Academia.eduAcademia.edu

Figure 13 – uploaded by Shanna Knights

See full PDFdownloadDownload figure
Fig. 14. Lifetime test of 17-cell stack comprised of 250kW natural gas power plant hardware. Coolant operating temperature: 82°C at stack outlet. Design of cell optimized to minimize localized high temperature effects at membrane in fuel inlet. Actual temperature at membrane is not measured.

Figure 14 Lifetime test of 17-cell stack comprised of 250kW natural gas power plant hardware. Coolant operating temperature: 82°C at stack outlet. Design of cell optimized to minimize localized high temperature effects at membrane in fuel inlet. Actual temperature at membrane is not measured.

Related Figures (12)

Fig. 1. Fuel starvation polarization. Humidified anode/cathode feed streams: nitrogen/air. 3 bar, 75°C, 4mg/cm? Pt on each of cathode and anode. A polarization of a complete fuel starvation (no hydrogen) with humidified nitrogen flowing on the anode is presented in Fig. 1 for an anode with a 4 mg Pt/cm? loading of Pt black catalyst on the anode. The cell voltage, as measured from the cathode to the anode, drops due to a rise in anode poten- tial, which reaches levels sufficient for water oxidation, i.e. >1.23 V. The values of the cathode and anode potential are
Fig. 1. Fuel starvation polarization. Humidified anode/cathode feed streams: nitrogen/air. 3 bar, 75°C, 4mg/cm? Pt on each of cathode and anode. A polarization of a complete fuel starvation (no hydrogen) with humidified nitrogen flowing on the anode is presented in Fig. 1 for an anode with a 4 mg Pt/cm? loading of Pt black catalyst on the anode. The cell voltage, as measured from the cathode to the anode, drops due to a rise in anode poten- tial, which reaches levels sufficient for water oxidation, i.e. >1.23 V. The values of the cathode and anode potential are
Fig. 4. Effect of reduced humidification levels on non-optimized cell de- sign. Air/hydrogen operation with specified humidification levels. Current density: 1 A/cm?. Water management is key to optimum reliability and dura- bility of fuel cells. Inadequate water content, either globally
Fig. 4. Effect of reduced humidification levels on non-optimized cell de- sign. Air/hydrogen operation with specified humidification levels. Current density: 1 A/cm?. Water management is key to optimum reliability and dura- bility of fuel cells. Inadequate water content, either globally
Fig. 3. Comparison of different anode structures in severe failure testing. Fach cell has equivalent cathode (~0.7 mg/cm? Pt, supported on carbon). Testing conducted at 200mA/cm?, fully humidified nitrogen on anode. Anode loading at ~0.3mg/cm2 Pt supported on carbon (varied materials and compositions). Each curve represents the results from a four-cell stack with each cell in the stack of identical composition. Four separate stack tests were run to generate the curves.
Fig. 3. Comparison of different anode structures in severe failure testing. Fach cell has equivalent cathode (~0.7 mg/cm? Pt, supported on carbon). Testing conducted at 200mA/cm?, fully humidified nitrogen on anode. Anode loading at ~0.3mg/cm2 Pt supported on carbon (varied materials and compositions). Each curve represents the results from a four-cell stack with each cell in the stack of identical composition. Four separate stack tests were run to generate the curves.
Fig. 6. Lifetime test showing effect of membrane damage resulting in gas crossover. Air/hydrogen operation. Fig. 7. Stable lifetime performance achieved using counter-flow operation. Ten-cell stack, hydrogen/air operation. Improved water management can be achieved through various strategies. One challenge faced by cell and stack
Fig. 6. Lifetime test showing effect of membrane damage resulting in gas crossover. Air/hydrogen operation. Fig. 7. Stable lifetime performance achieved using counter-flow operation. Ten-cell stack, hydrogen/air operation. Improved water management can be achieved through various strategies. One challenge faced by cell and stack
One of the easiest strategies to implement to redistribute product water within the cell is through operation with coun- tercurrent reactant flows, essentially using the MEA as an in-cell humidifier [2,11,12]. A study was conducted to deter- mine the optimum flow configuration for drier conditions (no air humidification, fuel humidification at 10°C lower than the inlet coolant temperature). The strategies tested included co- and counter-flow strategies for the three fluid streams: air, fuel, and coolant. In this case, the cell was operated with a 15°C temperature differential from inlet to outlet in the coolant stream. Operation with fuel counter-flow to air and coolant achieved the most stable performance and the lowest cell resistance. Fig. 7 shows a lifetime test run under these conditions, with no degradation observed. The other config- urations tested resulted in a steady decrease in cell voltage and the accompanying increase in cell resistance, indicating the effect was due to membrane drying, resulting in proton conductivity loss. Fig. 5. Time to significant gas crossover (>10cm3/min at 2 bar pressure differential) as a function of inlet gas humidification for a non-optimized cell design. Air/hydrogen operation, current density: 540 mA/cm2, coolant temperature: 75°C, humidification levels of reactant streams as indicated.
