Control of High Efficiency PEM Fuel Cells for Long Life, Low Power Applications Part 2

Transcription

1 Control of High Efficiency PEM Fuel Cells for Long Life, Low Power Applications Part 2 Jekanthan Thangavelautham Postdoctoral Associate Field and Space Robotics Laboratory

2 Outline Review PEM Fuel Cell Degradation Catalyst Dissolution Chemistry Reaction Kinetics Physical Implications Effect on Fuel Cell Life Effective Control to Mitigate Degradation 2

3 Review 3

4 Motivation: Present Mobile electronic devices are energy hungry, battery have limited energy density could benefit from backup or alternative power sources. 4

5 Motivation: Future Mobile sensor networks can be an important tools to monitor the environment over long durations Intel eko 5

6 Fuel Cells Fuel Cell Electrolyte Fuel & Oxidizer Efficiency Operating Temp Alkaline Potassium Hydroxide H 2 and O 2 (pure) PEM Polymer Membranes (typically Nafion ) % o C H 2 and O % o C Direct Methanol Polymer Membranes CH 3 OH and O % o C Phosphoric Acid Phosphoric Acid H 2 and O 2 40 % o C Solid Oxide* Oxide ion conducting ceramic F: Methane, Propane, Butane, H 2 O % o C * No need for catalyst. 6

7 Fuel Cell Polarization Curve 7

8 Power Supply Mass for 3 Years 8

9 Challenges PEM Fuel cells face 4 major problems: Unreliable due to degradation [Shao-Horn et al., 2007], [Rubio et al., 2004 ], [Wu et al., 2008], [Borup et al., 2009], [Madden et al., 2010] Inefficient fuel storage [Schlapbach & Zuettel, 2001] Low power density [Barbir, 2005], [O Hayre et al., 2005] High cost [Barbir, 2005] Significant progress being made in all these areas. 9

10 Key components that can degrade Catalyst Layer Membrane Gas Diffusion Layer Fuel Cell Degradation 10

11 Fuel Cell Degradation Model Fuel Cell Component: Gas Diffusion Membrane Catalyst Layer 11

12 Speeds up and lower energy required for reactions Site for ionizing protons Site for assembly of water in the cathode Key metric platinum surface area S x i S,L 2 N x,i 4 R x,i Role of Platinum Catalyst 12

13 Platinum Loss Loss of platinum is irreversible Starts as slow-steady degradation of FC power performance Accelerates membrane structure degradation leading to catastrophic loss [Wu et al., 2008] 13

14 Platinum Oxidation & Dissolution Platinum Dissolution: Precipitation: Pt Pt 2 2e [1.188 V] Pt 2 H 2 Pt 2H Platinum Oxidation: Pt H 2 O = PtO + 2H 2e [0.98 V] PtO 2H Pt 2 H 2 O Ignored 14

15 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 15

16 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 16

17 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 17

18 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 18

19 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 19

20 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 20

21 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 21

22 Platinum Dissolution - Anode The FSRL The MIT Field and Space Robotics Laboratory 22

23 Platinum Dissolution Cathode The FSRL The MIT Field and Space Robotics Laboratory 23

24 Platinum Dissolution Cathode The FSRL The MIT Field and Space Robotics Laboratory 24

25 Platinum Dissolution Cathode The FSRL The MIT Field and Space Robotics Laboratory 25

26 Platinum Dissolution Cathode The FSRL The MIT Field and Space Robotics Laboratory 26

27 Kinetics of Catalyst Degradation Rate of Platinum dissolution using Butler-Volmer equations [Darling & Meyers, 2003], [Bi & Fuller, 2008]: r Pt 2 i n F( V U ) c 2+ n F( V U ) RT c 2+ RT Pt ref 1a 1 1i Pt 1c 1 1i k1 arh a iv exp exp Rate of Platinum Oxidation [Darling & Meyers, 2003], [Bi & Fuller, 2008]: ( ) ( ) 2an2F V U2i 2 2cn2F V U2i rpto i k2crhc i exp k2cch+ i exp RT RT 27

28 Kinetics of Catalyst Degradation 28

29 Catalyst Degradation: Current Model [Bi & Fuller, 2008]

30 Catalyst Degradation: Current Model [Thangavelautham & Dubowsky, 2012]

31 Ostwald Ripening

32 Effects of humidity to the kinetic model Relates humidity with proton activity: c H Effect of Humidity Accounts for water content [Springer et al., 2008]: Change in Nafion density with [Sethuraman, 2008]: 1 EW Nafion M H 2O H2 O C 0 C 1 RH C 2 RH 2 C 3 RH 3 Nafion C 4 C 5 1 C 6 32

33 Kinetics of Catalyst Degradation Two competing reactions occurring, dissolution and oxidation. Dissolution results in breakup of platinum into ions that mix into the water Oxidation results in film of platinum oxide forming over the platinum. Both reduce electrochemically active surface area. Platinum has an affinity to both the oxygen molecules and protons However oxidation in fact also reduces the effect of dissolution 33

