130
ENERGY CENTER OF WISCONSIN Report Summary 210-1 Life-Cycle Energy Costs and Greenhouse Gas Emissions for Gas Turbine Power April, 2002 report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report report energy center

Review of State-of-the-Art Fuel Cell Technologies for Distri

  • Upload
    vanthu

  • View
    217

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Review of State-of-the-Art Fuel Cell Technologies for Distri

ENERGY CENTEROF WISCONSIN

Report Summary210-1

Life-Cycle Energy Costs andGreenhouse Gas Emissions forGas Turbine Power

April, 2002

report report report report report

report report report report report

report report report report report

report report report report report

report report report report report

report report report report report

report report report report report

report report report report report

report report report report report

report report report report report

report report report report reportenergy center

Andrea
Andrea
Review of State-of-the-Art Fuel Cell Technologies for Distributed Generation
Andrea
January 2000
Andrea
Andrea
193-2
Andrea
A Technical and Marketing Analysis
Page 2: Review of State-of-the-Art Fuel Cell Technologies for Distri

Report 193-2

Review of State-of-the-Art Fuel Cell Technologies for Distributed Generation

A Technical and Marketing Analysis

January 2000

Prepared by

Robert J. Braun, Sanford A. Klein & Douglas T. Reindl Solar Energy Laboratory

University of Wisconsin-Madison Madison, WI 53706

Prepared for

595 Science Drive Madison, WI 53711-1076

Phone: (608) 238-4601 Fax: (608) 238-8733 Email: [email protected]

WWW.ECW.ORG

Page 3: Review of State-of-the-Art Fuel Cell Technologies for Distri

Copyright © 2000 Energy Center of Wisconsin All rights reserved

This report was prepared as an account of work sponsored by the Energy Center of Wisconsin (ECW). Neither ECW, participants in ECW, the organization(s) listed herein, nor any person on behalf of any of the organizations mentioned herein:

(a) makes any warranty, expressed or implied, with respect to the use of any information, apparatus, method, or process disclosed in this report or that such use may not infringe privately owned rights; or

(b) assumes any liability with respect to the use of, or damages resulting from the use of, any information, apparatus, method, or process disclosed in this report.

Project Manager

Ruth Urban Energy Center of Wisconsin

Page 4: Review of State-of-the-Art Fuel Cell Technologies for Distri

i

Executive Summary

The research presented herein summarizes Phase 1 of the ESERC-funded project entitled, Distributed Fuel Cell Technologies Survey and Benefit Study. This work was performed by the Solar Energy Laboratory at the University of Wisconsin-Madison for the Energy Center of Wisconsin. The Executive Summary will highlight both the technical status of the “near-term” fuel cell technologies and the anticipated commercial markets for first-entry fuel cell power plants.

Stationary applications of fuel cells are being targeted for every sector of the world energy market, from small residential cogenerators to large central power generating stations1. The recent technological advancements discussed in Section III of this report have helped move fuel cells from a mythical, next generation power-generating device, to a realizable commercial product. The maturation of fuel cells has finally positioned the technology to capitalize on their inherently high efficiency and low emissions. However, as with any newly emerging technology, the “rules” have yet to be established and, thus, from developers’ perspective the opportunity to shape and lead the industry is great. This is particularly true for fuel cells as their numerous advantages open up the doors for many potential applications. In particular, the modular nature of fuel cells greatly enhances their ability to be scaled. The application range includes (i) on-site systems from perhaps 3 kW to 3 MW, (ii) distributed and substation systems ranging from about 300 kW to 50 MW, and (iii) central stations from 50 MW to several hundred MW2. For markets above 50 kWe, the current competing technologies provide the sector’s needs with capital costs well below that of today’s fuel cell systems. While it is generally agreed that fuel cell capital costs need to reach $1500/kW to penetrate the above markets, many proponents of fuel cell technology believe that because the potential market is so large and because there are some markets (however few) where $3000/kW is competitive, only a small percent penetration would be enough to succeed commercially3.

Additionally, fuel cells are being touted as the heir apparent to high-temperature combustion systems because of their extremely low emissions (well below anticipated regulatory levels), high fuel-to-electric “conversion” efficiencies, low noise levels, and fuel flexibility. Despite their advantages, fuel cells will receive the toughest competition from new generation gas turbines, which can obtain nearly as high efficiencies with relatively low emissions at capital costs in the range of $600 to $800/kW4. Due to the low capital costs of turbines for systems greater than 10 MW, fuel cells are believed to have the best initial market penetration in the 30 kW to 10 MW range.

External market forces that may impact fuel cell commercialization are federally (and/or internationally) regulated emission standards and, in the U.S., the deregulation of the power markets. In tightly regulated emissions zones, such as California, fuel cell systems may be mandated even with their currently prohibitive capital costs exceeding that of other competing technologies. Another external driver is that the ever-increasing demand for energy has created over-loads at existing transmission and distribution stations. Distributed generation technologies, such as fuel cells, are likely alternatives to alleviate congestion. In addition to the

Page 5: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

ii

currently high capital costs of fuel cell systems, the deregulation of the power market represents an additional hurdle for fuel cell sales, in that areas where fuel cells might be attractive may be negated by the potential customer’s ability to access lower cost energy suppliers. With elimination of monopoly status, the gas and electric utilities, which have traditionally helped pioneer new technologies, may be more inclined to avoid higher risk options like fuel cells. However, while more traditional utilities may adhere to lower risk options, other utilities, such as Southern California Edison, are aggressively pursuing alternative distributed generation options. Further viewpoints on distributed generation and its potential impact on fuel cell commercialization will be taken up later in this summary.

Fuel Cell Technology Status (Technical Summary) Fuel cells were first invented in 1839, but remained undeveloped until the late 1950s. Presently, there are at least six different fuel cell types in varying stages of development. In general, the various fuel cell types are differentiated by their electrolyte and operating temperature. There are currently four types of fuel cells that are considered “near-term” and that are receiving the most development attention. Listed in order of increasing operating temperature they are: the 80°C Proton Exchange Membrane (PEM) fuel cell, the 200°C Phosphoric Acid Fuel Cell (PAFC), the 650°C Molten Carbonate Fuel Cell (MCFC), and the Solid Oxide Fuel Cell (SOFC) currently at 1000°C. Table 1 gives a summary of fuel requirements and characteristics of the above fuel cell types. A summary highlighting the salient features of each fuel cell type follows and in-depth summaries are available in the separate chapters of this report.

Proton Exchange Membrane Fuel Cells (PEMFCs) The proton exchange membrane fuel cell (PEMFC) is receiving increasing attention as major technical milestones have been reached. In the last several years, cell power density (power out per unit weight or volume) has increased by an order of magnitude and substantial cost reductions have been achieved. These advances are largely the result of a technology push to market PEMFCs in the transportation sector.

The proton exchange membrane fuel cell typically operates at about 80-85°C (185°F). This temperature is lower than the PAFC, MCFC, and SOFC which, are in a more advanced state of development. The operating temperature is set by both the thermal stability and the ionic conductivity characteristics of the polymeric membrane5. To get sufficient ionic conductivity, the proton-conducting polymer electrolyte requires liquid water. Thus, temperatures are limited to less than 100°C. The low-operating temperature allows the PEMFC to be brought up to steady-state operation rapidly. This characteristic, coupled with the lightweight, high power density features makes PEMFCs attractive for transportation applications. However, the low-temperature operation results in low-grade waste heat that is not suitable for most cogeneration applications and makes thermal integration with fuel processing equipment difficult. As Table 1 shows, water is the medium typically used to cool PEM fuel cell stacks.

The PEMFC can operate at elevated pressures and it is often beneficial to do so in order to obtain higher power densities. Air pressures up to 8 atm have been used. Also, the solid polymer membrane can support substantial differential reactant pressures which enables some flexibility in the PEMFC system design and opportunities for maintaining water balance in the

Page 6: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

iii

membrane.6,7 The system efficiency of a stationary PEM power plant is expected to approach 40% (LHV).

As with many of the lower temperature fuel cells, the PEMFC requires a pure H2 source for operation. Since hydrogen is not readily available, it is typically obtained by reforming a hydrocarbon fuel, such as methanol or natural gas. The reformed fuel often contains other gas species such as CO, which are detrimental to fuel cell operation. CO ppm levels of 50 or greater poison the catalyst, causing severe degradation in cell performance. Poisoning of the platinum catalyst occurs because of “tenacious adsorption” of CO on the catalyst sites8,9. Fuel type and processing represents the single greatest challenge facing fuel cells in general. This is particularly true for PEMFCs considering their susceptibility to electrocatalyst poisoning from low-level CO content. Given sufficient fuel processing, PEMFCs are expected to operate using hydrogen, methanol, propane, and natural gas fuels (and eventually gasoline).

The main technical issues that PEMFC developers face are (i) electrocatalyst poisoning by low-level CO concentrations in the fuel feed gas, (ii) water management and membrane operating temperature limits, (iii) membrane and system balance-of-plant costs, and (iv) cell life.

The excitement surrounding PEMFCs is tempered by the reality that many technological barriers stand between failure and success in the ultra competitive automotive market. For this reason, it is expected that PEMFCs will first be marketed in stationary applications. The low-operating temperature, rapid startup, high power density, and simplicity characteristics that make the PEM fuel cells attractive for transportation, also make them attractive in remote, standby, and premium power onsite markets. Ballard Power Systems of Vancouver, British Columbia is the leader in PEM technology and is pursuing both transportation and stationary PEM applications. Plug Power of Latham, NY and H-Power, IFC, and Allied Signal will also compete in stationary PEM applications.

