be particularly applicable for correcting side-to-side and local variations in temperature, upgrading will be particularly applicable in cases where the original alloy was insufficient or where the tube has a gas-touched length that is longer than the design estimate.

3.3 Build-up of internal oxide scale (g). See items (a) - (c).

32-12

Long-Term Overheating/Creep

• Chemical cleaning to remove deposits.

• Address underlying cause of overheating or • Periodic chemical cleaning to mitigate effects. See Chapter 4, Volume 1 for additional detail about the methods and determining timing. • Determine remaining life of affected tubes based on actual temperatures, stress levels and materials properties. See Chapter 8, Volume 1 for additional information about the methods of oxide scale analysis. • See additional options on primary list from above.

Table 32-3 Major Root Cause Influences, Confirmation and Corrective Actions (continued) Major Root Cause Influences

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

3.4 Overheating because of restricted steam flow due to chemical or other deposits, scale, debris, etc. (h). Selective sampling of suspect locations to verify whether local blockage is leading to excessive temperatures.

• Clean out tubes and remove source of blockages.

• Introduce measures to prevent future blockages.

• Local repairs as appropriate. • Perform selective sampling and/or ultrasonic measurement to determine extent of problem. • Perform remaining life estimate of affected tubes. • See long-term actions.

• Optimization of operation and fireside conditions must be the emphasis. See compilations of the applicable methods (reference 12).

3.5 Operating conditions or changes in operation

3.5.1 Previous similar problems in adjacent SH/RH

(i). Check temperature distribution through • As above. the circuit by performing analysis of GTL and measured temperatures; see (e) and (f) above.

(j). Monitor gas temperatures with 3.5.2 Combustion conditions pyrometers or infrared instruments. can lead to tube overheating. • Excessive flue gas temperature • Displaced fireball • Delayed combustion • Periodic overfiring or uneven firing of fuel burners.

• As above, plus • Restore boiler design (or optimized) conditions.

• Optimization of fireside conditions. See reference 12.

3.6 Blockage or laning of boiler gas passages (k). Can be recognized using cold air velocity technique. See flyash erosion mechanism for a discussion of the technique. (l). Visual examination to identify local flow blockages.

• Controlled with flow distribution screens; in practice is difficult to implement because of high temperatures in SH/RH. See Chapter 14, Volume 2 on flyash erosion for control of high local velocities through the use of the cold air velocity technique.

3.7 Increases in stress due to wall thinning (m). NDE evaluation to determine the extent of wall thinning. (n). If another mechanism (corrosion, erosion) is suspected, initiate actions to confirm their involvement.

• Initiate procedures to identify source of tube wastage - of particular concern are fireside corrosion or flyash/sootblower erosion processes that may be contributing to an increased oxidation rate.

• Check long-term actions in wastage mechanism chapters particularly fireside corrosion (Chapters 33 and 34) and fly ash erosion (Chapter 14, Volume 2).

Volume 3: Steam-Touched Tubes

32-13

3.2 Influences of initial design and/or material choice

1052 °F 1085 °F

1035 °F

1028 °F

1055 °F

1017 °F 1033 °F

1105 °F 1069 °F

1108 °F

1014 °F

107 0 °F

Gas flow

1018 °F

Steam Fluid Temperature + 950 °F

1069 °F

1036 °F

1014 °F

Reheater tube outside metal surface

1033 °F

1105 °F 1017 °F

1055 °F 1028 °F 1085 °F

1035 °F 1052 °F

Figure 32-10. A reheater tube that contains steam at 510°C (950°F) will have varying metal temperature around its circumference and through its wall according to: gas flow direction, ash deposit patterns, and internal scale formations. Typically the highest metal temperature will be achieved on the surface facing the gas flow. Source: G.A. Lamping and R.M Arrowood, Jr.2

Steam Temperature, °F 747 MW 1020

575 MW

940

980 2

9

16

23

30

37

44

51

58

65

72

79

86

93

97

104

118 132 146 160 174 188 111 125 139 153 167 181

Tube Number

Figure 32-11. Typical thermocouple measurements across a superheater showing that the actual tube metal temperatures can vary significantly with tube location and boiler power level. G.A. Lamping and R.M Arrowood, Jr.2

32-14

Long-Term Overheating/Creep

The aim of the design process is to provide a structure that functions within maximum allowable stresses and oxidation rates. As the temperature capability of each grade of tube material is reached, the designer specifies a transition to a higher grade material. If the actual operating temperatures exceed those foreseen in the design, then overheating will result. Figure 32-7b illustrates several examples of how this may occur: • Differences in gas-touched length of tube. On Figure 32-7b, location “M” is the transition between T22 and the more resistant 321H in Tube “B”. Location “N” is the material transition for Tube “D”. Although the transitions occur in about the same place, the distance from the inlet header to location “N” is greater than to “M”, thus the tube temperatures at “N” will be greater and therefore the chances are greater that the lower grade material may be inadequate for the temperatures actually experienced. This difference in gas-touched length can sometimes be as large as 20-30 feet. • Radiant cavity heating effects. Tubes in a cavity or that surround a cavity, such as Tube “D” at location “P” in Figure 32-7b, or just past a cavity may experience excessive temperatures because of radiant absorption effects. • Other effects of location. Tube “A” in Figure 32-7b is the lead or wrapper tube and thus will pick up more heat than tubes “B” “H”. There can also be side-toside or local gas differences; tubes in the center run hotter.

For any of these reasons, the transition to a higher grade material should have been made earlier in the circuit. Actions to confirm this root cause will include (a) through (c) above, plus (e). Review temperature data taken from across the tubes in the vestibule or header space. The array of thermocouples attached to tubing coming out of the header (used by operators for steam temperature control) should be used. See Figure 32-11 for an example of the process. (f). Perform an analysis of gastouched length (GTL) and location of the tube failures. Review SH/RH circuit material diagrams and calculate the GTL to the position(s) of failure. Plot the manufacturers recommended temperature limits (based on acceptable oxidation rates as discussed in Section 2.1 above) for each material in the circuit. Superimpose any information on measured steam and tube metal temperatures to determine whether the actual temperatures are over the design limits. Figure 32-12 shows how such a plot has been used as a diagnostic in a unit that had experienced a significant number of tube failures by long-term overheating. The lower temperature band in Figure 32-12 indicates the manufacturers estimate of tube temperatures. The shaded boxes above represent

Maximum oxidation temperature for T22

621 593 565

1000

538

Maximum oxidation temperature for T11

510 design steam temperature

900 ManufacturerÕs design estimate of tube temperature range

800

482 454 427 399

Metal and Steam Temperature, °C

Actual range of measured temperatures

1100 Metal and Steam Temperature, °F

Economy measures to minimize the use of the higher grade material or poor original design estimates of tube temperature can also be at the root of why a material experiences temperatures above its normal limits. Inadequate heat treatment in the original alloy, resulting in too low a creep strength, may also lead to failure by creep. Many creep failures have been observed to be caused by inadequate heat treatment followed by failures due to additional stresses which were not known during the design process or were underestimated.4

371

700

600 0

20

40

60 80 100 120 Gas Touched Length (ft)

140

160

180

Figure 32-12. Illustrates design steam and tube temperatures in a superheater along with the maximum oxidation temperatures for T11 and T22 materials. Field measurements of actual tube temperature ranges at four locations are also shown.

the oxidation limits for the two materials, T11 and T22 in the sample SH section. Various temperature measurements are plotted as vertical ranges at the gas-touched length for which they were measured. Note that the T11 tubes with gas-touched lengths greater than 55 ft are experiencing maximum temperatures that exceed the allowable oxidation limits. At certain times, the temperatures experienced by the T11 material were as high as 50°C (90°F) over the oxidation limit. These locations were consistent with the locations of tube failures by long-term overheating in the unit and clearly indicated that the extra gas-touched length was the primary cause. It is recommended that this method be used to confirm this root cause when it is suspected. The process will also provide an alert of the potential for other boiler tube failure mechanisms that are exacerbated by excessive temperatures such as fireside corrosion.

3.3 Buildup of internal oxide scale To confirm that the buildup of steamside oxide is a primary influence on the damage observed: (g). Institute any of the actions (a) through (c) outlined above.

3.4 Overheating because of restricted steam flow. Partial blockages can sometimes occur from chemical deposits, scale, debris or deposits. Usually they will result in short-term overheating failures (Chapter 36). (h). Selective sampling of suspect locations can verify whether a local blockage condition is responsible for excessive tube temperatures. However, the usual sequence of identification is a tube failure and then assessment of adjacent circuits.

Volume 3: Steam-Touched Tubes

32-15

3.5 Operating conditions or changes 3.5.1 Previous similar problems in adjacent SH/RH. Overheating can occur in a SH/RH circuit if absorption patterns are changed as a result of a redesign in another circuit. Such redesign may have been necessitated by prior failures. (i). Check the temperature distribution through the circuit by performing an analysis of gas-touched length and measured temperatures, as described in (e) and (f) above. 3.5.2 Combustion conditions Combustion conditions can lead to tube overheating. Such conditions include: excessive flue gas temperature, displaced fireball, delayed combustion, periodic overfiring or uneven firing of fuel burners.

(j). Gas temperature monitoring with various suction pyrometers or infrared measurements.

3.6 Blockage or laning of boiler gas passages Actions to confirm: (k). The cold air velocity technique discussed at length in the chapter on flyash erosion can be used to identify locations of locally high gas velocities. (l). Visual examination can often identify local flow blockages.

3.7 Increases in stress due to wall thinning The previous root cause influences have dealt primarily with increases in tube metal temperature as a contributor to excessive damage accumulation by creep. Any of the

processes by which tube wall thinning can occur, such as excessive oxidation, fireside corrosion, or erosion, can lead to increased tube hoop stresses which will also increase creep damage. Specific confirmation of wall thinning can be conducted by: (m). NDE methods, such as ultrasonic testing, can determine the extent of wall thinning. (n). If another mechanism is suspected, e.g., fireside corrosion, erosion, etc., actions to confirm those processes should be implemented. Reference to the chapter of this book dealing with each mechanism will provide a list of recommended actions.

4. Determining the Extent of Damage Determining the extent of damage will involve identifying locations with the thickest steamside oxide and thinnest walls. Ultrasonic testing is the most applicable and inexpensive NDE method for measuring both oxide thickness and wall thinning. More detail about the processes involved is given in Chapter 9, Volume 1. Note that among the hundreds of tubes in each superheater/reheater bank in a boiler there will be a distribution of gas and metal temperatures. As a result there will be an equivalent distribution of creep and corrosion rates. Therefore the key to a successful prevention program will involve a search for the highest risk

32-16

Long-Term Overheating/Creep

locations, and assessment of their conditions, not of the mean condition. Wall thickness measurements around the entire tube will be useful, particularly to distinguish the development of tube “flats”. Sacrificial tube samples are removed from select locations and subjected to laboratory metallurgical analysis and/or isostress rupture testing to confirm NDE results and to refine the estimates of accumulated damage. Potentially affected areas can also be identified by reviewing data from existing thermocouples or by installing thermocouples in at-risk locations.

5. Background to Repairs, Immediate Solutions and Actions Long-Term Overheating/Creep: Immediate Solutions and Actions 1. Effect repairs locally as needed. 2. Determine extent of damage and initiate an assessment of expected remaining life along with consideration of desired remaining economic life for the unit. This will lead to the correct choice among longterm options. 3. Correct off-design conditions such as displaced fireball, uneven firing, or tube blockages and flow restriction.

5.1 Need for remaining life assessment Any immediate solution or repair strategy should consider that a remaining life methodology be implemented. Such programs are discussed in detail in Chapter 8, Volume 1, and are summarized below under long-term options.

5.2 Repairs If the problem is highly localized, it may be possible to replace the affected tubes either with an upgraded material or the same material, depending upon how long the existing tube lasted and how long the desired life is. Care should be taken with this quick replacement method. It is important to remove the full extent of the damaged area, as indicated by wall loss, tube “flats”, or “alligator hide”. Removing an insufficient amount of damaged area will result in a repeat boiler tube failure. Repair methods are reviewed in Chapter 11, Volume 1.

Pad welding should not be used for repair because of the uncertainty of tube conditions, particularly the presence and depth of creep cracks, and the condition of the internal surface of the tube. Coatings are not recommended as a long-term solution for creep. Coatings may be useful as a palliative if the problem is fireside corrosion; discussion of coatings for that purpose is described in Chapters 33 and 34.

5.3 Correction of Òoff-designÓ conditions In this category are those actions that can correct obvious conditions including tube blockages and fireside combustion problems such as displaced fireball, uneven firing or fuel burners, correctable excess flue gas temperatures, etc.

Volume 3: Steam-Touched Tubes

32-17

6. Background to Long-Term Actions and Prevention of Repeat Failures Long-Term Overheating/Creep: Long-Term Actions 1. Long-term strategies mostly fall into two major categories: (i) materials approaches, such as upgrading to a material with better creep performance or (ii) control of tube temperatures by either design changes, or in a limited number of cases, operating changes. 2. An analysis of remaining life and the development of a plan for monitoring and periodic re-evaluation will be central to long-term prevention.

Figure 32-13 provides an overview of the available options. The numbers in the figure are for reference to the text only and do not imply a ranking among options; however, items outlined in bold indicate the strategies that have been shown to be the most successful. Remaining life assessment is almost mandatory in order to determine the most economic strategy among the available options. This includes understanding (i) the extent of damage, (ii) the expected remaining life (RL) of affected areas, and (iii) the desired life (DL) of the unit, The two primary approaches to prevent repeat failures are to (i) accept the temperatures being experienced at the location of interest and upgrade the material used, or (ii)

institute through re-design or (less commonly) operating changes to lower the temperature at the susceptible location. The review of long-term options looks first briefly at the remaining life assessment of SH/RH tubes.

Remaining life assessment (option 1, Figure 32-13) The importance of a comprehensive remaining life assessment cannot be overstated. It is a prerequisite to the optimal choice of long-term solutions. During the past twenty years a number of techniques have been developed to assess remaining life of superheater/reheater tubing subject to high temperature creep and corrosion. The most universally applied analysis method utilizes

Overheating (creep) confirmed and extent determined

Remaining life assessment

Materials solutions

Replace component

Same material

More resistant material

2

3

1

Design solutions

Major SH/RH redesign and replace

Redistribute steam flow

4

5

Operating solutions

Steam side

Gas side

6

8

Chemical clean oxide

Adjust fireside conditions

Remove blockages

Minimize laning

7

9

Notes: a) Remaining life (1) assessment is almost mandatory to decide which option should be adopted b) Boxes outlined in bold indicate options that have been most successful c) Numbers refer to main text

Figure 32-13. Strategies for preventing repeat failures by long-term overheating/creep.

32-18

Long-Term Overheating/Creep

Replacement with same material (option 2) Replacement in-kind is a repair option. More generally, the actual service conditions are locally more extreme than anticipated, which suggests the need to use a more resistant material. The decision will be based on (i) the expected lifetime of a new tube constructed of the same material as the original design, (ii) the desired lifetime of the unit, and (iii) the length of time to the next major outage when an alternative could be evaluated and implemented.

Change to a more resistant material (option 3) Tubes can be changed to a more resistant material, such as a higher chromium content ferritic steel or an austenitic stainless steel if better temperature resistance is the most economic option.

Creep Damage Ratio

If significant wall thinning and the concomitant accumulation of creep damage is occurring, an acceptable rate of wastage should be established on the basis of the desired remaining life and the rate of damage accumulation. As a rule of thumb, rates of wastage above 25 nm/hr (~ 9 mils/yr) are generally considered to require some residual life activity, such as an annual evaluation of residual life, formalized periodic re-inspection and analysis, etc.; rates above 50 nm/hr (~ 18 mils/yr) indicate a serious decrease in the life of the tubing. In the latter case, remaining life should be estimated and the appropriate actions initiated if that life is unacceptable. If fireside corrosion is a contributing mechanism, see separate chapters for available options.

1.00 0.80

This region is expanded in the lower graph.

0.60 0.40 0.20 0 0

1.00 Creep Damage Ratio

measurements of steamside oxide scale thickness and tube wall thicknesses to predict the remaining creep life of superheater/reheater tubing. The methodology is discussed in detail in Chapter 8, Volume 1.

40

80 120 160 200 240 Operating Time (103 Hours)

-0°F -10°F

-20°F

280

-30°F

0.95

0.90

0.85

0.80 0

40 20 60 Remaining Life (103 Hours)

80

Figure 32-14. Creep damage accumulation showing life extension obtained by reducing the metal temperature. Source: K. Hara, et al.11

Major SH/RH redesign and replacement (option 4)

Redistribute steam flow (option 5)

If the overheat problem is directly related to poor initial design of the superheater or reheater, such as poor circuit design or different gastouched lengths of the tubes between headers, then re-design might be the optimal economic solution. The main criteria should be to equalize the gas-touched length of the tubes and to avoid large radiant cavities.

The technique of steam flow redistribution in superheaters has seen recent significant development.11 Superheaters and reheaters are candidates for this method if the tubes are mostly not at the end of expected life and there are significant temperature differences among the tubes. Redistribution of steam flow can then serve to equalize the tube metal temperatures across the superheater.

Volume 3: Steam-Touched Tubes

32-19

Inlet header

orifice but with a tapered section. Typically the SFC is installed just downstream from the inlet header in the penthouse, which allows easy access to tubes and places the SFC out of the heat flux region.

Outlet header

SFC location Roof

A roadmap of the overall process, which includes both technical and economic considerations, is shown in Figure 32-16. Steps in the process include11: • Analyzing existing information, sampling, and field measurement to determine the current distribution of tube metal temperatures. • Analysis of the steamside oxide scale to determine tube temperatures and creep life expended to date and remaining. • Determining the locations and dimensions of SFCs required to achieve the desired tube metal and steam temperatures. • Estimating the expected creep life improvement.

Wall thickness equal to superheater tubing wall thickness

Approximately 1 ft. in length

OD equal to superheater tubing OD 3:1 taper

Steam Flow Controller

Figure 32-15. Schematic of a steam flow controller (SFC) and typical location in a superheater tube bank. Source: K. Hara, et al.11

With steam flow controllers (orifices) of different sizes and lengths in the tubes coming out of the inlet header, the flow can be decreased in “cold” tubes, which increases their temperatures. This change will also increase the steam flow to “hot” tubes with a concurrent decrease of the metal temperatures. Figure 3214 shows that the decrease in tube metal temperature can lead to significant increase in the remaining life of tubes. This tube which has operated

32-20

Long-Term Overheating/Creep

for approximately 200,000 hours would have an expected life of about 27,000 more hours. Decreasing the tube metal temperature by 30°F (~ 17°C) would increase the expected life to over 75,000 hours. Steam flow redistribution occurs by modifying the tube-to-tube steam flow resistance through the use of steam flow controllers (SFCs), Figure 32-15. The tube inner diameter is increased or decreased for a specific length, not with a sharp-edged

• Evaluating the economic and technical tradeoffs of identified options. • Performing the optimal modifications. • Validation testing and long-term monitoring of the results. Technical factors to be considered include11: (i) analyzing the increase in pressure drop across the superheater because too much plugging or flow restriction can lead to a heat rate penalty, although this has been found to be very minor, (ii) the optimal inside diameter of each SFC including some minimum value to avoid plugging, typically 12 mm (0.5 in.) or so, (iii) the number of SFCs, (iv) the number of plugged and replaced tubes, and (v) the temperatures of tubes after redistribution.

(Step 1)

Obtain superheater design and test data

(Step 2)

Candidate for steam flow redistribution?

SH/RH circuits are found in Chapter 4, Volume 1. Precautions should be taken to avoid stress corrosion cracking of stainless steel tubes during cleaning. This option has the added advantage of reducing exfoliation that can lead to solid particle erosion of HP/IP turbines.

No

Clean tubes and remove sources of blockage (option 7)

Yes (Step 3)

Perform steam flow redistribution design and economic analysis

(Step 4)

Acceptable cost savings?

Similarly, if the source of overheating is a blocked or partially blocked tube, high pressure fluid flushing or chemical cleaning may be necessary, depending on the problem. This can be a time consuming and repetitive option, which involves radiography of the bends, removal and replacement of tubes, or cleaning of the accumulated scale, debris or deposits.

No

Yes (Step 5)

Perform superheater modifications (Step 6)

(Step 7)

Make run/repair/replace decision

This involves quantification of the problem through use of fireside testing and optimization. Reference should be made to compilations of the appropriate procedures.12

Perform validation testing and long-term monitoring

Figure 32-16. Road map for achieving improved superheater longevity by steam flow redistribution. Source: K. Hara, et al.11

An overall strategy for the superheater will include consideration of SFCs along with selective tube plugging and tube replacement. Plugging of tubes that are near the end of their expected lives will not only extend the time to first failure but will have the additional advantage of increasing flow that can be directed to decrease the temperature of “hot” tubes. Selective tube replacement will also extend the time to failure. For a typical 400 MW unit, the cost to perform a steam flow redistribution modification is around $210,000

Adjust fireside combustion conditions and/or burners (option 8)

($525/MW). A condition assessment based on the oxide scale technique in a similar size unit is around $25,000 (~ $60/MW).

Chemical cleaning to remove steamside oxide layer (option 6) If excessive temperature in the SH/RH tubes has been exacerbated by the presence of increasing steamside oxide scale thickness, a solution involving chemical cleaning may be in order. Typical benefits are illustrated in Figure 4-2, Volume 1. Details on chemical cleaning of

Minimize laning of gases (option 9) Laning or channeling of gases through certain tube sections can lead to overheated tubes. This is very difficult to overcome; it can be recognized and monitored by way of the cold air velocity technique which is described at length in the chapter on the flyash erosion damage mechanism. However, the application of flow distribution screens, a means to control locally high velocities, is made more difficult in the SH/RH by the high temperature environment.

Volume 3: Steam-Touched Tubes

32-21

7. Case Study Long-Term Overheating/Creep Case Study: Field Experience Table 32-4 summarizes the key characteristics for four units that have recently experienced boiler tube failures by long-term overheating. There are a number of important factors about the causes of the problem that are evident from the field experiences in these units: 1. The majority of the failures were in ferritic materials, with the most prominent failure location being adjacent to the transition to a material containing a higher Cr content.

2. There was evidence of very thick internal oxide scale. The scales were usually multi-laminated and suffering exfoliation.

5. Temperatures were found to have been higher than the manufacturer’s oxidation limits. 6. Damage was usually detectable by performing a gas-touched length/materials evaluation such as shown in Figure 32-12.

3. The affected tubes always manifested the physical appearance of tube flats, “alligator hide”, and longitudinal failures that originated in the middle of the flats. 4. There was no contribution from fireside corrosion; no low melting point compounds were detected in the fireside tube deposits.

Table 32-4 Comparison of Key Factors in Field-Observed Long-Term Overheating Were Tube Flats, Fireside Alligator Operating Corrosion Hide Hours (see Note 1) Found?

Oxide Scale Estimated Thickness Temperature, Corrective (µm) °C Action

Unit

Failure Location

Material

Adjacent (next) Material

A

RH

9Cr

304H

> 100,000

No

Yes

> 100

> 630

Replacement

SPE (see Note 3)

B

RH

T11

T22

80,000

No

Yes

> 250

610 - 620

Upgrade

SPE

C

RH

9Cr

321H

60,000

No

Yes

n/a

> 630

Replacement (see Note 2)

None

D

SH

T11

T22

85,000

No

Yes

N/a

565 - 615

Redesign

DMW (see Note 4)

Notes: 1. Was fireside corrosion involved and/or was a low melting point deposit/ash found? 2. Replacement plus operational changes were used to correct high reheat temperatures during sliding pressure operation. 3. Solid particle erosion (SPE) in turbine due to thick and exfoliating steamside oxide. 4. Dissimilar metal weld (DMW) failures.

32-22

Long-Term Overheating/Creep

Other Unit Problems

8. References 1Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 2Lamping,

G.A. and R. M Arrowood, Jr., Manual for Investigation and Correction of Boiler Tube Failures, Research Project 1890-1, Final Report CS-3945, Electric Power Research Institute, Palo Alto, CA, April, 1985. 3Grunloh, H.J. and R.H. Ryder, Life Assessment of Boiler Pressure Parts, Volume 7: Life Assessment Technology for Superheater/Reheater Tubes, Research Project 225310, Final Report TR-103377-V7, Electric Power Research Institute, Palo Alto, CA, November, 1993 4Personal

Communication from E. Tolksdorf (VGB) to R.B. Dooley, February 16,1995. 5Dooley,

R.B. and H.J. Westwood, Analysis and Prevention of Boiler Tube Failures, Report 83/237G-31, Canadian Electrical Association, Montreal, Quebec, November, 1983. 6American Society of Mechanical Engineers, Boiler and Pressure Vessel Code: Section I: Rules for Construction of Power Boilers, ASME, New York. Data for Table 32-2 from 1983 edition.

7Stultz,

S.C. and J.B. Kitto, Steam: Its Generation and Use, Babcock & Wilcox Company, Barberton, Ohio. Data for Table 32-2 from 1978 edition. 8Singer, J.G., ed., Combustion Fossil Power: A Reference Book on Fuel Burning and Steam Generation, Combustion Engineering, Inc., Windsor, Connecticut. Data for Table 32-2 from 1981 edition. 9French,

D.N., Metallurgical Failures in Fossil-Fired Boilers, John Wiley & Sons, Wiley-Interscience Publications, New York. Data for Table 32-2 from 1983 edition. 10Eberle,

F. and J.H. Kitterman, “Scale Formation on Superheater Alloys Exposed to High Temperature Steam”, The Babcock & Wilcox Company, Alliance, Ohio.

11Hara,

K., C. Lee, R. Moser, T. Rettig, and K. Clark, Improved Superheater Component Longevity by Steam Flow Redistribution, Research Project 1893-13, Final Report TR-101697, Electric Power Research Institute, Palo Alto, CA, December, 1992. 12Sotter, J.G., J.A. Arnot, and T.M. Brown, Guidelines for Fireside Testing in Coal-Fired Power Plants, Research Project 1891-3, Final Report CS-5552, Electric Power Research Institute, Palo Alto, CA, March, 1988.

Volume 3: Steam-Touched Tubes

32-23

ACTIONS for Long-term Overheating/Creep Two paths for the BTF team to take in the investigation of longterm overheating damage begin here. The goal of these actions is to see if further investigation of long-term overheating is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a BTF has occurred and long-term overheating/creep is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by long-term overheating/ creep.

Action 1a: If a BTF has occurred and long-term overheating/creep is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of long-term overheating/ creep: ➠ Review Figure 32-7a and b for typical boiler regions. ➠ Review main text, Section 1.2 for description of susceptible locations.

➠ Confirm that the macroscopic appearance of the failure includes such features as: • Thick-edged fracture surface. • Thick, external scales, often with Y-shaped grooves, “alligator hide” appearance. • Tube “flats” on outside of tube, particularly with fracture or the maximum amount of “alligator hide” occurring in the flats. • Thick layered internal scale.

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for steamtouched tubing (Table 31-1) to pick a more likely candidate.

32-24

Long-Term Overheating/Creep

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by long-term overheating/creep.

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit: • Excessive temperatures detected during monitoring of installed thermocouples in the vestibule or header area. • Observation of “alligator hide” during routine inspections. • Detection of an excessive level of steamside oxide either by ultrasonic measurements taken during periodic inspection or by analysis of removed tube samples. • A problem with solid particle erosion in the turbine. • Fireside conditions that could lead to overheating of the tubes.

➠ These precursors can signal the potential for tube failures by longterm overheating/creep. If one or more has occurred, go to Action 3 which reviews root causes and outlines the steps to confirm the influence of each.