One of the easiest strategies to implement to redistribute product water within the cell is through operation with coun- tercurrent reactant flows, essentially using the MEA as an in-cell humidifier [2,11,12]. A study was conducted to deter- mine the optimum flow configuration for drier conditions (no air humidification, fuel humidification at 10°C lower than the inlet coolant temperature). The strategies tested included co- and counter-flow strategies for the three fluid streams: air, fuel, and coolant. In this case, the cell was operated with a 15°C temperature differential from inlet to outlet in the coolant stream. Operation with fuel counter-flow to air and coolant achieved the most stable performance and the lowest cell resistance. Fig. 7 shows a lifetime test run under these conditions, with no degradation observed. The other config- urations tested resulted in a steady decrease in cell voltage and the accompanying increase in cell resistance, indicating the effect was due to membrane drying, resulting in proton conductivity loss. Fig. 5. Time to significant gas crossover (>10cm3/min at 2 bar pressure differential) as a function of inlet gas humidification for a non-optimized cell design. Air/hydrogen operation, current density: 540 mA/cm2, coolant temperature: 75°C, humidification levels of reactant streams as indicated.
Fig. 8. Water management through MEA design. (a) Plan view of MEA showing perforations in GDL with increased perforation density in wet- ter outlet region. (b) Performance of MEA at increased current density compared to a conventional, non-pierced MEA design.
Fig. 8. Water management through MEA design. (a) Plan view of MEA showing perforations in GDL with increased perforation density in wet- ter outlet region. (b) Performance of MEA at increased current density compared to a conventional, non-pierced MEA design.
Fig. 11. Lifetime performance of a direct methanol fuel cell using load cycling strategy to reduce mass transport related degradation. 30s of open circuit voltage operation/30 min of on-load operation. A wide range of operating temperatures is required for fuel cells, depending on the application. For automotive ap- plications, this may include sub-zero operation upon start-up in freezing temperatures [17]. Regardless of the start-up re- quirements, the stack and system will likely be required limitations. Quantification of this increased loss can be ac- complished through testing of the cell in the high current density region on each of two different oxidants, air and a mixture, termed “helox”, consisting of 21% oxygen and 79% helium. The voltage response as a function of current den- sity and oxidant used provides information on gas diffusion limited performance, with an increasing gap at high current densities indicating increased losses [2,16]. The change in gas diffusion limited performance of a DMFC run continu- ously for only 16h is compared in Fig. 10 against one run under load cycling for almost 2000h. These cells suffered a change in gas diffusion limited performance of ~370 and ~35%, respectively, an order of magnitude improvement in degradation. The load cycling strategy used in this case con- sists of removing the load from the cell for 30 s during every 30 min operational period. As can be observed given the re- sults in Fig. 10, this has proven to be very effective in main-
Fig. 11. Lifetime performance of a direct methanol fuel cell using load cycling strategy to reduce mass transport related degradation. 30s of open circuit voltage operation/30 min of on-load operation. A wide range of operating temperatures is required for fuel cells, depending on the application. For automotive ap- plications, this may include sub-zero operation upon start-up in freezing temperatures [17]. Regardless of the start-up re- quirements, the stack and system will likely be required limitations. Quantification of this increased loss can be ac- complished through testing of the cell in the high current density region on each of two different oxidants, air and a mixture, termed “helox”, consisting of 21% oxygen and 79% helium. The voltage response as a function of current den- sity and oxidant used provides information on gas diffusion limited performance, with an increasing gap at high current densities indicating increased losses [2,16]. The change in gas diffusion limited performance of a DMFC run continu- ously for only 16h is compared in Fig. 10 against one run under load cycling for almost 2000h. These cells suffered a change in gas diffusion limited performance of ~370 and ~35%, respectively, an order of magnitude improvement in degradation. The load cycling strategy used in this case con- sists of removing the load from the cell for 30 s during every 30 min operational period. As can be observed given the re- sults in Fig. 10, this has proven to be very effective in main-