Phosphoric Acid Fuel Cells (PAFCs) The PAFC is the only commercially available fuel cell today. PAFC technology is well ahead of the other fuel cell types and is lead by International Fuel Cell (IFC) Corporation, in Massachusetts, U.S.A. IFC has been selling (with DOD subsidies) 200 kWe packaged, PAFC cogeneration units since about 1993 and have filled about 160 orders, with over 130 units operating in the field10.

PAFCs typically operate near 200°C (400°F). Cooling of the fuel cell stack is accomplished with pressurized boiling water. As with all fuel cell types, the PAFC operates on hydrogen which is typically delivered from a natural gas-supplied reformer. Like the PEMFC, during operation hydrogen ions migrate from anode to cathode. PAFCs can also operate at elevated pressures (up to 8 atm); however, the current packaged unit offered by IFC operates at ambient pressures.

The electrolyte material consists of 100% phosphoric acid (H3PO4) which acts as a transport fluid for the migration of dissolved hydrogen ions from the anode to the cathode, and to conduct the ionic charge between the two electrodes in order to complete the electric circuit. Because the electrolyte is a liquid, evaporation and migration are carefully controlled. Like the PEM, the

Page 7: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

iv

PAFC also employs platinum electrocatalysts in the cell electrodes. This limits the amount of CO the cell can tolerate before performance degradation. The present limit is about 2% (by volume) before cell voltage begins to decay.

Although PAFCs have been demonstrated up to the 11 MW size range, they are initially targeted to penetrate commercial and light industrial markets. The commercially offered PC25 packaged PAFC cogeneration unit can provide 200 kWe (at 36% efficiency) and about 750,000 Btu/hr (at 150°F) thermal energy for an approximate selling price of $3000/kWe ($4000/kW installed). While technical barriers to PAFC market penetration are few, their high capital cost is prohibitive. In sufficient volumes, mature PAFC technology is expected to attain $1500/kWe.

Molten Carbonate Fuel Cells (MCFCs) Molten carbonate fuel cells (MCFCs) are often considered a 2nd generation fuel cell as they are anticipated to commercialize after phosphoric acid fuel cells. Like PAFCs, MCFCs are a liquid electrolyte-based fuel cell that makes use of flat, planar configured fuel cell stacks. However, at 650°C (1200°F), the operating temperature is substantially higher and the molten carbonate electrolyte makes for a more corrosive environment than that encountered in the PAFC. The higher operating temperature enables internal reforming of hydrocarbon fuels, improving system design and efficiency. Additionally, the elevated operating temperature combined with fast electrode kinetics eliminates the need for expensive noble metal electrocatalysts and results in the highest electric efficiency of all fuel cell types. These same features also allow fuel flexibility, as carbon monoxide is not poisonous to the MCFC. These advantages are somewhat offset by the more severe material requirements for high temperature operation in the corrosive and oxidizing environments. Cell materials that demonstrate the necessary corrosion stability at reasonable costs have been primarily stainless steel alloys, ceramic composites, and semiconducting oxides. Long cell life (~40,000 hours) with the current state-of-the-art cell components represents a significant and ongoing technical challenge to the development of a commercial MCFC product.

Molten carbonate fuel cells typically consist of a lithium-potassium or lithium-sodium based electrolyte. After the cathode reaction, carbonate ions migrate through the electrolyte to the anode side of the cell to complete the fuel oxidation. Because of the CO2 requirement at the cathode, and production of it at the anode, CO2 must be transferred from the anode exhaust to the cathode inlet. This is normally accomplished through mixing of the anode exhaust with incoming air (oxidant) or by physically separating the CO2 from the other exhaust gas species through a “product exchange device11”.

MCFCs and SOFCs are considered “high” temperature fuel cells that offer several advantages for distributed or utility-scale power generation. At an operating temperature of 650°C, high quality waste heat can be produced and used for (i) fuel processing and cogeneration, (ii) internal methane reforming, and (iii) to drive a bottoming cycle. The MCFC, like other fuel cell types, exhibits improved cell performance under pressurized conditions. Today, pressurized operation is being pursued by about half of the molten carbonate fuel cell manufacturers.

U.S. MCFC developers are targeting commercial product size ranges between 250 kW – 3 MW. The leader in MCFC technology is Energy Research Corporation in Danbury, Connecticut,

Page 8: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

v

followed by MC-Power in Burr Ridge, Illinois. Both manufacturers are currently in later stages of the demonstration phase of their respective technologies. Early production units are projected to be available beginning in the 2001-2002 timeframe.

Solid Oxide Fuel Cells (SOFCs) Solid oxide fuel cells (SOFCs) are rapidly approaching commercialization in both commercial and distributed generation market sectors. Two different solid oxide fuel cell geometries are being developed, tubular and planar. The tubular design is the most advanced and, under the development of Siemens Westinghouse, has reached the field unit demonstration phase of commercialization. To date, Siemens Westinghouse has demonstrated 1 kW, 25 kW, 100 kW, and 250 kW tubular solid oxide fuel cells stack technology. Planar SOFC technology has lagged tubular development, but is rapidly hurdling performance and cost barriers, and some developers state that they will be situated to offer a commercial product as early as 2001. Solid oxide fuel cells employ a solid state electrolyte and operate at the highest temperature (1000°C/1800°F) of all fuel cell types. The SOFC uses a solid yittra-stabilized zirconia ceramic material as the electrolyte layer. During operation, oxidant (usually air) enters the cathode compartment and after the electrode reaction, oxygen ions migrate through the electrolyte layer to the anode where hydrogen is oxidized. As with the MCFC, SOFCs do not require external fuel reformers, as the operating temperature is sufficiently high to provide the necessary heat for the endothermic reforming reaction.

An additional driving force for pursuing this technology is the potentially high system efficiencies that SOFCs offer. Because of the high operating temperatures of the SOFC, they are attractive for topping and bottoming cycles, cogeneration, and implementation at utility scales. When integrated with a gas turbine (SOFC-GTs), SOFCs are expected to achieve 70-75% (LHV) system efficiencies, representing a significant leap in efficiency over all energy technologies.

While SOFC performance advantages are significant, several development challenges must be overcome. Many of these pertain to matching the thermal expansion coefficients of mating construction materials and fabrication techniques. Mating construction materials present problems because the high operating temperature necessitates the use of exotic ceramics and metal-ceramic composites. Due to the problems of mating component interconnects at high temperatures, much development is also underway at lowering the operating temperatures to the 700-900°C range. While ceramic materials can be costly, the larger fraction of cost is associated with the labor and cost intensive manufacturing techniques necessary to fabricate stack components. SOFCs presently make use of electrochemical vapor deposition, sintering, and plasma spraying manufacturing processes. Tubular designs are more costly than planar geometry SOFCs. Both technologies are making headway, however, the tubular design is closer to commercialization. If manufacturing cost targets are achieved, SOFC systems are likely to be one of the cheapest fuel cell technologies available at $800-$1000/kW.

Tubular SOFC technology has reached the later stages of demonstration and Siemens Westinghouse plans on releasing commercial 1 – 5 MW plants by 2002. Planar technology, while technically behind, has a shorter development path to a commercial product and is also

Page 9: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

vi

targeting commercial release before 2002. Most planar SOFC technologies are beneath 200 kW in size and are initially targeting commercial-HVAC applications in the 25 – 100 kW size range.

TABLE 1

A Summary of Current Fuel Cell Requirements and Characteristics*

Fuel Cell Type Parameters PEM PAFC MCFC SOFC

(Tubular)

Operating Temp. < 210°F ~400°F ~1250°F ~1800°F

Operating Press. 1 – 5 atm 1 – 8 atm 1 – 3 atm 1 - 15 atm

Construction Materials

Graphitic Carbon

Graphitic Carbon

Ni & Stainless Steel

Ceramics and Metals

Power Density pounds/kW

DOE Goals 8-10

~25

~60

~40

Heat Rejection (kWtherm/kW)

~0.48 @ 0.8 V

~0.55 @ 0.74 V ~0.25 @ 0.8 V ~0.52 @ 0.6 V

Efficiency (LHV)

~40% ~40% 50-55% ~50+%

Cooling Medium Water Boiling Water Excess Air Excess Air

Fuel Requirements H2 Fuel Fuel Fuel Fuel CO Poison Poison at > 3% Fuel Fuel CH4 Diluent Diluent Fuel Fuel NH3 Poison Poison Diluent Fuel Cl2 Poison Poison Poison Poison ? S2 Poison Poison Poison Poison

Special Problems Moisture control in the membrane

High-voltage operation

High fuel utilization High fuel or oxidant utilization

* Adapted from “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995.

Page 10: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

vii

Prospective Commercial Markets and Applications (Market Summary) Figure 1 depicts the relative positioning of the various fuel cell technologies with respect to electric demand (kW) and a range of electrical efficiencies for each segment. The residential market considered is from 1 kW to 10 kW. This segment consists of single and multi-family residences and is likely to be served only by PEMs and SOFCs operating initially in electric-only configurations. The small commercial market is from 25 kW to 500 kW and is targeted by all fuel cell types. Examples of the commercial market segment include hotels, schools, small to medium sized hospitals, office buildings, and shopping malls. It is anticipated that higher temperature fuel cells will operate in a cogeneration mode, supplying heat and electricity for cooling and lighting loads. MCFCs and SOFCs are the only types expected to be employed in both distributed power (3 MW to 100 MW) and industrial applications (1 MW to 25 MW) . The traditional utilities, unregulated subsidiaries, municipal utilities, and energy service providers are considered customers for the distributed power segment. The industrial market includes the chemical, paper, metal, food, and plastic industries.

