The Polymer Electrolyte Membrane Fuel Cell (PEMFC) has been projected to be the fuel cell of choice for future automotive applications. Among the most challenging aspect of this application is the occurrence of severe and frequent changes in power demand. In previous years [1, 2], we have presented a dynamic PEMFC model that captured the interdependencies between the electrochemical, chemical, thermal and membrane hydration phenomena. We then illustrated how feedback control can be used to mitigate the occurrence of membrane dehydration and Gas Diffusion Layer (GDL) flooding while responding to changes in power demand from the load. However, after each of these presentations, the audience would raise concern about the appropriateness and characteristics of the load used in our analysis.

The current paper is in response to these concerns. Specifically, we will present a load model aimed at mimicking the load expected in a fuel cell vehicle, including a DC motor, a number of DC-DC converters and the option of including peak-shaving components such as a rechargeable battery and/or a super-capacitor. The model has also been extended to include the kinematics of the vehicle (rotational and translational inertia as well as a simple wind resistance model), and thus can be connected to standardized drive cycle scenarios available from the EPA, [3].

One reason that the proposed load model is infrequently connected and analyzed alongside a detailed dynamic PEMFC model is that most researchers believe that the time constants associated with each are disparate enough to assume pseudo steady-state operation with respect to the fuel cell. While this is certainly true for the electrochemical and chemical aspects of the cell, it is not true for the thermal and hydration aspects, both of which have time constants on the order of minutes, which is certainly on par with the time scale of a typical drive cycle.

In contrast to a simple resistive load, the moderate complexity loads we have used in the past (usually termed constant current, constant voltage or constant power loads) contain an internal feedback structure that manipulates the load impedance so as to track a set-point signal (of current, voltage or power depending in the load type). However, this presumption of being able to manipulate the load impedance (at will) is what we and others have called into question. In contrast, the manipulated variable within our newly proposed load is the well defined the gain signal to the DC-DC converter. The load controller's feedback loop is then closed by selecting vehicle speed as the control variable, whose set-point is defined by the drive cycle data. In the case of a peak-shaving configuration (i.e., including a rechargeable battery and/or super-capacitor), multiple DC-DC converters will be needed and thus will supply additional manipulated variables. In this case, design of the load controller is non-trivial and will be discussed is detail.

In summary the objective of this presentation is to illustrate the expected interactions between the load / load controller and the PEMFC (in closed-loop with respect to temperature and humidity and open-loop/feed-forward with respect to fuel/oxidant flow). Of particular interest is the identification and avoidance of operational scenarios and controller designs that result in membrane dehydration and/or GDL flooding.

[1] Chmielewski & Lauzze, “Power Control of a Polymer Electrolyte Fuel Cell,” paper 314g, 2005 Annual Meeting of the AIChE, Cincinnati, OH. (Also, available as Lauzze, K.C., and D.J. Chmielewski, “Power Control of a Polymer Electrolyte Membrane Fuel Cell,” Ind. Eng. Chem. Res., 45, pp 4661-4670 (2006).)

[2] Ahmed & Chmielewski, “The Impact of Hydration Dynamics on the Control of a PEM Fuel Cell,” paper 631g, 2006 Annual Meeting of the AIChE, San Francisco, CA

[3] http://www.epa.gov/nvfel/testing/dynamometer.htm