meeting report
Year 2002 Fuel Cell Seminar
Palm Springs, CA USA
November 18-21, 2002
- CHEMAC International
Sterling Heights, MI
This meeting has been held every two years since its inception in the mid 1980s. Primarily organized by the Department of Energy (DOE), it is the largest and most comprehensive meeting held on the subject of fuel cells. Ten years ago, in 1992, about 500 people, mostly scientists, attended; this year’s attendance, greater than 2,500, comprised developers, engineers, managers, organizers, and promoters. The stated policies were then, as they are now, reduce dependence on foreign oil, conserve energy, and promote a hydrogen economy. That’s the political position; developers are looking for a way to make money.
As one approaches Palm Springs by road the windmill farms are an unforgettable image. There is no better place for discussions of fuel cells than Palm Springs where residents are interested in energy conservation and clean air and the SunLine Transit Authority is a leader in providing good bus service using the most advanced technologies available (and winter conditions are not encountered). For a few attendees who signed up early, SunLine provided a tour of their facility where their two fuel cell-powered buses were maintained and where the beginnings of a hydrogen economy could be observed. They call it their clean fuels mall. There were solar arrays (40kW) and a water electrolyzer (5kW, 70% efficiency, high pressure H2), hydrogen and hythane (a mixture of about 20% hydrogen and 80% natural gas, lower emissions than NG alone) fueling stations, and “tube” trailers for storage of hydrogen (3600 psi). Most SunLine buses run on natural gas but a few small buses are using hythane. A turbine generating station running on natural gas was generating electricity not only for their facility but feeding electricity back into the grid.
Over the years, the focus of the meeting has changed. In 1996 engineers and systems developers were more numerous than scientists; by 1998, people unfamiliar with fuel cells were in attendance seeking to understand business opportunities. The 2000 meeting reached a peak in hoopla. It seemed that government and quasi-government promoters were out in full force and a large number of engineers/analysts from energy utility companies were there to learn more about the prospects for their industry. The question at the meeting two years ago was, “Are the 1,900 attendees being misled or are they, in fact, leaders in a brand new industry?” The answer was not found – some left believing the promised land to be in sight; others left feeling that solutions to many problems are well into the future.
In 2000, I wrote that “many new firms have appeared to test the market and at least a half dozen have test units in the field, but the market cannot support all the wannabes. Although there are many low volume specialty markets and a potential military market of about $10-20 million per year, which are not very sensitive to cost issues and can support several small companies, the two big markets, residential and transportation, pose difficult cost issues. The residential unit has potential in the off-grid market (which is not as large as proponents think) but the business case for on-grid markets looks very weak. The automotive propulsion unit is farther away than most admit and the auto companies are already turning attention to auxiliary power units. Targets simply are not being met. Truck transportable units for utility companies widely mentioned in 1998 are nowhere in sight. Operating points have not changed in years, about 0.5W/cm2 for PEM hydrogen fueled and 0.15W/cm2 for direct methanol PEM fuel cells.”
This year’s meeting, back in the familiar Palm Springs setting after 2000’s show in Portland, Oregon, was a bit more subdued as the reality of the task has become more apparent. Some frustration was evident as realizations have not met expectations, but overall it was still a high energy environment with thousands of people attending and more than 100 exhibitors. There were two parallel sessions most of the time over the three days. The PEMFC, SOFC and DMFC were well addressed in overview talks, but much of the technical content was in the 220 or so poster papers. Poster sessions were crowded. Fortunately, this meeting had all the papers on CD (paper copies were available as a rather heavy book) so one could peruse the material at leisure. Disk or book may be ordered at www.gofuelcell.com
The exhibitors were a varied lot. I counted 23 fuel cell developers, 24 component suppliers, 18 material suppliers, 7 test equipment manufacturers and 16 government/trade agencies. IdaTech and Avista Labs had real fuel cells on display. There were at least six MEA suppliers present (3M, DuPont, Ion Power, Johnson Matthey, Superior MicroPowders, W.L. Gore) and four bipolar plate suppliers (Bulk Molding, Microponents, Mitisubishi, Schunk Kohlenstofftechnik). Other exhibitors of components were showing pumps, fittings, sensors, etc. Material suppliers were offering ceramic powders, carbons, foamed metals, etc. Test equipment ranged from passive loads to dynamic profile loads for exercising fuel cells.
It has been a bit of a problem to measure state-of-health (SOH) of a fuel cell. A few years ago we heard that fuel cells never fail. It is only in the past year that we have been hearing that fuel cells really do fail. Anyone with experience in electrochemical devices knows how difficult these things are. Almost all measures of durability have depended on open circuit voltage or voltage under load measurements. That is hardly adequate. Only recently have battery engineers used impedance measurements to detect failing cells. This technique could be useful with fuel cells. Measurements of crossover currents, polarization curves, current-voltage profiles can also be useful. There is likely much more to be learned about how fuel cells are failing. (See discussion later of membrane decomposition.)