In addition to those listed above, fuel cells may also find penetration in other, non-traditional markets, such as niche power and renewable. Special niche market applications would include computer centers and other customers who require premium power quality and high reliability. Producers of renewable or “opportunity” fuels such as landfills, waste water treatment plants, and refineries are other likely applications.

In the commercial sector, fuel cells will see the most competition with one another and other emerging technologies, such as microturbines. Ballard Power is developing 250 kWe natural gas-fueled PEM units for stationary power and will seek to commercialize within the next two years. IFC has developed the only commercially available fuel cell power plant, the PC-25C. The IFC plant is a 200 kWe, packaged cogeneration system that is likely to be more efficient than PEM technology and more suitable for cogeneration needs. However, this advantage may be offset by lower cost PEMs (if PEMs realize two markets, stationary and transportation). Higher temperature fuel cells will also compete in this market. ERC and MC-Power plan 250 and 300 kWe MCFC systems, respectively. These systems may be highly efficient, but will have reduced siting capability due to larger footprints. SOFCs, on the other hand, are efficient and have reasonable power densities allowing smaller plant footprints than MCFCs. SOFCs, are also expected to be cheaper than the other fuel cell types.

High temperature fuel cells are expected to serve the distributed power and industrial cogeneration sectors. Initially these units will be small (<5 MW) but will gradually increase in size as operating experience is gained and costs are reduced. Larger plants (~100 MW) are envisioned in the future (ca. 2015) and may be integrated with coal gasification.

Distributed Generation Viewpoints The energy industry is experiencing rapid changes due to increased competition from deregulation, available new technologies, and emission requirements. Also, the ever-increasing energy market continues to outdistance the addition of new generation facilities. As one source noted, the world energy demand exceeds the planned addition of 1200 GW of electricity by 200512.

Page 11: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

viii

FIGURE 1 Markets for Fuel Cell Technologies∗

The traditional electric utility perspective has been that large central power plants will continue to provide the vast majority of electric power in the U.S. for the foreseeable future13. This viewpoint is based upon the large economies of scale that keep the cost of electricity (COE) low, making it hard for distributed power generation (DPG) to compete and displace the large central plants. It is argued that this point is underscored in a deregulated utility market. However, clearly not all utilities subscribe to this line of thinking. Many utilities (e.g., Unicom, Southern California Edison) are investing in alternative generation technologies to meet future energy demands. Unicom has made recent business deals with AlliedSignal, a manufacturer of 75 kWe microturbines. Southern California Edison is involved with the evaluation, development, and demonstration of at least seven different types of distributed generation technologies. Recent trends show that distributed generation technology is becoming viable and maturer. The largest growth in new generation has occurred in private power projects. In 1995, the distributed

∗ Adapted from “State-of-the-Art Fuel Cell Technologies for Distributed Power,” EPRI TR-106620, July 1996.

100,000

10,000

1

10

100

1,000

PEM

SOFC

PEM PAFC MCFC SOFC

MCFC SOFC

SOFC-GT

SOFC

SOFC-GT

Residential / Portable

Commercial IndustrialCogen

Distributed Power

(kW)

Page 12: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

ix

power demand for small generator installations (300 kW to 5 MW) was about 10 GW, the Far East accounting for nearly 30% of the total14. By the end of the decade the demand is expected to increase to about 12 GW. The demand in this size range is further evidenced by the bar graphs depicted in Figures 2 and 315. These figures illustrate the number of cogeneration installations installed in the U.S. and Japan in 1994, respectively. The largest number of installations in the U.S. was in the 500 kWe – 1 MWe range, just slightly edging the 100 - 500 kW segment. In Japan, the 100 –500kW segment clearly represents the largest market demand.

Figure 2 U.S. Cogeneration Market Segmentation (1994)*

* Utility Data Institute information adapted from Reference (15).

U.S. Cogeneration Installations in 1994

0

25

50

75

100

125

150

175

200

225

0 - 50 kW 50 - 100 kW 100 - 500 kW 500 kW - 1MW

1 - 3 MW 3 - 5 MW

Electric Capacity

Num

ber o

f Sys

tem

s

Page 13: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

x

Figure 3 Japan Cogeneration Market Segmentation*

Depending on market penetration, the COE of distributed power generation may continue to decrease while offering many advantages and non-traditional benefits that central utilities cannot match. Some of the often-stated advantages of distributed generation include:

1. Economy power or “peak-shaving” which allows the customer to take advantage of time-of-day pricing, effectively leveraging fuel costs against electricity prices.

2. Cogeneration and trigeneration

3. Premium power—uninterrupted power supply and high power quality

4. Little or no transmission and distribution (T&D) expansion costs

5. Utilities can meet energy demand incrementally with a lower cost, lower risk investment -essentially enabling a “just-in-time” philosophy

6. Niche markets, such as emerging countries or remote locations where there is little or no existing T&D infrastructure and limited fuel options, could be better served.

* Utility Data Institute information adapted from Reference (15).

Cogeneration Installations in Japan

0

200

400

600

800

1000

1200

1400

1600

0 - 100 kW 100 - 500 kW 500 kW - 1 MW 1 - 3 MW 3 - 5 MW

Electric Capacity

Num

ber o

f Sys

tem

s

Page 14: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

xi

Competing Technologies The Electric Power Research Institute (EPRI) defines distributed generation as the “integrated or stand-alone use of small modular resources by utilities, utility customers, and third parties in applications that benefit the electric system, specific customers, or both16.” It is synonymous with on-site generation and cogeneration. The size range that EPRI targets is large 3 kW – 50 MW. Given this size range, the competing technologies are numerous for distributed power generation as seen in Table 2.

The sizes tabled are subject to uncertainty as DPG technology is modular and the number of units operating in parallel will be determined by the specific site economics. While uncertainties for technology applications in the various market sectors exist, the last column in Table 2 reflects the current targets and expectations. Additionally, potential transportation applications are listed for fuel cells based on the results of private communications, and literature and internet search results.

Interestingly, PEMs are likely to be the only players in automobile and light truck applications. The operating temperatures, power densities, and weight are the key factors driving this. PAFCs, once thought to be a possible contender, are unlikely to be seen in transportation unless it is in heavier duty applications17. It addition to stationary applications, microturbines may also be developed for transportation. The additional market for both PEMs and microturbines will increase the likelihood of distributed generation market penetration as the increased production volumes for two markets enables larger manufacturing economies of scale.

Diesel engines have been receiving much research and development attention in the transportation sector. Many experts involved in the PNGV efforts believe that diesels are in fact the most likely candidate for 21st Century transportation. Efficiencies are approaching 40+% and emissions are being drastically reduced—some of the research has been accomplished here in Madison. At the moment, diesel stationary applications appear to be limited to standby power and small industrial markets. EPRI reports list Stirling engines as a possible power provider. Not enough is known to comment further on their status or performance expectations, except that they have been in development for over 100 years and have yet to establish a significant commercial niche.

Small turbines have been making steady advancements. Like some fuel cells, when configured in combined cycle operations, small gas turbine efficiencies of nearly 60% can be reached. The efficiencies listed are electric only. Cogeneration naturally improves the efficiencies of many of the technologies, but more importantly it affects the economics.

Renewables have also been listed. Performance and costs for solar applications, while variable depending on geographics, are assumed to be well established. Additional surveying is required to fill in information gaps regarding biomass and wind power.

Page 15: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

xii

TABLE 2 DISTRIBUTED GENERATION TECHNOLOGY∗

Distributed Technology

Types Size* Efficiency* Markets

Fuel Cells PEM (80°C)

PAFC (200°C)

MCFC (650°C)

SOFC (1000°C)

1 – 500 kW

200 kW – 1.2 MW

1 – 20 MW

3 kW – 25 MW

40%

40%

55%

45 – 65%

L&MT, Residential, PP, RP

MT, Commerc. Cogen, PP

HT, PP

Residential, Comm. Cogen, PP, RP

Engines Diesel

IC—Natural Gas

Stirling Cycle

50 kW – 6 MW

5 kW – 2 MW

1 – 25 kW

33 – 36%

33 – 35%

20%

SP for comm. and small indust., T&D support

PP and Commerc. Cogen.

Residential, RP

Combustion Turbines

Microturbines

“Small” turbines*

25 – 500 kW

1 – 100 MW

26 – 30%

33 – 45%

SP, RP, Commerc. Cogen.

Indust. Cogen, T&D support

Renewables Solar (PVs)

Wind

Biomass

1 – 1000 kW 10 – 20% RP, peak shaving, power quality, green power

RP, peak shaving, green power

*Notes: (i.) “small” turbines include cascaded humidified air turbines (CHAD), advanced turbine systems (ATS), and

intercooled aeroderivative Cycle (ICAD). (ii.) Sizes are uncertain as distributed generation is characterized by modularity and the economics have yet TBD. (iii.) Efficiencies = electric only (no heat recovery, HV basis unknown); PV efficiency is sunlight to ac power. L&MT – Light and Medium duty transportation applications (e.g., automobiles, trucks, buses) MT – Medium duty transportation applications (e.g., trucks, buses) HT – Heavy duty transportation applications (e.g., rail, marine—ships, naval vessels) PP – Premium Power RP – Remote Power SP – Standby Power

∗ Adapted from “State-of-the-Art Fuel Cell Technologies for Distributed Power,” EPRI TR-106620, July 1996.

