"Cell Conversion," Feature Article, April 2003

cell conversion

A DOE-sponsored program would ease fuel cells into cars—as backup power for all those electrical extras.

by Jeffrey Winters, Associate Editor

The stock in fuels cells rose—figuratively and literally—after President George W. Bush called for hydrogen-powered cars in his State of the Union address in January. After all, fuel cells and the so-called hydrogen economy go together like peanut butter and jelly.

You'd be excused, then, for thinking that the Department of Energy's Office of Fossil Energy would be the last place to promote fuel cell development. And yet, not only has Fossil Energy funded some research into large, stationary fuel cells, but the office is now sponsoring a $500 million public-private partnership, the Solid State Energy Conversion Alliance, or SECA, to develop a new generation of fuel cell. The goal is to make a durable and reliable unit suitable for stationary use or transportation by 2010. One of SECA's mandates is that the complete fuel cell system must come in under $400 per kilowatt.

Four different private companies and a number of government and university labs have taken on the challenge. The catch, if you can call it that, is that this fuel cell will have to run on diesel, gasoline, or methane. "It must work with the existing fuel structure, as well as with the sort of hydrogen infrastructure that the president proposed," said Wayne Surdoval, the SECA program manager.

There are four main fuel cell technologies, and Fossil Energy looked at all of them when it set up SECA. Two of the approaches—phosphoric acid fuel cells and molten carbonate fuel cells—work well, but they were judged to be fairly mature. Companies such as Fuel Cell Energy Inc. and UTC Fuel Cells already sell those kinds of fuel cells commercially, and Surdoval said that it seemed unlikely a research project could wring an order of magnitude cost reduction.

Another design, the proton exchange membrane, has been the technology of choice among those trying to develop fuel-cell-powered cars. But PEM cells are incredibly finicky about their fuel stream. "Just about everything contaminates a PEM," Surdoval said, "so the fuel processors get very expensive and complex when using traditional fossil fuels."

Almost by default, then, SECA turned to solid oxide fuel cells. Unlike phosphoric acid and molten carbonate fuel cells, SOFCs use a solid ceramic electrolyte, often based on a special form of doped zirconium oxide that conducts oxygen ions. Oxygen ions travel through the crystal lattice of the ceramic electrolyte to react with the fuel. Although solid oxide fuel cells have been around since the 1950s, they are believed to be far from mature. Siemens Westinghouse Power Corp. of Pittsburgh developed a tubular model in the late 1980s that showed tremendous promise. Siemens Westinghouse is developing SOFCs under the SECA program, as are Cummins Power Generation of Minneapolis, Delphi Automotive Systems of Flint, Mich., and GE Power Systems of Atlanta.

A Solid Option

Solid oxide fuel cells have a couple of inherent advantages over PEMs. First, they are much more tolerant of impurities in the fuel stream, greatly simplifying the fuel reforming process. Another, almost counterintuitive advantage is their operating temperature. PEMs work at around 70°C, far below the point where fuels like methane can be turned into hydrogen. That means heat must be bled away from the fuel stream before it reaches the membrane. SOFCs work closer to the reforming temperatures, which may lead to greater efficiency.

A critical edge might be potential cost. PEM cells depend upon platinum, and even tiny amounts of that precious metal can be expensive. By one estimate, a PEM fuel cell powerful enough to run a car would require $6,000 in platinum alone. The ceramic at the heart of an SOFC need not contain intrinsically expensive elements.

But SOFCs are still not cost-competitive. High operating temperatures—an advantage when it comes to fuel processing—create a constellation of problems. "You get very difficult materials interactions," said Steve Visco, a researcher who is leading fuel cell research at Lawrence Berkeley National Laboratory in California. "High operating temperatures also create problems in terms of seals and insulation." Manufacturing ceramic or ceramic-coated parts is far from trivial, and those parts must be treated with care. Visco said it is much like handling a 2-mm-thick dinner plate—when all-ceramic fuel cells fail, they fail catastrophically.

As a result, SOFCs are still too expensive to consider in most applications. "To get the cost down," Surdoval said, "it requires a new business model." The Office of Fossil Energy had often sponsored research into large, stationary power systems. With SECA, it now wanted modular units as small as 5 kW.

"We looked at what we could do with fuel cells that would provide more electricity onboard vehicles," said Jean Botti, chief technologist at Delphi's Dynamics, Propulsion, and Thermal sector and a leader of the Delphi SOFC effort. "We thought that the auxiliary power unit market would be much more attractive in terms of cost per kilowatt than the automotive market. The cost per kilowatt for an engine is a lot lower than the cost of onboard energy. Engines are extremely cheap."

Delphi's prototype auxiliary power unit converts gasoline into a hydrogen-rich gas and feeds it to a solid oxide fuel cell to generate electricity.

The onboard electrical needs of a car are no longer limited to the starter motor and a small radio. Heaters and air conditioning, power locks and windows, and even automatically adjustable seats have become standard. All draw electricity from the alternator. More specialized vehicles, such as semi-tractors and recreational vehicles, have even larger appetites that must be met through idling a diesel engine or switching on a separate diesel generator.

Those needs could be met just as easily by a 5 kW fuel cell system. But to develop such a system, Botti and the Delphi team would have to start from scratch. "The biggest stumbling block was to develop a stack that would handle the variation in temperature," Botti said. "Stationary stacks for commercial applications don't cycle; they work steadily for a long time. In a car, to the contrary, there is a lot of cycling. So we had to develop our own stacks."

Botti said his team has made great strides in tackling the cycling problem, though considerable work remains. Delphi has succeeded in removing almost all the precious metal from the cells, and in reducing the size and weight of the complete system stack, reformer, and electrical components. A 5 kW system that once took up 5 cubic feet now takes up less than a third of that space. Delphi claims that the progress it has made already has brought its system well along the path of commercial practicality in the auxiliary power market.

Bringing Down the Price

But if SOFCs are going to move beyond a niche role, Lawrence Berkeley's Visco believes they will need a radically different internal design. More specifically, Visco said, the bulk of the ceramic material in the fuel cell must be replaced by stainless steel, which is much less expensive.

The key, Visco said, is to lower temperatures. By reducing the thickness of the ceramic electrolyte (to between 10 and 20 microns deposited on a porous ceramic support structure), the operating temperatures for SOFCs have dropped dramatically. At 650°C to 800°C, metal can be used for many internal parts, including the framework that interconnects individual cells that, in series, form the stack.

To reduce costs, the amount of ceramic material in each cell has to be shaved to a bare minimum. That means it's necessary to find a substitute for the ceramic that supports the electrolyte. "Forget fabrication costs. Two millimeters of nickel zirconium puts you out of the ballpark in terms of cost," Visco said. "That's a problem."

To get those savings, the way the SOFCs are fired must change. Currently, the ceramics are formed in a traditional air-filled kiln. At the temperature needed to densify the ceramic—some 1,200°C—that air would quickly oxidize any exposed metal. To eliminate that oxidization, the Lawrence Berkeley team developed a firing process in a kiln filled with hydrogen that can produce a thin ceramic layer coating a porous structure made of stainless steel. Steel is stronger and more thermally conductive than ceramics (both of which are big pluses) and only one-sixth to one-tenth the cost.

By reducing the amount of ceramics in the cell, Visco said, the overall cost of an SOFC system can be brought down to the point where it can compete with diesel generators and gas turbines.

When SECA set its cost target at $400 per kilowatt, SOFCs were an order of magnitude too expensive. Surdoval suggested that the research to date has already driven down the costs by a factor of four. A revolutionary order of magnitude improvement seems well within reach.