hydrogen storage or an electric motor, and thus should not be compared directly to the costs of an
internal combustion engine drivetrain, costs this low would appear to make the fuel cell a viable
competitor with the ICE--and it would be several times as efficient. General Motors projects the
efficiency of a PEM cell to be 55 percent to 70 percent using hydrogen fuel or 40 percent to 55
percent with methanol as a hydrogen carrier. Energy density currently is about 200 W/kg, but GM
hopes to raise this to 333 to 500 W/kg.^96 Mercedes Benz has recently demonstrated a prototype
PEM cell that operates a van. Although the existing system occupies essentially all of the van’s
cargo space, Mercedes apparently believes it can have a production prototype ready within 5
years or so.^97
The PEM fuel cell stack has been the subject of extensive research over the last five years, and
some recent designs, especially by Ballard, have shown considerable promise. The current Ballard
cell has a specific power rating of only 200 W/kg, equivalent to that of advanced lead acid
batteries, and has demonstrated full load efficiencies in the range of 36 to 46 percent.^98 Although
there have been some assertions that commercial PEM fuel cells can be available relatively
quickly, most researchers suggest that a commercial model is still at least 12 years away, and such
swift commercialization would require both continued government finding of research and rapid
resolution of a number of remaining problems. Pessimistic assumptions on these factors leads to
an estimate of 20 to 25 years for commercialization.^99 The goals are to double the specific power
and reduce cost by an order of magnitude or more while increasing efficiency to more than 50
percent.Current PEM fuel cells have been built with relatively high platinum loadings for the catalyst,
and use expensive membranes which some believe are “over-specified” for automotive use.
Moreover, the graphite bipolar plates are expensive. Highly conducting, corrosion resistant
alternatives are needed to reduce costs in this area. Large reductions in platinum loadings--thus
far achieved only in small laboratory cells--and cheaper membrane technologies also are required
if the PEM fuel cell is to be manufactured at reasonable cost. Significant progress has been made
in these areas, especially in reducing platinum loading at the laboratory cell level, although much
remains to be done to scale up to an EV size stack. It is unclear whether cheaper membranes and
plates will result in efficiency reductions, creating tradeoffs between competing goals. Current
PEM fuel cells also require very pure water to hydrate the membrane, and, hence, startup at low
temperatures poses difficulties with freezing.
Although the PEM stack fueled by hydrogen itself can be quite efficient (about 60 percent at its
maximum efficiency point, about half of rated power), the accessory drives require power that
detracts from overall system efficiency.^100 As noted above, the drives provide hydrogen to the
anode, compressed air to the cathode, water to hydrate the membrane, and a cooling system to
remove waste heat, all of which requires substantial power. For example, a 25 kW stack that is 50
percent efficient at rated power will generate 25 kW of heat to be removed by the cooling system,(^96) General Motors briefing charts.
(^97) Daimler-Benz, High Tech Report, March 1994.
(^98) P. Howard, "Ballard Zero Emission Fuel Cell Bus Engine," paper presented at the 12th International EV Symposium, 1994.
(^99) H.F. Creveling, "PEM Fuel Cell for Transportation Applications,” paper presented at the Automotive Technology Development Contractors
Coordination Meeting, U.S. Department of Energy, October 1993. 100
C. Borroni-Bird, Chrysler Corp., "The Challenges Facing Fuel Cells for LDV Applications,” presentation to OTA, Sept. 19,1993.