Page 16: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

xiii

Market Segmentation When the prospective DPG size ranges (not including transportation) are indexed against the various end-use sectors, the following application market breakdown results:

Residential (1 – 15 kW)

Light Commercial (25 – 250 kW)

Commercial w/ Cogen.

(50 kW – 3 MW)

Industrial & Distributed (3 – 50 MW)

PEM PEM PAFC MCFC

SOFC PAFC MCFC SOFC

Solar PVs SOFC SOFC Small Turbines

Stirling Solar PVs IC Engines

IC Engines Microturbines

Microturbines

Stirling

*Diesel is not listed as it is most frequently targeted for standby power applications.

From the above, it is apparent that the most significant competition lies in the light commercial sector as fuel cells, PVs, engines, and microturbines are expected to be viable options. While not listed as such, light commercial markets are likely to have some cogeneration needs. PEM technology is targeting cogeneration markets, but will be able to serve only low-grade thermal energy (hot water) demands without external gas-fired equipment. For this reason, they have been left out of the commercial cogeneration market segment. The residential sector seems likely to be dominated by fuel cell and solar technology. At the other end of the spectrum, gas turbines will most likely command the industrial sector, with some competition anticipated by higher temperature fuel cells. Hybrid power generating systems incorporating fuel cells with microturbines are also possible. Several studies of fuel cell-integrated combustion turbine systems have been done. Most recently, EPRI has published several reports on “breakthrough” technology, one of which is comprised of solid oxide fuel cell technology integrated with microturbines to obtain high-efficiency, lower cost, low-emission performance.

Page 17: Review of State-of-the-Art Fuel Cell Technologies for Distri

Executive Summary

xiv

Conclusions Fuel cells and other competing technologies are gaining increasing interest as potential distributed generation solutions to meeting future energy demands and alleviating power congestion at existing distribution stations. Fuel cells are deemed a viable solution once sales volumes increase. Presently, with the numerous efficiency and environmental advantages that fuel cells offer, the expectations are great. Thus, given the nearly zero supply of fuel cells (PAFCs are the only currently available commercial product), the demand is high. The overlap of markets, particularly the commercial market, will make the various fuel cells in competition with one another. The success of one over another is difficult to predict, especially given uncertainties in cell performance and plant costs. However, with the various advantages of each type, it may be just as likely that each fuel cell technology will find its own unique and significant market segment. If cost targets are obtained, how and where fuel cells will be employed is only limited by the creativity of system designers and the vision of commercial developers.

References 1 W.R. Dunbar, “A General Overview of the Fuel Cell Industry,” Center for Energy Studies Report, Marquette University, 1993. 2 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995. 3 Ibid. 4 Ibid. 5 S. Srinivasan, B. Dave, K. Murugesamoorthi, A. Parthasarathy, and A. Appleby, “Overview of Fuel Cell Technology,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.65. 6 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995. 7 K.B. Prater, “Polymer electrolyte fuel cells: a review of recent developments,” J. Power Sources, 51, 1994, p. 133. 8 Ibid. p. 130. 9 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 430. 10 G. Scheffler, E. Hall, and D. Stein, “80 Months of Commercial Experience with the PC25 Fuel Cell Power Plant,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, November, 1998, pp. 24-27. 11 S. Srinivasan, B. Dave, K. Murugesamoorthi, A. Parthasarathy, and A. Appleby, “Overview of Fuel Cell Technology,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.65. 12 C.D. Maloney, “Distributed Generation-Opportuntiy for the 21st Century,” 1998, American Power Conference, Chicago, IL. 13 J.P. Leeper, and J.T. Barich, “Technology for Distributed Generation in a Global Market Place,” 1998, American Power Conference, Chicago, IL. 14 C.D. Maloney, “Distributed Generation-Opportuntiy for the 21st Century,” 1998, American Power Conference, Chicago, IL. 15 D.Dunnison, “PEM Fuel Cell Power Plant Development at Ballard Power Systems,” Proc. 2nd Annual Distributed Resources Conference, EPRI TR-107585, November, 1996. 16 Electric Power Research Institute, Distributed Generation webpage, www.EPRI.com, 1998. 17 Marshall Miller, Institute of Transportation Studies, UC-Davis, private communication.

Page 18: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-1

1. Introduction to Fuel Cell Technology

The development of the fuel cell covers nearly 160 years, since its invention by William R. Grove in 1839. After the invention of the fuel cell, development effort remained dormant for over 100 years. Over this large period of time, only one fuel cell type to date has been abandoned--the direct coal fuel cell1. Fuel cells first received widespread attention and development effort during the 1960s when Apollo missions employed alkaline-based fuel cells to provide electric power to their space vehicles. With the energy crisis in the 1970s, fuel cell development received sustained attention and sponsorship, mainly through the Department of Energy and various utility institutes. Today, these advanced power generating devices have come closer to providing a viable alternative to conventional power plant systems and have been the subject of increased attention due to the technological advancements and potential applications in the transportation market, and from national and international pressures to reduce greenhouse gas emissions (primarily CO2) and improve fuel conversion efficiencies.

Within the last two years, several fuel cell stationary demonstration power plants have either begun operation or are in the planning and construction phase. The most visible of these has been the 2 MW molten carbonate fuel cell (MCFC) demonstration plant in Santa Clara. Additionally, packaged phosphoric acid fuel cell (PAFC) onsite cogenerators are commercially available and currently have some 100 units operating in the field (and 185 units on order)2. Fuel cells have also received national attention in newspapers3 and magazines with recent developments in the transportation sector of a fuel cell which operates directly off of methanol, and the potential use of small fuel cells in cellular phones and laptop computers. These recent developments signify the advent of maturing fuel technologies, which may compete in the marketplace in the very near future.

This objective of this section is to provide an introduction to fuel cell technology. A description of fuel cell devices is first presented, followed by the basic operation of a fuel cell to illustrate the salient features of the fuel oxidation and electric generating processes. Operating principles and component functions are also discussed. Finally, the basic types of fuel cells expected to enter the marketplace are presented, followed by a section on how fuel cell energy systems are configured. Sections 2 through 5 detail the individual technological and commercial status of the various fuel cell types.

FUEL CELL DESCRIPTION The fuel cell is an electrochemical device that converts the chemical energy of the fuel directly into electrical energy. That is, no intermediate conversions of the fuel to thermal and then mechanical energy forms are required. All fuel cells consist of two electrodes (anode and cathode) and an electrolyte (usually retained in a matrix). Therefore, they operate much like a battery except that the reactants (and products) are not stored, but continuously fed to the cell.

The general operation of a fuel cell may be understood by examining the flows and reactions of a single cell depicted in Figure 1-14. Unlike ordinary combustion, fuel (hydrogen-rich) and

Page 19: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-2

oxidant (typically air) are delivered to the fuel cell separately. The fuel and oxidant streams are separated by an electrode-electrolyte system. In a typical fuel cell, gaseous fuels are fed to the anode (negative electrode) compartment and an oxidant is fed continuously to the cathode (positive electrode) compartment5. The electrochemical reactions take place at the electrodes to produce electric current and thereby, power. Thus, the electrochemical process is a direct conversion of the chemical energy bound up in the fuel by the ionization of reactant species at the electrodes and subsequent charge transport through an electrolyte. The primary product of fuel cell reactions is water.

As Figure 1-1 shows, the direction of ion transport may vary between fuel cell types. The ion can be negatively or positively charged. Different electrolyte types facilitate the conduction of positively or negatively charged species and may be found in liquid or solid phases. The fuel and oxidant are admitted to the fuel cell in gaseous phase and flow past the surface of the anode or cathode opposite the electrolyte and generate electrical energy by the electrochemical oxidation of the fuel, usually hydrogen, and the electrochemical reduction of the oxidant, typically oxygen from air6. The processes involved in fuel cell operation are made up of several steps that are physical, chemical, or electrochemical in nature. For instance, before the electrochemical reactions take place at an electrode, the gaseous reactants must first diffuse (physical) through the porous microstructure of the electrode and adsorb (chemical) to the electrode prior to the electrode reactions (electrochemical). In order for the electrochemical reaction to take place the gaseous species must be in physical contact with both the electrode and the thin electrolyte layer that wets the porous electrode. This region is called the three-phase

Page 20: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-3

interface and is absolutely critical to the performance of fuel cell systems. If, for example, the electrode contains too much electrolyte, the electrode may “flood” and restrict the transport of gaseous reactants to the electrochemical reaction sites7. The result is a decrease in cell performance. Much research and development has be expended at optimizing the design and manufacturing of electrolyte and electrode structures which will maintain an appropriate equilibrium balance of wetted electrolyte in the porous electrode and maximize the three-phase surface area to promote high current densities.

Cell Stack Assemblies In practice, a single cell will produce less than one volt of electrical potential and therefore, fuel cells are often stacked on top of each other and connected in electrical series to produce higher voltage levels. The number of cells stacked depends on the desired power output and individual cell performance; typically cells are stacked from a just a few (for less than 1 kWe of power) to several hundred (for 250+ kWe). As illustrated in Figure 1-2, cell stacks consist of repeating fuel cell units; each comprised of an anode, cathode, electrolyte, and a bipolar separator plate between cells. Reactant gases, which typically consist of desulphurized and reformed natural gas and ambient air, flow over the electrode faces (anode and cathode) in channels through the bipolar separator plates, generating electrical and thermal energy by the simultaneous electrochemical oxidation of fuel and reduction of oxygen. Reactant gases are often delivered from fuel preparation equipment piping to an external manifold arrangement as shown in Figure 1-38. Due to the intrinsic nature of establishing the cell electrical potential, not all the reactants can be consumed in the oxidation process. Thus, it is necessary that some hydrogen exit the fuel cell stack. Typically, about 20% of the hydrogen delivered to the fuel cell stack is unused and is often “burned” in ancillary equipment downstream of the fuel cell module.

