«EMR & inversion-based control of a multi-stack Fuel cell system»

Transcription

1 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «EMR & inversion-based control of a multi-stack Fuel cell system» Prof. Loïc Boulon, Khalid Ettihir, Neigel Marx, Prof. Daniel Hissel Université du Québec à Trois-Rivières Hydrogen Research Institute

2 1. Introduction Why FC system? - Outline EMR of FC system Representation Experimental Validation 3. Multi-FC system Why Multi-FC system? Energy management & simulation results EMR of a Multi-FC system Experimental results 4. On-line identification for energy management Why On-line identification Experimental results Integration into the EMR of the Multi-FC system

3 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «Introduction»

4 - Why FC systems- 4 H2 + O2 (air) Fuel Cell (electroc converter) Electricity + Heat + Water For vehicular application + No local emission / Fast Refueling - Cost & durability O2 supply Power Electronics Fuel Cell Stack Gas humidification H2 supply Temperature Management Némo Low Speed Vehicle (Hydrogen Research Center) Modeling & Simulation for Architecture design Energy management System performances 4

5 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «EMR of a FC system»

6 - EMR of a FCS - 6 EMR of the FC stack (ancillaries are summarized into the source elements) Dynamics are highlighted Control input are on the source elements Strong interactions: a FC stack cannot be seen has a cartesian system H2 Atmo H2 supply Thermal Therm Air supply Air Atmo T fc T fc T fc T fc T fc Δq P h2 q h2 q air P air i fc v fc Electrochemistry Elec

7 Inputs: measured values are imposed to the model Current steps Air flow variation (Constant H 2 flow) - Experimental validation - 7 Outputs: simulated values are compared to measurements H 2 input pressure Air input pressure

8 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «Multi-FC system»

9 Several fuel cells for: - Multi-Fuel Cell System - Energetic performance improvement H2 + O2 (air) Faulty mode operation Modular design (scale cost reduction) 9 This work is focused on power sharing in order to increase the energetic efficiency One fuel cell system Multi fuel cell system (4 fuel cells with a P/4 rated power) 9

10 - Power Distribution - 10 Single stack Equi-distribution Daisy-Chain Optimization-based W W 200 W 200 W 200 W 0 W 0 W 300 W 500 W??? FC 4????????? FC 3 FC 2 FC 1

11 Simulation results - Power Distribution - Optimisation based 11 Equi-distribution = same shape than 1 stack Optimisation-based is always the best distribution Daisy chain is not very relevant Daisy Chain Equi-distribution Efficiency vs Power curves Main improvements are realized at low power (up to 10-15%) Moreover, the Operating Range (OR) is increased (FC are generally note used between P=0 and P=Pmax/3) 11

12 - EMR of a multi Fuel Cell System - Power Electronics 12 FC EMR could be reduced to a simple multiphysics converter due to study assumptions P H2 P H2 v bat i fc2 v bat i fc1 v bat PT H2 P H2 i fc3 Σi P H2 v bat i fc4 v bat Bat EM provides requested current and distribution criterion from SoC Powertrain requested power i fc4 ref v bat i ref k d EM Distribution critieria is given by a look-up table obtained with the optimization algorithm

13 Experimental results are highly dependant of: - Results on a driving cycle DC bus power FC power 13 The driving cycle The FCS sizing Power(W) The vehicular energy management (power split between the FCS and the battery pack) Distribution Method Equidistribution Optimal distribution Consump tion Relative value kj 100 % kj % Fuel consumption(j) Time(s) 6 x 105 Daisy Chain Equi-distribution Optimal distribution Time(s)

14 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «On-line identification of FC system for global energy management»

15 - Good operation point seeking- 15 Multiphysics behavior & ageing are difficult to model but Many energy management are model based

16 - Experimental results - 16 The identified model fits with experimental results 3 modules EM organisation Experimental results show the method relevance

17 - Vectorial representation- 17 H2 PT Bat to save space & to reach a more syntetic scheme: vectorial representation Just a representation: No new assumptions EM

18 - Vectorial representation- 18 H2 Output measurement is requested for close loop estimation PT Bat EM to save space & to reach a more syntetic scheme: vectorial representation Just a representation: No new assumptions Model parameters Energy Management is a complicated block

19 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «Conclusion»

20 - Vectorial representation- 20 EMR is useful in order to organize several works in same way Fuel Cell modeling Power sharing between the 4 fuel cell systems Best operating point tracking How to organize the Energy Management block? 3 levels with strong interactions (the performances of each FCS should impact the vehicular energy management): Fuel Cell System (ancillaries) Multi Fuel Cell System (power sharing) Vehicular (power split)

21 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «BIOGRAPHIES AND REFERENCES»