Action 2: Determine (confirm) that the mechanism is long-term overheating/creep. A failure has occurred which the BTF team has tentatively identified as being long-term overheating (Action 1a). Action 2 should clearly identify long-term overheating as the primary mechanism or point to another cause. The actions listed will be executed by confirming the macroscopic appearance of the failure, removing representative tube sample(s) and subjecting them to detailed visual and metallographic analysis. A primary objective is to ensure that the mechanism is not fireside corrosion.

➠ Evaluate locations of failure. Are failed tubes in a location susceptible to long-term overheating such as near material transitions, where there is variation in gastouched length, in or just beyond a cavity, in highest temperature locations?

➠ Characterize failure surface. Is failure “thick-edged”?

➠ Analyze internal scales. Do sectioned tubes (of ferritic materials) show thick, internal oxide scales?

➠ Calculate ratio of wall thinning loss to internal oxide scale buildup. See Figure 32-6. Is the ratio greater than five?

➠ Calculate ratio of wall thinning loss to internal oxide scale buildup. Is the ratio greater than three?

➠ Analyze composition of external ash/ deposits. Are there low melting point ash components in external deposits?

Problem may be fireside corrosion. Review susceptible locations for that failure mechanism.

Ductile failures will be more indicative of a wastage mechanism. See fireside corrosion, an erosion mechanism, or short-term overheating.

Problem is probably not long-term overheating, see erosion and fireside corrosion.

Problem is almost certainly a wastage mechanism, either fireside corrosion or erosion.

Problem is likely to be influenced by wastage mechanism, either fireside corrosion or erosion.

Fireside corrosion is likely to be a primary contributor to the damage.

continued on next page

Volume 3: Steam-Touched Tubes

32-25

Action 2: Determine (confirm) that the mechanism is long-term overheating/creep (continued). Probable mechanism is long-term overheating/creep. Steps to confirm will include evaluation for signs of creep damage or overheating: • In ferritic materials spheroidization and decreased fireside surface hardness • In austenitic stainless steels the presence of sigma phase and grain boundary cavities. • Graphitization may be present in carbon or carbon molybdenum steels. • Note that creep voids will typically be found only adjacent to the cracking, not in the bulk material.

➠ Go to Action 3: Root Cause Determination

References to other sources of information: • Main text (this chapter) provides the background to the mechanism and the development of long-term overheating. • Summary of the steps and methods of metallurgical investigation of boiler tube failures can be found in Chapter 6, Volume 1. • Fireside corrosion is the mechanism most likely to be confused with longterm overheating. It is the subject of two separate writeups, Chapter 33 for coal-fired units and Chapter 34 for oil-fired units, which should be reviewed for clarification of the differences.

32-26

Long-Term Overheating/Creep

Action 3: Determine root cause of long-term overheating/creep A BTF failure has occurred and the mechanism has been confirmed as long-term overheating (Action 2) or a precursor has occurred (Action 1b). The goal of this Action is for the BTF Team to review the potential root causes of long-term overheating, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.1 All causes of overheating

➠ (a). Direct measurement of temperatures by thermocouples. ➠ (b). Metallurgical analysis of tube structure and oxide thickness and morphology of selected tube samples. ➠ (c). Ultrasonic testing for direct measurement of steamside oxide thickness. ➠ (d). Visual examination for evidence of slag buildup, laning, bowed or misaligned tubes acting as leading tubes.

3.2 Influences of initial design and/or material choice • Original alloy inadequate for actual operating temperatures. • Inadequate heat treatment of original alloy. • Tubes at failure location have gas-touched lengths longer than design estimate and/or row-to-row variation in gas-touched length. • Side-to-side or local gas temperature differences. • Radiant cavity heating effects. • Lead tube/wrapper tube material not resistant enough to temperature.

➠ (e). Review temperature data from thermocouples installed in vestibule or across the header. ➠ (f). Review SH/RH circuit material diagrams, calculate and plot GTL as a function of steam and metal temperatures, plot positions of failures.

3.3 Build-up of internal oxide scale

➠ (g). See items (a) - (c).

3.4 Overheating because of restricted steam flow due to chemical or other deposits, scale, debris, etc.

➠ (h). Selective sampling of suspect locations to verify whether local blockage is leading to excessive temperatures.

Volume 3: Steam-Touched Tubes

32-27

Action 3: Determine root cause of long-term overheating/creep (continued) Major Root Cause Influences

➠ Actions to Confirm

3.5 Operating conditions or changes in operation

32-28

Long-Term Overheating/Creep

3.5.1 Previous similar problems in adjacent SH/RH

➠ (i). Check temperature distribution through the circuit by performing analysis of GTL and measured temperatures; see (e) and (f) above.

3.5.2 Combustion conditions can lead to tube overheating. • Excessive flue gas temperature • Displaced fireball • Delayed combustion • Periodic overfiring or uneven firing of fuel burners.

➠ (j). Monitor gas temperatures with pyrometers or infrared instruments.

3.6 Blockage or laning of boiler gas passages

➠ (k). Can be recognized by way of cold air velocity technique. See flyash erosion mechanism for a discussion of the technique. ➠ (l). Visual examination to identify local flow blockages.

3.7 Increases in stress due to wall thinning

➠ (m). NDE evaluation to determine the extent of wall thinning. ➠ (n). If another mechanism (corrosion, erosion) is suspected, initiate actions to confirm their involvement.

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. A key for remaining life assessment, which is central to proper choice of long-term actions, is an evaluation of steamside oxide thickness. That determination is the primary objective of the survey of affected areas; an analysis of wall thickness and direct monitoring of tube temperatures are also important parts of this Action 4.

➠ Identify all locations to be examined. See typical locations in Figure 32-7a and b and listed in the main text, Section 1.2 Examine for indirect signs of overheating

➠ Perform UT survey to determine oxide scale thicknesses. See Chapters 8 and 9, Volume 1 for more detail.

Examine for direct signs of overheating

➠ Monitor tube temperatures. Install thermocouples (T/C) and/or monitor existing T/C for signs of local overheating.

Examine for signs of wall thinning as contributor to creep damage.

➠ Perform UT survey to determine extent and distribution of wall thinning. See Chapter 9, Volume 1 for more detail.

➠ Perform selective tube sampling to confirm NDE results, evaluate the nature and extent of oxide buildup.

➠ Use results interactively with Action 3 (Root cause evaluation).

➠ Begin remaining life assessment. Go to Action 5: Immediate Solutions and Actions

Volume 3: Steam-Touched Tubes

32-29

Action 5: Implement repairs, immediate solutions and actions The most important actions for the BTF team are to (i) initiate a remaining life assessment based on oxide thickness and wall thinning derived from NDE survey, and (ii) make a quick repair using the same material if, after calculating the expected degradation rate, there is assurance that it will survive to the next outage. Other actions that can be executed in the short-term are also noted.

➠ Gather sufficient information so that a remaining life assessment of affected tubes can be initiated.

➠ Implement repairs or replacement of affected tubes identified from the NDE Survey (Action 4). ➠ See Chapter 11, Volume 1 for summary of applicable tube repair techniques. ➠ Develop a plan to replace affected tubing, including an economic assessment that considers the estimated rate of future failures. ➠ Ensure that the full extent of damage is removed, as indicated by wall loss, the presence of “alligator hide” or tube “flats”. Failure to do so will result in immediate repeat failures. ➠ Temporary pad welds should not be used because of the uncertainty associated with the base metal condition (the depth and extent of creep cracks, and the condition of the inside tube surface).

➠ Adjust off-design conditions such as tube blockages, displaced fireball, uneven firing or fuel burners, excess flue gas temperatures, etc.

32-30

Long-Term Overheating/Creep

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will include the analysis of remaining life based on the oxide scale methodology and an economic evaluation to ensure that the optimal strategy has been chosen.

Major Root Cause Influences

➠ Long-Term Actions

All causes of overheating

➠ Determine remaining life of affected tubes based on actual temperatures, stress levels and materials properties. See Chapter 8, Volume 1 for additional detail. ➠ Make the change to a higher grade of steel (next higher chromium level). ➠ Tube/circuit realignment; consider steam flow redistribution. See main text for summary of methodology. ➠ Major SH/RH redesign and replacement. ➠ Redesign tube bank. ➠ Retube with same material, depending on expected and desired life of the SH/RH.

Influences of initial design and/or material choice • Original alloy inadequate for actual operating temperatures. • Inadequate heat treatment of original alloy. • Tubes at failure location have gas-touched lengths longer than design estimate and/or row-to-row variation in gas-touched length. • Side-to-side or local gas temperature differences. • Radiant cavity heating effects. • Lead tube/wrapper tube material not resistant enough to temperature.

➠ As above: steam flow redistribution will be particularly applicable for correcting side-toside and local variations in temperature; upgrading will be particularly applicable in cases where the original alloy was inadequate or where the tube has a gas-touched length that is longer than the design estimate.

Build-up of internal oxide scale

➠ Address underlying cause of overheating or ➠ Periodic chemical cleaning to mitigate effects. See Chapter 4, Volume 1 for additional detail about the methods and determining timing. ➠ Determine remaining life of affected tubes based on actual temperatures, stress levels and materials properties. See Chapter 8, Volume 1 for additional information about the methods of oxide scale analysis. ➠ See additional options on primary list from above.

Overheating because of restricted steam flow due to chemical or other deposits, scale, debris, etc.

➠ Introduce measures to prevent future blockages.

Operating conditions or changes in operation

➠ Optimization of operation and fireside conditions must be the emphasis. See compilations of the applicable methods (reference 12).

Volume 3: Steam-Touched Tubes

32-31

Action 6: Implement long-term actions to prevent repeat failures (continued) Major Root Cause Influences

➠ Long-Term Actions

Previous similar problems in adjacent SH/RH Combustion conditions can lead to tube overheating. • Excessive flue gas temperature • Displaced fireball • Delayed combustion • Periodic overfiring or uneven firing of fuel burners.

➠ Optimization of fireside conditions. See reference 12.

Blockage or laning of boiler gas passages

➠ Controlled with flow distribution screens; in practice is difficult to implement because of high temperatures in SH/RH. See Chapter 14, Volume 2 on flyash erosion for control of high local velocities through the use of the cold air velocity technique.

Increases in stress due to wall thinning

➠ Check long-term actions in wastage mechanism chapters, particularly fireside corrosion (Chapter 33 and 34) and flyash erosion (Chapter 14, Volume 2).

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components that might be implied by the presence of long-term overheating damage or its precursors.

32-32

Long-Term Overheating/Creep

Long-Term Overheating Aspect

Alert for Other Cycle Components

➠ Actions Indicated

Tube overheating as evidenced by build-up of internal oxide scale

• Potential for exfoliation of oxide which can carryover into turbine sections. • Exfoliating scale can lead to tube blockage and failures by short-term overheating (see Chapter 36).

➠ Chemical cleaning of SH/RH sections. See Chapter 4, Volume 1 for more detail. ➠ Monitoring plan to assess the severity of oxide buildup in affected tubes, including UT inspection for direct measurement of oxide scale, and tube sampling to confirm type and extent of scale.

Total redesign of the superheater or reheater.

May change absorption patterns through the SH/RH sections and may increase temperatures in other circuits.

➠ Check temperatures in the redesigned section, and other sections.

Chapter 33 • Volume 3

SH/RH Fireside Corrosion/Coal-Fired Units Introduction Superheater/reheater (SH/RH) fireside corrosion (also called “molten salt” attack or “coal ash” corrosion or “liquid-phase” corrosion) is not currently a major problem for U.S. units using 538°C or 541°C (1000°F or 1005°F) steam. Historically, it has been a significant problem for units operating in the U.S. at higher steam temperatures above 565°C (1050°F), and in those burning coals with high fractions of chlorine such as is common in the United Kingdom (U.K.). It is important to review the current understanding of this damage type as advanced coal-fired units may operate at higher main steam temperatures as one means of achieving higher unit efficiencies. Further, as the opportunity to burn

coals of various compositions develops, it is important to understand the effects of coal composition on fireside corrosion. A third reason for reviewing the fireside corrosion mechanism is that it is often confused with damage occurring by long-term overheating in superheater and reheater tubes (Chapter 32); field experience indicates long-term overheating is currently the more prevalent of the two. Both mechanisms eventually result in creep failure and share some common features of failure. However, corrective actions differ as do the underlying causes. Discussion of fireside corrosion in waterwalls of coal-fired units (Chapter 18, Volume 2), and in superheater/reheater tubing of oilfired units (Chapter 34) can be found in separate chapters.

Volume 3: Steam-Touched Tubes

33-1

a

1. Features of Failure and Typical Locations Superheater/Reheater Fireside Corrosion (Coal-Fired Units): Identification Keys 1. Macroscopically, fireside corrosion will generally be accompanied by tube wastage at the 10 and 2 o’clock positions (12 o’clock is the upstream position) and by the presence of multi-layered fireside scale and ash. 2. Distinctive deposits, containing low melting ash components and the presence of alkali iron trisulphates (Na, K)3Fe(SO4)3 are characteristic of the most common mode of fireside corrosion. 3. A second mode is a sulfidation mechanism which will result from carburization of stainless steel tubing and produce evidence of discrete iron sulfides in the grain boundaries of affected locations. 4. Comparing the amount of wall thinning to the buildup of steamside oxide can determine the extent to which fireside corrosion or erosion has been a contributor relative to long-term overheating.

or thick, friable ash layers are less likely to be covering corrosion sites.3

1.1 Features of failure

Macroscopically, there will usually be a multilayered fireside scale and ash deposit. The deposits found are generally tightly bound to tubes at room temperatures and will typically consist of three layers1, 2:

Fireside corrosion damage will be primarily distinguished from longterm overheating by the presence of low melting point ash compounds. The fireside scale and ash should be examined metallographically and chemically; the presence of low melting point (in the range 550°C to 620°C (~ 1025°F to 1150°F)) constituents in the ash is indicative of fireside corrosion. Similarly, the presence of the alkali iron trisulphates (Na, K)3Fe(SO4)3 , particularly in the middle layer, is a further indicator of an active fireside corrosion mechanism. The innermost layers will be superficially similar for both mechanisms.

1. A hard, brittle and porous outer layer, which makes up the bulk of the deposit and has a composition similar to that of boiler flyash.

2. A white intermediate layer consisting of compounds of complex alkali sulfates including alkali iron trisulfates. When this layer has a chalky consistency, corrosion has been found to be mild or nonexistent; when fused and semiglossy, corrosion has been found to be severe.

3. A black, glossy inner layer, composed primarily of oxides, sulfates and sulfides of iron. Deposits adjacent to the metal oxide are the ones that directly influence corrosion of the substrate. Such deposits, if tightly bonded, should be suspected of covering corrosion sites, whereas loosely bonded dust

Tube wastage will often be evident and manifested as flat spots on the tube at the 10 o’clock and 2 o’clock positions (12 o’clock is the upstream position). A typical cross-section showing this wastage and the presence of significant deposits can be seen in Figure 33-2. In this case

a

5. As fireside corrosion is superficially similar to long-term overheating in superheater/ reheater tubes, some care in diagnosis of the underlying mechanism is required.

Figure 33-1 shows the three deposit types schematically.

S team

Inner layer

I ntermediate layer (molten) Outer layer (fly ash)

Flue gas flow

Figure 33-1. Schematic representation of fireside corrosion development for superheaters and reheaters involving a molten intermediate layer (alkalis, sulfates). This case shows maximum wastage at the 10 and 2 o'clock positions.

33-2

SH/RH Fireside Corrosion/Coal-Fired Units

tube thinning will be most evident around the edges of the deposit. Alternatively, the maximum wastage can occur at the 12 o’clock position. Figure 33-3 shows two tube sections with fireside corrosion and long-term overheating/creep damage. Depending on the corrosion rate, removal of the fireside scale and ash deposit will reveal the smooth contoured appearance typical of a fluxed corrosion reaction (the right hand tube segment in Figure 33-3) or distinctive longitudinal grooving termed “alligator hide” as shown in Figure 33-4, and at the 10 o’clock and 2 o’clock positions in Figure 33-3. Greatest wall loss will generally be seen in tubes that have been operated at the highest temperatures over a period of time. The measurement of steamside oxide thickness by ultrasonic or laboratory measurement will usually confirm this observation. The ratio of maximum wall loss to oxide scale thickness at any location will give an indication of the degree to which fireside corrosion is a problem relative to long-term overheating. Figure 33-5 indicates that if the ratio is greater than five either fireside corrosion or erosion is definitely active. For ratios less than three, long-term overheating is the primary mechanism.4

Figure 33-2. Tube sample exhibiting fireside corrosion. Note the presence of multi-layered scale along with wastage flats at the 10 and 2 o'clock positions of the tube's circumference. Source: S.R. Paterson, et al.4

On austenitic steels, the external magnetite and spinel oxides become non-adherent when they reach a thickness of about 0.2 mm (~ 0.008 in.) and tend to spall off along with any deposits that have formed when the boiler cools. Bare metal can result and lead to rapid corrosion rates. Any oxide scale that is thicker than about 1 mm (0.04 in.) should be considered as evidence of a fast corrosion rate.3 There may also be localized, overlapping pits 0.5 - 2 mm (~ 0.02 - 0.08 in.) in diameter. Figure 33-3. Two tube sample segments showing fireside corrosion. The left shows the ash pattern as-removed; the right shows the tube with the ash removed. On this segment, the 12 o'clock position shows a smooth contour typical of a fluxing fireside corrosion reaction, and the 10 and 2 o'clock positions show alligator hide typical of long-term overheating/creep.

Volume 3: Steam-Touched Tubes

33-3

The 300 series stainless steels can also be rapidly corroded if attacked by hydrogen sulfide gas, H2S, which can be formed in flue gas that has high CO levels. Identification keys will include the carburization of the stainless steel, the presence of iron sulfide scale or discrete grain boundary sulfides. Fireside corrosion can be superficially similar to long-term overheating in superheater/reheater tubing. The key differences between these two mechanisms and short-term overheating are highlighted in Table 33-1. The most distinguishing characteristic will be the presence of low melting ash compounds in the deposits formed for fireside corrosion. The ratio of wall thickness loss to oxide scale thickness as shown in Figure 33-5 and described above can also help identify the prominent damage mechanism. Figure 33-4. Grooving of the tube's external surface, known as "alligator hide", associated with coal ash corrosion. The fireside oxide scale and ash deposit have been removed by glass bead blasting. Source: S.R. Paterson, et al.4

Superheater Tubes with Ash Corrosion 0.2

B F

Y = 5X

Wall Loss,Inch

0.15 A

0.1

Y = 3X A

F

B

E

C

0.05 E D

0 0

C

0.01 0.02 0.03 Steamside Oxide Scale Thickness, Inch

D 0.04

Figure 33-5. Nondestructive field measurements of wall loss and steamside scale thickness on a SA-213 T22 superheater tube sample are plotted. The greatest wall loss was almost ten times greater than the steamside oxide thickness consistent with ash corrosion or fireside erosion being dominant. Source: S.R. Paterson, et al.4

33-4

SH/RH Fireside Corrosion/Coal-Fired Units

Table 33-1 Comparison of Characteristics of Long-Term Overheating (Creep), Short-Term Overheating, and Fireside Corrosion (Coal-Fired Units) In Superheater/Reheater Tubing Characteristic

Long-Term Overheating

Short-Term Overheating

Fireside Corrosion

Fracture Surface and Appearance of Failure

• Generally thick-edged, brittle final failure. • Generally accompanied by external tube wastage at the 10 o’clock and 2 o’clock positions.

• Usually thin-edged, ductile final failures. • Swelling of tubes without ovalization. • “Fish-mouth” appearance of tube rupture.

• Tube wastage, particularly at the 10 and 2 o’clock positions. • Longitudinal cracking, final failure can be (but not necessarily) by overheating.

Internal Scale?

Yes, generally extensive, multilaminated and exfoliating.

• Not necessarily thick. • Depends on age of tube at failure.

Yes, particularly if tube metal overheating was an influencing factor.

External Scaling?

• Yes, thick, laminated and often longitudinally cracked. • Usually two layers - (i) a hard, porous outer layer with composition typically that of flyash, and (ii) a black glossy inner layer mostly oxide but may contain some sulfates and sulfides of iron.

Not necessarily thick.

Yes, with multi-layers: (i) a hard, porous layer - composition typically of flyash, (ii) an intermediate layer containing complex alkali sulfates, and (iii) a black, glossy inner layer mostly of oxides, sulfates, and sulfides of iron.

Outside surface appearance after removal of scale/deposits

Characteristic longitudinal grooving and pitting (“alligator hide”).

Swelling, stretch marks on tube metal.

Characteristic longitudinal grooving and pitting (“alligator hide”). Sometimes the corroded area is smooth and featureless. Sometimes “orange peel” appearance at extremities of severe corrosion.

Composition of External Scales/Deposits

Does not contain low melting point ash compounds such as alkali iron sulfates.

Not relevant.

Does contain low melting point compounds such as alkali-iron sulfates (coal-fired units).

Wall Thinning?

Typically wastage flats at 10 o’clock and 2 o’clock positions caused by accelerated oxidation. There is always a layer of oxide adjacent to the tube.

Only because of bulging of tube material.

Primary feature of failure, may be worse at the 10 and 2 o’clock positions. Depending upon the rate of corrosion, a protective oxide layer may remain on the tube or may have been fluxed off.

Ratio of wall loss to steamside oxide thickness?

Typically less than 3:1.

Not relevant.

Typically greater than 3:1; for ratios greater than 5:1 fireside corrosion or erosion is the dominant mechanism.

Tube Material Degradation

Yes, generally extensive signs of overheating and/or of creep damage, particularly near to the crack tip. Creep voids will not be found removed from crack tip.

Depends on the material and the maximum temperature reached.

If overheating has been a problem, yes; otherwise, no. Fireside corrosion can occur in a tube at design temperatures.

Change in material hardness

Localized softening near the rupture is typical.

Localized hardening near the rupture is likely.

Hardening is not necessary; if there has been no overheating, there will be no change in hardness.

Volume 3: Steam-Touched Tubes

33-5

1.2 Locations of failure Figure 33-6 shows typical locations of concern. Corrosion will generally be the worst in the highest temperature locations. Parts at highest risk therefore include:

• At bottom bends of platens especially those facing the fireball.

• Leading sides of all tubes in pendant platens, especially hottest (leading) tubes, and steam outlet tubes

• Tubes with a longer gas-touched length (GTL). GTL is the distance measured along the tube circuit from the inlet header to the point of corrosion.

• Tubes out of alignment that act as leading tubes

• Spacers and uncooled hangers, and the fins and studs on tubes.

• Tubes in the outlet (final) sections towards the header, because these are at the highest temperatures. • Just prior to a change of material, e.g., in T22 just prior to the austenitic material, as the lower Cr content material may be operating above its design point. • Wrapper tubes

Figure 33-6. Typical locations where fireside corrosion can occur.

33-6

• Tubes that surround a radiant cavity (i.e., they may pick up more heat)

SH/RH Fireside Corrosion/Coal-Fired Units

As noted above, on a particular tube, failure will often be at the 10 o’clock or 2 o’clock position (flue gas impinging at the 12 o’clock position). Alternatively, the maximum corrosion damage may be at the tube crown or on one side for a tube out of alignment.

2. Mechanism of Failure Superheater/Reheater Fireside Corrosion (Coal-Fired Units): Mechanism 1. Corrosion generally occurs as the result of the formation of low-melting point, liquidphase, alkali-iron trisulfates (Na, K)3Fe(SO4)3. Flue gas induced erosion may also play a role in removing protective oxides. Failure finally occurs by long-term creep. 2. Coal composition such as Na, K, and S, can be a primary determinant of the wastage rate; alkaline earth oxides (CaO, MgO) suppress corrosion. There is no universally applicable corrosivity index for coal composition that allows, a priori, a prediction of fireside corrosion rates, although several available indices provide qualitative guidance. 3. There is a peak in the rate of corrosion in the range of 650-750°C (~ 1200-1380°F). The severity of the corrosion is a function of the tube material, gas temperature, and coal composition. This observation has led to placing limits on tube metal temperatures by limiting main steam temperature. 4. Corrosion can also occur by a sulfidation mechanism whereby incomplete combustion leads to high CO levels in the flue gas and local carburization, resulting in the formation of hydrogen sulfide gas H2S, subsequent sulfidation, and as a result, rapid corrosion.

2.1 Mechanism of Attack by Alkali Salts Salts (alkali-iron trisulfates) of the type (Na, K)3 Fe (SO4)3 form by the reaction of alkali sulfates with iron oxide in the presence of SO3. These salts can have melting temperatures as low as 552°C (1025°F) but are generally in the range 600-650°C (~ 1110-1200°F). This compares with the much higher melting points of the simple sulfates, for example at 884°C (1623°F) for Na2SO4 and 1069°C (1956°F) for K2SO4. The formation of alkali-iron trisulfates has an upper temperature bound set by their stability, which is a function of the partial pressure of SO3 at the ambient temperature.4 In the superheater, combustion should be complete and the gas temperatures are lower than in the furnace, with typical final superheater maximum tube temperatures around 600650°C (~ 1110-1200°F). These metal temperatures fall in the range in which alkali metal salts, condensed from the flue gas, can react with the tube surface oxides in the presence of SO3, to form complex liquid sulfates, a result that leads to rapid corrosion. The precise mechanism by which corrosive attack occurs is still not fully understood. It is presently held that the alkali iron trisulfates disrupt the protective tube oxide by a fluxing mechanism and there is a subsequent transport of iron from the tube surface to form a non-protective oxide in the salt. Subsequently, accelerated oxidation occurs because of (i) removal of the protective oxide by mechanical loss, (ii) oxidation-sulfidation attack caused by sulfur in the slag, and (iii) via fluxing of the oxide and sulfidation attack of the substrate metal In a recent investigation of alloys for use in the SH/RH sections of advanced steam plant6, laboratory and field probe tests indicated that the mechanism of coal-ash corrosion in those tests was an accelerated oxidation/sulfidation process. In low-

corrosion-resistant, low-chromium alloys, oxidation/sulfidation occurred at a linear rate indicative of a fluxing action. In the higher chromium-containing alloys, an initial resistance was broken down by a diffusioncontrolled step - the formation of a chromium oxide layer which caused depletion of the underlying zone and rendered it more susceptible to oxidation/sulfidation. Although the precise mechanism is still unknown, the results are well established. The wastage rate is a function of metal temperature as shown in Figure 33-7 for 21/4 Cr ferritic steel and 18Cr-8Ni stainless steel and in Figure 33-8 for many other alloys. This characteristic bellshaped curve has been confirmed by both laboratory work and plant probe trials for a wide range of materials.2,3,8 Figures 33-7 and 33-8 show several important factors. There is a rapidly increasing rate of corrosion with temperature to a maximum around 700°C (~ 1300°F). If steam temperatures are controlled to around 538°C (1000°F), tube metal temperatures should be limited to approximately 595°C (~ 1100°F) which is outside the zone where the maximum attack occurs. Thus, the consideration of fireside corrosion has lead, in part, to placing limitations on superheater/reheater tube temperature as an operating consideration. A second response to fireside corrosion in SH/RH tubes has been to use a more resistant material. The increased resistance of the high Cr alloys like IN 671 (50 Cr - 50 Ni) is indicated in Figure 33-8. Because the wastage rate is a strong function of tube metal temperature, the propensity for rapid corrosion can be exacerbated by any of a number of problems that lead to even slightly elevated tube temperatures, such as the presence of thick internal oxide scale, or a variety of operating conditions such as excessive flue gas temperatures or local hot spots.