Seminar participants examine Toyota’s FCHV4 Highlander, one of several fuel cell vehicles on display.
Greeting a visitor to SGL Carbon’s booth is David Wood, key account and technical manager for the St. Marys, Pennsylvania, firm.
What has happened since the last meeting? The industry has moved to the sampling and demonstration phase. The PEMFC (polymer electrolyte or proton exchange membrane fuel cell) developers have perhaps a couple of hundred units in the 500W to 5kW range out in the field in realistic situations with now a year or so experience behind them. The next phase, customer acceptance, is by no means a certainty. DuPont described a nine-layer MEA which, upon inquiry, turned out to include everything from one electrode backing to the other, i.e., a full cell. What does the fuel cell company do? The answer is assemble the stack and the system. So, which company is the real full cell manufacturer? The day may come when a fuel cell manufacturer may have no knowledge of electrochemistry. Another consequence of this situation, of course, is inertia to change.
Except for a couple of large utility demonstration installations, the SOFC companies are just beginning to get a few units out in the field. At this meeting there seemed to be a shift in the winds in favor of the SOFC away from PEMFC, but whether this is the result of disenchantment with PEMFC as the work gets harder or enchantment with SOFC as progress is made with planar constructions (which are more adoptable for smaller units than the tubular constructions) is hard to say. When I asked one SOFC expert, his candid reply was, “Probably some of both.”
The DMFC is still a work in progress hardly beyond the laboratory phase but more on this later.
Two keynote speakers had some interesting views. David Freeman, chairman of the California Consumer Power and Conservation Financing Authority, has been a long-time supporter of fuel cells. At the meeting, he received the Fuel Cell Seminar award for outstanding support of the industry. In his presentation, he made a case for the hydrogen economy and expressed disappointment that fuel cells were not ready when the energy crunch hit California – “an opportunity lost.” Are fuel cells ready now? “We dare not be deluded by the P.R.” But he is clearly optimistic. Dr. Stephen Jacobson, Sarcos Corp., is a well-known professor in robotic engineering whose company and its offspring have been responsible for many of the familiar robotic dinosaurs of the movies and amusement parks. His theme was the introduction of the human exo-skeleton, a means of carrying out movements of the body with the use of sensors and actuators to relieve the body of much of its burden – the soldier able to support much more hardware than the body alone, a handicapped person able to do necessary activities. This all requires energy on board, 10W up to hundreds of watts on peak demand (such as stumbling), and, of course, the fuel cell offers more than batteries can provide. This is not just modeling; hardware is in development. Five or ten years from now robotics may be an important market.
On the overview talks, Jerry Leitman of FuelCell Energy had several points concerning the stationary market. FuelCell Energy produces large MCFC for utility applications and has recently moved into smaller 200kW size units for the customer side of the meter applications. He notes that most failures are the result of distribution failures of the grid, so the backup device needs to be at the customer. In his view, fuel cells are ready today to serve the customer. He projects 1600MW of installed capacity in the U.S. by 2012 producing electricity at a cost of about 10 cents/kW. Those are very optimistic numbers. (Even more optimistic numbers came from Rastler of EPRI projecting, for instance, thousands of MW in residential units by 2008.) Ferdinand Panik of DaimlerChrysler (Andreas Schict presented for Dr. Panik) provided an overview of the transportation sector. This is familiar territory and he reiterated that target dates for commercialization have slipped to 2010 or later. (This is, of course, the on-road sector of the transportation market.) They have a new hydrogen-fueled vehicle in Europe, the Anne Van, with a 150-mile range. Demonstrations in the transportation sector will continue to 2006 when the next decisions will be made. One decision to be made is who will pay for a hydrogen infrastructure. The message seemed to be that the auto companies will not. Interestingly, he said NECAR feasibility was demonstrated in 1997. It seems some years have been dropped from the history of Daimler’s fuel cell cars. Larry DuBois of SRI International and PolyFuel made the case for fuel cells in portable applications vis-à-vis batteries using the familiar arguments. He predicted 1.6K portable fuel cell units by the end of 2002, 15M units by 2007, and considers butane a possible fuel for portable SOFC units. (PolyFuel, a spinoff of SRI, is a developer of small direct methanol fuel cells.) Don Huberts, CEO of Shell Hydrogen B.V. and present chairman of the California Fuel Cell Partnership, covered the matter of fuel supply. He said the fuel companies will provide the infrastructure needed. He sees a mixture of packaged fuels and distributed fuels, central production, and on-site production, mainly of hydrogen. The Methanol Fuel Cell Alliance, of course, has a somewhat different vision. Papers by Dolan and Heffelfinger point out many of the possible problems with the hydrogen economy and potential advantages of methanol.