22 - Authors - 22 Prof. Loïc Boulon Université du Québec à Trois-Rivières, Canada Hydrogen Research Institute PhD in Electrical Engineering at Univ. of Franche-Comté (2009) Research topics: Energy Storage Subsystems, Fuel Cell Systems, Operation of vehicles in cold winter conditions

23 - References - 23 Neigel MARX, Loïc BOULON, Frédéric GUSTIN, Daniel HISSEL : "A review of multi-stack and modular fuel cell systems : interests, application areas and on-going research activities", Elsevier International Journal of Hydrogen Energy, Accepted paper. Loïc BOULON, Kodjo AGBOSSOU, Daniel HISSEL, Pierre SICARD, Alain BOUSCAYROL, Marie-Cécile PERA, "A macroscopic PEM Fuel Cell model including water phenomena for vehicle simulation, Elsevier Renewable Energy, Vol. 46, pp Oct Loïc BOULON, Daniel HISSEL, Alain BOUSCAYROL, Marie-Cécile PERA, "From Modeling to Control of a PEM Fuel Cell using Energetic Macroscopic Representation", IEEE trans. on Industrial Electronics, Vol. 57, no. 6, pp , Loïc BOULON, Marie-Cécile PERA, Philippe DELARUE, Alain BOUSCAYROL, Daniel HISSEL, "Causal fuel cell system model suitable for transportation simulation applications", ASME Fuel Cell Science and Technology, Vol. 7, Iss. 1, 2010 David TOQUICA, Neigel MARX, Loïc BOULON, Frédéric GUSTIN, Daniel HISSEL, "Degraded mode operation of multi-stack fuel cell systems", IEEE Vehicle Power and Propulsion Conference (VPPC) 2014, Coïmbra (Portugal), Accepted paper. Jorge E. GARCIA, Daniel. F. HERRERA, L. BOULON, P. SICARD, A. HERNANDEZ, "Power Sharing for Efficiency Optimisation into a Multi Fuel Cell System", IEEE International Symposium on Industrial Electronics (ISIE) Clément DEPATURE, Loïc BOULON, Pierre SICARD, Mickaël FOURNIER, "Simulation model of a multi-stack fuel cell system", European Conference on Power Electronic (ECCE Europe) 2013, Lille (France), Khalid ETTIHIR, Loïc BOULON, Kodjo AGBOSSOU, Sousso KELOUWANI, "Design of an Energy Management Strategy for PEM Fuel Cell Vehicles using a maximum efficiency and a maximum power operation modes", IEEE International Symposium on Industrial Electronics (ISIE) 2012, Hangzhou (China), Khalid ETTIHIR, Loïc BOULON, Kodjo AGBOSSOU, On-line Proton Exchange Membrane Fuel Cell Identification Based on a RLS algorithm,in proc. of Fundamentals and Development of Fuel Cells 2013, 6p., Karlsruhe, Germany, 2013.

24 EMR 14 Coïmbra June 2014 Summer School EMR 14 Energetic Macroscopic Representation «Appendix: new EMR pictograms»

25 - Colors - 25 Power source Power system System model System control Control strategy pale green gold violet sky blue blue orange background RGB = (255,215,0) «gold» red border RGB = (255,0,0) «red» light blue background RGB = (135,206,235) «sky blue» dark blue border RGB = (0,0,255) «blue» light green background RGB = (152,251,152) «pale green» dark green border RGB = (0,128,0) «greeen» purple background RGB = (238,130,238) «violet» dark blue border RGB = (0,0,255) «blue» dark blue background RGB = (0,0,255) «blue» dark blue border RGB = (0,0,255) «blue» Web X11 colour, standard colours on web pages

26 No equation number in slides - EMR pictograms a element name element name element name a Name x 1 y 1 (2a x a) (a/2 x a) (a x a) (radius= a) borders of power elements = b pt power vectors (size b, full arrows) element name element name signal vectors (size b/2, empty arrows)

27 - Control pictograms - 27 (same pictograms same size - with or without oblique bar) (support square a x a) borders of control elements = b/2 pt signal vectors (size b/2, empty arrows) stratégie

28 - estimation pictograms - 28 Source borders of estimation elements = b/2 pt signal vectors (size b/2, empty arrows)

29 - Example - 29 parallel connexion DC machine wheel environment battery choppers gearbox chassis u bat u ch1 i arm Bat. u bat i tot i ch1 u bat m ch1 i arm u ch2 e arm i field T em Ω gear T gear Ω wh F wh v ev v ev F res Env. i ch2 i field e field m ch2 Ω wh-est v ev-est u ch2-ref i field-ref SoC est u ch1-ref i arm-ref T em-ref T gear-ref F wh-ref v ev-ref strategy driver request