Volume 3: Steam-Touched Tubes

33-7

This sensitivity to temperature also explains why damage typically occurs at the 10 o’clock and 2 o’clock positions on the tubes. Deposits are usually thinner at these locations because of the flow around the tube (Figure 33-1) and, as a result, these locations are generally slightly hotter than the balance of the tube. Furthermore, a minimum level of SO3 (on the order of 250 ppm) is required for the corrosion process to occur, and access of SO2 to the tube surface is also easier in those locations of the tube because of the thinner deposits.

Weight Loss in 100 hrs, % 100

80

60 T22 (2 1/4% Cr) ferritic steel

40

18 CrÐ8 Ni stainless steel

20

0 538 (1000)

593 (1100)

649 (1200)

704 (1300)

760 (1400)

816 (1500)

The accelerated wastage results in loss of wall thickness which causes an increase in hoop stress and leads to eventual creep failure.

871 (1600)

Metal Temperature, °C (°F)

2.2 Sulfidation of Stainless Steels

Figure 33-7. Effect of metal temperature on fireside corrosion rate. Source: A.J. Blazewicz and M. Gold1

Weight Loss Ratio

4

3

2

316H 321H 347H 310S 17 Ð 14 CuMo Esshete 1250 Incoloy 800H AISI 314 TEMPALOY A-2 HK 4M Inconel 617 IN 671 Incoloy 807

Sulfidation is a second mechanism that may cause fireside corrosion in superheater/reheater tubes. Stainless steels are rapidly corroded at high temperatures by hydrogen sulfide gas, H2S; this can be formed in flue gas that has high CO levels, such as caused by incomplete combustion or low excess air levels. Sulfidation attack and carburization of the stainless steel will result in discrete iron sulfides at the grain boundaries of the corroded region.

1 2 3 4 5 6 7 8 9 10 11

5

12 13

1 4 6 3

1

9 10

2

8

7 11 13

It has also been proposed that a local carburizing atmosphere can be created during startup oil firing when unburned oil coats the tubes. When this deposited oil subsequently burns, a carburizing atmosphere is created which then makes the stainless steel more susceptible to sulfidation.

12

0 600 (1110°F)

650 700 (1210°F) (1290°F) Test Temperature °C (°F)

Synthetic ash: 37.5 mol% Na2SO4 37.3 mol% K2SO4 25 mol% Fe2O3

750 (1380°F)

Synthetic gas: 80%N2, 15% CO2, 4% O2 1% SO2, including saturated H2O Exposed time: 50 hrs Catalyst: V2O5

Figure 33-8. Weight loss ratio of various materials as a function of test temperature. Source: S. Kihara, et al.7

33-8

SH/RH Fireside Corrosion/Coal-Fired Units

2.3 Coal composition and corrosivity indices A basic introduction to the effects of coal composition on boiler tube failures was presented in Chapter 2, Volume 1. Some additional comments specific to fireside corrosion in superheater/reheater tubes is provided here.

The most important species responsible for fireside corrosion in coalfired boilers are alkali metal compounds and sulfur.

The mitigating properties of CaO and MgO have also been documented. Most recently in an investigation of alloys for advanced steam plant, it was found that a 21/2% addition of CaO was beneficial in reducing the rate of fireside corrosion, particularly for the less-resistant materials.10 Calcium oxide lowers the potential for sulfur at the tube surface by forming a layer of calcium sulfate in the deposit; this is expected to result in a lower corrosion rate. In the field experience, there is often such a layer found. The effect could also be due to increasing melting point. Several corrosion indices have been proposed to determine the likelihood that a given coal will lead to a potential fireside corrosion problem, including methods proposed by Borio, et al.11, Raask12, and Shigeta, et al.13 The index developed by Borio, et al.11 was based on an experimental study of U.S. coals from six mines, specifically for Eastern U. S. coals. The index produced a value ranged

sis)

coal ba

io n re du ct io n

co rro s

ion ros

50

50 %

cor

0

0 0.5 1.0 1.5 2.0 2.5 3.0 8.0 12.0 sis)

l ba

O3 Fe 2

(%

coa

10 12 14 16 18 20 Acid Soluble K20 (expressed as 2 Na2O Eq., ppm x 10 coal basis) red uct ion

0

25%

There is some evidence that the corrosiveness of these deposits increases with increasing coal sulfur content over about 2 percent.9 However, in general, for coals with sulfur levels above about 2 percent, changes in the content of volatile alkali metal compounds have a stronger influence on corrosion than do changes in sulfur content.2

(ppm e Na 2O

900 800 750 700 650 600 550 500 400 0-300

lubl

Acid so

0% Corrosion Reduction

As a general rule, for U.S. coals, eastern high-sulfur coals can be corrosive and prone to low melting sulfate or alkali-iron trisulfate formation, whereas western low-sulfur coals usually do not cause a coalash corrosion problem, but can cause slagging and fouling because of the high levels of sodium and calcium in the coal.6

Relationship of Coal Constituents to High Temperature Corrosion*

sio

%

75

o orr

c

0%

n

tio

uc

ed nr

on osi

ion

uct

red

r

cor

10

100 150 200 250 300 350 CaO Equivalent (% coal basis)

400

450

Corrosion Index 22 20 ) sis 18 l ba coa 16 % . eq 14 aO 12 sC da e 10 s res 8 exp O( 6 Mg + O 4 Ca 2 0

Example (coal F) Acid SOL K2O (expressed as Na2O eq.) Acid SOL Na2O CaO + MgO (expressed as CaO eq.) Fe2O3 Quantity of CaO eq. required to reduce corrosion rate by 100% (1 hour) (All values are measured on coal basis)

0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0

1410 ppm 540 ppm 0.50% 1.9% 160%

* (1) Probe temperatures maintained at average of 1100 °F (2) Probe metal Ð 321 SS

Figure 33-9. Nomograph for determination of corrosive index of coal. Source: R. Borio, et al.11

from 0-22, with the potential for corrosion increasing with higher index values. It considered the effect of alkali metals (Na + K), alkaline earth metals (Ca + Mg) and sulfur (as FeS) on the corrosion potential of the coal. For example, the presence of CaO and MgO in coals elevates the melting point of sulfates and lessens the possibility of molten salt attack. A nomograph (Figure 33-9) was developed that incorporated four coal composition factors: (i) acid soluble K2O, (ii) acid soluble Na2O, (iii) iron oxide Fe2O3, and (iv) CaO + MgO. This work lead to practical applications such as the blending of coal to adjust levels of CaO + MgO as a function of the Na, K and S content of coal. The peak temperature and the corrosion-temperature relationship at each temperature are likely to be different if coals are blended

or washed to meet emission standards, or if coals of different composition, such as Western coals are considered. The Borio index has had some success in ranking the potential for corrosion in units using Eastern U.S. coals, as most recently confirmed in field probe tests of seven materials under consideration for advanced plant conditions.6 Two units were part of the test program and burned high sulfur subbituminous coal with Borio index ratings of 2.5 and 3.5 For five of the seven alloys tested, the index ordered the corrosion behavior. However, an evaluation of the use of the index to correlate with units which had severe corrosion problems and burned a wider range of coal types met with only limited success.2

Volume 3: Steam-Touched Tubes

33-9

0.5 - 1.0

Medium High

A third index based on an extensive study of boiler deposits, test probe deposits, and laboratory corrosion coupons was developed by Shigeta, et al.13 It relates weight loss to the amounts of Na2O and K2O, and their ratio, the amount of SO2, and the inhibitive effects of CaO and MgO. The expression developed is shown in equation at bottom of this page. The four coefficients are read from Figures 33-10a, b, and c. The expression was validated for Type 347H stainless steel and Alloy 17-14 CuMo in 6,000 hour probe tests. Expected weight loss for other alloys was also determined and a ratio between the response of Type 347H and the alloy of choice can be determined from Figure 33-8 as a function of expected metal temperature. These indices can provide a qualitative guide to the effect of coal composition on the propensity for fire-

2 0.4 Example: Ash contains 12% Na2O and 8% K2O. Ratio is 6 to 4. Enter figure at that ratio on X-axis.

1

2 8

4 6 6 4 Na2OÐK2O Weight Ratio

8 2

10 0 1.0

4

0.8

MgO

3

0.6 2 0.4

CaO 1

0.2

0

4

0

1000

8 12 Contents of CaO, MgO (%)

16

20

1.2

0.8

0.4

2000

3000 SO2, ppm

4000

5000

Figure 33-10. Derivation of coefficients and effects of weight loss as a function of Na2O/K2O ratio, content of CaO, MgO and SO2. Source: J. Shigeta, et al.13

side corrosion. This is one of the tools necessary to confirm the root cause of SH/RH tube failures, as is outlined in Section 3 below.

Weight loss (mg/cm2) = -0.08 + 0.13 (wt % acid soluble Na2O + K2O) x CNa2O/K2O x CCaO x CMgO x CSO2 x time (hrs.)

33-10

0.2

1.6

c)

SO2 Content

> 1.0

Weight Loss (mg/cm2/hr)

Low

Coefficient (SO2)

< 0.5

CaO and MgO Content

Corrosiveness

0.6

Na2O 0 K2O 10 b)

Water-soluble Na + K, weight percent

0.8 3

Coefficient (K, Na)

proposed a simple ranking system based on the sum of the percentages of water-soluble sodium and potassium in coal as determined on the bomb residue from the measurement of coal calorific value. The three categories are:

1.0

4

Coefficient (CaO), and (MgO)

Raask12

Weight Loss (mg/cm2/hr)

a)

Na2O/K2ORatio

This limited applicability would indicate that other factors need to be considered when looking for predictive methods associated with fireside corrosion, and that use of the Borio index should probably be limited to bituminous coals of the type used to establish in the index.

SH/RH Fireside Corrosion/Coal-Fired Units

(33-1)

2.4 Effect of Cl in Coal Chlorine appears to play a role in releasing the alkalis from the coal, but levels above about 0.2 weight percent appear to be required.3 Corrosion rate has been observed to be a strong function of chlorine level for U.K. coals; increasing essentially

linearly for superheater/reheater alloys with increase in chlorine content of the coal over the range 0.1% to 0.5%.3 Past reasoning was that the correlation with chlorine was indirect. The release of sodium contained in sodium chloride in the coal triggered the release of potassium, so that the chlorine acted as a catalyst to the formation of sulfate or trisulfate attack. There is also some evidence that the formation of HCl in the flame can destroy the protective oxides such as Fe2O3. A review of the effects of chlorine in U.K. coals was made in 1991.5 U.K. coals generally have significantly higher levels of chlorine than typical U.S. coals of up to 0.7 weight percent (averaging around 0.25%). The review includes details on fireside corrosion and points to the linear dependence on chlorine for corrosion rates of austenitic superheater and reheater alloys. Additional key aspects from that review are:

• Corrosion rates in plants burning high chlorine coals can reach well over 100 nm/hr (~ 35 mils/yr) at surface metal temperatures of 630°C (~ 1165°F). • The source for chlorine in U.K. coals is mostly as weakly bound ions associated with organic matter; with the balance being NaCl. Chlorine is released as HCl, about 80 ppm for each 0.1% Cl in the coal. • HCl plays a key role in releasing both sodium and potassium from coal ash which leads to aggressive sodium-potassium sulfate deposition. • A correlation that relates corrosion rates to metal temperature, gas temperature, coal chlorine content, leading/non-leading tube position and alloy composition has been developed based on CEGB operating experience and plant corrosion measurements.14 It is of the form: R = K L B (Cl-a) (Tm - b)c (Tg - d)e

Corrosion Rate (nm/h)

150

100

50

0 0.05

Corrosion rate for leading tubes, normalized to metal temp 640°C, gas temp 1050 °C 0.15 0.25 0.35 0.45 0.55 Chlorine Content of Coal (% as received)

Figure 33-11. The linear dependence of corrosion rate on coal chlorine content for austenitic steels. 50 nm/hr is approximately 18 mils/yr. Source: W.H. Gibb and J.G. Angus15

where R = corrosion rate K, a, b, c, d, and e = empirically derived constants Cl = chlorine content in wt % Tm = surface metal temperature Tg = gas temperature L = lead edge tube factor B = alloy factor (unity for 300 series alloys) The relationship has been found to hold for austenitic alloys 316, 321, 347 and Alloy E1250. The B factor for more resistant alloys has also been evaluated and is: for Alloy 310 (B = 0.3) and 50% Cr-50% Ni (B= 0.1). The chlorine relationship for these more resistant alloys has not been confirmed. Figure 33-11 shows a plot of corrosion rate versus chlorine content which has lead to the observation that the increase in rate is approximately linear with chlorine content for U.K. coals.

It should be repeated that this correlation and the understanding of the effects of chlorine on high temperature corrosion is only applicable to U.K. utility operations. A number of U.S. utilities have been burning coals with elevated chlorine levels (up to 0.4 wt. %) from the Illinois Basin and have not, to date, experienced the extensive corrosion that has been observed in U.K. plants. Further field testing and laboratory work is currently being initiated to understand the basis for the emphasis on chlorine in U.K. utility experience, and the extent, if any, of chlorine effects in U.S. coals. There may be a difference in the way that the chlorine is bound in the coal and released in the flame between U.S. and U.K. coals. Alternatively, the coal chlorine content may simply be a useful measure of the availability in the coal of alkali metal salts that participate in fireside corrosion of superheater/reheater tubes.

(33-2)

Volume 3: Steam-Touched Tubes

33-11

3. Possible Root Causes and Actions to Confirm Superheater/Reheater Fireside Corrosion (Coal-Fired Units): Root Cause 1. Major influences on accelerated corrosion are overheating of the tubing and the deposition/formation of low melting-point, complex alkali salts. 2. Tube overheating can occur as a result of a number of design, operating, and/or maintenance-related conditions. 3. Fuel composition can affect the propensity for molten alkali salt formation.

3.1 Introduction Table 33-2 lists the primary root cause influences. In general, there are three groups of root causes: (i) overheating of tubes, (ii) fuel factors, and (iii) combustion factors. Independent of the root cause there are two actions that can be taken to confirm that the likely failure mechanism is fireside corrosion, or to monitor its progress: (a). Wall thinning can be identified using non-destructive examination techniques, primarily ultrasonic testing. As illustrated in Figure 33-5 and discussed previously, when the ratio of wall loss to the buildup of steamside oxide scale is large (typically greater than a factor of three and certainly greater than five), then fireside corrosion (or erosion), not longterm overheating, is the dominant damage mechanism. Additional detail about inspection and sampling methods is provided in Chapter 9, Volume 1. (b). Ash analysis to measure the melting point of ash layers, looking for low melting point constituents. Thermogravimetry (ASTM E1131)16 or differential thermal analysis (ASTM E794)17 can be performed to identify melting points of compounds in the deposit. Differential scanning calorimetry can also be used to determine melting points of specific species. The three groups of root causes are examined separately in the discussion that follows.

3.2 Influence of overheating of tubes There are a number of underlying causes of overheated tubes, any one of which can elevate the tube metal temperature into the highly corrosive regime. Independent of the underlying cause, the following actions can be used to confirm whether tube overheating is contributing to a problem with fireside corrosion in superheater/reheater tubes:

33-12

SH/RH Fireside Corrosion/Coal-Fired Units

(c). Oxide thickness measurements (ultrasonically) at typical locations provides an indication of tube temperatures within the element. See Chapter 8, Volume 1 for details of the interpretation of oxide scale thickness. (d). Selective tube sampling and metallurgical analysis can be used to confirm the results of NDE measures of steamside oxide thickness and tube metal wall thickness. Oxide composition may also provide additional insight about temperatures reached in the tubes. See additional discussion of this point in Chapter 2, Volume 1. (e). Monitoring of thermocouples permanently installed across the SH/RH outlet legs in the vestibule can provide an indication of the hottest platens across the boiler. This then pinpoints where steps (a) and (c) should be performed. Specific causes of overheating can include: 3.2.1 Poor initial design Ð choice of material. This aspect of poor design may be evidenced by overheating that leads to accelerated corrosion at or near changes to higher grade material, where the original design should have changed to the higher grade material sooner. Actions to confirm are: (f). See items (c) and (e) above. 3.2.2 Poor initial design Ð extra gastouched length. Poor initial design can also result in extra gas-touched length for similar (maybe adjacent) tubes. This means that one tube will run much hotter. Possible locations where this might be a consideration include accelerated corrosion of lead or wrap-around tubes (leading tubes are most affected), or a tube that surrounded a radiant cavity and therefore picks up excessive heat. A discussion of gas-touched length of tubes can be found in Chapter 32. Particular reference to Figure 32-12 can be made for an example of how a plot of location of tube failures and

Table 33-2 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

• Choose repair strategy based on severity of corrosion rate. • Implement long-term actions from choices in Figure 33-12 in conjunction with on-going program of remaining life assessment and monitoring.

• Perform remaining life assessment. • Set up long-term monitoring and re-evaluation program. • Evaluate full range of available options using roadmap in Figure 33-12.

(c). NDE of steamside oxide thicknesses. (d). Selective tube sampling and metallurgical analysis to confirm steamside oxide and wall thickness. (e). Monitoring of thermocouples installed across the SH/RH outlet legs in vestibule to identify hottest platens across the boiler.

• As above.

• As above.

3.2.1 Poor initial design - choice of material

(f). Items (c) and (e) above.

• As above, primary emphasis on upgrading to a more resistance material.

• As above, emphasis will be on identifying locations where material upgrading is required. • May involve redesign of circuit to extend the use of the higher grade material.

3.2.2 Poor initial design - extra gas-touched length.

(g). Evaluate temperatures across the element (via thermocouple or steamside oxide measurements) to determine if sections particularly near material changes are running too hot. See discussion of gas-touched length in Chapter 32 and sample plot in Figure 32-12.

• As above.

• As above, emphasis will be on identifying locations where material upgrading is required. • May involve redesign of circuit to extend the use of the higher grade material.

3.2.3 Internal oxide growth which occurs during operation.

(h). Items (c) and (d) above.

• As in primary list above (repairs, followed by long-term strategy) plus chemical cleaning of steamside scale.

• As in primary list above, also see actions for the long-term overheating/ creep of tubes (chapter 32).

3.2.4 High temperature laning.

(i). Monitor temperatures as in (e) above. • As in primary list above (j). Laning can be identified with cold air (repairs, followed by velocity technique. See Chapter 14, long-term strategy). Volume 2 on flyash erosion for a discussion of the technique.

• Controlled with flow distribution screens; in practice is difficult to overcome because of high temperatures in SH/RH. • Review primary list of alternatives in Figure 33-12 for options.

Actions to Confirm

3.1 Potential actions for all root causes of fireside corrosion (a). NDE measures (typically UT) to identify wall thinning. (b). Ash and deposit analysis to identify presence of low melting point constituents, particularly alkali iron trisulfates.

3.2 Influence of overheating of tubes.

Volume 3: Steam-Touched Tubes

33-13

Table 33-2 Major Root Cause Influences, Confirmation and Corrective Actions (continued) Major Root Cause Influences 3.2.4 Tube misalignment (out of bank)

Actions to Confirm (k). Visual examination.

3.2.6 Operational problems when coal type is changed

3.2.7 Rapid startups causing reheater to reach temperature before full steam flow

(l). Check startup probe and that initial gas is limited to 1000°F (538°C) prior to RH flow.

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

• Realign tubes, implement on-going program of remaining life assessment and monitoring.

• Perform remaining life assessment. • Set up long-term monitoring and re-evaluation program.

• Evaluate whether operating procedures such as sootblowing can be economically changed to protect SH/RH tubes.

• Perform remaining life assessment. See discussion of methods in Chapter 8, Volume 1. • Set up long-term monitoring and re-evaluation program. • Evaluate full range of available options using roadmap in Figure 33-12.

• Modify startup procedures if feasible.

• As above.

3.3 Root causes related to fuel factors • As above. 3.3.1 Change to fuel with unusually corrosive ash, particularly those with high S, Na, K, or Cl

(m). Evaluate coal composition using corro- • As in primary list above sivity index. (repairs, followed by (n). Analysis for low melting point of ash long-term actions). components using probes. (o). Analysis of metallurgical cross sections, particularly for Cl, S, C, Na, and K. (p). Install continuous readout corrosion sensors if unit switches coal or uses spot market coal.

• As above, plus • Develop a fireside testing program using guidance provided in fireside testing guidelines.18 • Investigate coal changes with Coal Quality Impact Model (CQIM)24-26 or equivalent, including economics evaluation.

3.4 Root causes related to incomplete or delayed combustion. (q). Monitor for levels of CO and O2. (r). Check for unburnt startup oil.

33-14

SH/RH Fireside Corrosion/Coal-Fired Units

• As in primary list above (repairs, followed by long-term actions.).

• Perform remaining life assessment. See discussion of methods in Chapter 8, Volume 1. • Evaluate full range of available options using roadmap in Figure 33-12. • Develop a fireside testing program using guidance provided in fireside testing guidelines.18

of gas-touched length can be used as a diagnostic. Actions to confirm: (g). Evaluate temperatures across the element via thermocouple or steamside oxide measurements to determine if sections are running too hot, particularly near material changes. 3.2.3 Internal oxide growth which occurs during operation. This buildup of internal oxide gradually increases the tube temperature until it rises into the regime where the corrosion rate rapidly increases.

3.2.7 Rapid startups can cause reheater to reach temperature or even overheat before full steam flow and thus cause overheating. An action to confirm is: (l). Check startup probe and check that the initial gas temperature is limited to 1000°F (538°C) prior to RH flow.

3.3 Root causes related to fuel factors

3.2.4 High temperature laning.

3.3.1 Change of fuel to a coal with an unusually corrosive ash, particularly those with high content of S, Na, K or Cl.

Actions to confirm consist of:

Actions to confirm include:

(i). Monitor temperatures as in (e) above.

(m). Evaluate coal composition using corrosivity index for Eastern coals.11 As noted above, there is no universally applicable, quantitative index to the corrosivity of coals, however for Eastern U.S. coals, the Borio index can give at least a qualitative indication of whether a fuel change has lead to a potentially more corrosive condition. Western coals generally do not cause corrosion problems of this type because of low sulfur and high alkaline earth oxide levels, and are not amenable to analysis by the Borio index.

(h). See items (c) and (d) above.

(j). Laning of gases can also be detected using a cold air velocity test (CAVT). The details of the test are described in Chapter 14, Volume 2 on flyash erosion. 3.2.5 Tube misalignment (out of the bank). In addition to above indicated actions: (k). Visual inspection can detect if there is a misalignment problem. 3.2.6 Operational problems when coal type is changed. This may change the relative absorption patterns between furnace and convective sections.

(n). If liquid ash corrosion is suspected, thermogravimetry (ASTM E1131)16 or differential thermal analysis (ASTM E794)17 can be performed to identify melting points of compounds in the deposit. Deposits can

be collected using high temperature probes; the methods and analysis are described in reference 18. (o). Energy dispersive x-ray and/or x-ray dot mapping of metallographic cross sections through damaged tubes can be used to detect the presence and distribution of S, C, Na, K, and Cl. This type of analysis is needed to confirm whether the corrosive wastage is simple molten salt attack or whether a local reducing or carburizing condition is also leading to attack via sulfidation. (p). Install corrosion sensors (continuous readout), especially if the unit often switches coal or uses spot market coal.

3.4 Root causes related to incomplete or delayed combustion Actions to confirm (q). Monitor for levels of O2, H2S, and CO. High levels of CO (> 1%) and low levels of oxygen (< 0.1%) are of particular concern.3 The level of CO can also be measured in the flue gas at the economizer outlet or after the ID fans. Oxygen levels can be sampled at the economizer exit to provide an overall indication of the combustion process in the boiler. (r). Check for unburnt startup oil deposits on tubes.

4. Determining the Extent of Damage Ultrasonic testing (UT) can be used to measure both wall thinning and steamside oxide thickness. Locations should be chosen which are the most susceptible to fireside corrosion as described in Section 1.2 above. On a particular tube, locations of maximum wastage should be identified. Selective sam-

pling is recommended to confirm the results of the NDE examinations and to evaluate internal scale and external deposits. Additional detail on UT methods can be found in Chapter 9, Volume 1. Methods of metallurgical evaluation are reviewed in Chapter 6, Volume 1.

Volume 3: Steam-Touched Tubes

33-15

5. Background to Repairs, Immediate Solutions and Actions Superheater/Reheater Fireside Corrosion (Coal-Fired Units): Immediate Solutions and Actions 1. Immediate solutions should be chosen in conjunction with: (i) a knowledge of the severity of corrosion, and (ii) an analysis of remaining life discussed under long-term actions below. 2. Repairs should occur on two levels depending upon the severity of the problem. For the short term, tubes can be replaced with the same material, or a palliative coating may provide somewhat better corrosion resistance; for the long term, change to a more resistant material is the preferred materials solution. 3. Other actions are possible including limiting metal temperatures, reducing main steam temperatures, chemical cleaning, tube alignment, minimizing laning of gases, fuel changes, or blending or washing the coal. The choice will depend on the root cause of the problem and at least a cursory economic analysis to identify the most likely costeffective solution.

33-16

5.1 Need for remaining life assessment Any immediate solution or repair strategy should consider that a longer term remaining life analysis methodology should also be implemented. Such programs are discussed in detail in Chapter 8, Volume 1, and are summarized under long-term actions below.

5.2 Repairs If the corrosion rate has been relatively modest (< 25 nm/hr (~ 9 mils/yr)) and is likely to continue to be so, an acceptable strategy is to retube with the same alloy and monitor closely the wastage rate. Another alternative, but not a preferred approach, is the use of a palliative coating or tube shields. An overview of options is provided under longterm actions. Pad welds should definitely not be used as a repair measure because of the uncertainty of the tube condi-

SH/RH Fireside Corrosion/Coal-Fired Units

tion, such as the presence of creep cracks and their depth, and the conditions of the internal tube surface. Further discussion about weld repairs can be found in Chapter 11, Volume 1. For higher corrosion rates that are resulting in rapid wastage of the existing alloy, the replacement should be with a more resistant material. In addition to the upgrading of materials, there are some steps that can be taken over a short term to minimize the severity of fireside corrosion. These include: taking interim steps to limit tube metal temperatures such as by limiting main steam temperature or chemical cleaning to remove steamside oxide, aligning tubes, increasing sootblowing operation, and others. However, because of the range of options and the interconnectedness of them, a complete overview of all options is included in the next section.

6. Background to Long-Term Actions and Prevention of Repeat Failures Superheater/Reheater Fireside Corrosion (Coal-Fired Units): Long-Term Actions Long-term strategies will generally fall into two major categories: (i) materials strategies that provide increased protection or replace the component, and (ii) operating strategies to try to control tube temperatures or corrosiveness of the coal.

As noted above, it may not be possible to remove the root cause for many fireside corrosion problems. Knowing how to minimize the wastage rate and the application of a predictive remaining life assessment process, including periodic inspection and monitoring, are the keys to economic handling of fireside corrosion problems.

rials options and (ii) operating options. If an initial design condition is at the root cause of a fireside corrosion problem there are three approaches that can be taken: correct the problem; accept the fault and the corresponding corrosion rates and expect an increased outage frequency; or accept the fault and seek a material solution.

Figure 33-12 outlines most of the available strategies for superheater/ reheater fireside corrosion. As shown in that figure, there are two primary, and not mutually exclusive, routes that can be followed: (i) mate-

The circled numbers in Figure 33-12 are used to identify options for the discussion that follows; no ranking is implied, however, boxes with bold outline indicate those options which have been the most successful.