Figure 1-2 Fuel Cell Stack Components8

Page 21: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-4

Each cell component has several functions. There are two primary functions to the electrolyte: (i) it must act as a transport medium for the migration of ionic charge (thus completing the electric circuit), and (ii) serve to separate gaseous fuel and oxidant streams. Thus, the electrolyte must exhibit high ionic conductivity and low electronic conductivity; that is, it must act as an insulator to the electric current, thereby preventing any short-circuiting of the cell. Liquid phase electrolytes are not employed as a free liquid, but are retained in an electrolyte matrix. Solid-phase electrolytes have no retaining matrix and are thus, simpler in design. The electrode functions are to (i) conduct electrons to and from the electrochemical reaction sites, (ii) to provide porous pathways for the diffusion of gaseous species to reaction sites, and (iii) have sufficient chemical activity to promote the electrode reactions. Separator plates are utilized to (i) act as current collectors, providing electrical connection between cells, and (ii) to separate the fuel and oxidant gas streams in adjacent cells.

Due to the functionality requirements of fuel cells, materials play a crucial role in their success. Critical aspects to the sustained performance of all components include corrosion, porosity, mechanical strength, conductivity, chemical and dimensional stability with both mating components and liquid and gaseous species at the cell operating temperatures and pressures, and manufacturability characteristics. Thus, the development, performance, and viability of fuel cells have been heavily dependent upon advancements from materials research, mechanical design, and manufacturing techniques.

Figure 1-3 Internal and External Manifolding Configurations9

Page 22: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-5

BASIC TYPES OF FUEL CELLS Fuel cells are typically classed by the electrolyte used. The type of electrolyte employed particularly affects, among many things, the cell operating temperature. Cell temperature has a significant impact on construction materials, manufacturing processes, fuel preparation requirements, and to some extent, the way in which the fuel cell is applied and its current state of development10. In addition to the type of electrolyte, the various fuel cells are further distinguished by the electrode materials and associated electrochemistry.

There are currently four types of fuel cells that are receiving the most development attention. Listed in order of increasing operating temperature they are: the 80°C proton exchange membrane (PEM) fuel cell, the 200°C phosphoric acid fuel cell (PAFC), the 650°C molten carbonate fuel cell (MCFC), and the solid oxide fuel cell (SOFC) currently at 1000°C. Table 1-1 gives a summary of the above fuel cell types and some salient characteristics of each. Other fuel cell types include alkaline and direct methanol fuel cells. Alkaline fuel cells were employed in Apollo space missions and are still used on the Space Shuttle, however, these types are not considered viable for terrestrial applications due to nearly zero tolerance to CO2 and CO constituents in the fuel. Direct methanol is receiving increased attention for transportation applications but is far behind the other types. These two types will not be surveyed in this report.

Proton Exchange Membrane Fuel Cells (PEMs) PEMs operate at the lowest temperature (80°-100°C) of all the fuel cell types. The electrolyte consists of a solid polymeric membrane fitted between two platinum-catalyzed porous electrodes. The limit on the operating temperature of the fuel cell is set by thermal stability and conductivity characteristics of the polymeric membrane11. The most common medium used to cool the fuel cell stacks is water.

During operation, hydrogen ions (protons) migrate through the electrolyte from the anode to the cathode and the fuel oxidation reaction occurs on the “air” side of the cell. Table 1-2 summarizes the electrode reactions for all fuel cell types explored in this report. The cell can operate at elevated pressures, and it is often beneficial to do so in order to attain high power densities. Air pressures up to 8 atm have been used. Like nearly all fuel cells, the PEM requires a pure H2 source for operation. Most notably, CO and S are poisons to the PEMs, thus extensive fuel processing is required for operation on hydrocarbon fuel sources.

Historically, PEM development has been behind other fuel cell types. However, significant technological advancements have been achieved in PEMs most recently and they have developed rapidly. Much of the recent development attention has been focused on PEMs for automotive applications, due to their low-temperature rapid transient capabilities, lightweight, and relative size. As with other fuel cells, PEMs will first be targeted for commercialization in stationary power applications in the year 2000 timeframe.

Phosphoric Acid Fuel Cells (PAFCs) Up until recently, the PAFC was the only fuel cell type to be commercially available (PEM cells are now available at small scales, but not at the scales of PAFCs yet.) The leader of this fuel cell

Page 23: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-6

technology is International Fuel Cell (IFC) Corporation, in Massachusetts. PAFCs typically operate near 200°C (400°F). Cooling of the fuel cell stack is accomplished with boiling water.

TABLE 1-1

Types of Fuel Cells and Some Characteristics

Proton Exchange

Membrane

(PEM)

Phosphoric Acid Fuel Cell

(PAFC)

Molten Carbonate Fuel Cell (MCFC)

Solid Oxide Fuel Cell

(SOFC)

Operating Temperature

80°C (200°F)

200°C (400°F)

650°C (1200°F)

1000°C (1800°F)

Year of Market Entry (Projected)

2000 1993 2001-2002 2003

Fuel-to-Electric Efficiency (HHV)

35-40% 36-40% 50-65% 50-60%

Size Range 50 W – 1500 kW

50 – 1200 kW (IFC Only)

250 kW – 100 MW

3 kW – 100 MW

Anticipated Commercial Cost* (2 MW)

$1000/kW $1200/kW $1000/kW $800/kW

Some Applications Vehicles

Stationary Power

Stationary Power

Vehicles

Stationary Power

Cogeneration

Stationary Power

Cogeneration

* These costs represent the mature product costs obtained from high volume manufacturing.

The electrolyte material consists of 100% phosphoric acid (H3PO4) which acts as a transport fluid for the migration of dissolved hydrogen ions from the anode to the cathode, and to conduct the ionic charge between the two electrodes in order to complete the electric circuit.

As with all fuel cell types, the PAFC operates on hydrogen that is typically delivered from a natural gas supplied reformer. Like the PEMFC, during operation hydrogen ions migrate from anode to cathode. PAFCs can also operate at elevated pressures (up to 8 atm), however, the current packaged unit offered by IFC operates at ambient pressures.

Although PAFCs have been demonstrated up to the 11 MW size range, they are initially targeted to penetrate commercial and light industrial markets. The commercially offered IFC PC25 packaged PAFC unit can provide 200 kWe (at 36% efficiency) and about 205 kW thermal

Page 24: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-7

energy for an approximate selling price of $3000/kWe12. A system description and thermodynamic analysis of the PC25 unit can be found in Section 2 of this report.

Molten Carbonate Fuel Cells (MCFCs) The MCFC operates at 650°C (1200°F). These fuel cells usually consist of a lithium-potassium or lithium-sodium based electrolyte. After the cathode reaction, carbonate ions migrate through the electrolyte to the anode side of the cell to complete the fuel oxidation. Because of the CO2 requirement at the cathode, and production at the anode, CO2 must be transferred from the anode exhaust to the cathode inlet. This is normally accomplished through mixing of the anode exhaust with incoming air or by physically separating the CO2 form the other exhaust gas species through a “product exchange device”13.

MCFCs and SOFCs are considered “high” temperature fuel cells that offer several advantages for distributed or utility scale power generation. At an operating temperature of 650°C, high quality waste heat can be produced and used for (i) fuel processing and cogeneration, (ii) internal methane reforming, and (iii) to drive a bottoming cycle. The MCFC, like other fuel cell types, has improved cell performance under pressurized conditions. Today, pressurized operation is being pursued by about half of the molten carbonate fuel cell manufacturers.

MCFCs are in the demonstration phase of their technology with commercialization plans for the year 2002. Several demonstration plants have been built and tested in the U.S. and abroad from 250 kW to 1.8 MW in capacity. The leaders in this technology are the U.S.-based Energy Research Corporation in Danbury, CT and MC-Power of Burr ridge, IL.

Solid Oxide Fuel Cells (SOFCs) SOFCs currently operate at 1000°C (1800°F). The SOFC employs a solid yittra-stabilized zirconia ceramic material as the electrolyte layer. During operation, oxidant (usually air) enters the cathode compartment and after the electrode reaction, oxygen ions migrate through the electrolyte layer to the anode where hydrogen is oxidized. As with the MCFC, SOFCs do not require external fuel reformers as the operating temperature is sufficiently high to provide the necessary heat for the endothermic reforming reaction.

Because of the high operating temperatures of the SOFC, they are attractive for topping and bottoming cycles, cogeneration, and implementation at utility scales. Additionally, a unique feature of the SOFC is that it is expected to be a feasible power-generating device at scales ranging from 1 kW to 100 MW. However, before market entry at these scales is realized, several development challenges must be overcome. Many of these pertain to matching the thermal expansion coefficients of mating construction materials and fabrication techniques. Due to the problems of mating component interconnects at high temperatures, much development is also underway at lowering the operating temperatures to the 800-900°C range.

In general, much of the SOFC development lags the other fuel cell types. Currently, several demonstration plants are under construction from sizes of 25-100 kWe. Because of the solid electrolyte, the SOFC can be manufactured in both planar and tubular configurations. Most manufacturers are pursuing planar development. Westinghouse leads in tubular SOFC technology.

