In the not so distant past, the fuel cell stack, just one part of the fuel cell power system, cost a commercially infeasible $5,000 per kilowatt (kw.) Much progress has been made since.
The California Air Resources Board (CARB) established the Fuel Cell Technical Advisory Panel (FCTAP) in 1996 to assess the status and prospect of fuel cells for auto use. The FCTAP’s 1998 study concludes:
"The Panel’s visits and discussions with leading developers and potential manufacturers of PEM fuel cell components and stacks made clear that the fundamental technical barriers to the development of automotive fuel cell stacks have been overcome by the advances achieved over the past 5–7 years. The large increases in the specific performance of PEM cells and stacks also have lowered the cost barriers to the point where future mass production may be able to meet the stringent cost goals for critical cell components and stacks intended for automotive applications." (Kalhammer et al., 1998, p. III-21)
"…[On] the basis of the performance already achieved with preprototype stacks, there appear to be reasonable prospects for meeting the $20/kW ($1000/50kW) cost target for automotive fuel stacks if production reaches about 100,000 to 200,000 units per year. At this level, the most critical stack components—membrane, MEA and separator plate—will reach production volumes that justify true mass manufacturing methods." (Kalhammer et al., 1998, p. III-54).
RAND 2002 review of Fuel Cell power system costs notes Ford Motor Company and Directed Technologies (DTI) concluded in a 2000 study that costs for PEM stacks will range from $19 to $28 per kilowatt in high-volume production. The range is due to the overall stack power rating (e.g., 60 to 80 kW). Costs include material, manufacturing, and assembly costs, as well as markups to reflect profit, scrap, R&D, and administrative costs.
RAND notes Ford and DTI anticipate further cost reductions require a reduction in platinum catalyst loading or in gas diffusion electrode costs, because the costs of other components are based on “mature” manufacturing technologies.
The RAND study itself concludes:
“the projections that fuel-cell system costs can fall to $35 to $60 per kilowatt in volume production are well grounded. Costs might be even lower if the very low platinum designs currently being investigated pan out. However, it is too early to tell if this will be the case.”
Assuming a mid-range price of $50/kw, a 50kw automotive power system would cost $2,500, comparable to the cost of an internal combustion engine.
It is important to note that the key cost driver with regard to fuel cells is the amount of platinum required. This
precious metal is used as a catalyst in the membrane electrode assembly (MEA). Studies that present higher cost
estimates assume very high platinum loadings. An Arthur D. Little (ADL) study for instance assumed a requirement
of 0.8 mg/cm2, of platinum. For comparison, the Ford/DTI study assumes 0.25 mg/cm2, and the FCTAP study cites
designs that range from 0.25 to 0.35 mg/cm2. DOE claims a breakthrough of 0.15 mg/cm2 for experimental MEAs in
laboratory quantities from a 3M design. A Los Alamos National Laboratory 2002 study describes PEM fuel cells with
low platinum loads (0.11 and 0.15 mg/cm2) that have been tested to 4,000 hours. Ballard Power has reduced
its platinum requirement to roughly that used in the catalytic converters of today’s vehicles.
The ADL study is based on a fuel cell that can accept hydrogen reformed from gasoline. This may account for the much higher catalyst loading, as such fuel cells likely require more platinum as well as rhodium to protect against impurities that remain after the reforming process (note that other studies based on hydrogen reformed from gasoline also present substantially higher costs.)