Corrosion rate confirmed Extent determined

Remaining life assessment

Materials solutions

1

Operating solutions 8

Provide protection

Modify fuel

Replace component

Control (limit) tube temprature

2

Shielding 3 Coating More resistant material

7

Ð Change fuel Ð Blending Ð Washing Gas side

Same material

5 Monolithic

12

Minimize laning of gases

Limit main steam temp.

Sootblowing operation

9

14

15

Excess air strategies

13

Tube alignment

6 Coextruded

Original thickness 4 Notes: a) Remaining life assessment (1) is almost mandatory to decide which option should be adopted b) Boxes outlined in bold indicate options that have been most successful c) Numbers refer to main text

Steam side

Chemical clean deposits 11

Redistribute steam flow 10

Figure 33-12. Strategies for preventing repeat failures by fireside corrosion in superheater/reheater tubes of coal-fired units.

Volume 3: Steam-Touched Tubes

33-17

Remaining life assessment (option 1, Figure 33-12) A remaining life assessment is required to relate the rate of corrosion wastage to the desired life or to determine the time available to implement the desired option. Therefore such an assessment should be undertaken in parallel with any of the other options. A systematic program will include: baseline measurement, monitoring rates of wastage, application and monitoring of control measures, and assessment of the effects on remaining life. Monitoring of the flue gas, metal and steam temperatures, combustion conditions, and fuel composition should also be considered, as these can determine corrosion rates while the unit is still on-line.3 Particularly important are step changes in key parameters. A critical step in determining the remaining life of a tube and in choosing the optimal solution is knowing the rate at which damage is accumulating. The lifetime of superheater/reheater tubes can be calculated using the oxide thickness technique as discussed in detail in Chapter 8, Volume 1, which includes the damage accumulation from both fireside corrosion and long-term overheating. A key goal of the remaining life assessment will be to establish the acceptable rate of wastage, given the desired remaining life of the unit and an analysis of damage accumulation. As a rule of thumb, rates of wastage above 25 nm/hr (~ 9 mils/yr.) indicate the need for some periodic activity such as annual inspection and for the explicit determination of re-evaluation periods. Rates above 50 nm/hr (~ 18 mils/yr.) indicate a serious decrease in life of the tubing; an estimate of remaining life should be made and the appropriate actions taken if that life is unacceptable.

33-18

Shielding (option 2) The most common temporary measure is the installation of shields to the leading edges of affected tubes. High chromium alloys such as Type 304 or 347 (austenitic stainless steels) or 309 (25 percent Cr) are often used. The shield is curved to fit the tube surface and tack-welded in place. As they are not cooled, the metal temperatures of the shields are above the range for extensive attack as indicated in Figure 33-8.

Coatings (option 3) A variety of processes have been suggested for applying metallized coatings to tubes in situ as a means of increasing corrosion resistance. The advantage of coatings is that very corrosion-resistant materials can be applied at specific susceptible sites, so it is not necessary to replace entire sections of tubing in order to overcome localized problems. Among the coating methods that have been tried for fireside corrosion resistance are chromizing and aluminizing. Flame- or plasmaspraying, with and without subsequent heat treatment, have also seen significant development work. The various coatings and application techniques have had variable and generally poor results in U.S. utilities. The primary problems have been adherence of the coating, the reproducibility of coating techniques, and the need to have very well prepared surfaces. In general, coatings should currently be regarded as a quick fix which will require continued maintenance. The former CEGB has tried a number of coatings for use in corrosion and erosion resistance.19,20 The primary use of coatings has been for the prevention of fireside corrosion in waterwalls of coal-fired units (see the description of results in Chapter 18, Volume 2). There is less field experience for coatings on superheater/reheater tubing in either coalor oil-fired units.

SH/RH Fireside Corrosion/Coal-Fired Units

If used, important features of coatings include19: • Importance of clear specification of the areas to be coated, materials to be used and thickness required. Time available and other restrictions to site applications should be clearly specified. • Importance of surface preparation to provide a contaminant-free, roughened surface. • Importance of producing samples throughout the application as a check that process parameters were producing consistent and acceptable quality. Currently, either replacement in-kind or with a material of high corrosion resistance, depending on the wastage rate which has been experienced, are preferred options to the application of coatings.

Replacement with same material, same thickness (option 4) If the corrosion rate is only slightly higher than that required to reach the desired life as calculated from the remaining life assessment, tube replacement can be made in-kind.9

Change to a more resistant material, ÒmonolithicÓ (option 5) An upgraded material can be used where unit operation is at high temperatures and fireside corrosion remains a problem despite the attempts at other fixes. A number of alternatives have been used successfully. The material chosen will depend on what is currently being used, and what the desired resistance is to be. If the damage is occurring on a ferritic material then an appropriate change is to a higher grade (Cr) or an austenitic stainless steel. If serious fireside corrosion of austenitic materials is taking place, then consideration may be given to more resistant materials. There have

310 stainless steel. Weight Loss mg/cm2 30

An extensive field evaluation of the resistance of various materials was also conducted using air-cooled, retractable corrosion probes constructed of various materials. Corrosion probes were exposed for 4,000 hours, 12,000 hours, and 16,000 hours at three facilities. Two of the units burned eastern high-sulfur subbituminous coals with Borio Index values of 2.5 and 3.5 respectively; the third burned western lowsulfur coal for which the Borio index was not applicable.

347HSS Esshete 1250 17-14 CuMo 253MA 310SS Alloy 800H 35Cr-45Ni

20

10

0

0

600

650 700 Temperature, °C

750

Figure 33-13. Effect of test temperature on corrosion loss for various alloys in 0.25 vol% SO and 5 wt% alkali sulfate. Source: W. Wolowodiuk, et al.10

been a number of laboratory and field investigations of superheater/ reheater corrosion of materials, particularly for advanced steam conditions. The following brief review of some of the laboratory and field investigations highlights some of the key issues as they might apply to materials choices in currently operating plant. A laboratory investigation of the response of eleven high-temperature alloys (7 cladding alloys for coextruded tubes and 4 chromized steels) to a simulated field environment provided the following general conclusions6:

• Corrosion rates increased with increasing levels of alkali (Na2SO4 + K2SO4). The effect of alkali level was more significant than SO2 for

the more highly resistant materials.

• The relation between tube material, temperature and corrosion rate is shown for these laboratory tests in Figure 33-13. The peak temperature for corrosion loss is different for different alloys, in part because the temperature range in which the molten salt is stable depends on the surface character of the metal. Corrosion products can also affect the melting temperature of salt. Generally, temperatures tend to peak at a lower level for more highly resistant alloys. For example, the peak temperature is about 700°C (~1300°F) for low-resistance alloys like 17-14 CuMo and 600°C (~ 1100°F) for highly resistant alloys like Type

Maximum wastage was observed on the high sulfur subbituminous probes at the 10 and 2 o’clock positions (flue gas impinged at 12 o’clock) with an appearance that confirmed the occurrence of liquidphase (coal-ash) corrosion. The following additional observations were noted: • Resistance to coal-ash corrosion increases with an increase in chromium level in the alloy. A plot of alloy chromium content versus average metal loss is shown in Figure 33-14. Results show that a Cr content over 25% Cr was required for stable corrosion resistance.21 This has important implications for the choice of replacement materials.

• The rate of corrosion experienced by Type 347 stainless steel could be reduced by a factor of two by using Type 310Nb stainless (SA213-T310 CbN). Further, some of the materials which show the most promise have been tested only in short-term tests, their longer-term performance is as yet relatively unknown. As a general rule, the more resistant the material, the more expensive it will be. For this reason, it may be preferable to use duplex tubing such as discussed next. Such co-

Volume 3: Steam-Touched Tubes

33-19

stainless steel). A typical application used 54 mm (2.13 in.) outside diameter 310 with 2.9 mm (0.11 in.) thickness on 2.0 mm (0.08 in.) Eshete 1250 inner. Performance improvement factors have ranged from 2-5 times better than the original installed materials, depending upon the particular application.22

Average Sound Metal Loss (mils) 100 Plant A Plant B Plant C

80

60

40

20

0 5

10

15

20 25 Cr Content Wt%

30

35

Figure 33-14. Effect of chromium content on resistance to attack after 16,000 hours exposure at three plants. Source: J.L. Blough, et al.6

extruded tubes can supply more reliable resistance, but they are rather expensive due to the complexity in manufacturing. There is an economic advantage in using high strength and resistant monolithic tubing, such as SA213-T310 CbN. A trial calculation has suggested that the increased cost of such tube materials is around 10% compared with the use of Type 347H for superheaters of advanced boiler designs, i.e., steam condition of 566°C (1050°F) and 31 MPa.10

applications when compared to mechanically bonded or bi-metallic tubes.22 Specifications for such tubes have existed for several years.23

Change to a more resistant material, composite (such as co-extruded) tubing (option 6)

The primary material for the outer layer of superheater/reheater tubing has been Type T310 stainless (25 Cr 20 Ni), or a modified Type 310 made by increasing the silicon content to 0.75 - 1.5%, the chromium to a minimum of 25%, and niobium to a minimum of 8X the carbon level. Where additional resistance has been needed a 50Cr-50Ni alloy has been used as the outer layer. The inner layer has typically been Eshete 1250 (a high-temperature, high-strength

Highly resistant materials such as alloy 310 and 50% Cr- 50% Ni can provide excellent resistance, but since they have insufficient creep strength their use is in the form of co-extruded tubing. Metallurgicallybonded processing provides superior tubes for use in utility boiler

33-20

The former CEGB has used such tubes on a large-scale since 1976. Over 230 km of tubing has been installed in twelve units. Such a materials solution has been found to be the most economical approach to dealing with fireside corrosion problems which are caused by the aggressive nature of the coals burned in U.K. boilers.22

SH/RH Fireside Corrosion/Coal-Fired Units

Tube bending characteristics of coextruded material have been found to be identical to monolithic materials.22 Weld procedures have generally matched weld metals to the base metal to maintain property levels. Conventional welding techniques and normal quality control have been found sufficient to ensure good weld quality. The CEGB experience base was over 70,000 welds through 1984 without weld failure. No preheat or post-weld heat treatments have been required. A discussion on welding of co-extruded tubing can be found in Chapter 11, Volume 1.

Fuel change, blending, washing (option 7) Fuel changing expressly for the short-term control of fireside corrosion is unlikely; most often the motivating factor is to reduce emission levels, particularly of sulfur. However, as discussed above, changes that adjust the presence of corrosive species, including fuel changes, blending or washing, or that increase the amount of mitigating species will have positive effects on the corrosion rate. Changing coals to some of the Western varieties or lignite will offer these benefits inherently. Analysis of the likelihood that a change in fuel or in fuel handling will be beneficial can be made by evaluating the fuel composition with a corrosivity index. As discussed at length above, however, there is still no absolute agreement on a particular index for all U.S. coal types. A method of investigating coal changes, blending and washing, such as the application of the Coal Quality Impact Model (CQIM)24-26 can be used. Such a method will provide information on the total eco-

nomic impact as well as on the potential for fireside corrosion.

Limit tube temperatures (option 8) Given the strong correlation between tube metal temperature and the potential for corrosion damage, several strategies can be implemented that directly address root causes of overtemperature in the tubes; they are discussed separately below.

Limit steam temperature (option 9) One means to minimize tube metal temperatures is to limit the main steam temperature; this has historically been a primary strategy for the control of superheater/reheater fireside corrosion. With tube surface temperatures approximately 30 to 80°C (~ 50 to 150°F) higher than steam temperatures, limiting the steam temperature to around 538°C (1000°F), or in some cases to 566°C (1050°F) should result in an upper limit on tube temperature that falls below the region of maximum corrosion as shown in Figure 33-7. This approach has governed operating temperatures in the U.S. since the late 1950s when extensive problems with corrosion were experienced in boilers operating at 621°C (1150°F) and 650°C (1202°F).27 It has been estimated that as a result, only about 8 percent of U.S. coalfired boilers are still operating at 1050°F (566°C) steam. There are several problems with this approach. The primary drawback is that it is not efficient; there is a severe heat rate penalty. Predicting exactly what temperature is required is difficult, given the considerable scatter in the corrosion rate as a function of temperature and its peak. Finally, in a field survey of 42 operating units2 the relationship between operating steam temperature and fireside corrosion was found not to be direct, but to involve the consideration of the numerous other factors such as coal ash composition and flue gas temperature.

For these reasons, it was recommended that if limits on main steam temperature were to be used as a control strategy, an empirical approach be used to set steam temperature by monitoring tube wastage rates and ensuring that the rate was less than 25 nm/hr (~ 9 mils/yr.).2

Redistribute steam flow (option 10) The technique of steam flow redistribution in superheaters has seen recent significant development.28 Significant nonuniformities in tube temperature can exist across a superheater from a variety of causes. This results in a wide range of tube metal temperatures and damage accumulation by creep, as well as making these tubes more susceptible to fireside corrosion because of the characteristic “bellshaped” curve of wastage. Redistribution of steam flow can serve to equalize the temperatures across the superheater. With steam flow controllers (orifices) of different sizes and lengths in the tubes coming out of the inlet header, the flow can be decreased in “cold” tubes, which increases their temperatures. This change will also increase steam flow to “hot” tubes with a concurrent decrease of the metal temperatures. In conjunction with selected tube plugging and replacement, such a method offers significant opportunity for extending the life of superheaters and reheaters. This method was designed primarily as a means of extending the creep life of tubes subject to overheating. If overheating of selected tubes is at the root cause of fireside corrosion, the technique could also be considered to selectively lower tube metal temperatures and thus decrease fireside corrosion. The method is discussed in more detail in Chapter 32 on long-term overheating/creep in superheater/reheater tubes.

Chemical cleaning to remove steamside oxide scale (option 11)

If excessive temperature in the superheater/reheater tubes is a contributing factor to the corrosion process, and if that condition has been exacerbated by the presence of increasing steamside oxide scale thickness, a solution involving chemical cleaning may be in order. An overview of chemical cleaning in SH/RH circuits can be found in Chapter 4, Volume 1.

Minimize laning of gases (option 12) Laning or channeling of gases through certain tube sections can lead to overheated tubes. This is very difficult to overcome; it can be recognized and monitored using the cold air velocity technique described at length in Chapter 14, Volume 2 on flyash erosion. However, the application of flow distribution screens, a means to control locally high velocities, is made more difficult in the SH/RH by the high temperature environment.

Correct misalignment of tubing (option 13) If misalignment of tubes has resulted in localized fireside corrosion, this problem should be corrected by realigning the affected tubes. This will decrease the number of tubes that are directly exposed to the gas flow.

Change frequency, check effectiveness of sootblowing (option 14) This can be an important action because it can stop the formation of excessive deposits which result in laning or channeling in adjacent areas.

Lowering excess air (option 15) Lowering excess air could reduce the amount of SO3 and possibly the stability of the alkali-iron trisulfates; however, such a change may also increase waterwall fireside corrosion (see a more complete discussion in

Volume 3: Steam-Touched Tubes

33-21

7. Case study Superheater/Reheater Fireside Corrosion Case Study: Field Experience Chapter 18, Volume 2). A utility survey conducted in 1987 reviewed the effect of steam temperature and coal sulfur level on the severity of SH/RH fireside corrosion. The primary conclusions from that work were: 1. The higher the main steam temperature the more likely that the unit would report a problem with fireside corrosion. Nine of the 16 boilers operating at 566°C (1050°F) reported severe problems. In contrast, only 12 of 36

2. The severity of corrosion was ranked qualitatively (none, light, moderate, severe) and was found to be related to coal sulfur level; the higher the sulfur content, the more severe the corrosion, particularly at sulfur contents greater than 2.5%. 3. The bell-shaped curve was found to be at best only a general guide to the likelihood of severe fireside corrosion, except at the lowest operating temperatures. In other words, tube metal temperature was not the only factor that determined whether fireside corrosion would occur. There also needed to be a corrosive and molten (low-melting

point) ash. Not all coals are corrosive. The need for a corrosive coal to cause damage is a major difference between fireside corrosion and long-term overheating failures. 4. If unit temperatures are to be set to avoid fireside corrosion, an empirical approach that includes monitoring of wastage rates should be implemented. If corrosion rates exceed a level of about 25 nm/hr (~ 9 mils/yr), steam temperatures can be lowered until that level is achieved or bettered. Source: D.N. Williams, et al.2

8. References 1Blazewicz,

A.J. and M. Gold, “High-Temperature GasSide Corrosion in Coal-Fired Boilers”, presented at the ASME Winter Annual Meeting, New York, December, 1979. 2Williams,

D.N., H.R. Hazard, H.H. Krause, L.J. Flanigan, R.E. Barrett, and I.G. Wright, Fireside Corrosion and Fly Ash Erosion in Boilers, Research Project 2711-1, Final Report CS-5071, Electric Power Research Institute, Palo Alto, CA, February, 1987. 3Laxton,

J.W., D.B. Meadowcroft, F. Clarke, T. Flatley, C.W. King, and C.W. Morris, The Control of Fireside Corrosion in Power Station Boilers, Third edition, Central Electricity Generating Board, 1987. 4Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 5Latham,

E., D.B. Meadowcroft, and L. Pinder, “The Effects of Coal Chlorine on Fireside Corrosion”, Chlorine in Coal, J. Stringer and D.D. Banerjee, eds., Elsevier Science Publishers, Amsterdam, 1991, pp. 225-246.

33-22

SH/RH Fireside Corrosion/Coal-Fired Units

6Blough,

J.L., M.J. Krawchuk, G.J. Stanko, and W. Wolowodiuk, Superheater Corrosion: Field Test Results, Research Project 1403-19, Final Report TR-103438, Electric Power Research Institute, Palo Alto, CA, November, 1993. 7Kihara,

S., A. Ohtomo, I. Kajigaya, and F. Kishimoto, “Recent Plant Experiences and Research Into Fireside Corrosion in Japan”, Werkstoffe and Corrosion, 39, 1988, pp. 69-83. 8Koopman,

J.G., E.M Marselli, J. Jonakin, and R.C. Ulmer, “Development and Use of a Probe for Studying Corrosion in Superheaters and Reheaters”, Proceedings of the American Power Conference, Volume 1, 1959, pp. 236-245. 9Holmes,

D.R. and D.B. Meadowcroft, “Fireside Corrosion and Problems of Tube Life Prediction”, Symposium on Thermal Utilities Boiler Reliability, McMaster University, Hamilton, Ontario, May, 1983. 10Wolowodiuk,

W., S. Kihara, and K. Nakagawa, Laboratory Coal Ash Corrosion Tests, Topical Report GS6449, Research Project 1403-19, Electric Power Research Institute, Palo Alto, CA, July, 1989.

11Borio,

R., et al., “The Control of High-Temperature Fireside Corrosion in Utility Coal-Fired Boilers”, U.S. Office of Coal Research, Research and Development Department, No. 41, April ,1969.

20Bennett,

A.P. and M.B.C. Quigley, “The Spraying of Boiler Tubing in Power Plants”, Welding and Metal Fabrication, November, 1990, pp. 485-489. 21Anon.,

E., Mineral Impurities in Coal Combustion Behaviour Problems and Remedial Measures, Hemisphere Publishing Company, Washington, D.C., 1985.

“Approaches for Predicting Corrosion Rates of Superheater Tubes in Coal-Fired Boilers”, Materials & Components in Fossil Energy Applications, U.S. Department of Energy and the Electric Power Research Institute, No. 106, October, 1993.

13Shigeta,

22Flatley,

12Raask,

J., Y. Hamao, H. Aoki, and I. Kajigaya, “Development of a Coal Ash Corrosivity Index for High Temperature Corrosion”, ASME Joint Power Conference, Portland, Oregon, October, 1986. 14Meadowcroft, D.B, in T.N. Rhys-Jones, ed., Surface Stability, The Institute of Metals, London. 15Gibb,

W.H. and J.G. Angus, J. Inst. Energy, Volume 56, 1983, p. 149. 16American

Society for Testing and Materials, Standard E1131-86, “Standard Test Method for Compositional Analysis by Thermogravimetry, 1992 Annual Book of ASTM Standards: General Methods and Instrumentation, Volume 14.02, American Society for Testing and Materials, Philadelphia, PA, 1992. 17American

Society for Testing and Materials, Standard E794-85 (1989), “Standard Test Method for Melting Temperatures and Crystallization Temperatures by Thermal Analysis, 1992 Annual Book of ASTM Standards: General Methods and Instrumentation, Volume 14.02, American Society for Testing and Materials, Philadelphia, PA, 1992. 18Sotter, J.G., J.A. Arnot, and T.M. Brown, Guidelines for Fireside Testing in Coal-Fired Power Plants, Research Project 1891-3, Final Report CS-5552, Electric Power Research Institute, Palo Alto, CA, March, 1988. 19Morgan-Warren,

E.J., “Thermal Spraying for Boiler Tube Protection”, Welding and Metal Fabrication, Jan/Feb, 1992, pp. 25-31.

T. and T. Thursfield, “Review of Corrosion Resistant Co-Extruded Tube Development for Power Boilers”, 1984 ASM Conference on Coatings and Bimetallics for Energy Systems and Chemical Process Environments, held at Hilton Head, South Carolina, November 12-14, 1984.

23CEGB

Standard 680 710, Engineering Documents Unit,

London. 24Davidson,

P.G, et al., Development and Application of the Coal Quality Impact Model: CQIMTM, Research Project 2256-2, Final Report GS-6393, Electric Power Research Institute, Palo Alto, CA, April, 1990. 25CQIM

Computer Code and Manuals: Volume 1: CQIM: User’s Manual and Theory Manual, Volume 2: Data Collection Manual and Theory Manual, Electric Power Research Institute, Palo Alto, CA, December, 1990. 26Pavlish,

J.H., P.R. Miller, N.C. Craig, and A.K. Mehta, “CQIM - An Analytical Tool Used to Evaluate Performance and Economic Issues”, Proceedings: Ninth Annual International Pittsburgh Coal Conference, October, 1992. 27Reid, W.T., External Corrosion and Deposits - Boilers and Gas Turbines, Elsevier, New York, 1971. 28Hara,

K., C. Lee, R. Moser, T. Rettig, and K. Clark, Improved Superheater Component Longevity by Steam Flow Redistribution, Research Project 1893-13, Final Report TR-101697, Electric Power Research Institute, Palo Alto, CA, December, 1992.

Volume 3: Steam-Touched Tubes

33-23

ACTIONS for SH/RH Fireside Corrosion Two paths for the BTF team to take in the investigation of SH/RH fireside corrosion damage begin here. The goal of these actions is to see if further investigation of fireside corrosion is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a SH/RH BTF has occurred and fireside corrosion is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by fireside corrosion.

Action 1a: If a SH/RH BTF has occurred and fireside corrosion is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of SH/RH fireside corrosion: ➠Review Figure 33-6 for typical boiler regions. ➠Review main text, section 1.2 for description of susceptible locations

➠ Confirm that the macroscopic

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit: • Any evidence of molten salts observed or measured. • Any evidence of wall loss observed or measured.

appearance of the failure includes such features as:

• Any evidence of “alligator hide” observed or measured.

• Multilayered fireside scale and ash. See Figure 33-1.

• Change to a more aggressive coal.

• Maximum attack (wastage) at the 10 o’clock and 2 o’clock positions (may be at other locations so that this is not definitive).

• Evidence that tubes may be overheating such as from thermocouple readings or evidence of excessive steamside scale build-up.

• Tube surface smooth or with “alligator hide” appearance (see Figure 33-3).

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for steamtouched tubing (Table 31-1) to pick a more likely candidate.

33-24

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by fireside corrosion.

SH/RH Fireside Corrosion/Coal-Fired Units

➠ These precursors can signal the potential for SH/RH tube failures by a fireside corrosion mechanism. If one or more has occurred, go to Action 3 which reviews root causes and outlines the steps to confirm the influence of each.

Action 2: Determine (confirm) that the mechanism is fireside corrosion. A SH/RH tube failure has occurred which the BTF team has tentatively identified as being fireside corrosion damage (Action 1a). Action 2 should clearly identify fireside corrosion as the primary mechanism or point to another cause. The primary identifier will be the presence of low melting point components in the region adjacent to the tube wall. Sample removal, metallographic analysis and deposits analysis will be used to make this determination. A primary objective is to make sure that the failure that has been experienced is not primarily long-term overheating/creep (Chapter 32) and Table 33-1 for distinctions.

➠ Characterize the extent of damage. Is there significant wall thinning across a number of tubes on the fireside?

➠ Plot wall thinning against measurements of steamside oxide (Figure 33-5). Is the ratio of wall thinning loss to steamside oxide buildup greater than three?

➠ Analyze deposits. Does metallographic and melting point analysis of deposits detect the presence of low melting point constituents such as alkali iron trisulfates, indicative of the most common mode of fireside corrosion?

➠ Evaluate microstructure. If tubing is stainless steel, is there evidence of carburization and discrete iron sulfides in the grain boundaries of affected locations (a sulfidation mechanism)?

If damage is localized, it may be flyash erosion damage (Chapter 14, Volume 2); however, continue through flow chart, particularly deposit analysis.

Mechanism more likely to be longterm overheating/creep (Chapter 32); review that mechanism, review analysis of deposits to ensure that there is no evidence of low melting point ash components which would be indicative of fireside corrosion.

Mechanism is probably not fireside corrosion. Review other fireside failure mechanisms, particularly long-term overheating/creep (Chapter 32).

Mechanism is probably not fireside corrosion. Review other fireside failure mechanisms, particularly flyash erosion (Chapter 14, Volume 2).

Probable mechanism is fireside corrosion.

➠ Go to Action 3: Root Cause Determination

References to other sources of information:

• Main text (this chapter) reviews the mechanism and the distinctive nature of external tube deposits formed.

• A summary of steps in a typical metallurgical examination can be found in Chapter 6, Volume 1.

Volume 3: Steam-Touched Tubes

33-25

Action 3: Determine root cause of the fireside corrosion A BTF failure has occurred and the mechanism has been confirmed as fireside corrosion (Action 2) or a precursor to fireside corrosion has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes of fireside corrosion, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.1 Potential actions for all root causes of fireside corrosion

➠ (a). NDE measures (typically UT) to identify wall thinning. ➠ (b). Ash and deposit analysis to identify presence of low melting point constituents, particularly alkali iron trisulfates.

3.2 Influence of overheating of tubes.

➠ (c). NDE of steamside oxide thicknesses. ➠ (d). Selective tube sampling and metallurgical analysis to confirm steamside oxide and wall thickness. ➠ (e). Monitoring of thermocouples installed across the SH/RH outlet legs in vestibule to identify hottest platens across the boiler.

3.2.1 Poor initial design - choice of material

➠ (f). Items (c) and (e) above.

3.2.2 Poor initial design - extra gas-touched length.

➠ (g). Evaluate temperatures across the element (via thermocouple or steamside oxide measurements) to determine if sections particularly near material changes are running too hot. See discussion of gas-touched length in Chapter 32 and sample plot in Figure 32-12.

3.2.3 Internal oxide growth which occurs during operation.

➠ (h). Items (c) and (d) above.

3.2.4 High temperature laning.

➠ (i). Monitor temperatures as in (e) above. ➠ (j). Laning can be identified with cold air velocity technique. See Chapter 14, Volume 2 on flyash erosion for a discussion of the technique.

3.2.5 Tube misalignment (out of bank)

➠ (k). Visual examination.

3.2.6 Operational problems when coal type is changed 3.2.7 Rapid startups causing reheater to reach temperature before full steam flow

33-26

SH/RH Fireside Corrosion/Coal-Fired Units

➠ (l). Check startup probe and that initial gas is limited to 1000°F (538°C) prior to RH flow.