Page 25: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-8

Table 1-2 Summary of Electrode Reactions in the Various Fuel Cell Types

Fuel Cell Type Anode Reaction Cathode Reaction

Phosphoric Acid H2 2H+ + 2e- ½ O2 + 2H+ + 2e- H2O

Proton Exchange Membrane H2 2H+ + 2e- ½ O2 + 2H+ + 2e- H2O

Molten Carbonate H2 + CO3= H2O + CO2 + 2e- ½ O2 + CO2 + 2e- CO3

=

Solid Oxide H2 + O= H2O + 2e- CO + O= CO2 + 2e-

CH4+ 4O= 2H2O + CO2 + 8e-

½ O2 + 2e- O=

GENERAL PERFORMANCE Fuel cells convert fuel to electricity at efficiencies greater than any conventional energy conversion technology. Figure 1-4 compares fuel cell efficiencies with conventional power producing equipment versus power output. From strictly an efficiency performance standpoint, it is easy to understand the excitement surrounding fuel cells; they represent a potential future in which the efficiency of power generating equipment could be doubled from some of today’s power plants.

Figure 1-4 Power Plant Efficiency Comparisons14

Fuel Cells

Diesel Engines

IC Engines

Steam and Gas Turbines

Page 26: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-9

As previously stated, the fuel cell is an electrochemical device which converts the chemical energy of the fuel directly into electrical energy, requiring no intermediate conversions of the fuel to thermal and then mechanical energy forms. Instead of being released entirely as heat, much of the chemical energy of the fuel is released in the form of an electron stream which, in turn, flows out of the cell through a load and is returned to the electrode15. The rate at which the fuel is oxidized is controlled (or limited) by the load. The primary thermodynamic advantage of fuel cells over conventional engines (such as spark-ignition and diesel) lies in the oxidation process, which is generally the largest source of inefficiency.

In conventional combustion, a fuel is directly mixed with oxygen and reacts to form products of combustion. The gas phase reactions in combustion are characterized by random collisions between molecules in which electrons are exchanged. From a thermodynamic standpoint, the driving force for the oxidation of the fuel is the difference between the chemical potentials of the reactants and products, that is, the chemical affinity of the reaction. The potential of the fuel to do useful work is irreversibly destroyed in the highly entropy producing combustion process. Dunbar and Lior16 state that there are two ways to reduce entropy productions (and therefore increase the fuel conversion efficiency) in the fuel oxidation process: (i) increase the temperature at which combustion occurs, and (ii) reduce the chemical affinity (or driving force) of the reaction. The fuel cell accomplishes the latter by first passing the reactant ions (H+ ions in the case of PEMs and PAFCs, CO3

= ions in MCFCs, and O= ions in SOFCs) through an electrolyte. This allows a less violent (in comparison to conventional combustion) chemical reaction to take place inasmuch as the force driving the reaction is lower17. Thus, both the high temperature feed of reactants to the fuel cell and the “controlled” oxidation process contribute to higher electric efficiencies than experienced by conventional oxidation equipment (e.g., stationary engines and combustion turbines).

However, fuel cells do experience losses in efficiency due to other effects. Temperature, pressure, gas composition, and fuel and reactant utilization influence their performance. These operating variables affect the magnitude of the irreversible voltage losses and hence, the cell potential. The losses when expressed in terms of voltages are called polarizations. There are three main types of polarizations -- ohmic, activation, and concentration. These polarizations may arise in any of the reaction steps of a real fuel cell involving mass transport, adsorption, chemical reaction, or charge transfer18. Activation losses occur at all current densities but are dominant at low current density and are associated with sluggish electrode reaction kinetics at the three-phase interface (i.e., they can be viewed as a resistance to initiating the electrode reactions). Ohmic losses arise from resistances to charge conduction through the various cell components and demonstrate a linear dependence with current. Lastly, concentration polarization is a resistance to diffusional transport of reactants to and from the electrochemical reaction sites. Like activation losses, concentration losses occur over the entire operating range of the cell but the contribution to the total loss is most significant at high current density operating points. When operating the fuel cell in this regime, the cell is essentially “starved” of reactants as they cannot be supplied at the high rate at which the electrode reactions demand. The cumulative effect of these inefficiencies is depicted Figure 1-5.

In general, a decrease in the operating cell voltage of a fuel cell results in a lower fuel cell efficiency. The relationship between the relevant parameters (e.g., temperature, pressure, reactant utilization, and gas composition) and the various polarizations is quite complex. While

Page 27: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-10

increases in temperature decrease the theoretical cell voltage, the numerous transport and kinetically constrained transport processes are improved with increasing cell operating temperature. The exploration of these transport processes along with other sources of cell irreversibilities is beyond the scope of this report. Of relevance herein is the current state-of-the-art fuel cell technology, which demonstrates electric efficiencies in the 36-60% range.

Figure 1-5 Actual and Ideal Cell Voltage vs. Current Density19

FUEL CELL SYSTEMS Fuel cell systems differ from conventional power generation systems in a single distinctive way. Since all fuel cells operate primarily on hydrogen fuel (which is not readily available), fuel cell systems must incorporate fuel-processing equipment. Typically, fuel processing equipment preheats the fuel, often near the operating temperature of the cell, and reforms it (or other primary hydrocarbon fuels) to a hydrogen-rich gas stream.

Figure 1-6 exhibits a schematic diagram of a generic fuel cell system. Fuel, often in the form of natural gas, enters the plant and is delivered to the fuel processing sub-system. In the fuel processor, the fuel is desulfurized, preheated, and reformed. Removal of sulfur from the fuel can take place at lower or higher temperatures depending on the desulfurization process. Reforming the natural gas, that is, converting the hydrocarbons to carbon dioxide and hydrogen via reforming and shift reactions, requires steam. Therefore, prior to reforming, superheated steam is injected into the natural gas supply (often at an approximate molar steam-to-carbon ratio of

Page 28: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-11

3:1). Steam-to-carbon ratios may differ for various fuel cell types and are most critical in determining fuel cell system performance.

After fuel preparation, the gas is delivered to the fuel cell where the fuel is electrochemically oxidized producing both thermal and electrical energy. Electric efficiencies range from 36 – 60% depending on the type of fuel cell and the system configuration. Fuel cells produce DC power which is converted to AC through power conditioning at efficiencies as high as 98%. By using conventional heat recovery equipment, first law efficiencies of 80-85% can be realized.

Figure 1-6 Generic Fuel Cell System Description20

OVERVIEW Fuel cells are being targeted for every sector of the world energy market, from small residential cogenerators to large central power generating stations21 The recent technological advancements have helped move fuel cells from a hypothetical, next generation power-generating device, to a realizable commercial product. Each fuel cell type has its advantages and disadvantages. Given the prospectively large power generating markets anticipated in the near future, it is very possible that all types (even some not discussed in this report) could find a substantial market niche in the 21st Century.

The objective of this report is to survey the fuel cell technologies that are receiving the most development attention and assess the relative status of each. This report presents extensive research results of each fuel cell technology. For each of the major fuel cell types, materials

Page 29: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-12

construction, operation, performance characteristics, manufacturers, anticipated markets, current development status and technical issues, and economic outlook are discussed. In general, PAFC technology leads the other fuel cell types, however, PEM technology is rapidly closing the distance. As recent as 1991, PEM technology was not considered a contender for larger scale stationary power generation. However, technological breakthroughs in power density and materials cost reduction coupled with the infusion of vast funding resources from automakers have situated PEMs as the foremost candidate in transportation markets, and a likely competitor in stationary power. MCFC technology, the most efficient of all types, has historically been considered a 2nd generation fuel cell technology as it will commercialize after PAFCs. Lastly, out of all fuel cell types, SOFCs have long been viewed as the farthest from commercialization. However, tubular SOFC technology developed by Siemens Westinghouse is challenging MCFC commercial timeframes. Except for PAFCs, which are already commercial, the remaining fuel cell types are expected to commercialize in the stationary power market at various levels within the next five years. Table 1-3 summarizes some of the technical aspects which will be discussed in further detail and will be referred to throughout this report.

In this report, the phosphoric acid fuel cell is first presented as it is the most advanced fuel cell type. Next, proton exchange membrane fuel cells are discussed, followed by the higher temperature fuel cells, molten carbonate and solid oxide.

Page 30: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-13

TABLE 1-3

A Summary of Current Fuel Cell Requirements and Characteristics*

Fuel Cell Type

Parameters PEM PAFC MCFC SOFC (Tubular)

Operating Temp. < 210°F ~400°F ~1250°F ~1800°F

Operating Press. 1 – 5 atm 1 – 8 atm 1 – 3 atm 1 - 15 atm

Construction Materials

Graphitic Carbon

Graphitic Carbon

Ni & Stainless Steel

Ceramics and Metals

Power Density (pounds/kW)

DOE Goals

8-10

~25

~60

~40

Heat Rejection (kWtherm/kW)

~0.48 @ 0.8 V ~0.55 @ 0.74 V ~0.25 @ 0.8 V ~0.52 @ 0.6 V

Efficiency (LHV)

~40% ~40% 50-55% ~50+%

Cooling Medium Water Boiling Water Excess Air Excess Air

Fuel Requirements

H2 Fuel Fuel Fuel Fuel

CO Poison Poison at > 3% Fuel Fuel

CH4 Diluent Diluent Fuel Fuel

NH3 Poison Poison Diluent Fuel

Cl2 Poison Poison Poison Poison ?

S2 Poison Poison Poison Poison

Special Problems

Moisture control in the membrane

High-voltage operation

High fuel utilization

High fuel or oxidant utilization

* Adapted from “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995.

