V 2 G [PDF]

Zitiervorschau

1

Autonomous Distributed V2G (Vehicle-to-Grid) considering Charging Request and Battery Condition Y. Ota, Member, IEEE, H. Taniguchi, Member, IEEE, T. Nakajima, Member, IEEE K. M. Liyanage, Senior Member, IEEE, J. Baba, Member, IEEE, and A. Yokoyama, Member, IEEE

Abstract—Integration of large scale renewable energy sources into power grid, battery energy storage performs an important role for smoothing their natural intermittency and ensuring gridwide frequency stability. Plug-in hybrid electric vehicle and electric vehicle have potential of alternative of the battery because of its high performance lithium-ion battery and longer plug-in time than driving time. Therefore, vehicle-to-grid is expected to be one of the key technologies for smart grids integrating renewable energy sources. In this paper, an autonomous distributed vehicle-to-grid control scheme is proposed. Grid-connected electric vehicles contribute frequency regulation and spinning reserve triggered by self-terminal frequency, which is a signal of supply and demand balance in the power grid. Proposed scheme also consider charging request for the next drive and battery condition during the vehicle-to-grid. Satisfaction of vehicle user convenience and effect load frequency control is evaluated through coupled analysis of vehicle-to-grid model and typical power grid model. Index Terms—Smart Grid, (Plug-in Hybrid) Electric Vehicle, Vehicle-to-Grid, Autonomous Distributed Control, Battery Stateof-Charge, Load Frequency Control.

I

electric vehicle (PHEV) and electric vehicle (EV) are expected toward the future low carbon energy system. Therefore, there is a large potential of vehicle-to-grid (V2G) [5]-[6] of grid-connected (PH)EVs. When (PH)EVs are aggregated to MW class, the aggregated V2G pool contributes the LFC by information exchange to load dispatching center or ISO (Independent System Operator) [7]-[9]. In this paper, an autonomous distributed V2G control scheme considering charging request, battery condition, and also contribution to the power grid is proposed. Self terminal frequency droop control realizes fast response of vehicle battery with synchronization among multiple vehicles. Battery state-of-charge (SOC) is managed by using SOC balance control, and then the V2G control is switched to one-way charging control, smart charging or V1G, satisfying charging request preset by vehicle user. Proposed V2G control scheme is explained in chapter II, then verified by a coupled analysis using the mathematical model of automotive lithium-ion battery, typical two area interconnected power grid model, and proposed V2G control model, in chapter III.

I. INTRODUCTION

ntermittent renewable energy sources require another dispatching resource such as thermal power generation, adjustable speed pumped storage, battery energy storage, and so on. Recent smart grid technology points out the ability of distributed generations and controllable loads in the demand side. Authors propose ubiquitous power grid concept [1] as shown in Fig. 1, and the blade pitch angle control of wind power generation [2], the heat pump water heater [3], and the battery energy storage [4] are integrated into the power grid load frequency control (LFC). Because of advanced lithium-ion battery and development of charging infrastructure, large amount of plug-in hybrid This work was supported by Grant-in-Aid for Scientific Research (KAKENHI) (B) (22360122), Specially Promoted Research Grant from Power Academy of Japan, Research Grant from TEPCO Research Foundation, and Ubiquitous Power Grid Endowed Chair of the Center for Advanced Power & Environmental Technology (APET) of the University of Tokyo. Y. Ota, H. Taniguchi, and T. Nakajima are with the University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo, 113-8656, Japan (e-mail: [email protected], [email protected], [email protected]). K. M. Liyanage is with University of Peradeniya, Peradeniya 20400, Sri Lanka (e-mail: [email protected]). J. Baba and A. Yokoyama are with the University of Tokyo, 5-1-5, Kashiwanoha, Kashiwa, Chiba, 277-8568, Japan (e-mail: [email protected], [email protected]).