Action 3: Determine root cause of the fireside corrosion (continued) Major Root Cause Influences

➠ Actions to Confirm

3.3 Root causes related to fuel factors 3.3.1 Change to fuel with unusually corrosive ash, particularly those with high S, Na, K, or Cl

➠ (m). Evaluate coal composition using corrosivity index. ➠ (n). Analysis for low melting point of ash components using probes. ➠ (o). Analysis of metallurgical cross sections, particularly for Cl, S, C, Na, and K. ➠ (p). Install continuous readout corrosion sensors if unit switches coal or uses spot market coal.

3.4 Root causes related to incomplete or delayed combustion.

➠ (q). Monitor for levels of CO and O2. ➠ (r). Check for unburnt startup oil.

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Evaluation will be based on detecting wall thinning. Wastage rates in excess of 25 nm/hr (~ 9 mils/yr) are of concern.

➠ Identify all locations to be examined. Refer to Section 1.2 of main text and Figure 33-6 for typical locations. Damage may be widespread and missed locations are sites for future failures.

➠ Perform UT survey to measure extent of damage via wall thinning and steamside oxide thickness. A review of UT methods is provided in Chapter 9, Volume 1.

➠ Perform tube sampling to measure wall thinning and steamside oxide buildup to confirm results of NDE inspection and to determine the degree of damage.

➠ Use results interactively with Action 3.

➠ Go to Action 5: Implement Repairs, Immediate Solutions and Actions. Begin remaining life assessment.

Volume 3: Steam-Touched Tubes

33-27

Action 5: Implement repairs, immediate solutions and actions The most important actions for the BTF team are to (i) initiate a remaining life assessment based on the wastage rate and extent of steamside oxide as derived from the NDE survey, (ii) choose a repair strategy based on remaining life assessment, and (iii) coordinate the long-term strategy from options outlined in Figure 33-12.

➠ Gather sufficient information so that a remaining life assessment of affected tubes can be initiated. See Chapter 8, Volume 1 for summary of the oxide methodology.

➠ Implement repairs or replacement of affected tubes as identified from the NDE Survey (Action 4). ➠See Chapter 11, Volume 1 for summary of applicable tube repair techniques. ➠Develop a plan to replace affected tubing including an economic assessment of the future possible failure rate and the resulting optimal extent of new tubing. ➠Ensure that the full extent of damage is removed as indicated by wall loss, or the presence of "alligator hide". Failure to do so will result in immediate repeat failures.

➠ Chemically clean, as needed to remove excessive steamside deposits.

➠ Adjust fireside conditions, as needed. References to other sources of detailed information: • Main text (this chapter) and Table 33-2 provide additional detail on repairs, immediate solutions and actions and the corresponding root causes. • Guidance on chemical cleaning can be found in Chapter 4, Volume 1. • Guidance on fireside testing can be found in reference 18.

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will include the analysis of remaining life and an economic evaluation to ensure that the optimal strategy has been chosen from those shown on Figure 33-12 and described in the main text.

33-28

Major Root Cause Influences

➠ Long-Term Actions

Potential actions for all root causes of fireside corrosion.

➠ Perform remaining life assessment. ➠ Set up long-term monitoring and re-evaluation program. ➠ Evaluate full range of available options using roadmap in Figure 33-12.

Influence of overheating of tubes.

➠ As above.

Poor initial design - choice of material.

➠ As above, emphasis will be on identifying locations where material upgrading is required. ➠ May involve redesign of circuit to extend the use of the higher grade material.

Poor initial design - extra gas-touched length.

➠ As above, emphasis will be on identifying locations where material upgrading is required. ➠ May involve redesign of circuit to extend the use of the higher grade material.

SH/RH Fireside Corrosion/Coal-Fired Units

Action 6: Implement long-term actions to prevent repeat failures (continued) Major Root Cause Influences

➠ Long-Term Actions

Internal oxide growth which occurs during operation.

➠ As in primary list above, also see potential actions for the long-term overheating/creep of tubes (Chapter 32).

High temperature laning.

➠ Controlled with flow distribution screens; in practice is difficult to overcome because of high temperatures in SH/RH. ➠ Review primary list of alternatives in Figure 33-12 for options.

Tube misalignment (out of bank).

➠ Perform remaining life assessment ➠ Set up long-term monitoring and re-evaluation program.

Operational problems when coal type is changed.

➠ Perform remaining life assessment. See discussion of methods in Chapter 8, Volume 1. ➠ Set up long-term monitoring and re-evaluation program. ➠ Evaluate full range of available options using roadmap in Figure 33-12.

Rapid startups causing reheater to reach temperature before full steam flow.

➠ As above.

Root causes related to fuel factors.

➠ As above.

Change to fuel with unusually corrosive ash, particularly those with high S, Na, K, or Cl.

➠ As above, plus ➠ Develop a fireside testing program using guidance provided in fireside testing guidelines18. ➠ Investigate coal changes with Coal Quality Impact Model (CQIM)24-26 or equivalent, including economics evaluation.

Root causes related to incomplete or delayed combustion.

➠ Perform remaining life assessment. See discussion of methods in Chapter 8, Volume 1. ➠ Evaluate full range of available options using roadmap in Figure 33-12. ➠ Develop a fireside testing program using guidance provided in fireside testing guidelines18.

Volume 3: Steam-Touched Tubes

33-29

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components implied by the presence of fireside corrosion in the SH/RH tubes, or by its precursors.

33-30

Aspect of SH/RH Fireside Corrosion

Alert for Other Cycle Components

➠ Actions Indicated

Corrosive coal.

• Potential for waterwall fireside corrosion • Potential for back-end corrosion

➠ Develop a fireside testing program using guidance such as provided in the fireside testing guidelines.18 ➠ Investigate coal changes with Coal Quality Impact Model (CQIM)24-26 or equivalent, including economics evaluation. ➠ Mitigate negative aspects of coal composition if possible by fuel switch, blending or washing.

Tube overheating because of excessive steamside oxide.

• Potential for additional tube failures by longterm overheating/creep (Chapter 32). • Exfoliation of scale with subsequent carryover into turbine could lead to solid particle erosion. • Exfoliation could lead to tube blockage and additional SH/RH failures by short-term overheating/creep (Chapter 36).

➠ Chemical clean unit if necessary. See guidance in Chapter 4, Volume 1.

Poor combustion conditions.

• Low unit efficiency • Poor mill performance

➠ Combustion adjustments to improve unit efficiency. See guidance in reference 18. ➠ Correct mill performance.

Total redesign of the superheater or reheater.

• Would change absorption patterns through the SH/RH sections and may increase temperatures in other sections.

➠ Check temperatures in the redesigned and other areas.

SH/RH Fireside Corrosion/Coal-Fired Units

Chapter 34 • Volume 3

Temperature (°C) 900 PO2 (atm)

851

1.0 0.2 800

Liquid 720

700 670 645 605

602

600

500 V2O5

10

525

20

30

40 50 MOL. % Na2O

60

N3 V

562

N2 V

NV6

NV

575

70

80

SH/RH Fireside Corrosion/Oil-Fired Units Introduction Fireside corrosion in oil-fired units (also termed “oil ash corrosion” or “liquid phase corrosion”) is generally confined to the higher temperature sections of the superheater (SH) and reheater (RH) where metal temperatures can exceed 600°C (~ 1100°F). The melting points of ash components, responsible for the corrosion process are not low enough to cause problems in lower temperature sections such as waterwalls.

Damage produced by fireside corrosion is often confused with that from long-term overheating/creep (see Chapter 32). As several of the key features of the two are similar, some care is needed to correctly identify the primary mechanism, and thus take the proper corrective action. Two other Chapters specifically address fireside corrosion in coalfired units: that in waterwalls (Chapter 18, Volume 2) and in SH/RH tubes (Chapter 33).

Volume 3: Steam-Touched Tubes

34-1

1. Features of Failure and Typical Locations Superheater/Reheater Fireside Corrosion (Oil-Fired Units): Identification Keys 1. Macroscopically, fireside corrosion in oil-fired units will generally be uniformly located on the side of the tube experiencing the highest temperatures and highest heat flux. In general it will have an undulating surface or the appearance of unevenness. 2. Distinctive fireside deposits containing low melting ash components are key in the identification of fireside corrosion. There is generally a loose, blue/black outer deposit and strongly adherent, black oxide adjacent to the tube. The latter will look molten and shiny. For high vanadium residual oils, typical deposit compounds are vanadium sodium complexes such as sodium vanadyl vanadate (Na2O • V2O4 • 5V2O5). 3. In the case of low vanadium residuals, the deposits that form are essentially sulfatic (sodium sulfates, Na2SO4 and derivatives) with small amounts of NV6 (Na2O • 6V2O5), sodium vanadates, of V/Na ratio lower than 6, and nickel vanadates.

1.1 Features of failure Macroscopically, tube wastage will often be evident and manifested as undulations and unevenness of the tube surface. A typical cross-section showing this wastage and the presence of significant deposits can be seen in Figure 34-1. Tube thinning will be most evident around the edges of the deposit. Greatest wall loss will generally be seen in tubes that have been operated at the highest temperatures over a period of time. The measurement of steamside oxide thickness (by ultrasonic or laboratory measurement) will usually confirm this observation. The ratio of maximum wall loss to oxide scale thickness at a location will give an indication of the degree to which fireside corrosion is a problem relative to longterm overheating. If the ratio is greater than five then fireside corrosion (or erosion) is the more active mechanism. On austenitic steels, the magnetite and spinel oxides become nonadherent when they reach a thickness of about 0.2 mm (0.008 in.) and tend to spall off along with any deposits that have formed when the boiler cools. Bare metal can result and lead to rapid corrosion rates. Any oxide scale that is thicker than about 1 mm (0.04 in.) should be considered as evidence of a fast corrosion rate.1 There may also be localized, overlapping pits 0.5 - 2 mm (~ 0.02 - 0.08 in.) in diameter. Unburnt oil on the upstream surfaces of tubes, such as caused by excess oil during startup, is a source of carbon contamination and subsequent carburization of the coated surface. This is particularly critical for austenitic tube materials which have a higher solubility for carbon than ferritic materials.1

34-2

SH/RH Fireside Corrosion/Oil-Fired Units

The tube will usually be coated with a multilayered fireside scale and ash deposits typically consisting of two layers. Deposits adjacent to the metal oxide are the ones that directly influence corrosion of the substrate. A black, glossy inner layer, is typical and may be superficially similar to that which occurs as a result of long-term overheating. If this layer is glassy or shows signs of having been molten against the tube metal, then a very fast corrosion rate has probably occurred. The absence of a layer of protective oxide adjacent to the tube surface is indicative of the fastest corrosion rates. A hard, brittle, and porous outer layer may have alternating dark/black/blue and light bands. Fireside corrosion will be primarily distinguished by the presence of low melting point ash compounds in these deposit layers. The fireside scale and ash should be examined metallographically and chemically for the presence of low melting point constituents. Note that oil ash deposits tend to be soluble in water and are more easily removed than coal ash deposits. For high vanadium residuals these constituents will be vanadium oxides and/or vanadium sodium complexes such as those listed in Table 34-1 and the eutectics shown on the V2O5- Na2O phase diagram in Figure 34-2. For low vanadium residuals, these compounds will be essentially sulfatic (sodium sulfates Na2SO4 and derivatives) with small amounts of NV6 and nickel vanadates. An example of the distinctive layering and the presence of an alkali vanadyl vanadate in the inner layer as detected by energy dispersive x-ray is shown in Figure 34-3.

Figure 34-1. General appearance of a 9% Cr final superheater tube containing fireside corrosion deposits (top of the figure) in an oil-fired boiler after 50,000 hours service. The bottom of the figure shows a section of tubing having been acid cleaned to remove the deposits and the ring section taken through the cleaned section shows the general appearance of tube wastage on the outside surface. Source: J. Hickey, Irish Electricity Supply Board

Table 34-1 Melting Point of Some Commonly Observed Slag Deposits Compound

Melting Point, °C

V2O5

670

V2O3

1,970

V2O4

1,970

Na2O • V2O5

605

2Na2O • V2O5

650

3Na2O • V2O5

850

Na2O • V2O4 • 5V2O5

625

Na2O • V2O4 • 11V2O5

575

Na2SO4

884

Na3Fe(SO4)2

624

Temperature (°C) 900 PO2 (atm)

851

1.0 0.2 800

Liquid 720

700 670

Note: See also the V2O5-Na2O phase diagram (Figure 34-2) which shows eutectics down to 525°C (977°F).

645 605

602

600

500 V2O5

10

525

20

30

40 50 MOL. % Na2O

60

N 3V

562

N 2V

NV6

NV

575

70

80

Figure 34-2. Equilibrium phase diagram for the V2O5-Na2O system showing PO2 dependence. (N) represents Na2O and (V) is V2O5. Source: J.R. Wilson2

Volume 3: Steam-Touched Tubes

34-3

d)

a)

b)

c)

Figure 34-3. (a) Shows a secondary electron image of an oil ash scale and deposit on 21/4 Cr - 1 Mo steel reheater tube that exhibited oil ash corrosion. (b, c, and d) Show energy dispersive X-ray spectra from three layers in the ash. The spectra indicate that the inner layer (b) may be an alkali vanadyl vanadate. Absence of significant iron in the inner layer suggests that the vanadium compound is either directly attacking the tube material or is fluxing away the indigenous iron oxide as rapidly as it forms. The middle layer (c) is probably iron oxide which has precipitated at the external surface of the molten layer. The outer layer (d) is comprised of ash deposit constituents.

34-4

SH/RH Fireside Corrosion/Oil-Fired Units

Removal of the fireside scale and ash deposit will reveal an undulating surface and distinctive longitudinal grooving and pitting appearance (“alligator hide”) as shown in Figures 33-3 and 33-4, Chapter 33. Note that the appearance of alligator hide is indicative but not definitive as it can also occur if the metal has been thermally shocked such as by sootblower condensate.3 Final failure will typically be as a longitudinal crack associated with the alligator hide. If failure has been long-term the fracture surface will appear brittle, rapid corrosion rates will result in fracture surfaces that are ductile. The key differences between fireside corrosion and long-term overheating/creep are highlighted in Table 342. The most distinguishing characteristic will be the presence of low melting ash compounds in the deposits formed as a result of fireside corrosion, and the ratio of wall thickness loss to oxide scale thickness.

1.2 Locations of failure Typical regions of the boiler subject to tube failures by fireside corrosion are shown in Figure 34-4. These locations are basically the same as those where long-term overheating/ creep and fireside corrosion in coalfired units can occur. However, in oilfired units, fireside corrosion tends to occur in higher temperature components, particularly those tubes with metal temperatures in excess of about 600°C (~ 1110°F). As a result,

failures may be more prevalent in outlet sections and can occur in either austenitic materials or in ferritic materials exposed to the higher temperatures. Specific locations at the highest risk therefore include: • Leading sides of all tubes in pendant platens, especially hottest (leading) tubes, and steam outlet tubes • Tubes out of alignment that act as leading tubes • In the outlet (final) sections towards the header because these are at the highest temperatures. • Just prior to a change of material, e.g., in T22 just prior to the austenitic material, as the lower Cr content material may be operating above its design point. • Wrapper tubes • Tubes that surround a radiant cavity (i.e,, they may pick up more heat)

Figure 34-4. Typical boiler locations were oil ash fireside corrosion can occur.

• At bottom bends of platens especially those facing the fireball. • Tubes with a longer gas touched length (GTL). GTL is the distance measured along the tube circuit from the inlet header to the point of corrosion.

• Spacers and uncooled hangers, and the fins and studs on tubes.

Volume 3: Steam-Touched Tubes

34-5

Table 34-2 Comparison of Characteristics of Long-Term Overheating/Creep, Short-Term Overheating, and Fireside Corrosion (Oil-Fired Units) In Superheater/Reheater Tubing Characteristic

Long-Term Overheating

Short-Term Overheating

Fireside Corrosion

Fracture Surface and Appearance of Failure

• Generally thick-edged, brittle final failure. • Generally accompanied by external tube wastage at the 10 o’clock and 2 o’clock positions.

• Usually thin-edged, ductile final failures. • Swelling of tubes without ovalization. • “Fish-mouth” appearance of tube rupture.

• Tube wastage in high temperature/heat flux portions of the tube. • Longitudinal cracking, final failure can be (but not necessarily) by overheating.

Internal Scale?

Yes, generally extensive, multilaminated and exfoliating.

Not necessarily thick. Depends on age of tube at failure.

Yes, particularly if tube metal overheating was a root cause.

External Scaling?

• Yes, thick, laminated and often longitudinally cracked. • Usually two layers - (i) a hard, porous outer layer with composition typically that of flyash, and (ii) a black glossy inner layer, mostly oxide but may contain some sulfates and sulfides of iron.

Not necessarily thick.

Yes, with multi-layers: (i) a hard, brittle and porous outer layers, which may have alternating dark blue/black bands and (ii) a black, glossy inner layer strongly bonded to tube; has an appearance of molten deposits and is shiny.

Outside surface appearance after removal of scale/deposits

Characteristic longitudinal grooving and pitting (“alligator hide”).

Swelling, stretch marks on tube metal.

Characteristic undulations or unevenness of surface. In worst areas there might be some “alligator hide” and longitudinal cracking.

Composition of External Scales/Deposits

Does not contain low melting point ash compounds such as alkali iron sulfates.

Not relevant.

Does contain low melting point compounds. In high vanadium oil will be vanadium sodium complexes. In low vanadium oil will be sulfatic (sodium sulfates and derivatives).

Wall Thinning?

Typically wastage flats at 10 o’clock and 2 o’clock positions caused by accelerated oxidation. There is always a layer of oxide adjacent to the tube.

Only because of bulging of tube material.

Primary feature of failure, worse on portion of tube subject to high temperatures and/or high heat flux.

Ratio of wall loss to steamside oxide thickness?

Typically less than 3:1.

Not relevant.

Typically greater than 3:1; for ratios greater than 5:1 fireside corrosion or erosion is the dominant mechanism.

Tube Material Degradation

Yes, generally extensive signs of overheating and/or of creep damage, particularly near to the crack tip. Creep voids will not be found removed from crack tip.

Depends on the material and the maximum temperature reached. For example, for the most rapid overheating failures, there will be relatively little microstructural change.

If overheating has been a problem, yes; otherwise, no. Fireside corrosion can occur in a tube at design temperatures. Can be a carburized band adjacent to the fireside deposits.15

Change in material hardness

Localized softening near the rupture is typical.

Localized hardening near the rupture is likely.

Hardening is not necessary; if there has been no overheating, there will be no change in hardness. If carburization is associated with corrosion, then an increase in hardness may be observed.15

34-6

SH/RH Fireside Corrosion/Oil-Fired Units

2. Mechanism of Failure Superheater/Reheater Fireside Corrosion (Oil-Fired Units): Mechanism 1. In residual oils containing a high level of vanadium, molten vanadates increase the corrosion of boiler tubes. The mechanism is thought to be a fluxing of the normally protective oxide and/or an increase of the diffusion of oxidant to the protective layer. Greater wastage may occur if the tube becomes carburized. 2. In the case of low vanadium residuals, the deposits that form are essentially sulfatic (sodium sulfates Na2SO4 and derivatives) with small amounts of NV6 and nickel vanadates, which also flux the protective oxide. 3. Because of the low melting points of ash constituents, tube metal temperatures around 600°C (~ 1110°F) or lower are in the critical regime to experience corrosion by this mechanism.

2.1 Introduction Low melting point compounds produced by the firing of residual oils are aggressively corrosive to the protective oxides which form on SH/RH tubes. As liquid phases begin to deposit and solidify on tube surfaces, the temperature of the outside surface increases, leading to additional deposition and melting of the deposits. The primary compositional determinants from the fuel are the levels of vanadium and sodium in the residual oil. The effects of two general classifications of residual oils are described here: high vanadium, low sodium fuel typical of Venezuelan sources, and low vanadium, low sodium fuels typical of some of those from the Persian Gulf. Typical compositions of each are shown in Table 34-3 along with a typical Mexican fuel oil.15 For either fuel, corrosive slag occurs when two preconditions have been met: (i) operation occurs with a residual oil that forms low melting point compounds, particularly those containing vanadium and sodium, as well as sulfur, and (ii) tube metal temperatures that are above 600°C (~ 1110°F). Final failure occurs as a stress rupture when thinned tube walls are no longer able to contain the requisite pressure. One of the primary effects of oil-firing is that there is relatively little ash produced (typically around 0.2%) in comparison with coal-firing (around 20%) and therefore the beneficial effects of ash components are not present. It has been reported that when both oil and coal are fired, that the coal ash can dilute the vanadium content of deposits produced.1 Greater tube wastage has been observed if the tubes become carburized during service.16

2.2 Effect of fuel composition on the rate of corrosion 2.2.1 High vanadium residual oils. In residual oils containing a high level of vanadium, it is generally agreed that the molten vanadates increase the corrosion of boiler tubes by fluxing the normally protective metal oxide layers, or by increasing the diffusion of oxidant to the protective layer.4 Although the exact mechanism of attack is still being refined, Figure 34-5 shows a ten-step process of oxygen transport that illustrates the relevant factors. Table 34-1 summarizes the most commonly observed tube deposits that result from high vanadium oils. Many have melting points that are below the expected temperatures in the hotter sections of the SH/RH. Vanadium pentoxide (V2O5) is the highest oxidation product of vanadium and is listed for completeness although it is rarely observed in operating boilers. Sodium vanadyl vanadate (Na2O • V2O4 • 5V2O5) is one of the many complex compounds usually present in the ash and is the most commonly reported because there is X-ray data for it. In fact, it is not present in the corroding deposit because it absorbs oxygen to become Na2O • 6V2O5 (referred to as NV6) depending on the Na:V ratio. Figure 34-2 is the equilibrium diagram for the Na2O - V2O5 system showing how the melting points vary for these families of compounds. The dotted line on the figure illustrates the liquidus for 0.2 atmosphere oxygen partial pressure; the solid line for 1.0 atmosphere. The substantial differences lead to a corresponding significant difference in corrosion rates in pressurized fired equipment.2 It also indicates how operation at low levels of excess oxygen can help mitigate the effects of fireside corrosion by raising the melting point of the vanadates.

Volume 3: Steam-Touched Tubes

34-7

The greatest corrosion is not related to the absolute levels of V and Na but the ratio of their oxides and temperature.14 The fastest corrosion rate (for temperatures up to about 900°C (~ 1650°F)) is at a ratio of V2O5: Na2O of 6:1 (NV6), which corresponds to a Na:V ratio of about 0.1:1. Thus for a typical Venezuelan residual containing 300 ppm vanadium, only about 30 ppm sodium is required to form a very corrosive medium. Other oxides at low concentrations exert little influence on corrosion rates. The corrosivity can also depend on the sulfur content of the oil-ash deposit in cases where the sulfur content in the oil is greater than 3%.15 The presence of refractory oxides (MgO, CaO, Al2O3, and SiO2) inhibit corrosion when present at high concentrations in the ash, e.g., a ratio of 2:1, additive:ash by weight.2 2.2.2 Low vanadium, low sodium residuals. With a typical Persian gulf residual oil, for example, containing 50 ppm V, 70 ppm Na and 2.5 weight percent S, the deposits or slags are essentially sulfatic (sodium sulfates, Na2SO4 and derivatives) with small amounts of NV6 and nickel vanadates. The fireside corrosion rate is controlled by the dissolution of otherwise protective scales in the liquid alkali sulfates. Laboratory and corrosion probe trials have shown that the deposition rate of sodium is proportional to its concentration, independent of the presence of vanadium. The deposition of vanadium increases with vanadium content in the flue gas, but large amounts of sodium can depress the rate. No realistic reductions in the fuel oil impurities will markedly improve the rate of corrosion which is of the same order as those for vanadate corrosion. The same sharp increase in rate for austenitics above 600°C (~ 1110°F) is evident.

34-8

Table 34-3 Impurity Levels of Residual Fuel Oils Fuel Source

Vanadium, ppm

Sodium, ppm

Sulfur, Wt %

Venezuelan

280-300

10-20

2.5

Persian Gulf

45-50

35-70

2.5

150-340

1-100

3.2-4.5

Mexican15

Atmosphere 1. Diffusion of O 2 through porous oxide

Porous Oxide

10. Formation of porous oxide

2. Adsorption of O 2 by liquid vanadate

9. Transport of metal ions through melt

3. Transport of oxygen species through melt

Liquid Vanadate 4. Adsorption of oxygen species by metal oxide

8. Dissolution of metal oxide in liquid vanadate

Metal Oxide 5. Transport of oxygen ion 7. Transport of metal ion through metal oxide(s) through metal oxide 6. Reaction of oxygen ion with metal Metal

Figure 34-5. Possible reactions that take place in liquid vanadate (oil-fired) corrosion. Source: J.R. Wilson2

SH/RH Fireside Corrosion/Oil-Fired Units

2.3 Effect of tube temperatures on the rate of corrosion Because of the much lower operating temperatures, fireside corrosion in the waterwalls of oil-fired units is unlikely, however, superheater/ reheater temperatures are in exactly the right range during operation to experience a problem. The higher the temperature, the faster the corrosion rate in these temperature ranges. As a result, if steamside oxide buildup leads to higher tube temperatures, or if unit firing rates are increased to compensate for insulated tubes, the resultant increase in tube temperature can lead to an outbreak of tube failures by this mechanism.

2.4 Effect of tube material composition The corrosion rate increases sharply for austenitic materials (less for ferritic) between 600 and 650°C (~ 1110 to 1200°F). With a Venezuelan residual, for temperatures less than 600°C (~1110°F), the rate is less than 50 nm/hr (~ 18 mils/yr), whereas at 650°C (~ 1200°F), it can be up to 200 nm/hr (~ 70 mils/yr). Clearly, these higher rates can cause a considerable problem if left unchecked. High chromium materials generally have better corrosion resistance than ferritic materials. Type 310 has about a factor of three better resistance than ferritic steels; an alloy of 50Cr 50 Ni, which has seen significant development for applications where corrosion has proven to be particularly troublesome, has

extremely low corrosion rates particularly below 700°C (~ 1300°F). Its use is generally in duplex or coextruded tubing so that the superior corrosion resistance is combined with a creep resistant core.

2.5 Effect of unit operation on corrosion rates The amount of excess air is a contributing factor to the rate of corrosion. In high vanadium residual oil, a level of excess air that is just above the minimum stoichiometric requirement will reduce the corrosion potential by preventing the oxidation of vanadium to its highest oxidation state (V2O5). The higher melting point oxides V2O3 and V2O4 will be more likely to form, and will not deposit as corrosive compounds. However, the addition of Mg additives will probably still be required to reach acceptable corrosion rates. The effect of such additives is described in the next section. Note that ashes containing vanadium complexes are toxic and are slightly water soluble which may lead to groundwater contamination if used for land fill. In the case of low vanadium residuals, reasonable success has been obtained by the correct control of oxygen to just above stoichiometric combustion requirements. Neither further excess oxygen reductions, nor the use of Mg additives has had significant additional benefit. These effects tend to confirm that the active process is corrosion due to sulfatic compounds; thus control of SO3 production is an important criterion.