34 Kinetics of Catalyst Degradation The reactions are dependent on: Operating Voltage Temperature Humidity 34

35 Mass Balance Equation c 2 c (1 kd) D 1.5 Ar j j 2 i Pti t x i S, L Change in Platinum Ion Concentration Diffusion and Migration Losses Dissolution of Platinum Ions The FSRL The MIT Field and Space Robotics Laboratory 35

36 Net Mass of Platinum Left c M ( t) M ( t 0) A(1 kd) t D dt 0 M x x L Diffusion and Migration Losses Total mass of platinum at time t is the initial mass minus the mass that diffuses and migrates away. 36

37 Migration vs. Dissolution Migration movement of particles due to external electric field. Importance of migration dependent on primarily on current and but also temperature. k N N migr diff 2FiR RT FC 37

38 End of Life End of life is taken as when the platinum surface area decreases by 25 % of original. Rate of area decrease is expected to be linear followed by a non-linear crash 38

39 Effect of Migration 39

40 40 Models vs. Data

41 41 Models vs. Data

42 FC Degradation Polarization Curve 42

43 FC Degradation Power Performance 43

44 Fuel Cell Efficiency * * Manyapu, K., Chapter 3: PEM Fuel Cell Steady State Model SM Thesis,

45 Effect of Voltage on Life 45

46 Effect of Voltage Oscillation 46

47 Effect of Temperature on Life 47

48 Effect of Relative Humidity 48

49 Kinetic Model: Conclusions Avg. degradation rate under ideal conditions between 1-9 V/hr Agrees with compiled experimental evidence [Wu et al., 2008] Catalyst degradation simulations highlight the severe effects of temperature and operating voltage Operating at high voltage increase fuel cell efficiency but drastically shortens life Very low humidity also found to shorten life. 49

50 The Effects of Degradation on Control For the field sensor network, we consider the effect of degradation, when the fuel cell is directly connected to a load We also analyze other suitable configurations to connect a fuel cell to a load. Point of comparison is expected life. End of life is when catalyst surface area decrease 25 % from original. 50

51 Fuel Cell Power Configuration 1 51

52 Feasibility of Degradation Control Fuel Cell System Configuration Operating Voltage Power Output [w] Fuel Cell Ambient Environment Life [Months] Config. 1 Fuel Cell Direct o C, RH %

53 Fuel Cell Power Configuration 2 53

54 Feasibility of Degradation Control Fuel Cell System Configuration Operating Voltage Power Output [w] Fuel Cell Ambient Environment Life [Months] Config. 1 Fuel Cell Direct Config. 2 Fuel Cell Battery Hybrid o C, RH % avg o C, RH %

55 Fuel Cell Power Configuration 3 55

56 Feasibility of Degradation Control Fuel Cell System Configuration Operating Voltage Power Output [w] Fuel Cell Ambient Environment Life [Months] Config. 1 Fuel Cell Direct Config. 2 Fuel Cell Battery Hybrid Config. 3 Fuel Cell Hybrid + Power Conditioning o C, RH % avg o C, RH % avg o C, RH %

57 Fuel Cell Power Configuration 4 57

58 Fuel Cell System Configuration Operating Voltage Feasibility of Degradation Power Output [w] Fuel Cell Ambient Environment Control Life [Months] Config. 1 Fuel Cell Direct Config. 2 Fuel Cell Battery Hybrid Config. 3 Fuel Cell Hybrid + Power Conditioning Config. 4 Fuel Cell Hybrid + Power Conditioning + Environment Control o C, RH % avg o C, RH % avg o C, RH % avg o C, RH %

59 Discussion Our models shows fuel cells directly connected to typical loads in the field have short lives High Operating Voltage High Temperature Extremely Low/High Humidity Use of fuel cells in direct configuration suggest short lives. This reinforces the notion the devices are unreliable. The key is to prevent high voltages, off nominal temperatures, humidity and oscillations, 59

60 Discussion Introducing voltage oscillations using DC-DC convertor is expected not do much better [Gallardo, 2010]. Hybrid fuel cell-battery combination is a solution that significantly increases life. Requires power conditioning that suppresses any oscillation from the fuel cell. Use of environmental control system to control humidity and temperature results in further increase in life 60

61 Control of Fuel Cell Degradation Catalyst Degradation can be substantially reduced through effective control Need to avoid deep cycling/fluctuation of voltage, temperature, humidity to maximize life and Balance conditions to maximize efficiency Expected Lifetime with Control Scenario Efficiency Voltage Humidity Temp. Life [years] 1 65 % o C % o C % o C % o C

62 FC Operating Condition for Long Life Life: 3+ years Anode Pressure: 1.1 Bar Cathode Pressure: 1 Bar Voltage: 0.8 V ( 65 %) Temperature: o C Relative Humidity: % Further increase in life expected if temperature and humidity is held constant within stated range. 62

63 Conclusion Our fuel cell degradation research suggests that to achieve long life the operating and environmental conditions for the fuel cell must be carefully controlled. 63

64 Acknowledgements Professor Steven Dubowsky Prof. Paolo Iora (Univ. of Brescia, Politecnico Milano) Prof. Yang Shao Horn Dr. Igal Klein Dr. Alex Schecter Daniele Gallardo, Dan Strawser, Kavya K. Manyapu Financial support by Israel s MOD Basic Science Office 64