Effect of Controlled Anode Flow Release on Dead-Ended Anode Proton Exchange Membrane Fuel Cells

Monday, May 12, 2014: 08:40
Bonnet Creek Ballroom II, Lobby Level (Hilton Orlando Bonnet Creek)
Q. P. G. Meyer (University College London), S. Ashton, O. Curnick, T. Reisch, P. Adcock (Intelligent Energy), P. R. Shearing, and D. J. L. Brett (University College London)
Proton Exchange Membrane Fuel Cells (PEMFCs) offer the possibility of zero-emission electricity generation. The technology has shown tremendous advances in terms of performance and durability and wide-scale commercialization in a range of applications is imminent.

      In order to make further improvements to the key performance metrics, it is important that we know as much as possible about the way the fuel cell is working in space and time. Electrochemical impedance spectroscopy (EIS) is a powerful in situ diagnostic technique for monitoring fuel cell performance; however, its use is limited to steady state conditions and typically requires costly hardware.

      A low-cost multichannel impedance analyzer developed during this project using a commercial data acquisition National Instrument hardware and LabVIEW in house software[1]. This novel diagnostic system is applied to a commercial Intelligent Energy air-cooled open-cathode 5-cell PEMFC stack.

Fig. 1. (a) Voltage loss at 0.75A cm-2 in Through flow and Dead ended conditions. (b) Reconstructed Nyquist plots at 0.75A cm-2 in Through flow and Dead ended conditions.

This system was used to distinguish the causes of the voltage loss in dead ended anode conditions. A novel reconstructive impedance strategy was developed. The same frequency of interest was measured every 20 seconds in through-flow and dead-ended conditions. The events were repeated, but changing the frequency of interest (1 kHz to 1 Hz) at each repeat. It was found possible to associate Nyquist plots with different stages of the voltage loss over time (Fig 1 a). These results (Fig 1 b) highlight an increase of the resistance in the low frequency region during dead-ended operation, an indication of flooding. Thus the reconstructive impedance technique enables the study of dynamic processes, unsuitable for investigation using steady-state EIS.

                 The influence of the current density on the voltage loss during a purge was also investigated (Fig 2). This result highlights the role of water in causing the voltage drop, since the amount of accumulating water is proportional to the current density. However, off gas analysis, inserting the probe between the hydrogen exhaust and the purge valve, also highlighted that the nitrogen accumulation is also proportional to the current density. This result suggests that the accumulation of nitrogen does not only come from permeation from the air through the membrane, but also from the small quantity contained in the hydrogen (50 ppm at BOC zero grade), with an accumulation rate proportional to the hydrogen consumption rate and to the volume of the anodic compartment.

Fig. 2. Evolution of the voltage losses (a) and nitrogen accumulation (b) from 0.4 to 1 A cm-2in dead ended anode.

In addition, a 16-segment S++ plate, capable of current, temperature and impedance mapping enabled to move from global to localized in situ measurements. The local and temporal evolution of the ohmic resistance, current density and temperature was studied as the current density was increased in order to provide more information on the process described Fig.2. Also, the heat distribution profile, obtained simultaneously from the S++plate and the thermal imaging camera provided a grid of the temperature distribution inside the stack and at its edges.


[1]  Q. Meyer, S. Barass, O. Curnick, T. Reisch, and D. J. L. Brett, “A multichannel frequency response analyser for impedance spectroscopy on power sources,” Journal of Electrochemical Science and Engineering, vol. 3, no. 3, pp. 107–114, 2013.