2.6 Effect of additives on propensity for corrosion A variety of additives, most notably magnesium, but also including manganese, aluminum, silicon, and calcium, alone or in combination with each other, have been used to control corrosion. It is important to understand how they affect the mechanism so that proper additive choice can be made. The three objectives of an additive are (i) to raise the fusion point of the ash, (ii) to combine chemically with all the corrosive agents present, and (iii) to not cause plugging or wear in fuel systems. The use of magnesium, as MgO, is based on increasing the melting point of the vanadium complexes and to eliminate the formation of low melting point eutectics. Table 34-4 shows the success of that strategy as the melting point of typical complexes plus MgO increases. Figure 34-6 shows the phase diagram for the addition of MgO. The aim is to form 3MgO•V2O5 which has a melting point of 1071°C (1960°F). The situation is complicated by the presence of Na2O which reduces the melting point, or requires more MgO to control. Table 34-4 Melting Points of MgO V2O5 Compounds Compound

Melting Point, °C

2MgO • 3V2O5

640

MgO • V2O5

742

2MgO • V2O5

980

3MgO • V2O5

1071

Volume 3: Steam-Touched Tubes

34-9

To prevent corrosion, it is necessary that the deposit on the tubes has a melting point above the metal temperature. It is difficult to control vanadate corrosion by the use of additives without accumulating large volumes of modified fly ash; thus, it is important to add only enough MgO to counteract the corrosive vanadate. This entails fuel and slag analysis, reference to the phase diagrams for the expected behavior of the proposed combinations, and analysis of oil and additive flow rates. Excessive magnesium slag buildup will cause sintering and fusing of the outer layers, which may be more difficult to remove by sootblowing or water washing.

Temperature (°C) 1100

1071

1000

980

Liquid 900

800 742

700

670

10

20

30

40 50 MOL. % MgO

60

70

Mg3(VO4)2

Mg2V2O7

500 V2O5

604

Mg V2O6

600

Mg2V6O17

640

As noted above, the benefits of Mg additives for use with low vanadium residuals has not been as marked as that for the high vanadium oils. 80

Figure 34-6. Equilibrium phase diagram for the V2O5-MgO system showing that there is no PO2 dependence. Source: J.R. Wilson2

34-10

SH/RH Fireside Corrosion/Oil-Fired Units

3. Possible Root Causes and Actions to Confirm 3.1 Introduction Table 34-5 lists the primary root cause influences. In general there are three groups of root cause influences: (i) fuel factors, (ii) excessive tube metal temperatures, and (iii) the effects of operating conditions. Independent of root cause, the two most important actions are: (a). Analyze the composition and melting point of deposits. A key confirmation that fireside corrosion is the active mechanism is the presence of low-melting point ash constituents. Metallographic sampling and analysis will be key identification tools. Chemistry and constituents of the deposits can be determined using EDX mapping, spectrochemical analysis, x-ray diffraction or atomic absorption spectroscopy. Melting temperatures of ash are measured by the use of thermogravimetric analysis and/or differential thermal analysis. More information about such analyses is provided in Chapter 6, Volume 1. (b). Monitor progression of wall thinning on a periodic basis. When the ratio of wall loss to the buildup of steamside oxide scale is large (typically greater than a factor of three and certainly at a ratio greater than five), then fireside corrosion rather than long-term overheating/creep is dominant. Additional detail about inspection and sampling methods is provided in Chapter 9, Volume 1.

3.2 Influence of oil composition There is a significant effect of oil composition on corrosion rate as discussed at length in the mechanism section above. Changing to a more aggressive fuel can significantly increase corrosion damage. Actions to confirm that this is a primary factor in the appearance of fireside corrosion are:

(c). Monitor oil corrosiveness by placing controlled-temperature corrosion probes or deposition probes into susceptible locations in the SH/RH. (d). Analyze ash deposits using methods outlined in (a) above to periodically determine whether significant deposition of low-melting point constituents has begun or has accelerated.

3.3 Influence of tube metal temperatures Unfortunately, the higher temperature sections of the SH/RH are well within the temperature regime of significant corrosion susceptibility. Furthermore, increases in tube metal temperatures, such as caused by the buildup of steamside scale, can increase the rate at which corrosion occurs and increase the number of tubes at risk. As deposits form on the external tube surface, they will insulate the tube metal but result in higher ash temperatures. Combustion conditions that lead to higher flue gas temperatures can cause localized tube overheating. The influences are reviewed below. 3.3.1 Excessive temperatures caused by steamside oxide buildup Actions to confirm that steamside oxide buildup and the resultant overheating is a contributing factor to the appearance of fireside corrosion include: (e). Oxide thickness measurements (ultrasonically) at typical locations provides an indication of tube temperatures within the element. See Chapter 8, Volume 1 for details of the interpretation of oxide scale thickness and its relation to tube metal temperatures. Chapter 9, Volume 1 reviews the use of UT for oxide thickness measurements.

Volume 3: Steam-Touched Tubes

34-11

Table 34-5 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

3.1 Potential actions for all root causes of fireside corrosion (a). Ash and deposit analysis to identify • Choose repair strategy presence of low melting point conbased on severity of stituents, particularly vanadium/ corrosion rate. vanadium-sodium and sodium sulfate • Implement long-term complexes. strategy from choices in (b). NDE measures (typically UT) to identify Figure 34-7 in conjuncwall thinning and steamside oxide scale tion with on-going buildup. program of remaining life assessment and monitoring.

• Perform remaining life assessment. • Continual check on the use of Mgbased additives. • Optimization of excess oxygen levels. • Set up long-term monitoring and reevaluation program. • Evaluate full range of available options using roadmap in Figure 34-7.

3.2 Influence of oil composition (c). Monitor oil corrosiveness using corrosion or deposition probes. (d). Analyze ash deposits as in item (a) above.

• As above, plus • On-going consideration of the use of Mg-based additives.

• As above with emphasis on additives.

3.3 Influence of overheating of tubes. 3.3.1 Excessive temperatures caused by steamside oxide buildup.

(e). NDE of steamside oxide thicknesses. (f). Selective tube sampling and metallurgical analysis to confirm steamside oxide and wall thickness. (g). Monitoring of thermocouples installed across the SH/RH outlet legs in vestibule to identify hottest platens across the boiler.

• Choose repair strategy based on severity of corrosion rate. • Implement long-term strategy from choices in Figure 34-7 in conjunction with on-going program of remaining life assessment and monitoring. • Institute periodic chemical cleaning. See additional detail in Chapter 4, Volume 1.

• Perform remaining life assessment. • Consider program of periodic chemical cleaning. See Chapter 4, Volume 1. See also options for long-term overheating of tubes (Chapter 32). • Set up long-term monitoring and re-evaluation program. • Evaluate full range of available options using Figure 34-7.

3.3.2 Excessive temperatures as caused by operating conditions. - high temperature laning of gases - changes in absorption patterns between furnace and convection sections, - RH overheating because of rapid startups - tube misalignments

(h). For high temperature laning: monitor • Modify operation to temperatures as in (g) above and concorrect the specific sider the use of the cold air velocity test problem. (CAVT). Details of the latter can be • Implement long-term found in Chapter 14, Volume 2 on strategy from choices in flyash erosion. Figure 34-7 in conjunc(i). For reheater overtemperature during start tion with on-going prosequences: check the startup probe and gram of remaining life limit temperatures to 538°C (1000° F) assessment and prior to RH flow. monitoring. (j). Visual inspection can be used to detect tube misalignments.

• Evaluate full range of available options using Figure 34-7.

34-12

SH/RH Fireside Corrosion/Oil-Fired Units

Table 34-5 Major Root Cause Influences, Confirmation and Corrective Actions (continued) Major Root Cause Influences

Actions to Confirm

Immediate Actions and Solutions

Long-Term Actions and Prevention of Repeat Failures

3.4 Influence of operating factors 3.4.1 Operation with high levels of excess oxygen and/or (k.) Check operating logs for typical excess oxygen levels. periods of very low excess oxygen.

3.4.2 Poor sootblowing operations

(l). Check sootblowing frequency, effectiveness, and superheat level of blowing medium.

• Modify operating procedures, if economically feasible to reduce levels of excess oxygen. • Implement long-term strategy from choices in Figure 34-7 in conjunction with on-going program of remaining life assessment and monitoring. • Evaluate whether operating procedures such as sootblowing can be economically changed to protect SH/RH tubes. • Implement long-term strategy from choices in Figure 34-7 in conjunction with on-going program of remaining life assessment and monitoring.

• Perform remaining life assessment. See discussion of methods in Chapter 8, Volume 1. • Set up long-term monitoring and re-evaluation program. • Evaluate full range of available options using Figure 34-7.

Volume 3: Steam-Touched Tubes

34-13

(f). Selective tube sampling and metallurgical analysis can be used to confirm the results of the NDE measures of steamside oxide and tube metal wall thickness. The same samples can be subjected to external deposit analyses as outlined in item (a) above. (g). Monitoring of thermocouples permanently installed across the SH/RH outlet legs in the vestibule can provide an indication of the hottest platens. This action can help prioritize where inspections and sampling should occur. 3.3.2 Excessive temperatures caused by operating conditions A variety of operating conditions might cause excessive tube temperatures: high temperature laning, changes in absorption patterns between the furnace and convective sections, rapid startups that cause the reheater to reach temperature before full flow starts, tube misalignments, etc. The actions to confirm these conditions are:

(h). For high temperature laning: monitor temperatures as in (g) above. Laning of gases can also be detected using a cold air velocity test (CAVT). The details of the test are described in Chapter 14, Volume 2 on flyash erosion. (i). For reheater overtemperature during start sequences: check the startup probe and limit initial gas temperature to 538°C (1000°F) prior to RH flow. (j). For tube misalignment (out of the bank): visual examination can be used to detect a problem.

3.4 Influence of operating factors 3.4.1 High levels of excess oxygen and/or periods of very low excess oxygen. The amount of excess air is a contributing factor to the rate of corrosion. In high vanadium residual oil, a level of excess air that is just above the minimum stoichiometric requirement (< 1% excess oxygen) will encourage the formation of

higher melting point constituents that will not deposit as corrosive compounds. There is relatively less effect of excess air levels when burning low vanadium residuals. Actions to confirm include: (k). Check operating logs for typical excess oxygen levels. 3.4.2 Poor sootblowing operations Typically, the first layer of deposit on clean tube surfaces is dry and loose because the tube metal temperature is low enough that the ash deposits are in solid form. Proper sootblowing can easily remove these early deposits. However, if sootblowing operations are executed improperly, deposits can remain on tubes, raising the outside deposit temperature, resulting in the deposition of liquidphase ash which is difficult to remove, and as described above, begins the rapid corrosion of tube surfaces. An action to confirm is: (l). Check sootblowing frequency, efficacy and superheat level of blowing medium.

4. Determining the Extent of Damage Ultrasonic testing (UT) can be used to measure both wall thinning and steamside oxide thickness. Locations should be chosen which are the most susceptible to fireside corrosion as described in Section 1.2 above. Locations on a particular tube should be investigated to find maximum wastage. Selective sampling is recommended to confirm the results of the NDE examinations and to evaluate internal scale and exter-

34-14

SH/RH Fireside Corrosion/Oil-Fired Units

nal deposits. Additional detail on UT methods can be found in Chapter 9, Volume 1. Methods of metallurgical evaluation are reviewed in Chapter 6, Volume 1. Monitoring thermocouples, either permanently placed or temporary installations, can be used to detect excessive metal temperatures that might foster future accelerated wastage or tube failures.

5. Background to Repairs, Immediate Solutions and Actions Superheater/Reheater Fireside Corrosion (Oil-Fired Units): Immediate Solutions and Actions 1. Immediate solutions should be chosen in conjunction with: (i) a knowledge of the severity of corrosion and (ii) an analysis of remaining life discussed under long-term actions below. 2. A principal means to overcome fireside corrosion in oilfired units is the use of Mgbased additives. 3. Repairs can be considered on two levels depending upon the severity of the problem. For the short term, or for mild corrosion, tubes can be replaced with same material, or a palliative coating may provide somewhat better corrosion resistance; for the long term or severe corrosion, replacement should be made with a more resistant material.

5.1 Need for remaining life assessment Any immediate solution or repair strategy should consider that a longer term remaining life analysis methodology should also be implemented. Such programs are discussed in detail in Chapter 8, Volume 1, and are summarized under the long-term actions below.

Note that considerable care must be taken in any repairs, inspections or movements within oil-fired units because of the extreme toxicity of vanadium dust. Breathing equipment or masks must be used for any repairs. Note that as with other vanadium complexes, ashes containing these oxides are toxic. They are also slightly water soluble which may lead to groundwater contamination if used for land fill.

5.2 Repairs If the corrosion rate has been relatively modest (< 25 nm/hr (~ 9 mils/yr)) and is likely to continue to be so, an acceptable strategy is to retube with the same alloy and monitor closely the wastage rate. Another approach, although not preferred, is the use of a palliative coating or tube shields An overview of options is provided under long-term actions. Pad welds should definitely not be used as a repair measure because of the uncertainty of the tube conditions such as the presence of creep cracks and their depth, and the conditions of the internal tube surface. Further discussion about weld repairs can be found in Chapter 11, Volume 1.

5.3 Use of Mg-based additives Utilities that are burning, or know that they will be burning, high vanadium residual oils will generally use an additive as the most cost-effective means to prevent fireside corrosion.

5.4 Other short-term options Several other short-term options that can be considered depending on the root cause might include (i) interim steps to limit tube metal temperatures such as by cleaning steamside oxide, (ii) aligning tubes, and (iii) improving sootblowing operation. A complete overview of all options is included in the next section.

For higher corrosion rates that are resulting in rapid wastage of the existing alloy, the replacement should be with a more resistant material.

Volume 3: Steam-Touched Tubes

34-15

6. Background to Long term Actions and Prevention of Repeat Failures Superheater/Reheater Fireside Corrosion (Oil-Fired Units): Long-Term Actions Corrective strategies will depend on the type of oil burned. For high vanadium residuals, consideration should be given to (i) the use of Mg-based additives, and (ii) operation with low excess air levels. For either high or low vanadium residuals, additional options generally fall into two major categories: (i) materials strategies that provide increased protection or replace the component, and (ii) design or operating strategies to try to control tube temperatures.

As noted above, it may not be possible to remove the root cause for many fireside corrosion problems. Knowing how to minimize the wastage rate and the application of a predictive remaining life assessment process including periodic inspection and monitoring, are the keys to economic handling of fireside corrosion problems. Figure 34-7 outlines most of the available corrective actions for superheater/reheater fireside corrosion. As shown in that figure, three primary, and not mutually exclusive, routes that can be followed are: (i) fuel options, primarily additives, (ii) materials options, and (iii) operating options. The circled numbers used in Figure 34-7 are used to identify options for the discussion that follows and no ranking is implied; however, boxes with bold outline indicate those options which have been the most successful.

Remaining life assessment (option 1, Figure 34-7) A remaining life assessment is required to relate the rate of corrosion wastage to the desired life or to determine the time available to implement the desired option. Therefore such an assessment should be undertaken in parallel with any of the other options. A systematic program will include: baseline measurement, monitoring rates of wastage, application and monitoring of control measures, and assessment of the effects on remaining life. Monitoring of the flue gas, metal and steam temperatures, combustion conditions, and fuel composition should also be considered as these can determine corrosion rates while the unit is still on-line.1 Particularly important are step changes in key parameters.

34-16

SH/RH Fireside Corrosion/Oil-Fired Units

A critical step in determining the remaining life of a tube and in choosing the optimal solution is knowing the rate at which damage is accumulating. The lifetime of superheater/reheater tubes can be calculated using the oxide thickness technique as discussed in detail in Chapter 8, Volume 1 which includes the damage accumulation by both fireside corrosion and long-term overheating. A key goal of the remaining life assessment will be to establish the acceptable rate of wastage, given the desired remaining life of the unit and an analysis of damage accumulation. As a rule of thumb, rates of wastage above 25 nm/hr (~ 9 mils/yr) indicate the need for some periodic activity such as annual inspections and for the explicit determination of re-evaluation periods. Rates above 50 nm/hr (18 mils/year) indicate a serious decrease in life of the tubing; an estimate of remaining life should be made and the appropriate actions taken that life is unacceptable.

Shielding (option 2) The most common temporary measure is the installation of shields to the leading edges of affected tubes. High chromium alloys such as Type 310 or 50Cr-50Ni are typically recommended. The shield is curved to fit the tube surface and tack-welded in place. As they are not cooled, the metal temperatures are above the range for extensive attack.

Coatings (option 3) A variety of processes have been suggested for applying metallized coatings to tubes in situ as a means of increasing corrosion resistance. The advantage of coatings is that very corrosion-resistant materials can be applied at specific susceptible sites, so it is not necessary to replace entire sections of tubing in order to overcome localized problems.

therefore be necessary.1 It should be noted that there may not be significant improvement in the resistance to fireside corrosion per se, but longer times to failure may result because the upgraded material has a lower creep rate and higher strength (so that a thinner failure thickness is achieved).1 As a general rule, the more resistant the material, the more expensive it will be.

Corrosion rate confirmed Extent determined

Remaining life assessment

Materials solutions

1

Operating solutions

Fuel solutions

15

9

Provide protection 2

Replace component

7

3

Shielding

Monitoring fuel change

Gas side Same material

13

Tube alignment

14

Sootblowing operations

5 Monolithic 6 Coextruded

Limit tube temperatures

Excess air strategies

8

Coating More resistant material

Use of additives

4 Same thickness

Steam side Redistribute 11 steam flow Chemical clean SH/RH

12

Limit main steam 10 temperature Notes: a) Remaining life assessment (1) is almost mandatory to decide which option should be adopted b) Boxes outlined in bold indicate options that have been most successful c) Numbers refer to main text

Figure 34-7. Strategies for preventing repeat failures by fireside corrosion in superheater/reheater tubes of oil-fired units.

Among the coating methods that have been tried for fireside corrosion resistance are chromizing and aluminizing. Flame- or plasma spraying, with and without subsequent heat treatment, have also seen significant development work. The former CEGB has tried a number of coatings for use in corrosion and erosion resistance.5,6 The primary use of coatings has been for the prevention of fireside corrosion in waterwalls of coal-fired units (see the description of results in Chapter 18, Volume 2). There is less field experience for coatings on superheater/reheater tubing in either coalor oil-fired units. Currently, either replacement in-kind or with a material of high corrosion resistance, depending on the wastage rate which has been experienced, are preferred options to the application of coatings.

Replacement with same material, same thickness (option 4) If the corrosion rate is only slightly higher than that required to reach the desired life as calculated from the remaining life assessment, tube replacement can be made in-kind.

Change to a more resistant material, ÒmonolithicÓ (option 5) An upgraded material can be used where unit operation is at high temperatures and fireside corrosion remains a problem despite the attempts at other fixes. The material chosen will depend on what is currently being used and what the desired resistance is to be. The corrosion rate of 300 series austenitic stainless steels, because of their nickel content, is greater than for ferritic stainless steels. Upgrading to an alloy such as 12 Cr Mo V may

Change to a more resistant material, composite such as co-extruded tubing (option 6) Co-extruded tubing, originally developed for SH/RH applications in coalfired units, could be used for severe corrosion in oil-fired units. A 50Cr50Ni alloy is used as the outer layer and is metallurgically bonded to a creep-resistant, but not corrosion resistant core, typically Eshete 1250 (a high-temperature, high-strength stainless steel). Metallurgicallybonded processing provides superior tubes for use in utility boiler applications when compared to mechanically bonded or bi-metallic tubes.7 Tube bending characteristics of coextruded material have been found to be identical to monolithic materials.7 Weld procedures have generally matched weld metals to the base metal to maintain property levels. Conventional welding techniques and normal quality control have been found sufficient to ensure good weld quality. The CEGB experience base was over 70,000 welds through 1984 (mostly in coal-fired units) without weld failure. No preheat or post-weld heat treatments have been required. A discussion about welding coextruded tubing can be found in Chapter 11, Volume 1.

Monitoring of fuel changes (option 7) Monitoring for increases in vanadium and sodium content whenever fuel changes are made will highlight a

Volume 3: Steam-Touched Tubes

34-17

potential increase in corrosion rates.

Use of Mg-based additives (option 8) A number of forms of magnesium and combinations with other metals have been found to be effective in reducing fireside corrosion in oilfired units, particularly the use of magnesium hydroxide Mg(OH)2 or magnesia, MgO. The use of magnesium has been found to form a magnesium vanadate complex (3MgO • V2O5) which is desirable because of it’s high melting point of 1071°C (1960°F) as shown in Figure 34-6. The type of additive to be used will depend on fuel chemistry and boiler design. The most effective product forms have been either a slurry or a solution. The form will determine the point of delivery and kind of delivery system. Treatment rates will depend on a number of factors including: particle size and distribution, reactivity of chosen additive, design of boiler, mode of application, and some trial and error.8 Choice and proper application of additives is a complicated subject.8,9 The most successful or optimized additives for a number of fuel oils are shown in Tables 34-6 and 34-7 for high vanadium, high sulfur and for a medium vanadium, low sulfur oil respectively.8 These are meant to be starting points as the optimum additive choice for each boiler and type of operation will differ. It is important to adjust for conditions different from those presented in these tables. Deposition probes which can provide useful short-term information on the quantity and corrosiveness of depositing ashes should be used to optimize the use of additives.10 The disadvantages of additives may include the relatively high continuous operating cost and a substantial increase in ash volume leading to additional downtime for cleaning.2 Increased abrasive wear of oil gun components can also occur. Thus, there is a need to continually check

34-18

Table 34-6 Additives for High-Temperature Corrosion and Fouling from High-Sulfur Oila Boiler Design Coal converted to oil

Feed Point with Fuel

Furnace Injection

Sootblower Feed Point

MgO-oil dispersion 2 to 7 microns

MgO dispersion (H2O or oil) 2 to 7 microns

Not recommended

As above but less than 1 micron size reactive forms

As above but higher in furnace to avoid water wall coating

MgO-H2O slurry

Oil-soluble Mg/Mn

MgCl2 solutionemulsion

Oil dispersion of MnO or MgO/MnO 2 to 7 microns Oil, or oil with coal potential

MgCl2 solutionemulsion Notes: a. High sulfur (2-3.5%) and vanadium (300-700) ppm, low sodium/vanadium ratio. b. Additives listed in preferred order of recommendation. Source: J.E. Radway and M.S. Hoffman8

the addition rate and its efficacy through the use of probes.

larly in coal-fired units when reducing the main steam temperature from 565 to 538°C (1050 to 1000°F).

Limit tube temperatures (option 9)

There are several problems with this approach. The primary drawback is that it is not efficient; there is a severe heat rate penalty. Predicting exactly what temperature is required may also be difficult.

Given the strong correlation between tube metal temperature and the potential for corrosion damage, several strategies can be implemented that directly address root causes of overtemperature in the tubes; they are discussed separately below.

Limit steam temperature (option 10) A specific strategy to minimize tube metal temperatures is by limiting main steam temperatures; this has historically been a primary strategy for the control of superheater/ reheater fireside corrosion, particu-

SH/RH Fireside Corrosion/Oil-Fired Units

For these reasons, it has been recommended that if limits on main steam temperature were to be used as a control strategy that an empirical approach be used to set steam temperature by monitoring tube

Table 34-7 Additives for High-Temperature Corrosion and Fouling for Intermediate-Sulfur Oila Feed Point with Fuel

Furnace Injection

Coal converted to oil

Dispersion of Mg/Al; Mg/Al Mn; Mg/Si 2 to 7 microns

Dispersion of Mg/Al 2 to 7 microns

Not recommended

Oil, or oil with coal potential

As above but less than 1 micron and more reactive forms

As above but higher in furnace to avoid water wall coating of MgO/Al2O3

MgO-H2O slurry

Boiler Design

Sootblower Feed Point

MgCl2 solution-emulsion Combustion catalyst

High reactivity MgO dispersion

Soluble additive (water or oil)

Soluble additive (water or oil)

Not usually available

High-reactivity MgO dispersion Notes: a. 1 to 2% sulfur, 150 ppm vanadium, sodium/vanadium greater than 0.15. b. Additives listed in preferred order of recommendation. Source: J.E. Radway and M.S. Hoffman8

wastage rates and ensuring that the rate was less than 25 nm/hr (~ 9 mils/yr).11

Redistribute steam flow (option 11) A technique of steam flow redistribution in superheaters has seen recent significant development.12 Redistribution of steam flow can serve to equalize the temperatures across the superheater. The method was

If excessive temperature in the superheater/reheater tubes is a contributing factor to the corrosion process, and if that condition has been exacerbated by the presence of increasing oxide scale thickness, a solution involving chemical cleaning may be in order. An overview of chemical cleaning in SH/RH circuits can be found in Chapter 4, Volume 1.

Correct misalignment of tubing (option 13)

Oil-soluble Mg/Mn

Gas-oil

(option 12)

designed primarily as a means of extending the creep life of tubes subject to overheating; however, if overheating of selected tubes is at the root cause of fireside corrosion, the technique can selectively lower tube metal temperatures and thus decrease fireside corrosion. The method is discussed in more detail in Chapter 32 on long-term overheating/creep of SH/RH tubes.

Chemical cleaning to remove steamside oxide scale

If misalignment of tubes has created localized fireside corrosion, this problem should be corrected by realigning the affected tubes. This will decrease the number of tubes that are directly exposed to the gas flow.

Change frequency and check effectiveness of sootblowing (option 14) This can be an important action because it can stop the formation of excessive deposits which result in laning or channeling in adjacent areas.

Lowering excess air (option 15) Operating with as low an excess air as possible can help keep constituent ash melting points high and therefore limit deposition of the liquid phase and the resulting rapid corrosion. The optimal level of excess oxygen has been found to range from around 0.8%10 to 1%2 depending on local CO generation. There may be problems holding to these levels in older units because of leakage or incomplete combustion leading to unburnt fuel as soot to the stack. Care must also be taken to avoid lowering the excess

Volume 3: Steam-Touched Tubes

34-19

7. Case Study None for this mechanism.

air too much as carburization of the SH/RH tubing could result.15

8. References 1Laxton,

J.W., D.B. Meadowcroft, F. Clarke, T. Flatley, C.W. King, and C.W. Morris, The Control of Fireside Corrosion in Power Station Boilers, Third edition, Central Electricity Generating Board, 1987. 2Wilson,

J.R., “Understanding and Preventing Fuel Ash Corrosion,” Corrosion 76, Paper No. 12, held March 2226, 1976, Houston Texas, 1976. 3Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 4Pantony,

D.A. and K.I. Vasu, J. Inorg. Nucl. Chem., Vol. 30, 1986, p. 423. 5Morgan-Warren,

E.J., “Thermal Spraying for Boiler Tube Protection”, Welding and Metal Fabrication, Jan/Feb, 1992, pp. 25-31. 6Bennett,

A.P. and M.B.C. Quigley, “The Spraying of Boiler Tubing in Power Plants”, Welding and Metal Fabrication, November, 1990, pp. 485-489. 7Flatley,

T. and T. Thursfield, “Review of Corrosion Resistant Co-Extruded Tube Development for Power Boilers”, 1984 ASM Conference on Coatings and Bimetallics for Energy Systems and Chemical Process Environments, held at Hilton Head, South Carolina, November 12-14, 1984. 8Radway,

J.E. and M.S. Hoffman, Operations Guide for

the Use of Combustion Additives in Utility Boilers, Research Project 1839-3, Final Report CS-5527, Electric Power Research Institute, Palo Alto, CA, December, 1987. 9Krause, H.H., Action of Fuel Oil Additives Containing Magnesium and Manganese on Superheater and Reheater Surfaces, Final Report CS-3281, Research Project 1839-1, Electric Power Research Institute, Palo Alto, Ca, December, 1983. 10Dooley, R.B. and H.J. Westwood, Analysis and Prevention of Boiler Tube Failures, Report 83/237G-31, Canadian Electrical Association, Montreal, Quebec, November, 1983. 11Williams,

D.N., H.R. Hazard, H.H. Krause, L.J. Flanigan, R.E. Barrett, and I.G. Wright, Fireside Corrosion and Fly Ash Erosion in Boilers, Research Project 2711-1, Final Report CS-5071, Electric Power Research Institute, Palo Alto, CA, February, 1987. 12Hara,

K., C. Lee, R. Moser, T. Rettig, and K. Clark, Improved Superheater Component Longevity by Steam Flow Redistribution, Research Project 1893-13, Final Report TR-101697, Electric Power Research Institute, Palo Alto, CA, December, 1992. 13Sotter, J.G., J.A. Arnot, and T.M. Brown, Guidelines for Fireside Testing in Coal-Fired Power Plants, Research Project 1891-3, Final Report CS-5552, Electric Power Research Institute, Palo Alto, CA, March, 1988. 14Wong-Moreno,

A., Y. Mujica Martinez, and L. Martinez, “High Temperature Corrosion Enhanced Residual Fuel Oil Ash Deposits”, Corrosion 94, Paper 185.

