International Symposium on Diagnostic Tools for Fuel Cell Technologies Trondheim, June 24th, 2009 On- -line diagnostic of a PEM fuel cell stack line diagnostic of a PEM fuel cell stack On based on the electrical power converter based on the electrical power converter Dr. Abdellah NARJISS, Dr. Frédéric GUSTIN, Dr. Daniel DEPERNET, Prof. Daniel HISSEL University of Franche-Comte (France) Événement - date Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Introduction General architecture of a PEM Fuel Cell system Gases Fuel Cell System Fluidic supervision Fuel Cell Stack Ancillaries Power converter Load Control Electrical supervision On-line diagnostic Événement - date How to optimize the FC behavior ? Objective : design an on-line diagnostic tool dedicated to automotive applications Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Outlines 1) Choice and sizing of the power converter 2) Real time control strategy 3) On-line impedance spectrometry 4) Conclusions Événement - date Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Choice and sizing of the power converter DC Bus : PEM FC Stack : 540V 115V (0A) – 55V (560A) Power converter High voltage elevation ratio ECCE Hybrid Electrical Vehicle Transportation applications ≈ 80kW Événement - date Improvement of durability and reliability of PEMFC in embedded applications Control of DC-link with high transformation ratio, low switch constraints and high efficiency Optimization of size and cost Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Choice and sizing of the power converter General description of the energy conversion scheme Événement - date Semi-conductors : Mosfets + Diodes F PWM =50kHz Planar technology transformer Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Choice and sizing of the power converter Sizing of the power converter for impedance spectrometry F SPECTRO F PWM FC Stack Power converter DC bus Δ Ie Δ Is Événement - date Sizing of the passive filters 1 V 1 Is 1 F PWM = 50kHz f 0 S L α C α Hz f f Δ I 2.f 2. . . Δ Vs 2.f L f C F SPECTRO : 1Hz to 2,5kHz S f Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Outlines 1) Choice and sizing of the power converter 2) Real time control strategy 3) On-line impedance spectrometry 4) Conclusions Événement - date Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Real time control Impedance spectrometry done by the power converter Current injection α = α 0 +A.sin( ω t) α 1 1 A Δα α 0 0 0 t t α 0 : duty cycle => normal behavior A : sinusoïdal modulation amplitude sinus de 20 points 40000 F Événement - date 30000 20 High resolution quality PWM F 20000 injection k 10000 0 -10000 F PWM = 50kHz -20000 => F injection_max = 2,5kHz -30000 -40000 Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Real time control Impedance measurement extraction of ripple PWM rejection filtering and phase referencing of desired ripple with DFT (Discrete Fourier Transform) harmonic impedance Fuel Cell Fuel Cell Voltage Voltage Re(Vk) Re(Vk) Im(Vk) Im(Vk) Zk Zk Fuel Cell Fuel Cell Événement - date Current Current Re(Ik) Re(Ik) Im(Ik) Im(Ik) Sampling Sampling Module ratio Module ratio Analogical Analogical DFT DFT extraction of ripple extraction of ripple (PWM rejection) (PWM rejection) Phase difference Phase difference Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Real time control Electrical behavior Electrical behavior good DC bus voltage stability power flows control ability experimental results (1kW prototype) Événement - date Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Outlines 1) Choice and sizing of the power converter 2) Real time control strategy 3) On-line impedance spectrometry 4) Conclusions Événement - date Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
On-line impedance spectrometry Injection strategy of fuel cell stimulus duty cycle sinusoidal modulation control of the current ripple amplitude Constant nominal Constant nominal duty cycle duty cycle Current ripple Current ripple reference reference + + DC/DC DC/DC Ripple Ripple + + + + Fuel Cell Fuel Cell converter converter controller controller - - Current Current Current ripple Current ripple measurement measurement Événement - date measurement measurement + + - - Continuous Continuous component of component of Continuous Continuous current current component component extractor extractor Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
On-line impedance spectrometry UBZM PEM 20-cell stack Experimental values On-line impedance spectrometry result 0,05 0,05 Gas flows 20A 0,045 0,045 FC stack temperature 48°C 0,04 0,04 FC air dew point 45°C 0,035 0,035 0,03 0,03 FC stack voltage 14,18V -Im (Ohm) -Im (Ohm) 0,025 0,025 essai_1 essai_1 FC current 12,15A 0,02 0,02 essai_2 essai_2 0,015 0,015 Air hygrometry level 85% 500Hz 500Hz 0,01 0,01 Hydrogen hygrometry 20% 2Hz 2Hz Événement - date 0,005 0,005 level 0 0 Load current 1,6A 0,1 0,1 0,15 0,15 0,2 0,2 0,25 0,25 0,3 0,3 -0,005 -0,005 Re (Ohm) Re (Ohm) Results reproducibility highest quality Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
On-line impedance spectrometry Influence of the spectrometry on the DC bus voltage Effects of spectroscopy on continuous voltage bus Effects of spectroscopy on continuous voltage bus Effects of spectroscopy on continuous voltage bus Effects of spectroscopy on continuous voltage bus Impact de la spectroscopie d’impédance sur la stabilité du bus continu Impact de la spectroscopie d’impédance sur la stabilité du bus continu Impact de la spectroscopie d’impédance sur la stabilité du bus continu Impact de la spectroscopie d’impédance sur la stabilité du bus continu 120 120 120 120 100 100 100 100 80 80 80 80 Vbus (Volts) Vbus (Volts) Vbus (Volts) Vbus (Volts) F = 2 Hz F = 2 Hz F = 2 Hz F = 2 Hz 60 60 60 60 F = 30 Hz F = 30 Hz F = 30 Hz F = 30 Hz F = 360 Hz F = 360 Hz F = 360 Hz F = 360 Hz 40 40 40 40 20 20 20 20 Événement - date 0 0 0 0 phase (radians) phase (radians) phase (radians) phase (radians) Δ Vmax < 10% Vbus Δ Vmax Fmin and Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
On-line impedance spectrometry Effect of the gas hydration level Fuel cell operating point : 8A, =48 ° C, Igases=20A T FC 0,07 Humidification 0,06 level increases 0,05 Tr=25°C, HR_air=30%, 0,04 HR_H2=21% -Im (Ohm) Tr=35°C, HR_air=50%, 0,03 Z stack diminishes HR_H2=20% Événement - date 0,02 Tr=45°C, HR_air=87%, HR_H2=20% 0,01 0 0,05 0,15 0,25 0,35 0,45 -0,01 Re (Ohm) Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
On-line impedance spectrometry Effect of the H 2 gas flow Fuel cell operating point : 8A 0,07 H 2 flow increases 0,06 0,05 -Im (Ohm) Igas=20A 0,04 Igas=15A 0,03 Igas=10A Z stack diminishes Événement - date in LF 0,02 0,01 0 0,1 0,15 0,2 0,25 0,3 0,35 Re (Ohm) Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Outlines 1) Choice and sizing of the power converter 2) Real time control strategy 3) On-line impedance spectrometry 4) Conclusions Événement - date Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Conclusions Many advantages : Important power flows transfer ability Ability to high transformation ratios High compactness On-line diagnosis ability Ability to modify in real time the FC control laws to improve durability / efficiency Next to do : Événement - date Higher power FC have to be considered Go from the impedance spectrometry to the FC stack on-line diagnosis Consider SOFC power plants Int. Symp. On Diagnostic Tools for FC Technologies – Trondheim – June 2009
Recommend
More recommend