«Fuel cell and ultracapacitor hybrid power system»

Transcription

1 EMR 12 Madrid June 2012 Joint Summer School EMR 12 Energetic Macroscopic Representation «Fuel cell and ultracapacitor hybrid power system» Dr. Olivier BETHOUX LGEP, University Paris 11, MEGEVH network,

2 Outline 1. Context Environmental and social constraints Fuel cell behaviour Ultracapacitor behaviour Power architecture 2. EMR of FCUC hybrid power system 2 3. Inversionbased Control of FCUC hybrid power system 4. Securing the system 5. Conclusion and perspectives

3 EMR 12 Madrid June 2012 Joint Summer School EMR 12 Energetic Macroscopic Representation «Context»

4 World primary energy demand Gtoe environmental and social constraints Oil 4 100% 2009 Coal Gas Biomass (wood) Nuclear Renewable (solar, wind, ) 80% (2009) Hydrocarbon sources Source : CO 2 Capture Project 2009 Time Social acceptance : Evolution of emission standards for pollutant in the European Union Time Source : ADEME

5 Hydrogen : a convenient energy carrier Hydrogen combined with FC provides electric power without local pollution Energy density [kj.kg 1 ] hydrogen : kj.kg 1 oil : kj.kg 1 coal : kj.kg 1 lithium battery : 540 kj.kg 1 leadacid battery: 50 kj.kg 1 H 2 O 2 Fuel Cell 5 electricity water heat Fuel Cell : electrochemical power converter H 2 Anode : H 2 2 H 2 e Acid Polymer Electrolyte Electric Load electrons 0 2 or air Cathode : ½ O 2 2 e 2 H H 2 O I electrons

6 Cathode canal Anode canal «Title of the presentation» FC system Compressor Gas supply humidifier Fuel cell behaviour H 2 tank Fuel Cell limitations Influence of current frequency Current frequency Air flow rate Time delay magnitude 6 Gas outlet V cell E Th rj Aln j j0 mexp( nj) E th and A are pressure dependent Main Fuel Cell limitations Fast load variations Stop/Start conditions Environment conditions

7 Electric Vehicle : a Fuel cell application Weight and Energy Storage for 500 km Range 7 Fuel Cell / Battery EV market

8 Power Puissance [kw] «Title of the presentation» A Fuel cell application Electric vehicle : the power train supply is a demanding application 8 PEMFC power train demand (ECE15 EU cycle) 10 5 P Max P Av 0 5 Power assistance Temps [s] Time [s] Ultracapacitor Battery PSA

9 voltage v UC [V] current i UC [A] «Title of the presentation» Ultracapacitor behaviour Porous electrodes made of activated carbon C anions cations Liquid electrolyte A C d (Hermann Von Helmholtz in 1853) Capacitive property of the interface between the electronic solid conductor and the ionic liquid conductor C with C Electrostatic effect : C C. C C A 3000 m 2.g 1 < 1 nm high specific power (6 kw.kg 1 or more) high cycle efficiency (95% or more) high number of chargedischarge cycles UC system dynamic response : current i UC and v UC (C = 26F & R = 32mOhms) simulation measured V UC 9 C UC I UC ESR

10 The different possible architectures to combine both sources 10 DC PàC PàC DC SC SC Charge électrique Electric load Charge électrique Electric load [GRC2007] M. García Arregui and al, Clean Electrical Power Conference, Capri,2007. [Blun2009] B. Blunier and al, VPPC DC PàC PàC DC SC DC DC Charge électrique Electric load SC DC DC Charge électrique Electric load [AZIB2010] T.AZIB and al, IEEE Transactions on Industrial Electronics, Vol. 57, Issue.12, 2010 [THOU2008] P. THOUNTHONG and al, IEEE Transactions on Industrial Electronics, Vol. 54, N. 6, Dec

11 EMR 12 Madrid June 2012 Joint Summer School EMR 12 Energetic Macroscopic Representation «EMR of FCUC hybrid power system»

12 General scheme of the system and its associated control structure Acquisition of measurements energy exchanges 12 State State State DC demand requested energy power Low level control DC interpretation power management distribution control Signal DC Unit of control and energy management DC Energetic Macroscopic Representation (EMR) systemic approach, energy exchanges, decomposition of complex system Control structure directly deduced from a graphical symmetry of the EMR model (inversion based control) EMR Modeling Command Structure Strategy

13 EMR of FCUC hybrid power system two converters architecture i Cbus i Load 13 C Bus i Coupl Load i FC i FC DC/DC i FC i UC DC/DC i UC i UC UC FC V FC d PàC d SC V FC V UC V UC R 3 A 3 R A 1 A 1 A 1 R R 1 1 Source R 1 A m R 2 R 2 R 2 A 2 A 2 A 2 Conversion without Conversion storage Power Source with Parallel storage coupling EMR modeling DC Bus Parallel Coupling Electric Load Load i Cbus i Load Fuel Cell Inductance Converter i Coupl Converter Inductance Ultracapacitor FC V FC i FC i FC i UC i UC V UC UC V UC i FC V FC V UC i UC d FC d UC