What is happening in the rest of the world? The European Union is investing $60-70 million in fuel cell research and technology development per year. In addition, several countries, particularly the Netherlands and Germany, are investing in specialized developments. The Japanese government supports fuel cell development through its NEDO and MITI programs in the tens of millions of dollars per year. All the major Japanese companies active in related industries (such as batteries) have programs similar to those found in the U.S. For instance, Toyota’s Highlander is equipped with the company’s own fuel cell. Honda has an in-house program as well as outside sources. Sanyo, Toshiba, Fuji Electric, etc., all have developmental activities comparable to those found in the U.S.
An attendee has a question for Emma Farndon (right), research scientist at INEOS Chlor of the United Kingdom.
Demonstrating Avista Laboratories’s modular fuel cell units are application engineers Ken Hydzik (left) and David Holmes.
One of the fuel cell-powered buses at SunLine was produced by Thor Industries and equipped with an ISE Research power train. This was described in a presentation by Paul Scott. It is a hybrid power system with a hydrogen fueled UTC PEM fuel cell and Panasonic lead acid batteries. About 14kWh of battery capacity at 576 volts are on board providing for power assist above 60kW. They expect to replace the batteries at two-year intervals, although the bus has been in service only two months so far. They also mention planning a Zebra battery in the future. (Metal hydride batteries in Toyota’s electric RAV4 and hybrid Prius have already routinely demonstrated over 100K miles of service but that’s another story.) They will not reveal voltage of the fuel cell stack but acknowledge DC/DC up-converters. The bus gets 8-11mpg, depending on air conditioner usage, about double that of a conventional bus. The other SunLine fuel cell bus is a product of the Georgetown University program. It has two Ballard Mark 5 units in series providing 100kW using methanol fuel and reformer. In addition to the Thor bus, three fuel cell-powered vehicles were on display at the meeting: DaimlerChrysler’s NECAR 4 using a 300V drive train and a Ballard 70kW fuel cell running on high pressure H2; Ford’s P2000 Th!nk vehicle with a Ballard Mark 700 fuel cell running on high pressure H2; and a Toyota FCVH4 vehicle (the Highlander) with a hydrogen fuel cell and a Prius-type hybrid battery. It was easy to see why the Highlander rather than the Prius was chosen; the fuel cell system (no one wanted to talk about the voltage or number of cells in the stack or stacks) took most of the space under the hood of this big engine vehicle.
MEA development for PEMFC is key to advancing this art, but radical departures from prior art or breakthroughs were not mentioned. The best combination of conductivity, mechanical strength, chemical stability, and costs determines what membrane is used. The perfluorosulfonic acid membranes of the Nafion type hold the fort. Developers have learned to handle thinner films, disperse lower catalyst loadings, and improve bonding to gas diffusion layers. Operating point is typically about 0.6W/cm2 at 80EC with H2/air. Durability is still a question as operating experience is limited. Cleghorn of W.L. Gore and Associates discussed decomposition of the membranes by loss of fluorine which has not been previously mentioned. A well-known failure mechanism, sintering of catalyst particles, seems to have been nearly eliminated. But durability seems to be measured in the 10K to 20K hour range (one to two years) rather than the >40K range needed for stationary applications. Degradation is related to fuel composition, the longest lifetime is obtained with pure hydrogen. Katagiri and other authors from Tokyo Electric Power described degradation by loss of sulfonic acid groups. Mentioned was the PBI-PA (polybenzimadazole-phosphoric acid) membrane which could permit operation at temperatures as high as 160EC where less poisoning of electrodes by CO occurs and lighter catalyst loadings can be used. This was mentioned by Zawodzinski as a development from the Case Western group. Plug Power is having some success with this and sees it as a major advance, but others are more skeptical. And there’s the problem – to get the advantages of higher temperature operation, there is a need to get away from aqueous, proton-conducting electrolytes only to enter the world of corrosive environments, lower conductivity, and/or solid state, oxide ion conductors. DuPont was talking of a new solution casting process expected to lower costs of Nafion membranes. They talked of 1mil films, even 0.5mil, (thinner aids water balance in the stack and decreases stack resistance) and potential costs of $50/m2. Operating point was about 0.5W/cm2, at 65EC, with H2/air, using 0.7mg/cm2 total loading.
Another key component is the bipolar plate. A few years ago, resin-bonded graphite plates were subject to swelling but advances seem to have been made in long-term dimensional stability. Several presentations and exhibitors described compression-molded thermoplastic resin plates which seem to have good dimensional stability. The costly machined graphite plates now seem to be past history. There is still some use of coated metal plates which are thinner and have higher conductivity, just as a fuel stack with metal plates has higher power density. PemCoat was exhibiting its coating system. Years of development seem to be paying off with higher quality, lower cost bipolar plates but, because of requirements for precision manufacturing and redox-resistant materials, they are not dirt cheap. With advances in MEAs and bipolar plates, some optimistic projections of costs are about $500/kW for a PEM stack, although it should be noted there are those who do cost projections and there are those who make a product and the latter usually come in two to three times higher.
Frano Barbir, Proton Energy, made the case for fuel cells in the telecom power backup application. (Proton Energy makes electrolyzers as well as fuel cells.) The lead acid battery is not perfect but it is tough competition. His numbers were: installed cost of PEM fuel cell $5000/kW, operating life <5K hours, system efficiency 30% (regen included). Goals are <$1000/kW, >40K hours (almost all standby time), efficiency 40%.
An interesting note of this meeting is how the hoopla about fuel processors has died down. Not that they are unimportant – in fact, they are a necessary part of the stationary PEM fuel cell system – but with growing interest in the hydrogen economy and perhaps a shift to high temperature fuel cells there seems to be reduced interest. Komiya of Tokyo Gas and Shinke of Osaka Gas described fuel reformers for residential use, the former describing a 5kW, 19 liter unit working at 65-75% efficiency with natural gas and the latter a 50 liter unit, about 75% efficient (92% methane conversion) which had achieved 8000 hours of operation. They have targeted a life of 90K hours including a monthly shutdown and startup.
Zizelman described the SOFC auxiliary power unit for vehicles which is in development by Delphi under a DOE program. Delphi is quick to point out the apparent advantages of SOFC over PEMFC in the APU application: higher power density, higher efficiency, temperature compatibility with exhaust gas, can burn “dirty” fuel, etc. The report covered experience with the Generation 2 device which is designed as a 42V, self-contained 5kW unit capable of running independent of the engine. The stack (actually two 30-cell stacks) was provided by Battelle Pacific Northwest National Laboratory. A partial oxidation reactor provides hydrogen from gasoline. It appeared the box was 50 liters in volume, mass 25kg. The operating point seemed to be about 0.35W/cm2 of electrode area at 750EC. Operating below about 800EC permits use of metallic interconnects. Start-up time was about 30 minutes.
The MCFC drew little attention. FuelCell Energy covered the bases mentioning a commercialization plan for their 200kW units and their multi-MW utility sites. They have installed 50MW/year manufacturing capacity.
The direct methanol fuel cell was addressed in poster papers and several oral presentations, but the most interesting was an exhibit by Forschungszentrum Juelich of a 2.5kW, 135-cell, working unit. Fz-juelich was feeding the stack 0.5 molar methanol. It was operating at about 60 mW/cm2 and achieving about 160W/l. There is considerable interest in the progress of small (1-30W) direct methanol fuel cells for portable devices. Bostaph, Motorola Labs, tackled a description of the issue by describing a 1W desktop charger for telephone batteries and a proposed unit for two-way radios. This is a group that has worked on these miniature fuel cells for more than five years and has now reached the demonstration phase of a 1W unit. It is a nice effort, but there seems to be growing realization of how tough the battery competition is. It will take a very competitive unit to displace Li-ion batteries in cell phones, and Bostaph seemed to say they were not optimistic. Perhaps a laptop computer (700Wh at 30W?) might be a better target. He noted a DMFC for this might well need to be a separate unit because of heat dissipation, safety, and contamination reasons. After hearing this talk, the laptop application appears to be a difficult undertaking, but one is still left with the thought that within the portable electronic devices market there is something for the fuel cell to fit.
In summary, there has been progress in engineering but the marginal rate of improvement is declining as the easier tasks are completed. State of health monitoring is in its infancy. The hydrogen economy raises questions and the fuel problem remains difficult. Several speakers were saying their companies were ready now to deliver, although PEMFCs are barely in the demonstration phase and SOFCs are further back. There are niche applications now – either military or very specialized – including some off-road transportation applications and some off-grid residential/industrial applications. Projections of costs even as low as the $600/kW range for planar SOFC are there, but buyers are paying from $2,000 to as much as $20,000/kW right now. At least 30 fuel cell companies can be identified. In any market, the fuel cell is likely to experience tough competition, and business plans seem to be widely varied.
Showing some fuel cell hardware to FCT’s photographer is Richard Hodge, regional sales manager for Parker Hannifin Corp.
DHolding a resin-bonded bipolar plate is Henrik Neuweger of Schunk Koblenstofftechnik GmbH, a German R&D firm.
A 2.5kW direct methanol fuel cell (DMFC) working unit by Germany’s Forschungszenthum Juelich GmbH attracts attention.