Page 31: Review of State-of-the-Art Fuel Cell Technologies for Distri

1-14

REFERENCES 1 Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993. 2 Zink, J.C., Power Engineering, McGraw-Hill, December, 1996. 3 New York Times, October 21, 1997. 4 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, Figure 1-1, p.1-1. 5 Ibid. 6 Ibid., p. 1-2. 7 Ibid., p. 1-2. 8 R.J. Braun, R.A. Gaggioli, W.R. Dunbar, "Improvements of A Molten Carbonate Fuel Cell Power Plant via Exergy Analysis," Proc. of the ASME Advanced Energy Systems Division, AES-Vol. 36, ASME Winter Annual Meeting, 1996, Atlanta. 9 D.T. Hooie and M.C. Williams, “Overview of Commercialization of Stationary Fuel Cell Power Plants in the United States,” Proceedings of the 30th IECEC, paper no. EC-254, ASME, 1995, p. 172. 10 Hirschenhofer, J.H., and McClelland, R.H., “The Coming of Age of Fuel Cells,” Mechanical Engineering, October 1995. 11 Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993. 12 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996. 13 Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993. 14 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996. 15 Dunbar, W.R., and Lior, N., “Sources of Combustion Irreversibility,” 1994, Comb. Sci. and Tech., Vol. 103, pp. 41-61. 16 Ibid. 17 Dunbar, W.R., and Lior, N., Gaggioli, R.A., “The Exergetic Advantages of Fuel Cell Systems,” 1990, Proc. Of the Florence World Energy Research Sym., S.S. Stecco and M.J. Moran, editors, Pergamon Press, New York. 18 Grubb, W.T., and Niedrach, L.W., “Fuel Cells,” Direct Energy Conversion, Inter-University Series, G.W. Sutton, editor, McGraw-Hill, New York. 19 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, Figure 2-2, p.2-5. 20 www.fuelcell-eur.nl 21 Dunbar, W.R., “A General Overview of the Fuel Cell Industry,” Center for Energy Studies Report, Marquette University, 1993.

Page 32: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-1

2. Phosphoric Acid Fuel Cells (PAFCs)

The phosphoric acid fuel cell (PAFC) is the only fuel cell technology that is commercially available today. Refinement of PAFC technology is far ahead of other fuel cell technologies, being in an advanced state of development for some 10-15 years. The leader of PAFC technology is International Fuel Cells (IFC) Corporation (a U.S.-based partnership with United Technologies and the Toshiba Corporation). The goal of IFC is to develop economically viable onsite PAFC cogeneration systems by the end of the decade. IFC formed the ONSI Corporation to market and sell their cogenerator. In 1993, ONSI went commercial with a 200 kWe PC25 unit and to date, has received orders for 185 units with 100 of these currently operating in the field1. Each unit provides 200 kilowatts of electric power; the byproduct heat that is produced is of sufficient grade that it can also be used for building space heating. Worldwide, PAFC technology has been demonstrated at scales ranging from 50 kWe to 11 MWe, with most demonstrations between 50 and 200 kWe.

The PAFC developed by IFC is the only fuel cell to consistently achieve demonstrated lifetimes of 40,000 hours or better when operating in production configurations. The PC25 units have operated on natural gas, propane, butane (expected), landfill gas, hydrogen, and anaerobic digester gases2. Emissions have been so low; thereby exempting the plants from air pollution permitting in southern California and the Bay Area-which have the most stringent air quality limits in the U.S. Field units have been subjected to ambient conditions of –32°C to 49°C and operated at altitudes of 1 mile. While the technical performance of PAFCs has been well demonstrated, the technology has yet to achieve any significant market penetration. The greatest barrier facing PAFC technology is the high capital cost of the unit, currently at $3000/kWe ($4000/kWe installed).

Relative to the other fuel cell types, little fundamental research is ongoing in PAFCs as development is appropriately focused on cost reduction. In general, PAFCs achieve electric efficiencies between 37-42% (LHV), employ high cost platinum electrocatalysts, and require external reformers to produce a hydrogen-rich gas feed from hydrocarbon feedstock. The near 200°C operating temperature of PAFCs is sufficient to provide low-grade thermal output in the form of 140°-250°F hot water or low-pressure (15 psi) steam. However, some technical challenges surrounding carbon component corrosion, cathode performance, and increasing the power density remain yet.

ELECTROCHEMISTRY Phosphoric acid fuel cell (PAFC) stacks, like other planar fuel cell types, consist of repeating fuel cell units, each comprised of an anode, cathode, electrolyte (retained in a matrix) and a ribbed separator plate (either of substrate or bipolar type) between cells, such as depicted in Figure 1-1. One additional repeating stack element is the cooling plate which removes the heat generated from the cell reactions and, depending on design and cooling fluid, may be spaced every 4-7 cells3,4. Additional stack components include manifolds, end plates, and other small components. In a typical PAFC, the basic structure consists of the phosphoric acid electrolyte in

Page 33: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-2

the matrix sandwiched between the anode and cathode as shown in Figure 2-1. The electrodes contain two layers: a catalyst layer in which the electrochemical reactions take place, and a porous substrate layer which mechanically supports the thin catalyst layer. As depicted in the figure, ribbed flow fields are machined out of the electrode substrate, providing reactant gas pathways. The matrix retains the H3PO4 electrolyte, and both materials are ionically conductive, but electronically nonconductive5. The separator plate prevents the mixing of anode and cathode reactant gases and acts as current collector by providing the electrical connection between adjacent cells.

Figure 2-1 Cell Structure in a PAFC6

The PAFC is maintained at a temperature near 200°C (400°F) and is capable of generating low-pressure steam for export. Phosphoric acid offers excellent thermal, chemical, and electrochemical stability and lower vapor pressures compared with other inorganic acids7. However, at this operating temperature and due to the inherently slow oxygen reduction kinetics in acid electrolytes, noble metal electrocatalyts are required to produce sufficiently fast electrochemical reaction rates8.

Figure 2-2 illustrates the operating principle of the PAFC. During operation, reactant gases, which typically consist of desulphurized and reformed natural gas and ambient air, flow over the electrode faces in passages through the separator plates. Hydrogen gas entering the anode compartment is drawn to the electrode by its electrochemical affinity for oxygen and adsorbed to the anode where it diffuses through the electrode to the electrocatalyst sites. At the anode/electrolyte interface and in the presence of an electrocatalyst, hydrogen dissociates into 2H+ and 2e-. The hydrogen ions then migrate through the electrolyte to the cathode/electrolyte interface carrying the positive charge and ultimately accomplishing useful work. At the cathode, the hydrogen diffuses to the electrocatalyst for oxygen reduction. Similarly, oxygen entering the cathode compartment is adsorbed to the cathode, diffuses to the electrocatalyst sites, and is reduced by hydrogen protons that have migrated through the phosphoric acid electrolyte. The electrode reactions that occur in the presence of the platinum electrocatalysts are identical to the proton exchange membrane fuel cell and are summarized as follows:

Anode: H2 2H+ + 2 e-

Cathode: O2 + 4H+ + 4e- 2H2O

Overall: 2H2 + O2 2H2O

Page 34: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-3

Cooling Plate Water, Air, or Oil

180 – 210 °C Operation

Fuel 80% H2

20% CO2

Depleted Fuel16% H2 20% CO2

Air 21% O2 Depleted Air

10% O2 H2O

H2

O2

Typically the oxygen electrode reactions are rate limiting and much research effort has been expended to improve the electrode performance. However, the reduction kinetics are generally poorer in PAFCs than in both PEMs and alkaline fuel cells.

Figure 2-2 Principles of Operation in a PAFC9,10

CELL CONSTRUCTION AND MATERIALS The type of materials employed in PAFCs have experienced relatively little change in the last 25-30 years. The most significant advancement surrounded the choice of carbon as the electrode support in the early 1970s11. This advancement has been claimed as the single greatest breakthrough for PAFCs enabling high surface area carbon blacks to support Pt at low-cost and without reduction in performance12. Subsequent progress in PAFC materials has largely been accomplished by adjusting the amounts (i.e., composition) of the various elements employed in the cell components. As Table 2-1 illustrates, slight changes have been made; a reduction in the amount of Pt electrocatalyst and an increase in the phosphoric acid content in the electrolyte from 95% to 100% H3PO4. More recent development efforts have focused on improving cathode performance, carbon corrosion stability, and manufacturability. In general, the relatively low operating temperature (200°C) of PAFCs does not present as stringent an operating environment as it does in molten carbonate and solid oxide fuel cells. However, the noble metal electrocatalyst particles that are deposited on the carbon support do experience sintering over time, and hence, reduction in catalyst surface area*. Also, the phosphoric acid is volatile and *Sintering is a process of bonding particles by heat. It involves not only the bonding of powder particles, such as ceramics, metals, and polymeric materials (e.g., Teflon), but also the elimination of the initial porosity to give a more dense product (cf. L.H. Van Vlack, Elements of Materials Science and Engineering, 6th Ed., Addison Wesley, 1989, pp. 284-5.)

Electrolyte H3PO4

2 H+

Anode H2 2H+ + 2e-

Cathode O2 + 4H+ + 4 e- 2 H2O

Page 35: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-4

escapes from the electrolyte matrix, coating stack components and creating corrosion problems (most notably for the carbon electrodes.)

Electrolyte / Electrolyte Matrix

The electrolyte in PAFCs consists of 100% phosphoric acid (H3PO4) which is a colorless, viscous hygroscopic liquid. Phosphoric acid demonstrates numerous attractive attributes including tolerance to CO2, low vapor pressure, high oxygen solubility, and good ionic conductivity at 200°C13. The electrolyte is not a free liquid, but is retained in a porous matrix structure made of silicon carbide (SiC)14. The function of the matrix is to hold the acid by capillary action, preventing escape of the liquid into the anode and cathode compartments, and lend some mechanical strength to the cell assembly. The freezing point (or solidification point) of the acid is high at 42°C. Below this temperature the acid will solidify with a subsequent increase in volume. The volume change behavior is a limitation of phosphoric acid fuel cells as phase change may occur between on-load and off-load conditions. Thus, the PAFC must be maintained above 42°C at all times as frequent volume changes can damage the electrodes and electrolyte matrix15.