II. VEHICLE-TO-GRID CONTROL SCHEME Supply and demand imbalance of the power grid can be observed from frequency deviation at home outlet [10]-[12]. V2G power output is controlled as droop characteristics against the frequency deviation shown in Fig. 2. V2G gain (Kmax) is adjusted considering a tradeoff between the effect of V2G and the fluctuation range of SOC according to additional charge-discharge cycles. Maximum V2G power (Pmax) is decided by specifications of the home outlet. When the battery SOC is near to full (empty), high-power charge (discharge) should not be required for preventing overcharge (over-discharge). During long-term V2G cycles, it is concerned that the battery SOC gradually shifts from its nominal value. Considering these battery features, SOC balance control is designed as following equation on the premise that accurate SOC estimation is realized. K V2G

⎧ ⎛ SOC − SOC low(high) ⎪ = K max ⎨1 − ⎜ ⎜ SOC max(min) − SOC low(high) ⎪⎩ ⎝

⎞ ⎟ ⎟ ⎠

n

⎫ ⎪ ⎬ ⎪⎭

(1)

Where, SOCmin, SOClow, SOChigh, SOCmax, and n are parameters. Fig. 3 shows a sample case (SOCmin, low, high, max = 10, 20, 80, 90[%], n=2). When battery SOC is away from 50[%], V2G power is suppressed to return to 50[%].

2 Hydro

Nuclear

Thermal

Battery

Ubiquitous Power Grid

Power Grid Wind

Power Grid Load Dispatching Center Regional Load Dispaching Center

ECU / BMU

Battery

DC-DC Converter

Smart Charging Vehicle-to-Grid (V2G) Inverter Motor Battery

Plug-in Hybrid Electric Vehicle Electric Vehicle Conditioning Battery Power Grid Contribution to LFC

Electric Vehicle Charging request

Wind Photovoltaic

Small Power Grid Microgrid

Load Heat Pump Heat Storage

Distributed Generator

Load

Controllable Load

Gas-engine Gas-turbine

H2 Local Network

Reversible Fuel Cell

Fig. 1 Autonomous distributed V2G in ubiquitous power grid.

Finally, V2G power (PV2G) is designed according to self terminal frequency deviation (Δf) and own SOC as follows. (2)

(3)

Necessary energy from present SOC to destination SOC (E) is calculated by using battery model as equation (A-4) in Appendix. On the other hands, we can estimate the expectation value of V1G charging (PV1G*) by using past tendency of the monitored frequency deviation. Therefore, the estimated duration of the V1G charging (TV1G*) is estimated as follows. E * (4) T = V1G

PV1G

*

III. SIMULATION USING V2G AND POWER GRID MODEL A. Power Grid Model Fig. 4 shows a two area interconnected power grid model, and simulation parameters are summarized in Table.1. The transfer function model expresses deviations from nominal nighttime conditions in Japanese 50Hz power systems [13]. Inertia constant is aggregated as one thermal power generator with turbine and governor dynamics. Power grid damping consists of load damping and droop characteristics of governor-free control of the aggregated thermal power generator. Synchronizing power coefficient is adjusted as recreating the oscillation period 2-3[s].

Charge

K max

V1G 0

Discharge

PV1G

( = f (Δ f )) 　　 ( 0 < K max Δ f ≤ Pmax ) ⎧ K max Δ f 　　 ⎪ = ⎨ Pmax 　　　　　　　　( Pmax < K max Δ f ) ⎪ 0 　　　　　　　　　( K Δ f < 0 ) max ⎩

V2G power [kW]

When it remain duration of the V1G charging to destination SOC for plug-out preset by vehicle user, V2G control is switched to V1G control, which is one-way charging with droop characteristics against frequency deviation as follows.

P max

Δf max

V2G

Frequency Deviation [Hz] Fig. 2 V2G control against frequency deviation.

V1G Charge

K max V2G gain [kW/Hz]

⎧⎪ K V2G Δ f　　 ( = f (Δ f , SOC )) 　　 ( K V2G Δ f ≤ Pmax ) PV2G = ⎨ ⎪⎩ Pmax 　　　　　　　　　　 ( Pmax < K V2G Δ f )

V2G Discharge

V2G Charge SOClow 0

10

SOCmin

20

30

SOChigh 40

50

60

Battery SOC [%]

Fig. 3 Battery SOC balance control .