Fig. 10. Comparison of change in gas diffusion limited performance in a DMFC depending on operational strategy used. Gas diffusion limited performance is defined as the difference in cell voltage on air as compared to a mixture of 79% 02/21% No at 0.5A/cm?. Fig. 9. Performance of a DMFC after 16h of steady state operation compared to initial performance.
Fig. 10. Comparison of change in gas diffusion limited performance in a DMFC depending on operational strategy used. Gas diffusion limited performance is defined as the difference in cell voltage on air as compared to a mixture of 79% 02/21% No at 0.5A/cm?. Fig. 9. Performance of a DMFC after 16h of steady state operation compared to initial performance.
Fig. 13. Comparison of lifetimes achieved in 250kW natural gas power plant hardware. Air/simulated reformate operation with 2% air bleed. One mitigation strategy that can be used for operation when sub-zero temperatures are expected, is purging of flow channels. This is an effective technique to clear the chan- nels of water prior to freezing [19]. Fig. 12 shows the per- formance of a cell after each of 55 freeze-thaw cycles, with operation and purging during each cycle. No significant per- formance loss is observed at 0.5A/cm?, and at 1.0 A/cm? a loss of only 0.2 mV /cycle is observed. This is consistent with data reported in [20], in which no degradation was observed when the cell was cooled to —78°C, and with that reported in [21], in which three freeze-thaw cycles to —10°C were completed with no observable degradation. An example of the improvement achievable using this ap- proach is represented in Fig. 13. In this case, the applica- tion was the Ballard 250kW Natural Gas Power plant. In an early design, the cells began to fail very early for this type of application; at 5000h for MEA type 1 and in less than 2000h for MEA type 2. In this case, the MEA type 2 was the preferred design in all other respects. Thus, the strategy mentioned above of failure analysis and modeling was accomplished to highlight key contributors to the early failures. Subsequent redesign and testing resulted in very significant improvements to durability, particularly with the target MEA type 2. In laboratory simulation trials, a short stack (17-cells) operated for over 13,000h without failure, with an average cell degradation rate of only 0.5 pV/h, as presented in Fig. 14. In a field trial under dynamic opera- tion, the full power plant operated for 7400h, to the end of the field trials, with no MEA failures.
Fig. 13. Comparison of lifetimes achieved in 250kW natural gas power plant hardware. Air/simulated reformate operation with 2% air bleed. One mitigation strategy that can be used for operation when sub-zero temperatures are expected, is purging of flow channels. This is an effective technique to clear the chan- nels of water prior to freezing [19]. Fig. 12 shows the per- formance of a cell after each of 55 freeze-thaw cycles, with operation and purging during each cycle. No significant per- formance loss is observed at 0.5A/cm?, and at 1.0 A/cm? a loss of only 0.2 mV /cycle is observed. This is consistent with data reported in [20], in which no degradation was observed when the cell was cooled to —78°C, and with that reported in [21], in which three freeze-thaw cycles to —10°C were completed with no observable degradation. An example of the improvement achievable using this ap- proach is represented in Fig. 13. In this case, the applica- tion was the Ballard 250kW Natural Gas Power plant. In an early design, the cells began to fail very early for this type of application; at 5000h for MEA type 1 and in less than 2000h for MEA type 2. In this case, the MEA type 2 was the preferred design in all other respects. Thus, the strategy mentioned above of failure analysis and modeling was accomplished to highlight key contributors to the early failures. Subsequent redesign and testing resulted in very significant improvements to durability, particularly with the target MEA type 2. In laboratory simulation trials, a short stack (17-cells) operated for over 13,000h without failure, with an average cell degradation rate of only 0.5 pV/h, as presented in Fig. 14. In a field trial under dynamic opera- tion, the full power plant operated for 7400h, to the end of the field trials, with no MEA failures.
Fig. 12. Effect of freeze-thaw cycles on fuel cell performance. Simu- lated reformate operation. Operation and purging conducted between each purge/thaw cycle.
Fig. 12. Effect of freeze-thaw cycles on fuel cell performance. Simu- lated reformate operation. Operation and purging conducted between each purge/thaw cycle.
Related topics:
EngineeringDirect methanol fuel cell
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved