Vehicle Fuel Efficiency Modeling

The results of the reformer analysis were integrated with a fuel cell electric vehicle model in order to assess the effect of reformer efficiency, reformer weight, and refórmate quality upon the fuel economy of the vehicle. The model, recently developed by Steinbugler (13), determines the power required for a given vehicle to cover an arbitrary driving cycle. It determines the size and weight of the fuel cell, fuel storage, and peaking device (if included) necessary to achieve a given performance specification. For this preliminary analysis, we have chosen parameters corresponding to a midsize vehicle having 1) a traditional steel chassis (Ford Taurus; 1050 kg glider weight), or 2) a lightweight body (aluminum-intensive Sable; 750 kg glider weight). The weight of two passengers (136 kg) was also included. A fairly high power-to-weight ratio (-0.06 kW/kg) was required to obtain performance comparable to contemporary automobiles, in particular the ability to accelerate 3 mph per second at 65 mph. City and highway driving were modeled using the federal urban and federal highway driving schedules (FUDS and FHDS, respectively). A peaking device, which can significantly alter the vehicle design and performance, was not employed in this initial analysis but is actively being studied.

The vehicle model was used to compare FCEV performance using 1) direct (compressed) hydrogen, 2) methanol steam reforming, and 3) gasoline/diesel POX. The reformers and fuel cell (both anode and cathode) were operated at 3 atm and a cathode stoichiometry of 2.0. The fuel cell polarization curve (14) was modified using the calculated anode overpotentials of Gottesfeld (15) to account for the effects of hydrogen dilution by C02 and N2. These effects are significantly larger than simple Nernst corrections, particularly at high current densities where mass transport effects are important For example, at 1 A/cm2, the fuel cell potential drops by an estimated 9.4% on SRM reformate and 20% on POX reformate. It was assumed that loss of fuel cell performance due to the presence of C02 could be prevented by the addition of a small amount (-2%) of air to the reformate (7). In the direct H2 system, the weight of the hydrogen storage (7.5% H2 by weight) was chosen as 50 kg, corresponding to a range of 200-300 miles. The specific power of the ADL POX reformer (4), 0.6 kWe/kg, is roughly equal to projections by General Motors (6) for SRM. Therefore, both processors are assumed to have the same weight calculated by scaling up 0.6 kWe/kg to obtain the required power and adding 15 gal of fuel. With a steel chassis, the vehicle requires a 150 kWe, 320 kg fuel processor/storage system; only 130 kWc, 270 kg are needed by the aluminum-intensive vehicle. It is assumed that the reformers can follow the load with an efficiency that is load-independent; furthermore, the shift reactors are assumed to convert all CO to H2 and CO2. The energy efficiency of the fuel processors, defined as the HV of H2 used by the fuel cell divided by the HV of the incoming fuel, is taken to be 0.76/0.79 (LHV basis/HHV basis) for SRM (10) and 0.71/0.78 for gasoline POX (12). In the highly endothermic SRM process, the system efficiency is only weakly dependent on fuel cell utilization (80%) because the anode exhaust is profitably used as fuel in the steam reformer burner. In the exothermic POX process, however, the assumed 90% H2 utilization directly yields a 10% reduction in system efficiency because the anode exhaust while useful for vaporizing fuel, for example, should not significantly increase the efficiency of a POX reformer having good thermal integration.

Table II. Preliminary estimates of fuel economy, in miles per equivalent gallon of gasoline (MPEG) for combinations of glider weight hydrogen source, and driving cycle.

Glider Weight (kg)

H2 Generation/ Storage System

Curb Weight of Vehicle (kg)




0 0

Post a comment