0
RESEARCH PAPER

Discharge of a Segmented Polymer Electrolyte Membrane Fuel Cell

[+] Author and Article Information
P. Berg

 Faculty of Science, UOIT, 2000 Simcoe Street N, Oshawa, ON L1H 7K4, Canada Phone: +1 905 721 3111 ext. 2457, Fax: +1 905 721 3304, peter.berg@uoit.ca

K. Promislow

 Department of Mathematics, Michigan State University, East Lansing, MI 48824 kpromisl@math.msu.edu

J. Stumper

 Ballard Power Systems, 4343 North Fraser Way, Burnaby V5J 5J9, Canada, jurgen.stumper@ballard.com

B. Wetton

 Department of Mathematics, University of British Columbia, 1984 Mathematics Road, Vancouver, BC, V6T 1Z2, Canada, wetton@math.ubc.ca

J. Fuel Cell Sci. Technol 2(2), 111-120 (Dec 10, 2004) (10 pages) doi:10.1115/1.1867977 History: Received August 16, 2004; Revised December 10, 2004

We present a transient model for an electrically segmented polymer electrolyte membrane (PEM) fuel cell which is run until extinction from a finite oxygen supply. The experimental cell is divided into 16 electrically isolated pucks which are fed oxygen from a small reserve and hydrogen from a conventional flow field. The experimental voltage and through-plane current in each puck, and puck-to-puck currents are recorded and compared to computed profiles. Seven qualitative characteristics of the current profiles during discharge are identified. These are used as targets for parameter tuning, from which puck-to-puck water distribution within the membrane electrode assembly (MEA) is inferred. The model is sensitive to system parameters, and holds promise as an in situ diagnostic tool for tracking this distribution by using MEA oxygen transport characteristics.

Copyright © 2005 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Figure 1

The electrical network, cathode, and membrane of the segmented cell. Each component is divided into 16 pucks. Through-plane currents Ij flow into the jth puck and Icj flow through the jth collector resistor. In-plane currents Iink flow between kth and k+1th pucks. The total current is IT, while the current tool has resistance R(t). Rc,Rin and Rmj denote the collector, in-plane, and jth membrane resistances, respectively. The volumes of headers are Vin and Vout. Arrows indicate definition of positive current direction.

Grahic Jump Location
Figure 2

The flow fields and MEA of the segmented cell

Grahic Jump Location
Figure 3

Experimental data: Constant-current tool characteristic for discharge at initial load of I0=300A. The total current IT is a function of the external voltage U. The function IT(U;I0) is extrapolated from this curve to enter the model. Note that the total current collapses at U≈0.4V.

Grahic Jump Location
Figure 4

Experimental data: Voltages of the 16 pucks over time when discharging at I0=300A. The total current starts to drop at t≈3.25s, as exhibited by Fig. 5. The puck voltages monotonicly decrease over a time scale of about 2.5 s. Note that at t=3.25s, the average puck voltage is about U¯j=0.5V, which differs by 0.1 V from the external voltage at that point (see Fig. 3). The difference is due to the voltage drop at Rc, the collector resistances, about 300A×4.5mΩ∕16≈0.1V.

Grahic Jump Location
Figure 5

Experimental data: Puck (or through-plane) currents over time when discharging at I0=300A (IT: thick line). Pucks starve in succession and remain at very low currents, about 2 A. The remaining currents increase as the total current has to be matched. Ultimately, all pucks starve and the total current collapses. Center pucks tend to starve first as they are the furthest from the replenishing headers.

Grahic Jump Location
Figure 6

Experimental data: In-plane (IP) currents between pucks over time when discharging at I0=300A. In-plane currents arise as a result of voltage differences between adjacent pucks. This in turn occurs when pucks begin to starve. Note that the in-plane currents are cut off artificially at Iink<−30A due to the data recording equipment.

Grahic Jump Location
Figure 7

(This and all remaining plots: Numerical results.) Diffusion coefficients puck dependent: Kc=3.45×108Pas∕m3,Dj parabolic shape down the channel (see text), T=343K, and remaining parameters as in the text. IT: Thick dotted line, Ij: Thick solid lines, Iink: Thin dotted lines. All scales have desirable magnitudes. Arrows indicate order of pucks: 8→1 and 9→16 for puck currents, 7→1 and 7→15 for in-plane currents.

Grahic Jump Location
Figure 8

Puck voltages (solid lines) and external voltage (dashed line) over time during the run shown in the previous figure. In comparison to the data, the scales are correct past the point t=2.5s, as well as the initial voltages, but the voltages do not drop sufficiently from the start. This mismatch is related to the choice of kinetic parameters.

Grahic Jump Location
Figure 9

Linearly increasing diffusion coefficient Dj down the channel: Kc and T as in Fig. 7. Tth and Ath have right orders of magnitude unlike T0. Also, the order and time scale of puck starvation do not match experimental data. Puck currents 1, 2, 8, 9, 15, and 16 are marked.

Grahic Jump Location
Figure 10

Identical diffusion coefficients in all pucks: Kc=3.45×108Pas∕m3,Dj≡7.2×10−7m2∕s,T=343K, and remaining parameters as in the text. Note that due to symmetry, through-plane currents of pucks, which are symmetric about the centre of the channel, fall on top of each other so that only 8 different puck currents are visible. Through-plane currents Ij reach the maxima too fast. Puck currents 1, 8, 9, and 16 are marked.

Grahic Jump Location
Figure 11

Identical diffusion coefficients in all pucks: Variation of the channel parameter Kc (see also next two figures) mainly shows in the time scale of the quasi-steady state, Ts. Note that Ts cannot vary more than by a factor of 4 (for given Dj and experimental gas volumes), as shown in Eq. 42. Here, we choose Dj≡7.2×10−7m2∕s and Kc=3.45×108Pas∕m3.

Grahic Jump Location
Figure 12

Identical diffusion coefficients in all pucks: Here, we choose Dj≡7.2×10−7m2∕s and Kc=3.45×109Pas∕m3

Grahic Jump Location
Figure 13

Identical diffusion coefficients in all pucks: Here, we choose Dj≡7.2×10−7m2∕s and Kc=3.45×107Pas∕m3

Grahic Jump Location
Figure 14

Identical diffusion coefficients in all pucks: Variation of the GDL oxygen diffusivities (see also next two figures) mainly shows in the rate of change from quasi-steady state, Tth, and the orders of the current amplitudes, unless we go to extreme low values of Dj as in Fig. 1. Here, we choose Kc=3.45×108Pas∕m3 and Dj≡7.2×10−7m2∕s.

Grahic Jump Location
Figure 15

Identical diffusion coefficients in all pucks: Here, we choose Kc=3.45×108Pas∕m3 and Dj≡7.2×10−6m2∕s

Grahic Jump Location
Figure 16

Identical diffusion coefficients in all pucks: Here, we choose Kc=3.45×108Pas∕m3 and Dj≡7.2×10−8m2∕s

Grahic Jump Location
Figure 17

Identical diffusion coefficients in all pucks (see also next two figures): Change of time scale of drop in total current (top dotted line) at large channel flow resistance Kc=3.45×1014Pas∕m3 (see also next two figures). Due to decoupling of the (symmetric) pucks, all puck and in-plane currents fall on top of each other. Here, we choose Dj≡7.2×10−7m2∕s.

Grahic Jump Location
Figure 18

Identical diffusion coefficients in all pucks: Here, we choose Dj≡7.2×10−8m2∕s

Grahic Jump Location
Figure 19

Identical diffusion coefficients in all pucks: Here, we choose Dj≡7.2×10−6m2∕s

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In