15Private

Communication from D. Lopez Lopez (IIE, Mexico) to R. B. Dooley, May, 1995.

16Lopez

Lopez, D., A. Wong Moreno, and L. Martinez, “Unusual Superheater Tube Wastage Associated with Carburization”, Materials Performance, Volume 33, No. 12, December, 1994, pp. 45-48.

34-20

SH/RH Fireside Corrosion/Oil-Fired Units

ACTIONS for SH/RH Fireside Corrosion Two paths for the BTF team to take in the investigation of SH/RH fireside corrosion damage begin here. The goal of these actions to see if further investigation of fireside corrosion is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a SH/RH BTF has occurred and fireside corrosion is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF by fireside corrosion.

Action 1a: If a SH/RH BTF has occurred and fireside corrosion is the likely mechanism.

➠ Determine whether the failure has occurred in a location that is typical of SH/RH fireside corrosion: ➠Review Figure 34-4 for typical boiler regions. ➠Review main text, Section 1.2 for description of susceptible locations

➠ Confirm that the macroscopic appearance of the failure includes such features as: • Multilayered fireside scale and ash (see Figure 34-1). • General wastage, undulating and uneven surface appearance. • Tube surface with grooving consistent with “alligator hide” appearance (see Figures 33-3 and 33-4, Chapter 33).

➠ If the BTF seems to be consistent with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ If the BTF does not seem to have features like those listed, return to the screening Table for steamtouched tubing (Table 31-1) to pick a more likely candidate.

Volume 3: Steam-Touched Tubes

34-21

Action 2: Determine (confirm) that the mechanism is fireside corrosion. A SH/RH tube failure has occurred which the BTF team has tentatively identified as being fireside corrosion damage (Action 1a). Action 2 should clearly identify fireside corrosion as the primary mechanism or point to another cause. The primary identifier will be the presence of low melting point components in external ash. Sample removal, metallographic analysis and deposits analysis will be used to make this determination. A primary objective is to make sure that the failure that has been experienced is not primarily long-term overheating/creep; See Chapter 32 and Table 34-2 for distinctions.

➠ Characterize the extent of damage. Is there significant wall thinning across a number of tubes on the fireside?

➠ Plot wall thinning against measurements of steamside oxide. Is the ratio of wall thinning loss to steamside oxide buildup greater than three?

➠ Analyze deposits. Does metallographic and melting point analysis of deposits detect the presence of low melting point constituents such as vanadium complexes and/or vanadium-sodium complexes (for high V oils) or sodium sulfates and derivatives (for low V oils)? Such deposits are indicative of the most common mode of fireside corrosion in oil-fired units.

➠ Evaluate microstructure. If tubing is an austenitic stainless steel, is there evidence of carburization and/or a sulfidation mechanism?

If damage is localized, it may be flyash erosion (Chapter 14, Volume 2); however, continue through flow chart, particularly deposit analysis.

Mechanism more likely to be longterm overheating/creep (Chapter 32); review that mechanism, review analysis of deposits to ensure that there is no evidence of low melting point ash components which would be indicative of fireside corrosion.

Mechanism is probably not fireside corrosion. Review other fireside failure mechanisms, particularly long-term overheating/creep.

Mechanism is probably not fireside corrosion. Review other fireside failure mechanisms, particularly flyash erosion (Chapter 14, Volume 2).

Probable mechanism is fireside corrosion.

➠ Go to Action 3: Root Cause Determination

Action 1b: If a precursor has occurred in the unit that could lead to future BTF by fireside corrosion.

➠ Determine whether one or more of the following precursors has been found or is likely to have occurred in the unit: • Any evidence of molten salts observed or measured.

34-22

SH/RH Fireside Corrosion/Oil-Fired Units

Action 3: Determine root cause of the fireside corrosion A BTF failure has occurred and the mechanism has been confirmed as fireside corrosion (Action 2) or a precursor has occurred (Action 1b). The goal of this Action 3 is for the BTF Team to review the potential root causes of fireside corrosion, identify probable ones, and take those actions that are needed to confirm which are operative in the unit. This step must be taken so that the proper actions can be taken to prevent future BTF from occurring by this mechanism. Execute, in parallel, Action 4 to determine the extent of damage.

➠ Review list of major root cause influences in first column, below ➠ Take indicated actions to confirm the applicability of that influence in unit. Major Root Cause Influences

➠ Actions to Confirm

3.1 Potential actions for all root causes of fireside corrosion.

➠ (a). Ash and deposit analysis to identify presence of low melting point constituents, particularly vanadium/ vanadium-sodium and sodium sulfate complexes. ➠ (b). NDE measures (typically UT) to identify wall thinning and steamside oxide scale buildup.

3.2 Influence of oil composition.

➠ (c). Monitor oil corrosiveness using corrosion or deposition probes. ➠ (d). Analyze ash deposits as in item (a) above.

3.3 Influence of overheating of tubes. 3.3.1 Excessive temperatures caused by steamside oxide buildup.

➠ (e). NDE of steamside oxide thicknesses. ➠ (f). Selective tube sampling and metallurgical analysis to confirm steamside oxide and wall thickness. ➠ (g). Monitoring of thermocouples installed across the SH/RH outlet legs in vestibule to identify hottest platens across the boiler.

3.3.2 Excessive temperatures as caused by operating conditions. • high temperature laning of gases • changes in absorption patterns between furnace and convection sections, • RH overheating because of rapid startups • tube misalignments

➠ (h). For high temperature laning: monitor temperatures as in (g) above and consider the use of the cold air velocity test (CAVT). Details of the latter can be found in Chapter 14, Volume 2 on flyash erosion. ➠ (i). For reheater overtemperature during start sequences: check the startup probe and limit temperatures to 538°C (1000° F) prior to RH flow. ➠ (j). Visual inspection can be used to detect tube misalignments.

3.4 Influence of operating factors. 3.4.1 Operation with high levels of excess oxygen and/or periods of very low excess oxygen.

➠ (k.) Check operating logs for typical excess oxygen levels.

3.4.2 Poor sootblowing operations.

➠ (l). Check sootblowing frequency, effectiveness, and superheat level of blowing medium.

Volume 3: Steam-Touched Tubes

34-23

Action 4: Determine the extent of damage or affected areas In parallel with Action 3 (root cause analysis), the BTF Team should determine the extent of damage. Evaluation will be based on detecting wall thinning. Wastage rates in excess of 25 nm/hr (~ 9 mils/yr) are of concern.

➠ Identify all locations to be examined. Refer to Section 1.2 of main text and Figure 34-4 for typical locations. Damage may be widespread and missed locations are sites for future failures.

➠ Perform UT survey to measure extent of damage via wall thinning and steamside oxide thickness. A review of UT methods is provided in Chapter 9, Volume 1.

➠ Perform tube sampling to measure wall thinning and steamside oxide buildup and to determine the degree of damage.

➠ Use results interactively with Action 3.

➠ Go to Action 5: Implement Repairs, Immediate Solutions and Actions. Begin remaining life assessment.

Action 5: Implement repairs, immediate solutions and actions The most important actions for the BTF team are to (i) initiate a remaining life assessment based on the wastage rate and extent of steamside oxide as derived from the NDE survey, (ii) choose a repair strategy based on remaining life assessment, and (iii) coordinate a long-term strategy from the options outlined in Figure 34-7. It is important that proper precautions be taken prior to repairs, inspections, or whenever movements in an oil-fired boiler are contemplated because of the toxicity of vanadium and derivative compounds, and the health risk thus posed.

34-24

• Any evidence of wall loss observed or measured.

➠ Consider the use of Mg-based

• Any evidence of “alligator hide” observed or measured.

➠ Consider operation at low levels

• Change to a more aggressive oil.

➠ Chemically clean, as needed to

• Evidence that tubes may be overheating such as from thermocouple readings or excessive steamside scale build-up. • Any periods of high emission of partially burnt particles.

➠ Determine whether operation has occurred with high levels of excess oxygen.

➠ These precursors can signal the potential for SH/RH tube failures by a fireside corrosion mechanism. If one or more has occurred, go to Action 3 which reviews root causes and outlines the steps to confirm the influence

SH/RH Fireside Corrosion/Oil-Fired Units

additives. of excess oxygen. remove excessive steamside deposits.

➠ Adjust fireside conditions, as needed. References to other sources of detailed information: • Main text (this chapter) and Table 34-5 provide additional detail on repairs, immediate solutions and actions and relate them to underlying root causes. • Guidance on additive use.8 • Guidance on chemical cleaning can be found in Chapter 4, Volume 1. • Guidance on fireside testing.13

Action 6: Implement long-term actions to prevent repeat failures The correction of the underlying problem(s) and the prevention of repeat failures are priorities for the BTF team. The proper choice of long-term actions will include the analysis of remaining life and an economic evaluation to ensure that the optimal strategy has been chosen from those shown in Figure 34-7 and described in the main text.

Major Root Cause Influences

➠ Long-Term Actions

Potential actions for all root causes of fireside corrosion.

➠ Perform remaining life assessment. ➠ Continual check on the use of Mg-based additives. ➠ Optimization of excess oxygen levels. ➠ Set up long-term monitoring and re-evaluation program. ➠ Evaluate full range of available options using roadmap in Figure 34-7.

Primary options will be the use of Mg-based additives and operation at low levels of excess oxygen.

Influence of oil composition.

➠ As above with emphasis on additives.

Influence of overheating of tubes. Excessive temperatures caused by steamside oxide buildup.

➠ Perform remaining life assessment. ➠ Consider program of periodic chemical cleaning. See Chapter 4, Volume 1. See also options for long-term overheating of tubes (Chapter 32). ➠ Set up long-term monitoring and re-evaluation program. ➠ Evaluate full range of available options using Figure 34-7.

Excessive temperatures as caused by operating conditions. • high temperature laning of gases • changes in absorption patterns between furnace and convection sections, • RH overheating because of rapid startups • tube misalignments

➠ Evaluate full range of available options using Figure 34-7.

Influence of operating factors. Operation with high levels of excess oxygen and/or periods of very low excess oxygen. Poor sootblowing operations.

➠ Perform remaining life assessment. See discussion of methods in Chapter 8, Volume 1. ➠ Set up long-term monitoring and re-evaluation program. ➠ Evaluate full range of available options using Figure 34-7.

Volume 3: Steam-Touched Tubes

34-25

Action 7: Determine possible ramifications/ancillary problems The final step for the BTF team is to review the possible ramifications to other cycle components implied by the presence of fireside corrosion in the SH/RH tubes, or by its precursors.

34-26

Superheater/ Reheater Fireside Corrosion Aspect

Alert for Other Cycle Components

➠ Actions Indicated

Use of additives.

• Mg-based additives can coat the waterwalls of the furnace and cause a reflection of heat into the convective passes. This could lead in turn to higher temperatures for SH and/or RH tubes and an increase in boiler tube failures by long-term overheating (see Chapter 32). • Additives can also cause increased erosion of burner components and additive transport lines.

➠ Monitor unit for signs of detrimental effects of additives.

Tube overheating because of excessive steamside oxide.

• Potential for additional tube failures by long-term overheating mechanism. • Exfoliation of scale with subsequent carryover into turbine could lead to solid particle erosion. • Exfoliation could lead to tube blockage and additional SH/RH failures by a short-term overheating mechanism (Chapter 36).

➠ Chemical clean unit if necessary. See guidance in Chapter 4, Volume 1.

Total redesign of the superheater or reheater.

• Would change absorption patterns through the SH/RH sections and may increase temperatures in other sections.

➠ Check temperatures in the redesigned and other areas.

SH/RH Fireside Corrosion/Oil-Fired Units

Chapter 35 • Volume 3

Dissimilar Metal Weld Failures

Introduction Dissimilar metal welds (DMWs) are used to join ferritic and austentitic steel tubing in the final outlet sections of superheaters (SH) and reheaters (RH). Prior to the mid1980s numerous outages occurred as a result of failures via low ductility cracking in the low alloy ferritic steel immediately adjacent to the weld fusion line. The problem was widespread; it was estimated at the time that 50% of all boilers in North

America with austenitic tubing had DMW failures.1 There was wide variability in the times to first failure ranging from as few as 30,000 hours to over 150,000 hours. A comprehensive effort over the last ten years has characterized the problem of DMW failures, provided a set of analytical tools for analysis, and provided appropriate solutions to the problem.2-4

Volume 3: Steam-Touched Tubes

35-1

1. Features of Failure and Typical Locations Dissimilar Metal Weld Failures: Identification Keys 1. DMW failures are found adjacent to the welds in the ferritic alloy in a circumferential orientation, and have thick-edged, low ductility features. 2. The manner in which failure develops will depend on whether the weld metal used was austenitic stainless steel or nickel-based; damage accumulation in either case will be by creep with fatigue assistance.

1.1 Features of failure Dissimilar metal welds (DMWs) are found in the superheater/reheater where austenitic materials, generally Types 304H, 321H or 347H, are joined with ferritic materials, generally T22, T5, or T9. There are three types of DMWs. The most common weld is made with austenitic stainless steel filler (E309). A second type of weld is made with a nickelbased filler such as Inconel 132, Inconel 182, Inconel 82, or Inco Weld A; Inconel and Inco are trademarks of the International Nickel Company. The third type is a pressure induction weld; welds made by this process will have features similar to fusion welds made with stainless steel filler metals.

Table 35-1 provides some key means to distinguish between the major types of DMW failures. Table 35-2 summarizes macroscopic features that will point to dissimilar metal weld failures in contrast to other SH/RH tubing failure mechanisms. Failure is generally seen macroscopically as a fusion line crack occurring at or near the heataffected zone on the low-alloy side of the joint (Figure 35-1). Cracking is typically oriented circumferentially around the tube. For induction welds, the fracture may be flat and featureless; for fusion welds, the fracture will tend to follow the contours of the fusion line. Final failures occur with low-ductility and manifest thick-edged fracture surfaces.

Table 35-1 Distinguishing Features (Microscopic) of Failures in Fe-Based Stainless Steel and Ni-Base Filler Metals in DMWs Characteristic

Iron-Base Stainless Steel Filler Metal

Nickel-Base Filler Metal

Location of Cracking within HAZ (generally)

Along prior austenite grain boundaries approximately 1-2 grain diameters from fusion line.

Immediately along weld interface associated with carbide precipitation and creep cavitation.

Carbide Morphology (See Section 2.2 starting on page 35-6)

Generally Type II.

Predominantly Type I.

Nature of Carbide

Diffuse array of smaller carbides.

Planar array of globular carbides.

Do carbides encourage interfacial growth

No.

Yes.

Creep voids associated with this carbide type?

No.

Yes.

Carbon activity gradient of filler with ferritic material?

Higher than for Ni-base fillers.

Lower than Fe-base fillers.

Thermal expansion with ferritic materials

Worse than Ni-base fillers.

Better than Fe-base fillers.

Time to final failure

About 1/3 to 1/5 of times for Nibase filler metal welds.

Three to five times longer than Fe-base filler metals.

Note: Induction welded DMW will have similar properties to those listed for Fe-based fusion welds above.

35-2

Dissimilar Metal Weld Failures

Table 35-2 Macro-Features Common to all Dissimilar Metal Weld Failures Macro-features • Thick-edged fractures with signs of low-ductility. • Circumferential cracking in the ferritic material. • Located near a dissimilar metal weld. • Formation of an "oxide notch" on the outside surface of the tube in the ferritic material. • Flat, featureless fracture surface (typical of induction weld). • Cracking following fusion line (typical of fusion welds). • Failures may be associated with bent tubes or other signs of overstressing.

Damage accumulates primarily by creep, possibly with some contribution by fatigue. Three microscopically distinct manifestations of the basic creep mechanism in DMWs have been identified; they can occur singly or in combination, as follows: • Development of cracks along prior-austenite grain boundaries in the low-alloy steel heat-affected zone (HAZ) one to two grains away from the fusion line. This is commonly observed in welds made with stainless steel filler metal and occasionally in nickelbased filler metal DMWs. Figure 35-2 shows a typical cross-section. • Development of cracking immediately at the weld interface on the low-alloy side of the weld, along a planar array of globular carbides. This is commonly observed for

Figure 35-1. Typical appearance of a cracked dissimilar metal weld.

DMWs made with nickel-base filler metal. Figure 35-3 shows this type of cracking. • Propagation of an oxide notch from the external surface. Oxide notches are almost universally seen in DMWs pulled from service; however, in many cases the notches do not propagate. Those that do are most commonly seen in thin-walled tubing and can be in either stainless steel or nickel-base filler metals. Failures in DMWs are generally accompanied by, but not caused by, carburization of the weld metal as indicated by increased microhardness. The degree of decarburization on the ferritic side of the weld can be seen in Figure 35-4 where the dramatic increase in hardness at the weld interface is also shown in DMWs removed from service after 24 years.

Early stages of creep damage, such as microvoid formation, particularly in nickel-base filler metals, may not be detectable by optical microscopy, but can be detected by scanning electron microscopy (SEM) techniques. DMW failures may have microstructures that show signs of overheating. This can be detected by an analysis of the oxide scale thickness on the ferritic side of the joint, somewhat removed from the DMW itself. A detailed discussion of the use of oxide scale measurements for analysis of tube temperatures can be found in Chapter 8, Volume 1.

1.2 Locations of failure DMWs are located in the superheater, reheater, vestibule and penthouse regions of the boiler in transitions between austenitic and ferritic materials.

Volume 3: Steam-Touched Tubes

35-3

Figure 35-2. Typical cross-sectional appearance of a dissimilar metal weld failure after long-time boiler service. This example is a DMW with stainless steel filler metal. Note the oxide notch on the OD and the intergranular cracking adjacent to the weld line. Source: S.R. Paterson, et.al.7. Inset shows further detail of the intergranular creep cracking adjacent to a pressure weld. Note the cracking is oriented normal to the hoop stress. Source: D. French.

35-4

Dissimilar Metal Weld Failures

Figure 35-3. Detailed metallographic appearance of cracking along the weld fusion line associated with a line of carbides. This is typically observed in dissimilar metal welds made with nickel-base filler metals.

Hardness (DPH)

Distance from Interface (mm) 2.03 1.52 1.02 0.51 0 0.51 1.02 1.52 2.03 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100

Weld Interface Weld metal T22 ▲ ▲▲ ▲

▲ ▲

Unit A Unit B

▲

▲ ▲

▲

▲ ▲ ▲

▲

▲

▲

▲ ▲ ▲ ▲ ▲ ▲ ▲▲

▲ ▲

▲

▲

▲

0.08 0.06 0.04 0.02 0 0.02 0.04 0.06 0.08 Distance from Interface (in.)

Figure 35-4. Microhardness profiles of dissimilar metal welds after 24 years of service showing an increase in hardness near the weld interface. Source: K.H. Holko, et al.2d

Volume 3: Steam-Touched Tubes

35-5

2. Mechanism of Failure Dissimilar Metal Weld Failures: Mechanism 1. Failures in DMWs are driven by service conditions arising from higher than expected service temperatures and three types of stresses: (i) primary system stresses such as caused by pressure and dead weight, (ii) secondary system stresses such as caused by the constraint of thermal expansion, and (iii) intrinsic or self-stress caused by differential thermal expansion of the joined materials. 2. In fusion welds with austenitic weld metals and in induction pressure welds, creep damage accumulates at prior austenite grain boundaries. 3. In the dominant mechanism for fusion welds with nickel-base weld metal, carbon diffusion during service leads to a planar array of globular carbides located at the weld interface on the ferritic side of the joint. Failure occurs when both a sufficient density of carbides has occurred and creep void growth has initiated.

2.1 Introduction Superheater/reheater design was discussed in Chapter 2, Volume 1. Design temperatures can range from 400 to 600°C (~ 750 to 1110°F) depending upon location. Increasingly higher tube metal temperatures demand either increased wall thicknesses and/or a material change. Carbon steel is used in the primary stages, whereas low alloy steels are used for most of the SH/RH, except for the finishing stages where austenitic stainless tubes are normally used. The DMWs of concern are those that join the ferritic materials to the austenitic stainless steel. As noted above, either fusion or induction welding processes are used. Filler metals are either nickel-based or iron-based austenitic stainless steels. Welds made by an induction process have properties that are similar to those for fusion welding with austenitic filler metals thus the comments made throughout the balance of this section pertaining to austenitic filler metals will also apply to induction welds. Differences in thermal expansion and creep behavior of the joined materials, and local metallurgical changes at the low-alloy steel to weld metal interface make the DMW more susceptible to failure than likematerial welds.

2.2 Microstructural changes in DMWs during service A brief review of microstructural changes in DMWs is presented here; additional detail can be found in references 3-6. The degradation of DMWs after longterm service includes a number of observable features, including (i) oxidation of the ferritic steel, including oxide notching, (ii) softening of the ferritic steel HAZ, (iii) migration of carbon from the HAZ into the weld metal, (iv) precipitation and growth of

35-6

Dissimilar Metal Weld Failures

carbides at the weld interface and the HAZ prior-austenite grain boundaries, and (v) the formation and growth of creep voids. These processes are strongly influenced by stress, temperature and time. The times to failure for field DMWs are strongly influenced by service conditions. The two modes of damage accumulation in DMWs were introduced above: austenite grain boundary cracking (dominant in iron-based filler-metal welds), and weld interface cracking (dominant in nickelbase filler-metal welds). There is a strong correlation between creep cracking and carbide morphology. Two types of carbide morphologies form in DMWs in service. Type I is a planar array of globular carbide. This type is prevalent in nickel-base filler metal welds, occurring along about 80% of the weld interface in a typical weld, and encourages interfacial cracking. Type II, is a diffuse array of generally smaller carbides found in a wider band. It is predominant in stainlesssteel filler metal welds and also occurs in about 20% of the nickelbase DMWs. Creep voids are only associated with Type I precipitates.3 Type II carbides inhibit interfacial growth, to the point where a crack developing in Type I material will stop or transfer to grain boundaries when Type II material is encountered. Decarburization occurs as a result of the migration of carbon from lowalloy steel to the weld metal. This gross decarburization however, is not thought to be a major factor in DMW failures. The ductility of low-alloy steels in the presence of a triaxial stress field is low. There is strong indication that the development of a triaxial stress state occurs within the weld when an axial stress, such as caused by bending, is imposed on the radial and circumferential stresses caused by differential thermal expansion.

Oxide notches that form at the outside surface in both nickel-base and iron-base welds were previously thought to initiate DMW failures. However, it is now established that damage can initiate at the outside surface, mid-wall or inside surface and that oxide notching is not a root cause of the problem.

2.3 Influence of welding variables Various filler metals and weld configurations/geometries can have significant effects on the life of a DMW. Table 35-3 provides the key results of a comparison of different filler types.2h

Other factors such as post weld heat treatments, use of backing rings, and presence of pre-existing weld defects can, in some cases, influence performance, but are generally not factors in the development of DMW failures per se. Sources for the design of, and procedures for, improved dissimilar metal welds and filler metals have been compiled.2h A summary of those findings is provided here. Under service conditions, nickel based filler materials have longer lives than stainless steel filler metals by up to five times because2h (i) their thermal expansion better matches ferritic materials than stainless steel fillers, and (ii) the carbon activity gra-

dients between ferritic steel and nickel base weld metal are lower than for stainless steel fillers, which limits the carbon migration from the ferritic material. Because the carbon activity is significantly lower, the time for a critical density of carbides to form along the weld line is much longer than the time for creep damage to form in the ferritic material adjacent to the stainless steel weld metal, leading to longer times to failure in the nickel-base filler metals. Service experience also indicates that repairs made with nickel-base fillers for welds originally made with stainless steel fillers show considerable life improvement, even if part of the original weld is left in place.

Table 35-3 Summary of Performance Characteristics of DMW Made With Various Filler Metals2h Filler

Tendency to Expansion Difference with Form Type 1 Interfacial 21/4 Cr - 1Mo Carbides (RT-1000°F)

Needs PWHT?

Thermal Stability

DMW Performance Observations

Conclusions Relative to DMW Use

Commercial alloys E309

27% greater

None

No

Fairly stable.

Gives poorest performance.

Use only in least arduous applications.

Inco 92

5% greater

Slight

No

Marked age hardening.

Shows little tendency to interfacial failure.

Use limited by age hardening.

Inco 132

7% greater

Marked

No

Fairly stable.

Significantly better than E309 (factor of 3-5X) but shows interfacial failure.

Better than E309 in most cases, widely used.

Inco 182

10% greater

Marked

No

Considerable age hardening.

Significantly better than E309 (factor of 3-5X) but shows interfacial failure.

Better than E309 in most cases, widely used.

Inco 82

3% greater

Marked

No

Fairly stable.

Significantly better than E309 (factor of 3-5X) but shows interfacial failure.

Better than E309 in most cases, widely used where TIG welding employed.

Inco A

3% greater

Marked

No

Fairly stable.

Significantly better than E309, last longer than most nickel welds but shows interfacial failure tendency.

Some indications that this is the best of the commercial nickel fillers.

None

No

Excellent.

No tendency to interfacial: shows best life

Microfissuring tendency needs control before widespread use possible.

Experimental Alloy HFS6 (See ref 2h)

7% greater

Volume 3: Steam-Touched Tubes

35-7

One potential drawback to the nickel-base filler metals was noted above. Once Type I carbides form at the fusion line, the rupture strength of the weldment is equal to or less than that for the stainless steel filler metal. Under usual service conditions, the time required for such carbides to form is long relative to the time for damage to accumulate in the stainless steel filler. However, if the weld is exposed to high temperatures, for example, above 595°C (~ 1100°F), even for relatively short times, then the carbides can form. The relative life of a number of standard nickel-base fillers compared to a standard E309 stainless steel weld in accelerated “discriminatory” tests is shown in Figure 35-5. Pressure welds have similar or shorter lifetimes than conventionally welded stainless steel (E309). Weld configuration was also found to be a significant variable. Evaluation of four configurations, shown in Figure 35-6, indicated that cracking in service welds is more prevalent in regions of the interface normal to the tube axis than those parallel to the tube axis. Under laboratory conditions, the longest lives in accelerated tests have been obtained with DMWs having interfaces at the greatest weld angles (Figure 35-6). Weld interface angle is therefore considered to be an important parameter in DMW performance and repair.

Standard S/S (E309)

4.6

INCO 132

Life ratio relative to standard S/S

INCO A

5.1

INCO 82(a)

5.1

INCO 82(b)

5.6 5.9

INCO 182 INCO 132 + wide cap

7.7 (a) High heat input (b) Low heat input

Figure 35-5. Relative performance of dissimilar metal welds with different commercial filler metals and geometries under accelerated discriminatory testing. Source: D.I. Roberts, et al.2h

0°

Filler

SS

T22

0° weld angle CL

37-1/2° Filler SS

T22

37-1/2° weld angle (standard) CL

2.4 Influence of stress and temperature The most important factors governing the life of DMWs are service conditions. Stresses of three general types contribute to premature failure in DMWs: (i) primary system stresses such as caused by pressure and dead weight loads, (ii) secondary system stresses such as caused by the constraint of thermal expansion, and (iii) intrinsic or self-stress caused by the differential thermal expansion of the joined materials.

Scatter band

1

60° Filler SS

T22

60° weld angle CL

Wide cap Filler SS

T22

Wide weld cap CL

Figure 35-6. Possible weld geometries for dissimilar metal welds. Source: D.I. Roberts, et al.2h

35-8

Dissimilar Metal Weld Failures

Note that the last of these, differential thermal expansion of the joint materials is not sufficient, by itself, to cause failures during normal unit operation. For failures to occur, additional, and usually abnormal, loading or temperatures must generally exist at the DMW. The field experience shows many examples of the effect of local secondary system stresses, such as constrained thermal expansion, leading to DMW failures. Cases have been observed of two sets of DMWs, one with noticeable constraint of thermal expansion which led to early tube failures, and a nearby tube without such constraint but identical in all other ways, which had not failed. A means to demonstrate the effects of stress level, service temperature, and composition of weld metal is shown in Figure 35-7. That figure shows the calculation of a weld performance factor (WPF) which is the ratio of the rupture stress of T22 ferritic material remote from the weldment to the rupture stress of the DMW. The larger the value of WPF the lower the performance of that weldment compared to the ferritic base material.

Weld Performance Factor 2.2 Austenitic DMWs 2.1 2.0 1.9 1.8 Inconel DMWs

1075°F

1050°F

1025°F

1075°F

1050°F

1025°F

1.7 1.6 1.5 1.4 1.3 3

4

5

6

7 8 9 Axial Stress, ksi

10

11

12

13

Figure 35-7. Estimated weld performance factors for dissimilar metal weld joints made using austenitic and nickel-based filler metals. Source: S.R. Paterson, et al.7

All of these influences have been combined into several computer codes, such as PODIS and DMW LIFE, which are available to predict the life of DMWs. Additional information about such procedures are included in the section on long-term actions below.

Volume 3: Steam-Touched Tubes

35-9

3. Possible Root Causes and Actions to Confirm Dissimilar Metal Weld Failures: Root Causes 1. Improper location or design of the original weld that does not account for actual service conditions of temperature and stress are the main causes of DMW failures. 2. Improper initial fabrication, particularly in pressure induction joints, or changes in operation that result in higher stresses, higher temperature or increased cycling, are also contributors to DMW failures.

3.1 Introduction Table 35-4 summarizes the potential root causes, actions to confirm, and corrective actions. One of the primary questions to be answered by the boiler tube failure team is whether the DMW BTF is driven primarily by excessive stresses or excessive temperatures.

3.2 Excessive tube stresses such as caused by improper initial design of tube supports, or supports that have stopped functioning in the designed manner One cause of DMW failures is if the initial design did not anticipate and properly accommodate loads on tubes. For example, excessive restraint on tubes, such as caused by locating the DMW near to the roof, furnace wall, or other fixed points, can be a source of high stresses. Improper tube supports can also cause excessive weld stresses. Support design faults can include (i) weld placement in the middle of a long span, and (ii) design that does not allow adequately for thermal expansion of the tube. In addition to poor design of supports, restraint on tubes can result during operation if there are support failures or if slag and other debris accumulates so as to constrain tube thermal expansion. Actions to confirm will include: (a). Visual examination of unit to determine whether there are likely susceptible locations or evidence of a problem such as bent tubes, warpage, misalignment of tubes, missing or broken supports, or other visual signs of overstressing.

35-10

Dissimilar Metal Weld Failures

(b). Perform a stress analysis of suspect locations. One such analysis5 used a piping stress code to calculate primary loads (pressure and dead weight) and secondary loads, such as constrained thermal expansion. Although this can be expensive and time-consuming, the information can be invaluable, and is a required input for all damage assessment models.

3.3 Excessive local tube temperatures Excessive local temperatures will increase thermally-induced stresses and will also accelerate damage accumulation in DMWs. These causes include (i) variation of temperatures across the SH/RH, (ii) tube temperatures higher than design values, (iii) partial tube blockages resulting in lower steam flows and subsequently to higher tube metal temperatures. Tube failures by longterm overheating/creep can also occur in the ferritic alloy near to the transition between material types (see Chapter 32). A knowledge of operative temperatures is required by damage assessment codes. A number of methods can be used to estimate temperatures, and thus determine whether this is a primary root cause. These include: (c). Review available thermocouple data from locations such as outlet tubes in the penthouse, etc. (d). Perform an evaluation of the oxide scale. In addition to addressing the temperature needs for the DMW analysis, this will also provide a life assessment estimate of the ferritic tubing. Such methods are discussed in more detail in Chapter 8, Volume 1. Oxide thickness, the key parameter, can be measured nondestructively by ultrasonic methods as outlined in Chapter 9, Volume 1.

Table 35-4 Major Root Cause Influences, Confirmation and Corrective Actions Major Root Cause Influences

Actions to Confirm

Long-Term Actions and Prevention of Repeat Failures

Immediate Actions and Solutions

3.2 Excessive tube stresses such as caused by improper initial design or improper tube supports. • locating the DMW near the roof, furnace wall or other fixed points or near to the header • weld placement in the middle of a long span • inadequate allowance for thermal expansion • support failures or slag accumulation leading to constraint of thermal expansion

(a). Visual examination of the unit to determine whether there are suspect locations or evidence of a problem such as bent tubes, warpage of tubes, misalignment, missing or broken supports. (b). Perform a stress analysis of suspect locations. Piping stress codes can be used to determine both primary and secondary stresses.

• Repair damaged locations using either a "dutchman" (preferred) or in-situ weld repair with nickel-based filler metal. • Determine the extent of damage through (i) visual examination to detect adjacent locations with obvious signs of distress, (ii) specialized radiography, (iii) oxide scale measurements and analysis, (iv) selective sampling, as required for confirmation.

• Implement a damage assessment code, such as PODIS, to optimize a program of control and prevention of DMW failures. Actions may include predicting remaining life, relocating welds, upgrading to higher grade materials. • Implement a periodic inspection program for hangers, supports and spacers, and a temperature monitoring program. • Redesign SH/RH to locate DMWs in areas of lower stress or lower temperature.

(c). Review of available thermocouple data for indications of overheating. (d). Perform oxide scale thickness evaluation including ultrasonic measurement and analysis of results.

• As above.

• Institute a program to measure and interpret oxide scale thickness periodically as a means of understanding tube temperature trends. • Apply damage assessment code to determine whether temperature is the predominant factor. • Redesign SH/RH so that the DMW is in a lower temperature regime.

(e). Review operating records with an eye toward conditions that may have increased either tube stresses or temperatures.

• As above.

• Implement a damage assessment code, such as PODIS, to optimize a program of control and prevention of DMW failures. Actions may include predicting remaining life, relocating welds, upgrading to higher grade materials. • Determine what effects changes in operation will have on DMWs. • Implement a periodic inspection program for hangers, supports and spacers, and a temperature monitoring program.

(f). Metallographic samples should be used to evaluate whether initial weld defects such as incomplete fusion or lack of penetration are a contributing cause.

• As above.

• Monitor for outbreaks of similar problems in other locations; a damage assessment method such as PODIS may provide guidance about general DMW life, although it cannot specifically pinpoint initial fabrication defects.

3.3 Excessive local tube temperatures. • tube temperatures above those anticipated in the design • variation across the SH/RH

3.4 Changes in unit operation • to increased unit cycling • change of fuel causing increased tube temperatures • redesign of adjacent SH/RH that results in higher tube service temperatures

3.5 Initial fabrication defects.

Volume 3: Steam-Touched Tubes

35-11

3.4 Changes in unit operation Changes in operation can be at the root of DMW failures. These might typically include (i) a change from base load to cycling of the unit, (ii) a change of fuel causing increased tube weld temperatures, and (iii) a redesign of an adjacent SH/RH that causes an increase in service temperatures experienced by the tube. Unit cycling in particular, can be a

major contributor to damage accumulation in DMWs. Actions to confirm this root cause include: (e). Review operating records to determine if it is likely that changes have caused an increase in primary or secondary stresses, or the temperatures of suspect tube locations.

3.5 Initial fabrication defects Initial fabrication defects can be a direct cause of DMW failures, particularly improperly fabricated pressure welds. To confirm: (f). Metallographic samples of the failed tubes should indicate an incomplete or improper weld, such as a lack-of-fusion defect.

4. Determining the Extent of Damage Determining the extent of damage is the first step in an assessment of remaining life analysis for DMWs; however, it can be tricky. Visual inspection should note locations of obvious failures and general information such as support locations and condition, platen distortion caused by support failure or slag/ash buildup, repairs or replacements, alterations to the boiler. The crux of the challenge is determining the degree of damage within the DMW itself. Surface techniques such as dye penetrant that detect only the formation of an oxide notch will be inadequate, particularly in thick-walled tubes. Often such notches may stop propagating and final failure will be induced mid-wall. A technique that can assess the full extent of damage is therefore needed. Another complicating factor is that the damage occurs at the interface between austentitic and ferritic material, and therefore it can be difficult to separate damage from the interface signal with either ultrasonic or eddy current testing.

35-12

Dissimilar Metal Weld Failures

A specialized radiographic technique, termed the Union Electric technique for its source of origin, can detect well-developed damage in DMWs made with stainless steel filler metal.5,8 The basics of the method are shown in Figures 35-8 and 35-9. Although the method can detect damage levels down to about 5% of the interface, film interpretation was found to be difficult below about 15% damage. Good correspondence between the damage predictions by the technique and actual damage was confirmed for over 50 DMWs that had damage levels ranging from 5 to 90% damage.5 For detecting creep damage with nickel-base filler metal manifested as small voids along the interface, and for small amounts of damage in all materials, sampling and destructive examination will be required. This may be performed on complete tube samples or boat samples of selected areas. The use of SEM and optical techniques to characterize the extent of creep damage is required. Chapter 10, Volume 1 reviews some of these methods.

a a Source - Iridium 192 centered on weld. Source to film distance 16 in.

Radiation beam

¥

a a

Oxide scale thickness has been measured by ultrasonic testing to find high-temperature tubes that are then subject to the more detailed damage assessment available from the Union Electric technique described above. The use of oxide scale for temperature analysis in ferritic tubes is described in detail in Chapter 8, Volume 1; UT measures of oxide thickness are discussed in Chapter 9, Volume 1.

Lack of fusion and/or HAZ cracking

Weld

Weld

Weld image

Weld image

B

A

B

Weld image

Single film - Dupont NDT 55 ASTM E 94, type 1 or double film - Dupont 70 and Dupont 45

Figure 35-8. Union Electric special radiographic technique for dissimilar metal welds. Source: H.J. Grunloh, et al.5

Etching of DMWs is difficult because of the variety of materials and the range of their chemical reactivities. One successful reagent is dilute aqua regia. Its composition by volume is 42% HCl, 17% HNO3 with the balance H2O. Success has also been achieved with a double etch, once with Nital, followed by electrolytic oxalic acid.9

Radiation from isotope source

Weld metal

Quarter section of DMW (three dimensional)

Crown Weld interface T-22

Cracked regions near interface

Projections of cracked regions on film

Film

Overlap

Two dimensional radiographic image

Figure 35-9. Union Electric technique and multiple radiographic defect images. Source: H.J. Grunloh, et al.5

Volume 3: Steam-Touched Tubes

35-13

5. Background to Repairs, Immediate Solutions and Actions Repair methods that are relatively simple to implement and have been found to provide extended lives of DMWs are briefly summarized here. The optimal approach has been found to be replacement of each damaged DMW with a “dutchman”, which is a short section containing a dissimilar metal weld made in the shop by an automatic TIG process. Then two field welds to insert the dutchman can be made in like material: ferritic-to-ferritic and austeniticto-austenitic. If it is not possible to obtain a shopwelded dutchman, then the next best approach is to insert a short piece of new tube by making two new field welds. This is a temporary measure; the field-welded insertion should be replaced with a proper dutchman at the first opportunity.

In-situ weld repairs of the damaged location are not recommended, however, if necessary for an emergency, the following steps are used: (i) remove all of the damaged weld metal, leaving as little original weld metal as possible; (ii) grind out base metal to effect a large included angle so that a large capping pass can be used, see the bottom geom-

etry in Figure 35-6, for example; (iii) make the repair weld with nickelbase filler metals, independent of the composition of the original weld metal. Weld geometry should follow the recommendations contained in Section 2.3 above and illustrated in Figure 35-6. Further, the in situ repair should be replaced with a dutchman at the next outage. Nickel-base welds are more difficult to make than those with stainless steel filler, however the upgrade to such welds that have more margin against failure is recommended, particularly (i) if stresses and temperatures are higher than prescribed limits, (ii) if significant boiler cycling is anticipated, or (iii) if there is significant uncertainty about operating conditions. Post-weld heat treatment (PWHT) has been found to have only a secondary effect on the lifetime of DMW, but as that effect is detrimental, PWHT should be avoided. If PWHT is used with stainless steel filler metal, limits similar to those in Figure 35-10 should be imposed. If used with nickel-base filler metals, care is required to avoid the formation of Type I carbides.

760

1400

740 1350 1300

700

Decarburization

680

1250 1200

660

No decarburization

640 620

1150 1100 0.2

Temperature (°C)

Temperature (°F)

720

600 0.5

1.0

5.0 Time (hr)

10.0

50.0

Figure 35-10. Permissible time/temperature conditions for post-weld heat treatment of dissimilar metal welds joining 21/4 Cr -1 Mo to 300-series stainless steel filler metal. Source: D.I. Roberts, et al.2h

35-14

Dissimilar Metal Weld Failures

6. Background to Long-Term Actions and Prevention of Repeat Failures 6.1 Overview of long-term actions A long-term program to ensure the integrity of DMWs should be a part of every BTF Team agenda. The use of a damage assessment code such as PODIS can be used to predict DMW lifetimes, to identify where new or replacement welds are required, and to anticipate where future actions may be needed. Long-term actions may include steps such as relocating DMWs, e.g., placing them away from fixed supports or in locations subjected to lower temperatures. Where there is uncertainty about the temperatures that are being experienced, prudent practice is to substitute with higher grade material. If a total replacement of the SH or RH is required, the optimum DMW locations, weld metal composition, and weld configurations should be “designed-in”. A program of periodic inspection of hangers, supports and spacers to ensure that secondary loads do not develop is also indicated. Periodic checking of unit temperatures with either direct measurement by thermocouples or by measuring oxide scale buildup is also recommended.

6.2 Damage assessment codes As noted previously, there are several codes available to (i) predict the level of damage that may be anticipated in a particular DMW, (ii) to

highlight potential failure locations, and (iii) to evaluate options for corrective actions. The features of one such code, developed by EPRI and designated Prediction of Damage in Service (PODIS)2g are discussed next as a means to introduce the nature of the calculations that are typically performed for such assessments. The PODIS code is an empirical method based on the field experience and accelerated test results. Life estimates are calculated from a knowledge of, or estimate of, the loading history to which the tubing was and will be subjected. Loading history is defined in terms of time, weld metal temperature, weld metal temperature change, number of cycles of temperature change, axial stress at the weld due to pressure, deadweight, and restrained thermal expansion loads within the tube assembly. Creep damage is imposed by three types of stresses: • Intrinsic damage (self damage). • Damage caused by primary (load-controlled) components of the system loading on the weld, including axial pressure loads and dead weight. • Damage caused by secondary (strain-controlled) components of the system loading on the weld such as those from restricted thermal expansion.

Volume 3: Steam-Touched Tubes

35-15

Any method to evaluate the accumulation of damage must provide for estimates of these stresses so that total damage can be estimated. A number of steps are typically required to accumulate and analyze the needed information including2g: • Plant inspection to determine tube condition and contributing factors to DMW failures, such as support conditions, slag buildup, misalignment, and tube warpage. • Review operating history to obtain background information necessary to estimate tube temperature levels and cycling. • Analyze system load histories at DMWs. • Assessment of actual DMW damage. See section on determining the extent of damage above. • Predict damage from a code such as PODIS.

35-16

Dissimilar Metal Weld Failures

• Correlate actual damage measurement with that predicted from PODIS to determine whether it is likely that all significant effects have been incorporated into the analysis. • Predict residual life based on projection of future unit operations. Some knowledge is required of the conservatism in each code. For example, the availability of field failure experience to calibrate the empirical constants in the PODIS model have led to a conservative treatment, particularly for Ni-based welds. The degree of conservatism has been explored in research funded by Empire State Electric Energy Research Corporation (ESEERCO).10,11 As a result, a program called DMW LIFE was developed which removes some of the conservatism in the PODIS analysis of Ni-based welds via an analysis of industry experience.

Further, the ability to analyze fireside corrosion and/or long-term overheating may be necessary to analyze properly the remaining tube life. For example, PODIS does not take account of the effects of fireside corrosion which will reduce tube wall thickness and thus increase stress levels. Modifications to PODIS have used data from the literature and limited field corrosion test results to include corrosion factors in the DMW LIFE code. Finally, neither the effects of long-term overheating/ creep in tubing nor oxide notch propagation are considered in PODIS.

7. Case Study Dissimilar Metal Weld Failures Case Study: Field Experience and Integrated Analysis Unit Description. The unit is a base-loaded, 350 MW, coal-fired unit put on line in 1970 with 165,000 operating hours. Superheater outlet design conditions are 2,500 psig at 540°C (1005°F); reheater outlet design conditions are 594 psig at 540°C (1005°F). Weld Locations. The unit contains approximately 1,675 DMWs located in the division panels, superheater platen, reheater front pendant, and superheater front pendant. Original welds were made with stainless steel filler metal. Although no failures had yet occurred, a similar unit had experienced DMW failures so that a program of analysis and preventive measures was considered prudent. Four regions were evaluated: SH penthouse, SH furnace, RH penthouse, and RH furnace. NDE and Metallography. Ultrasonic testing was used to determine oxide scale thickness in three of the four regions, the exception was the SH furnace tubes owing to access constraints. Every third tube containing a DMW in the fer-

ritic material was measured close to the DMW so as to provide an indication of joint temperatures, but removed from the weld geometry and associated discontinuities. Radiographic testing by the Union Electric Technique8 was performed on 65% (1,081) of the unit’s DMWs. The method had previously been validated by comparison with removed samples. Twenty-five percent of the DMWs radiographed indicated interfacial damage ranging from 5 to 90%. The majority of the damage was found in the RH furnace. Metallographic samples were taken of selected DMWs with extensive damage to confirm the mechanism and further confirm the accuracy of the RT. Temperature Estimates. Oxide thickness measurements and subsequent analysis of the results were used in conjunction with available thermocouple data to estimate tube temperatures. Boiler-wide temperature estimates were made. Stress Analysis. A piping stress analysis code was used to estimate primary (dead weight) and

secondary (primarily restrained thermal expansion) stresses. Pressure stresses were calculated and added to the results. Results for SH tubes found a range in secondary stresses from 14 to 18,501 psi; for primary stresses the range was 1,522 to 3,016 psi. Damage Analysis. Current temperatures were estimated from the inspection results and stress levels. A damage assessment code was used to formulate a strategy for DMW actions. DMWs with more than 50% interfacial damage were repaired immediately. A total of 13 DMWs in the RH and 5 DMWs in the SH were repaired. Repairs were performed by grinding out the original weld metal as nearly completely as possible, grinding a 60° angle on the ferritic side, and welding with either Inconel 182 or Incoweld A filler metal. Plans for future replacements with a “dutchman” were formulated. Source: This case study is a summary of an evaluation first reported by H.J. Grunloh, R.H. Ryder, and R. Hellner.5

Volume 3: Steam-Touched Tubes

35-17

8. References 1Dooley,

R.B. and H.J. Westwood, Analysis and Prevention of Boiler Tube Failures, Report 83/237G-31, Canadian Electrical Association, Montreal, Quebec, November, 1983. 2Prager, M., et al., Dissimilar-Weld Failure Analysis and Development Program, Research Project 1874-1, Final Report CS-4252, 8 Volumes, Electric Power Research Institute, Palo Alto, CA.

a. Prager, M., Volume 1: Executive Summary, November, 1985. b. Roberts, D.I., C.C. Li, and R.D. Nicholson, Volume 2: Metallurgical Characteristics, November, 1985. c. Ryder, R.H., H.J. Grunloh, R.F. Stetson, K.J. Tong, K.H. Holko, D.I. Roberts, F.V. Ellis, M.P. Borden, S.N. Cato, and B.W. Roberts, Volume 3: Accelerated Discriminatory Tests, November, 1986. d. Holko, K.H., C.C. Li, R.H. Ryder, D.I. Roberts, and C.F. Dahms, Volume 4: Utility Plant Results; November, 1985. e. Roberts, D.I., H.J. Grunloh, and K.H. Holko, Volume 5: Evaluation of Acoustic Emission and Enhanced Radiography, November, 1985. f. Prager, M., H.J. Grunloh, J.R. Foulds, R.H. Ryder, C.F. Dahms, and M. Krishnan, Volume 6: Weld Condition and Remaining Life Assessment Manual, August, 1988 g. Ryder, R.H., C.F. Dahms, M. Krishman, H.J. Grunloh, and J.R. Foulds, Volume 7: Prediction of Damage in Service (PODIS) Code - Background Document, May, 1988. h. Roberts, D.I., R.H. Ryder, H.J. Gurnloh, and B.E. Thurgood, Volume 8: Design and Procedure Guide for Improved Welds, November, 1989. 3Roberts,

D.I., R.H. Ryder, and R. Viswanathan, “Performance of Dissimilar Welds in Service”, Journal of Pressure Vessel Technology, Volume 107, August, 1985. 4Li,

C.C., R. Viswanathan, and R.H. Ryder, “The Microstructure and Remaining Life of Dissimilar Metal Weldments After Service in Fossil-Fired Boilers”, Proceedings of the ASME International Conference on Advances in Life Prediction Methods, April 18-20, 1983, Albany, New York.

35-18

Dissimilar Metal Weld Failures

5Grunloh,

H.J., R.H. Ryder, and R. Hellner, “Damage Assessment and Predictive Maintenance of Dissimilar Metal Welds in Superheater and Reheater Tubes”, in B. Dooley, ed., Proceedings: International Conference on Boiler Tube Failures in Fossil Plants, held in San Diego, California November 5-7, 1991, Proceedings TR-100493, Electric Power Research Institute, Palo Alto, CA, April, 1992, pp. 7-51 through 7-76. 6Ryder,

R.H., D.I. Roberts, and R. Viswanathan, “Dissimilar Metal Weld Failure Reduction” in B. Dooley and D. Broske, eds., Boiler Tube Failures in Fossil Power Plants: Conference Proceedings, Conference held in Atlanta, Georgia, November 10-12, 1987, CS-5500-SR, Electric Power Research Institute, Palo Alto, CA, 1988, pp. 2-63 through 2-72. 7Paterson,

S.R., T.A. Kuntz, R.S. Moser, and H. Vaillancourt, Boiler Tube Failure Metallurgical Guide, Volume 1: Technical Report, Volume 2: Appendices, Research Project 1890-09, Final Report TR-102433, Electric Power Research Institute, Palo Alto, CA, October, 1993. 8Gurnea,

R.F, “Radiographic Technique for Detecting Cracks in Dissimilar Weld Joints”, in R. Viswanathan and D.A. Roberts, eds., Proceedings: Seminar on Dissimilar Welds in Fossil-Fired Boilers, held in New Orleans, LA., February 23-24, 1984, Research Project 1874-1, Proceedings CS-3623, Electric Power Research Institute, Palo Alto, CA, July, 1985, pp. 4-47 through 4-60. 9Personal

Communication from S. Paterson (Aptech Engineering Services, Inc.) to R.B. Dooley, February, 1995. 10Sherlock,

T.P. and D.N. French, Dissimilar Metal Weld Assessment Program, Final Report, ESEERCO Project 91-20, Empire State Electric Energy Research Corporation, New York, NY, December, 1991. 11Sherlock,

T.P. and D.M. Conklin, Dissimilar Metal Weld Assessment Program, Final Report, ESEERCO Project 91-20, Phase II, Empire State Electric Energy Research Corporation, New York, NY, May, 1994.

ACTIONS for Dissimilar Metal Weld Failures Two paths for the BTF team to take in the investigation of DMW failures begin here. The goal of these actions is to see if further investigation is warranted or whether another BTF mechanism should be investigated.

➠ Follow Action 1a: If a SH/RH BTF has occurred and DMW failure is the likely mechanism.

➠ Follow Action 1b: If a precursor has occurred in the unit that could lead to future BTF in DMWs.

Action 1a: If a SH/RH BTF has occurred and DMW is the likely mechanism.

➠ Determine whether the failure is adjacent to a weld between dissimilar metals.

➠ Confirm that the macroscopic appearance of the failure includes such features as: • Circumferential cracking adjacent to the heat affected zone on the low- alloy (ferritic) side of the joint. • Thick-edged fracture with other signs of low ductility.

Action 1b: If a precursor has occurred in the unit that could lead to future SH/RH BTF in DMWs:

➠ Determine if there has been a change to more unit cycling.

➠ Determine whether a life extension of the unit is anticipated which would require an analysis of the effect of DMW on the desired unit lifetime.

➠ Determine if oxide dating of ferritic materials has indicated higher than anticipated operating temperatures.

➠ If the BTF seems to be consistent

➠ Determine if there have been indi-

with these features of failure, go to Action 2 for further steps to confirm the mechanism.

➠ Determine whether one or more of

➠ If the BTF does not seem to have features like those listed, return to the screening Table for steamtouched tubing (Table 31-1) to pick a more likely candidate.

cations of wide variations in tube temperatures. the following precursors has been found or is likely to have occurred in the unit: • Evidence of tube bowing or other signs of excessive tube distortion. • Evidence of tube rubbing. • Repetitively failed supports. • Slag buildup and hot spots induced by resulting hot gas flow. • Slag buildup at movable supports. • Slag or ash buildup beneath vertical platens. • DMWs located close to fixed supports, such as furnace walls and roofs.

➠ Determine whether there have been the addition of supports without consideration of their impact on stresses in DMWs.

➠ If one or more has occurred, go to Action 3 which outlines the steps to confirm the influence of each.

Volume 3: Steam-Touched Tubes

35-19

Action 2: Determine (confirm) that the mechanism is a DMW failure. A failure has occurred which the BTF team has tentatively identified as being specifically related to a DMW (Action 1a). Action 2 should clearly identify this as the primary mechanism or point to another cause. The actions listed will be executed by removing representative tube sample(s), followed by visual examination and detailed metallographic analysis.

➠ Analyze the macroscopic damage. Does damage have features similar to those listed in Table 35-2 indicative of DMW failures, such as circumferential cracking near the ferritic material’s HAZ, located near a DMW, thick-edged cracks, and/or the presence of an “oxide notch”?

➠ Analyze microstructure. Is there evidence of creep damage and cracking at a distance of about 1-2 grain diameters away from the fusion line along prior austenite boundaries for austenitic weld metals? Is there evidence of Type I carbides and cracking associated with creep cavitation immediately adjacent to the weld fusion line for Nibased welds?

Probable mechanism is related to the dissimilar metal weld.

➠ Go to Action 3: Root Cause Determination

35-20

Dissimilar Metal Weld Failures

A crack that is oriented axially (longitudinal to the tube) is more likely to be caused by overheating (see Chapter 32). Excessive wall thinning which would likely lead to a ductile, thin-edged frac