14 EMR of FCUC hybrid power system two converters architecture i Cbus i Load 14 C Bus i Coupl Load i FC i FC DC/DC i FC i UC DC/DC i UC i UC UC FC V FC d PàC d SC V FC V UC V UC energy flows EMR modeling DC Bus Parallel Coupling Electric Load Load i Cbus i Load Fuel Cell Inductor Converter i Coupl Converter Inductor Ultracapacitor FC V FC i FC i FC i UC i UC V UC UC V UC i FC V FC V UC i UC d FC d UC

15 Inversionbased control deduced from EMR of FCUC hybrid system DC Bus Parallel Coupling Electric Load 15 Load i Cbus Fuel Cell Inductor Converter i Coupl Converter i Load Inductor Ultracapacitor Model FC V FC i FC i FC i UC i UC V UC UC V UC i FC V FC V UC i UC d FC d UC Low level control V FCref V UCref i UCref i FCref i FCref i UCref i UC Comp V UC_ref Control i Coupl_ref k Strategy State of charge management _ref Losses estimation i Cbus_ref Energy distribution i Load_ref demand interpretation

16 FC DC Bus «Title of the presentation» A basic strategy to split power Parallel Coupling LF MF HF i FC V FC V UC i UC UC Electric Load frequency decomposition FC UC BUS i Cbus i Load P Cutoff frequency i Coupl V Fuel Cell Inductor Converter Bus Converter Inductor choose Ultracapacitor quence de A few 100mHz V FC i FC i FC i UC i UC V UC P load (t) Strategy Load V UC Model 16 d FC d UC Low level control V FCref V UCref The weighting coefficient K depends on frequency so that the FC reacts slowly and the UC bank responds to the high transient. i UCref i FCref i FCref Lowpass filter i Coupl_ref i Coupl i Cbus_ref i SCref k 0 Strategy Max i UC Comp i FCref i CH_ref V UC_ref State of charge management Control _ref Losses estimation Energy distribution demand i interpretation FCref

17 Inversionbased control «Title of the presentation» Toward the control implementation i Cbus i Load 17 Implemented control scheme C Bus i Coupl Load Controllers parameters tuning Inversionbased control FC i FC i i FC i FC UC i UC i UC DC/DC DC/DC UC V FC V FC V UC V UC d PàC d SC d FC d UC Current loops d FC d PI PI UC (fsw/10) V FCref V UCref i FC i FCref i UC i UCref i UCref T1 T2 Compensation loop (f < 100 mhz) i FCref i FCref i Coupl_ref i SCref k Strategy i UC Comp V UC_ref i FCref Low passfilter i UCref i FC Comp PI V UCref V UC _ref i Cbus_ref i CH_ref Voltage loop (f sw /100) ref i Ccoupl_ref PI I LOAD Energy distribution

18 a c c P CH CH [W] Pload [W] i PàC iifc / i/ PàCref iifcref [A] Simulation results b d IUC [A] i SC i SC [A] i SC i SC / / i SCref i [A] IUC / IUCref [A] i PàC i [A] IFC [A] ref [V] e f e VBusref [V] f V SCref [V] VUCref [V] Bus [V] VBus [V] Temps [s] time [s] [s] V SC SC [V] VUC [V] Temps [s] time [s] [s] Currents and voltages trajectories are suitably controlled Constraints are respected and load specifications are satisfied.

19 Experiment setup 19 Bidirectionnal load H 2 sensor motors dissipators H 2 supply Choppers & coils drives dspace 1103 Ultracapacitor bank Active Load Fuel Cell

20 a a c c P CH P CH [W] [W] Pload [W] iifc PàC i PàC / IFCref i PàCref / i [A] PàCref [A] ref V [V] Busref [V] e VBusref [V] f V [V] Bus [V] VBus [V] Temps Temps [s] time [s] [s] Experiment results b b 40 d d e f 20 i PàC iifc PàC [A] [A] i SC IUC i[a] SC [A] iiuc SC i/ SC / iiucref SCref / i SCref [A] [A] V SCref VUCref V [V] SCref [V] V SC VUC V[V] SC [V] Temps Temps [s] time [s] [s] 20 Currents and voltages trajectories are suitably controlled constraints are respected and load specifications are satisfyied.

