Operational cost, primarily power consumption, is the one of the main obstacles for hydrogen produced by polymer electrolyte water electrolyzers (PEWE). Roughly 70% of the hydrogen production cost stems from electricity [1], thus overall electrochemical efficiency has a direct influence on the hydrogen price. Dominating sources of electrochemical loss in PEWE cells are the kinetic, electric, and mass transport overpotentials. Conventional research and diagnostic techniques, such as polarization curves and EIS measurements, predominantly probe macroscopic performance characteristics without any spatial resolution of electrochemical performance. These offer limited insight into the in-situ behaviors of the PEWE and its overpotentials. The work of Schuler, et al. [2] gives a detailed understanding of transport mechanisms inside a PEWE by drawing extensively on ex-situ techniques. Work carried out by Zhang et al. [3] has successfully shown the bubble formation of water electrolysis inside a PEWE by means of an in-situ, high-speed micro-scale visualization system (HMVS). However the HMVS does not explicitly capture the local electrochemical behavior of the cell. Therefore, complex CFD models have been implemented to shed light on the inner workings of PEWE; while insightful, such models need to be validated with experimental data. This work aims to provide a reliable source of experimental data by developing a technique that allows for in-situ spatial and temporal current distribution measurements which can be used to validate existing modeling efforts.Current distribution measurements are described extensively in the works of Bender et al. [4, 5]; a major innovation was the inclusion of a printed-circuit board (PCB) with an array of shunt resistors attached to each segment as shown in Figure 1a (4.5 mm x 4.5 mm segments in this work). The measured voltage drop across each shunt resistor is used to calculate the current in each segment. Data acquisition on the distributed current board is done simultaneously to electrochemical control of the electrolyzer. Thus the distributed current measurement occurs passively while a potentiostat controls the electrolyzer, without interrupting its operation. This setup is capable of using flow fields of both 5cm2 and 25cm2 active area. The 5 cm2 flow field has 16 segments while the 25cm2 has 100 segments that are electronically isolated, yielding high spatial resolution of the current distribution. The distributed current is then interpolated to obtain smoothed contours as shown in Figure 1b and 1c. Preliminary current distribution experiments were carried out at 85⁰C, water flow rate of 0.5 to 3 mL/min/cm2, at a cell compression under bolt torques of 5 Nm and 10 Nm. Experiments were also carried out with two sets of liquid-gas diffusion layers: iridium-coated titanium and bare titanium felts. The results show a relatively uniform current distribution across the electrolyzer with no significant impact of flow rate. The experiments also revealed that the spatial current distribution is highly sensitive to local cell compression, which is not captured in the overall cell performance of a polarization curve. This work demonstrates the benefit of a spatially-resolved current distribution measurement technique and the insights it offer which would otherwise be impossible to probe.
Read full abstract