70

80

90 100

SOCmax

3

Regarding as the thermal power generator, ±5 [%] of its rated output is reserved for governor-free operation, and ±1.5 [%] of load capacity is reserved for LFC. Dispatching center allocates AR (Area Requirement) calculated by frequency and tie-line power flow to each thermal power generator. Both the frequency and the tie-line power flow are maintained by FFC (Flat Frequency Control) for grid-A and TBC (Tie-line Bias Control) for grid-B. Delay of frequency detection and AR calculation is modeled as a first-order lag. And communication delay from the dispatching center is assumed as a dead time. Outputs of renewable energy sources (RES) and load demands are fluctuated according to normal distributions, respectively. Frequency bands of the fluctuation are limited by low pass filter (LPF). B. V2G Model An automotive lithium-ion battery pack is assumed by the simplified battery model as shown in Appendix. Specifications of the battery cell and pack, which is consisted by 88 series of cells, are summarized in Table I. Parameters of the V2G control are summarized in Table II. Maximum V2G power (Pmax) is 5[kW] based on the assumption of 200[V]/25[A] home outlet, and maximum V2G gain (Kmax) is 200[kW/Hz], that is, maximum V2G power 5[kW] is supplied when frequency deviation is 0.025[Hz]. SOC balance control is the same condition in Fig. 3. The duration of the V1G charging (TV1G*) is estimated by using the monitored frequency deviation during past 4 hours. And 1 hour margin is considered because there is uncertainty such as current dependent loss by the internal resistance in calculation the necessary energy (WV1G) to the destination SOC. EVs are plugged into the grid-A at 3 hours with initial SOC as 30[%]. Then they are plugged out all together with destination SOC as 90[%] after 12 hours. This scenario would be severest condition for the power grid. On the other hands, in the grid-B, there is V2G pool having initial SOC balanced condition and not plugged-out. Number of EVs with V2G is forty thousand in the grid-A and ten thousand in the grid-B. Total V2G ability is 250[MW]. C. Simulation Results Fig. 5 shows results of coupled analysis using the proposed V2G model and the power grid model. There is no V2G control at first 4 hours. Frequency fluctuations are appeared with the effect of the supply and demand imbalance by RES and load fluctuations. From 4 to shortly before 9 hour, fluctuations are compensated by the V2G control in both grids. Battery SOC in grid-A is gradually lifted up by the SOC balance control. That in grid-B is maintained around 50[%] as supplying V2G charge and discharge cycles. Then the V2G control in grid-A is once switched to V1G charging, and V2G/V1G mode is chattered until 12 hour. Finally, the battery SOC in grid-A is arrived to the destination SOC (90[%]) by the V1G charging during last 4 hours. Ability of compensating the frequency fluctuation obviously decreases when the frequency deviation is staying minus value, because there is no V2G discharging in grid-A.

Δfa

ΔPV2Ga

Load_a

Grid-A

V2G_a

Dispatching Center

RES_a

Δfa ARa

ΔPtha

ARa

Inertia_a

Thermal_a

FFC

ΔPt Tie-line ΔPt ARb Δfb

Δfa

1 Ma.s

Δfa

Kab s

ARb ΔPthb

Δfb

TBC

Δfb

Thermal_b

1 Mb.s

RES_b

Inertia_b

Δfb

ΔPV2Gb

V2G_b

Load_b

Grid-B

Fig. 4 Power grid model. TABLE I PARAMETERS OF POWER GRID MODEL Parameters Load capacity [MW] Thermal power rated output [MW] Inertia constant [s]

Grid-A

Load damping coefficient [puMW/puHz] Tie-line synchronizing power coefficient Speed variation of GF [puHz/puMW] GF capacity [MW]

7090 5560 9.02

2

2 14

0.05 ±1213

Grid constant for AR calc. [puMW/puHz] Proportional and integral gain of LFC LFC capacity [MW]

Grid-B

33090 24252 8.58

0.05 ±278

5

5

1, 0.1 ±496

1, 0.1 ±106

Noise power of RES [MW] Time constant of LPF for RES fluctuation [s] Noise power of load [MW] Time constant of LPF for load fluctuation [s]

1212.6

Time constant for frequency detection [s] Time constant for AR calculation [s] Communication delay from control center [s]

0.1 4 1

278.0 300

109.1

50.5 30 0.1 4 1

TABLE II PARAMETERS OF BATTERY CELL (PACK) AND V2G CONTROL Parameters

Grid-A

Grid-B

Nominal voltage (Vnom) [V] Nominal capacity (Cnom) [Ah] Energy capacity [kWh]

3.7 (325.6) 50 (50) 0.185 (16.28)

Parameter of OCV-SOC (α) Efficiency of charging and discharging cycle (η) Internal resistance (Rint) [Ohm]

15 1.0 0.004 (0.352)

Maximum V2G power (Pmax) [kW] Maximum V2G gain (Kmax) [kW/Hz] Parameters for SOC balance control : n SOCmin, SOClow, SOChigh, SOCmax [%]

5 200 2 10, 20, 80, 90

Initial and destination SOC [%] Duration to plug-out [h] Duration for estimating V1G charging (TV1G*) [h]

30, 90 12 Past 4

No plug-out

50, 50

Number of V2G vehicles

40000

10000

No plug-out

4

(a) Frequency deviation of grid-A (Δfa)

Time [hours]

(b) Frequency deviation of grid-B (Δfb)

Time [hours]

(c) Tie-line power flow deviation (ΔPt)

Time [hours]

RES Load

V2G

Thermal

(d) Power outputs and load demands of grid-A RES

Load

V2G

Thermal

Time [hours]

(e) Power outputs and load demands of grid-B

Time [hours]

(e) V2G power and battery SOC of grid-A

Time [hours]

(f) V2G power and battery SOC of grid-B

Time [hours]

Fig. 5 Results of coupled analysis of V2G model and power grid model.

5 TABLE III QUALITY OF FREQUENCY DEVIATION IN GRID-A 1-4 hour (no V2G) 0.1311 0.0244 -0.1344 0.0225

Max [Hz] RMS+ [Hz] Min [Hz] RMS-[Hz]

4-8 hour (V2G) 0.1311 0.0145 -0.0955 0.0132

8-12 hours (V2G/V1G) 0.1186 0.0170 -0.1215 0.0119

12-16 hour (V1G) 0.1161 0.0130 -0.1779 0.0214

Table III summarizes maximum values, minimum values, RMS+, that are RMS (Root Mean Square) values when the frequency deviation is staying plus value, and RMS-, that are RMS values when the frequency deviation is staying minus value. Their values are separated in every 4 hours, and each period show almost no V2G, V2G, V2G/V1G and V1G control mode, respectively. Compensation effects are numerically confirmed especially in the RMS values with same tendency shown in Fig. 5. In this paper, total capacity of V2G 250[MW] is not sufficient against the RES fluctuation beyond 1000[MW]. Therefore, the maximum value of the frequency deviation during 4-8 hour with V2G is same as that during 1-4 hour without V2G.

Necessary energy (E) from present SOC (SOCi) to destination SOC (SOCd) is calculated by integration of OCV as following equation. SOC d

E =

= V nom (SOC d 　 ⎛

　　 + α SOC d ln ⎜ ⎜

nom

F

− SOC d

⎞ ⎟ − α SOC i ⎟ ⎠

⎛ SOC i ⎜C ⎝ nom − SOC i

ln ⎜

⎞ ⎟⎟ ⎠

VI. REFERENCES [1] [2]

[3]

[4]

[5]

[6]

[7]

[8]

[9] [10]

[11]

⎜C ⎟ ⎝ nom − SOC ⎠

Where, Vnom and Cnom are nominal voltage and capacity, respectively. R, F, T is gas constant, faraday constant, and battery temperature, respectively. α is a parameter about voltage change. Battery CCV (Closed Circuit Voltage) is also calculated considering internal resistance (Rint) as follows. (A-3) CCV = OCV + R int I

SOC d

⎝ C nom

V. APPENDIX Firstly, battery SOC is updated by integrating chargedischarge current (I) from initial SOC. d SOC (A-1) = ηI dt Where, η is charge-discharge efficiency. Then battery OCV (Open Circuit Voltage) is defined as following Nernst equation. ⎞ SOC RT ⎛ (A-2) ⎟ OCV = V +α ln ⎜

)

− SOC i + α C nom

(A-4)

⎛ C nom − SOC d ⎞ ln ⎜ ⎜ C ⎟⎟ ⎝ nom − SOC i ⎠

In charging, there is additional current dependent loss caused by the charge efficiency and the internal resistance.

IV. CONCLUSION The proposed autonomous distributed V2G control scheme satisfied battery SOC conditioning, severe charging request, and also additional charge and discharge against the RES and load fluctuations. It is necessary to implement the proposed V2G control to the automotive power electronics circuits, and evaluate impact on actual automotive batteries. In the future low carbon energy systems under large-scale integration of intermittent renewable energy sources and electric vehicles, the proposed scheme is expected to maintain quality of frequency as a distributed spinning reserve without own SOC fluctuation, and without interfering the conventional load frequency control. As an additional control, dispatching LFC signals to electric vehicles could be effective when those signals are properly coordinated with thermal power generator, pump storage, and stationary battery energy storage.

∫ OCV d SOC

SOC i

[12]

[13]

Center for Advanced Power & Environmental Technology (APET) of the University of Tokyo [Online]. Available: http://www.apet.t.utokyo.ac.jp Y. Nishizaki, H. Irie, A. Yokoyama, and Y. Tada, “Coordinated Control of Blade Pitch Angle of Wind Turbine Generators and Battery for Frequency Regulation and the Battery Capacity Reduction,” in Proc. 2008 International Conference on Electrical Engineering (ICEE), No.O052. H. Irie and A. Yokoyama, “Modeling for Frequency Control Analysis of Power System with a Large Penetration of Wind Power Generation by a lot of Controllable Heat Pump Systems and Battery Systems,” in Proc. 2008 International Conference on Power System Technology (POWERCON), No.0455. K. M. Liyanage, A. Yokoyama, Y. Ota, H. Taniguchi, and T. Nakajima, “Coordinated Control of Elements in Ubiquitous Power Networks to Support Load Frequency Control”, in Proc. 2009 International Conference on Industrial and Information Systems (ICIIS). W. Kempton, V. Udo, K. Huber, K. Komara, S. Letendre, S. Baker, D. Brunner, and N. Pearre, “A Test of Vehicle-to-Grid (V2G) for Energy Storage and Frequency Regulation in the PJM System”, Publications of MAGICC (Mid-Atlantic Grid Interface Cars Consortium) [Online]. Available: http://www.magicconsortium.org/_Media/test-v2g-in-pjmjan09.pdf , Jan. 2009. “Smart Garage Charrette Report”, Rocky Mountain Institute [Online]. Available: http://move.rmi.org/files/smartgarage/SmartGarageCharretteReport_2.10 .pdf, Dec. 2008. K. Shimizu, T. Masuta, Y. Ota, and A. Yokoyama, “Load Frequency Control in Power System Using Vehicle-to-Grid System Considering the Customer Convenience of Electric Vehicles”, in Proc. 2010 International Conference on Power System Technology (PowerCon) K. M. Liyanage, A. Yokoyama, Y. Ota, T. Nakajima, H. Taniguchi, “Impacts of Communication Delay on the Performance of a Control Scheme to Minimize Power Fluctuations Introduced by Renewable Generation under Varying V2G Vehicle Pool Size”, in Proc. 2010 IEEE SmartGridComm. A. Brooks, E. Lu, D. Reicher, C. Spirakis, and B. Weihl, “Demand Dispatch”, IEEE Power & Engineering Magazine, Vol.8, Issue.3, pp.2029, May. 2010. Z. Zhong, C. Xu, B. J. Billian, L. Zhan, S-J. Steven Tsai, R. W. Conners, V. A. Centen, A. G. Phadke, and Y. Liu, “Power System Frequency Monitoring Network (FNET) Implementation”, IEEE Trans. Power Systems, Vol.20, No.4, pp.1914-1921, 2005. O. Samuelsson, M. Hemmingsson, A. H. Nielsen, K. O. H. Pedersen, and J. Rasmussen, “Monitoring of Power System Events at Transmission and Distribution Level”, IEEE Trans. Power Systems, Vol.21, No.2, pp.1007-1008, 2006. Y. Ota, T. Hashiguchi, H. Ukai, M. Sonoda, Y. Miwa, and A. Takeuchi, “Monitoring of Interconnected Power System Parameters using PMU based WAMS”, in Proc. 2007 IEEE PowerTech Conference, pp.17181722. “Japanese Power System Models”, The Institute of Electrical Engineers of Japan [Online]. Available: http://www2.iee.or.jp/ver2/pes/23st_model/english/index.html, 2007

6

VII. BIOGRAPHIES Yutaka Ota (M’04) was born in Nagano, Japan on August 19, 1975. He received the B.S., M.S., and Ph.D.Eng. degrees from Nagoya Institute of Technology, Japan, in 1998, 2000 and 2003, respectively. He is currently a Project Assistant Professor of Ubiquitous Power Grid Endowed Chair in the Center for Advanced Power and Environmental Technology (APET) of The University of Tokyo. His research interests include vehicle-to-grid technology, modeling of batteries, and application of phasor measurement unit based wide area measurement system to power system monitoring, protection, and control.

Jumpei Baba was born in Japan on June 3, 1973. He received the B.Eng., M.Eng., and Ph.D.Eng degrees from The University of Tokyo, Tokyo, Japan in 1996, 1998, and 2001, respectively. He has been with the Department of Electrical Engineering, Tokyo University of Science, since 2001, and with the Department of Advanced Energy, Graduate School of Frontier Sciences, The University of Tokyo, since 2003. He is currently an Associate Professor of Department of Advanced Energy, Graduate School of Frontier Sciences, The University of Tokyo.

Haruhito Taniguchi was born in Japan on August 23, 1950. He received the B.S, M.S., and Ph.D. degrees in Electrical Engineering from Kyoto University, Kyoto, Japan in 1973, 1975 and 1994, respectively. In 1975, he joined Central Research Institute of Electric Power Industry (CRIEPI). He was Director of Power System Department, Director of System Engineering Research Laboratory, CRIEPI. He is currently a Project Professor, Ubiquitous Power Grid Endowed Chair, Center for Advanced Power and Environmental Technology (APET), the University of Tokyo, since 2008. He has been engaged in research mainly on planning, operation and control of power systems as well as new technology development. He is a distinguished member of CIGRE.

Akihiko Yokoyama (M’78) was born in Osaka, Japan on October 9, 1956. He received the B.Eng., M.Eng., and Dr.Eng. degrees from The University of Tokyo, Tokyo, Japan, in1979, 1981, and 1984, respectively. He has been with the Department of Electrical Engineering, The University of Tokyo, since 1984 and is currently a Professor in charge of Power System Engineering. He was a Visiting Research Fellow at the University of Texas, Arlington, and the University of California, Berkeley, during the period of February 1987 to February 1989.

Tatsuhito Nakajima (M’87) was born in Tokyo, Japan on December 13, 1962. He received the B.S., M.S., and Dr.Eng. degrees from the University of Tokyo in 1985, 1987, and 1990, respectively. He joined Tokyo Electric Power Company (TEPCO) in 1990. He has been with Power Engineering R&D Center of TEPCO. He is currently a Project Associate Professor in the Center for Advanced Power and Environmental Technology (APET) of the University of Tokyo. His research interests include application of power electronics for power systems. Kithsiri M. Liyanage (M’93, SM’10) was born in Sri Lanka on July 10, 1961. He obtained B.Sc.Eng (Hons), M.Eng. and Dr.Eng. all in Electrical Engineering from University of Peradeniya in 1983 and from University of Tokyo in 1988 and 1991 respectively. He has held positions at the University of Tokyo, Japan, University of Washington, USA and Universities of Ruhuna and Peradeniya in Sri Lanka. From September 2008 to August 2010 he was with the Center for Advanced Power and Environmental Technology (APET) of the University of Tokyo as a Visiting Research Fellow on sabbatical leave from University of Peradeniya where he is a Senior Lecturer currently.