Performance Analysis of Non-Humidified High-Temperature PEFC by Fuel Cell System Simulator

Abstract

Polymer electrolyte fuel cells (PEFCs) are widely used for automobiles and other applications due to their excellent start-ability and load followability. For further popularization in the future, it is expected to be applied to commercial vehicles such as large trucks and buses. However, insufficient cooling of the PEFC becomes a major issue, for example when heavy-duty vehicle needs to operate on uphill roads at high-loads and high- temperatures. This is because conventional electrolyte membranes use water as a proton-conducting carrier, so PEFC’s operation temperature must be controlled below 100°C.In this study, we focus on high-temperature PEFC (HT-PEFC), which is expected to improve cooling performance by high-temperature operation (over 100°C). HT-PEFC can operate under non-humidified high-temperature environment by doping phosphoric acid as a proton-conducting carrier in the electrolyte membrane. Then, we evaluate the usefulness of the HT-PEFC by performance analysis using a FC system simulator.We use FC-DynaMo(1),(2), which is researched and developed by NEDO's industry-academia-government collaboration project, as the FC system simulator. FC-DynaMo is based on the FC stack system of Toyota motor corporation's 2nd generation MIRAI. As shown in Figure 1, in addition to the FC stack, hydrogen system, air system, and cooling system are also modeled, making it possible to simulate unsteady operation.The target vehicle for the simulation is a heavy-duty truck (Class8), and the results of a survey conducted by Zhang et al.(3) on the various uses of in the United States were used as a reference. From the results including output power, vehicle speed, and road gradient, we extracted and used the driving data of long-haul with continuous uphill and downhill roads, which is considered to be the high-load and high-temperature operation.For the membrane electrode assembly (MEA) of HT-PEFC, we used APM STD (PBI-MEA) manufactured by Advent technologies, which uses a polybenzimidazole membrane. We give the characteristic that the exchange current density increases exponentially as the temperature rises, referring to the study of Korsgaard et al.(4). In addition, proton conductivity is given as 10S/m published by Advent technologies. Figure 2(a) shows the calculated IV characteristic of MIRAI-MEA. And Figure 2(b) shows the that of PBI-MEA. As the temperature rises, the effect of reducing the activation overpotential by increasing the exchange current density can be confirmed. These results are in good agreement with the study which investigated the operation characteristics of PBI-MEA by Chen et al.(5) and Waller et al.(6). However, compared to the IV characteristic of MIRAI-MEA shown in Figure 2(a), the overpotential is too large, the maximum output is less than half, and the heat generation is nearly double. This is because it is nearly twice as large as the MIRAI-MEA’s activation overpotential, which causes a decrease in output power and an increase in heat generation. With this current PBI-MEA performance, it is difficult to evaluate the system compared with MIRAI-MEA. Therefore, in this study, we hypothetically set up a high-temperature MEA assuming future development (Virtual HT-MEA). Virtual HT-MEA is given MIRAI-MEA’s low activation overpotential and the characteristics of PBI-MEA that don’t dry-out at high-temperature. Figure 2(c) shows the Virtual HT-MEA’s IV characteristics. Here, at high current density, the performance at 120°C and 180°C is lower than that at 70°C because the concentration overpotential increases due to the decrease in the oxygen molar concentration at high-temperature.Figure 3 shows the results of Class8 driving simulation with each MEA adapted to the FC stack. In the MIRAI-MEA shown in Figure 3(a), a combination of 4-FC stack systems and 5-cooling systems is optimal. With this combination, the FC cell temperature can be controlled below 95°C. Figure 3(b) shows the analysis results of the FC stack using PBI-MEA. As shown in Figure 2(b), the PBI-MEA has a large overpotential, so a combination of 10-FC stack systems and 5-cooling systems is required to meet the request power. The current PBI-MEA’s performance cannot show the merits of non-humidified high-temperature operation. Figure 3(c) shows the analysis results of the FC stack using the Virtual HT-MEA. Stable operation is possible with 4-FC stack systems and 2-cooling systems. This means that 3-cooling systems can be reduced compared to the MIRAI-MEA’s system. This is because the cooling performance is improved due to the larger temperature difference between the outside air and the FC stacks.Reference(1)S.Hasegawa, et al., ECS Transactions,104, (2021)(2)S.Hasegawa, et al., ECS Transactions,109, (2022)(3)C.Zhang, et al., Transportation Research,Part D95, (2021)(4)A.R.Korsgaad, et al., Journal of Power Sources,162, (2006)(5)C.-Y.Chen, et al., Journal of Power Sources,195, (2010)(6)M.G.Waller, et al., International Journal of Hydrogen Energy,41, (2016) Figure 1