21 Flow rate [l/min] «Title of the presentation» Experiment results FC FC / UC Time [s] H 2_FC = l H 2_FCUC = l 10% saving First results show a better global efficiency of the hybrid system

22 EMR 12 Madrid June 2012 Joint Summer School EMR 12 Energetic Macroscopic Representation «Securing the system»

23 New scheme Many applications require high load voltage compared to FC voltage two stages 23 i Diss i DissSW i ChSW i CH Système de dissipation i Coupl3 i Cbus2 CHARGE C Bus2 2 icoupl2 Bus DC 2 Bus DC n 2 Etage «haute tension» interface élévatrice dispositif dissipatif DC/DC d 2 V Bus1 i Cbus1 i L2 CBus1 1 i Coupl Bus DC 1 Bus DC n 1 PàC i PàC i i i PàC i PàC SC i SC SC DC/DC DC/DC SCs V PàC V SC d PàC d SC V PàC V SC

24 Securing the system 24 P FC P Ch a Requested power is temporally too large PàC Load Strategy a Load power modulation P Load > P FC P UC SC P UC b Delivered power is temporally too large PàC P FC P Ch Load Stratégie b Auxiliary device modulation to dissipate overenergy P Load < P FC P UC SC P UC

25 Securing the system k 2 Strategy 25 Dissipation device V Bus2 d Diss 2 Parallel coupling 2 d CH V Bus2 Electric load SD Load i Diss i DissSW i ChSW i CH switch 2 i Coupl3 switch 2 Parallel coupling i Cbus2_ref 2 i Cbus2 2_ref Bus2 DC i Coupl 2 2 i Coupl2_ref Converter d 2 i L2 V Bus V Bus2_ref Inductance 1 2 i L2 i L2_ref

26 Securing the system P Load [W] i FC [A] i UC [A] V UCref [V] V UCma x 20 V UC [V] V UCmi V BUS [V] V BUSref [V] i Brake [A] Time [s] Time [s] [s] n

27 EMR 12 Madrid June 2012 Joint Summer School EMR 12 Energetic Macroscopic Representation «Conclusion and perspectives»

28 Conclusion : Conclusion and perspectives 28 Using REM concept allows a systemic study of the FC/UC hybrid system. * It leads to easy tuning control structure; * It reveals degree of freedom where to define strategy. Perspectives : Model level: Taking into account humidity and temperature may be important to drive more gently the FC. Control structure: As the airflow rate has a low time constant, directly driving the airflow rate may be very important to optimize the system. Control structure: Comparing different control techniques to implement local loops (passivity, sliding modes, etc) Strategy: Modify strategy to optimize fuel consumption,

29 EMR 12 Madrid June 2012 Joint Summer School EMR 12 Energetic Macroscopic Representation «BIOGRAPHIES AND REFERENCES»

30 Authors 30 Dr. Olivier BETHOUX University Paris 11, LGEP, MEGEVH, France Assistant Prof. at IUT de Cachan, University Paris 11 (2006) PhD in Electrical Engineering at University of Cergy (2005) Research topics: electrochemical systems, energy efficiency, energy flexibility

31 References 31 [Azib 11] T. Azib, O. Bethoux, G. Remy, C. Marchand, Saturation Management of a Controlled FuelCell/Ultracapacitor Hybrid Vehicle, IEEE Transactions on Vehicular Technology, Vol. 60, Issue: 8, 8 December 2011, pp [Azib 10] T. Azib, O. Bethoux, G. Remy, C. Marchand, E. Berthelot, An Innovative Control Strategy of a Single Converter for Hybrid Fuel Cell/Supercapacitors Power Source, IEEE Transactions on Industrial Electronics, Vol. 57, Issue: 12, 1 December 2010, pp [Tiefensee 11] F. Tiefensee, M. Hilairet, D. NormandCyrot, O. Bethoux, Sampleddata energetic management of a fuel cell/supercapacitor system, IEEE Vehicle Power and Propulsion Conference VPPC, Lille, FR, September 2011, pp. 16, Proceedings of IEEE Vehicle Power and Propulsion Conference VPPC [Ghanes 11] M. Ghanes, M. Hilairet, J.P. Barbot, O. Bethoux, Singular perturbation control for coordination of converters in a fuel cell system, Electrimacs, CergyPontoise, FR, September 2011, pp. 16, Proceedings of Electrimacs [Azib 10] T. Azib, R. Talj, O. Bethoux, C. Marchand, Sliding Mode Control and Simulation of a Hybrid FuelCell Ultracapacitor Power System, IEEE International Symposium Industrial Electronics, ISIE 10, Bari, Italie, 1 July 2010, pp , Proceedings of IEEE International Symposium Industrial Electronics, ISIE 10 [Ramirez 12] Victor Ramirez, Romeo Ortega, Antonio SanchezSquella, Roberto Grino, Olivier Bethoux, Theory and experimental results of two dynamic energy routers, ACC 2012, Montreal, Canada, june 2012 [Mariéthoz 12] Sébastien Mariéthoz, Olivier Bethoux and Mickaël Hilairet, "A distributed model predictive control scheme for reducing consumption of hybrid fuel cell systems", IECON 2012, Montreal, Canada, sept [Bethoux 09] O. Bethoux, M. Hilairet, T. Azib A new online diagnosis technique for PEM fuel cell with integration perspective, IECON 2009 The 35th Annual Conference of the IEEE Industrial Electronics Society, 35 November 2009, Porto, Portugal. [De Bernardinis 12] A. De Bernardinis, E. Frappé, O. Béthoux, C. Marchand and G. Coquery, Simulation with faulttolerant strategy, The European Physical Journal Applied Physics / Volume 58 / Issue 02 / 2012, (15 pages)