The solidification point of phosphoric acid is dependent on the density of the acid16. Decreases in acid concentration suppress the freezing point, but reduce performance. Thus, for 100% concentrate acid, the PAFC stack is typically maintained above 45°C with a heater*.

Electrodes The gas diffusion electrodes consist of a catalyst layer in which the electrochemical reactions take place, and a substrate that provides mechanical support for the catalyst layer (see Figure 2-1). In the catalyst layer, both electrodes contain a mixture of platinum electrocatalyst supported on high surface area carbon black (or carbon “paper”) and are held together with a PTFE-polymeric binder. The high surface area is obtained by utilizing small carbon particles (~nanometers in diameter) that are impregnated by fine Pt particles within the porous microstructure of the carbon paper. The hydrophobic additives, such as PTFE (Teflon), also serve as wetproofing agents to prevent flooding of the electrode pores. The carbon paper is often an acetylene black type, of which the most commonly employed is Vulcan XC-7217. The carbon paper serves to (i) support the dispursement of fine Pt catalyst particles in high surface area fashion, (ii) provide numerous micropores in the electrode for gaseous diffusion, and (iii) increase the electrical conductivity of the catalyst layer18. The catalyst layer, which ultimately facilitates the electrochemical anode and cathode reactions, is only about 1mm in thickness19.

* As one might surmise, a new PAFC unit is shipped with lower density H3PO4 that is replaced with concentrated acid before start-up.

Page 36: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-5

Table 2-1 Evolution of PAFC Hardware

The substrate is a reinforced porous material adjacent to the catalyst layer that is electrically conductive, permitting both the passage of electrons and reactant gases20. It is constructed to about 60-65% porosity and is about 1-2 mm in thickness21. The substrate contains ribbed passages that provide no more than a 1 mm high passage for reactant gases. The substrate for each electrode is in physical contact with the flat separator plate.

As Table 2-1 shows, the state-of-the-art anode is PTFE bonded with approximately 0.1 mg/cm2 of Pt electrocatalyst supported on Vulcan XC-72 carbon black. Unlike higher temperature fuel cells, the only difference between anode and cathode in a PAFC is the amount of Pt present. With a loading of 0.5 mg/cm2, the cathode contains about five times as much platinum as the anode*. Electrode performance generally decays slowly with operation. Performance decay is associated with sintering of Pt particles and obstruction of gas diffusion pores due to flooding of the catalyst layer by electrolyte and water22. Separator Plates The function of separator plates is to act as current collectors by providing the electrical connection between cells and (ii) to separate the reactant and oxidant gas streams in adjacent cells23. In some designs it also contains the gas channels for introducing the reactant gases and removing the products and inerts. The separator plates employed in PAFCs are either of ribbed * All-in-all, some 54g Pt/kWe were required for the 11 MWe GOI demonstration PAFC power plant in Japan (Ref 2, page 3-3).

Page 37: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-6

separate (bipolar) or ribbed substrate type as shown in Figure 2-3. The latter type has become the more common type employed by U.S. and Japanese developers as it promotes a more uniform gas diffusion to the electrodes and enables storage of phosphoric acid in the porous substrate24. The material utilized for the plates is thin glassy carbon plate whose thickness is typically less than 1 mm25. The glassy carbon plate consists of graphite-resin mixtures that are carbonized at temperatures around 1650°F and are then heat-treated to 5,000°F to thermally strengthen their ability to withstand continuous operation at higher than ambient temperatures26. Presently, the separator plates have been assessed as sufficiently corrosion resistant for the anticipated 40,000 hour lifetime, but are costly due to manufacturing processes.

Figure 2-3 Types of Cell Structure27 (a) Ribbed Separate (Bipolar); (b) Ribbed Substrate

PAFC Stack Construction The design parameters for a PAFC stack are characterized by operating pressures, electrolyte management techniques, and methods of stack cooling. Operating pressures vary between atmospheric pressure operation for onsite applications to 8.2 atmospheres for utility and industrial applications. Electrolyte management techniques may vary between two different design philosophies. One design is for 40,000 hours only operation, i.e., only the necessary components and acid volume are used. The second design philosophy is characterized by the employment of an electrolyte addition system that allows vaporized electrolyte to be replenished during the life of the stack. The PAFC stacks are composed of the following components:

(1) Electrodes (2) Electrolyte and Electrolyte matrix (3) Separator plates (4) Cooling plates (5) Manifolding (6) Other small components

An example of a cutaway view of a fuel cell stack is illustrated in Figure 2-4. State-of-the-art PAFC stacks, depending on power output, may consist of 50 to 500 cells stacked vertically. Assembly of the stack involves stacking the unit cells (electrodes, and electrolyte matrix), inserting cooling plates, assembling current collecting plates, insulating clamping plates at the top and bottom of the assembly, and clamping the total assembly with clamping bolts28. The

Page 38: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-7

clamping pressure with which the stacks are subject to should be sufficient to minimize the electrical and thermal resistances between layers but within the limit of material strength29.

PERFORMANCE In comparison with other fuel cell types, the electrical efficiency of PAFC systems (37-42% LHV) is at the relatively low end efficiency goals for fuel cell power plants. Nearly all fuel cell types exceed the electric efficiency of PAFCs (with the possible exception of the developing Direct Methanol cells). This apparent disadvantage is offset by other considerations, such as tolerance levels to fuel contaminates and cogeneration potential. Like PEMs, the PAFC employs a high cost metal catalyst (Pt) and requires an external reformer to produce a hydrogen-rich gas feed. However, unlike the alkaline fuel cell, which tolerates no CO or CO2, and the PEM in which CO is prohibitive, the PAFC sustains no performance degradation to CO2 and can tolerate CO levels up to 3% (by volume) in the fuel. Additionally, operation at 200°C enables the production of low-pressure steam or high grade hot water (250°F).

Figure 2-4 Cutaway View of a PAFC Stack40

Page 39: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-8

Voltage-Current Performance Cell performance for any fuel cell type is a function of pressure, temperature, reactant gas composition and utilization. The major voltage loss in the PAFC occurs at the cathode, and polarization is further increased when operating on air rather than with pure oxygen because of the dilution of the reactant; for example, cathode polarization operating with air is 560 mV versus 480 mV on oxygen at a current density of 300 mA/cm2 (30). In contrast, voltage losses at the anode are at least one order of magnitude lower (4 mV/100 mA/cm2)31. The ohmic losses in PAFCs are also relatively small, albeit larger than the anode polarizations at 12 mV/100mA/cm2.

Figure 2-5 depicts the progress of pressurized-PAFC performance from 1977 to 1992. Atmospheric pressure operating fuel cell stacks demonstrate reduced voltage performance. Published data indicates that a 32-cell short stack obtained 0.65V/cell at 215mA/cm2 (32). Typical PAFCs will generally operate in the range of 100 to 400 mA/cm2 at 600 to 800 mV/cell33. More recent performance specifications are unavailable as much of the development efforts are now considered proprietary.

Figure 2-5 PAFC Voltage Performance since 197734

a - 1977: 190°C, 3 atm, Pt loading of 0.75 mg/cm 2 on each electrode (35) b - 1981: 190°C, 3.4 atm, cathode Pt loading of 0.5 mg/cm 2 (36) c - 1981: 205°C, 6.3 atm, cathode Pt loading of 0.5 mg/cm 2 (36) d - 1984: 205°C, 8 atm, electrocatalyst loading was not specified (37) e - 1992: 205°C, 8 atm, 10 ft 2 short stack, 200 hrs, electrocatalyst loading not specified (37) f - 1992: 205°C, 8 atm, subscale cells, electrocatalyst loading not specified (37)

Page 40: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-9

Cell performance improves with higher pressures and higher current density. This can be attributed to lower diffusion (concentration) polarization at the cathode and an increase in the theoretical cell potential. Also, activation polarizations at the cathode are also reduced due to the increased partial pressures of oxygen and water. The increase in water partial pressure is significant because it affects the acid concentration which results in a concomitant rise in ionic conductivity38.

As cell temperature increases, the reversible cell potential decreases. However, in most fuel cells, this thermodynamic disadvantage is typically offset by reductions in cell polarizations due to improved reaction kinetics. This is true for PAFCs as the cathode reaction kinetics on Pt improve. The anode reaction registers relatively little improvement with increased temperature, but CO poisoning is reduced due to the equilibrium shift of the CO adsorption reaction on platinum. This effect increases PAFC tolerance to CO poisoning and is quantitatively illustrated in Figure 2-6.

Figure 2-6 Effect of Temperature on Cell Voltage with Impurities39

Current Density The current density in fuel cells is dependent upon several factors, primarily the activation, concentration, and ohmic polarizations. Each of these factors is associated with transport processes within cell components. As previously described, the electrochemical reactions take place at the surface of the so-called three-phase zone in the electrode. The physical picture of this zone is depicted in Figure 2-7. To increase the current density (and thereby the cell power output), each of the polarizations must be addressed. Methods to reduce these polarizations include maximizing the number of contact sites, increasing the partial pressure of the reactant gases, and minimizing the diffusion pathways40.

Page 41: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-10

Figure 2-7 Model of 3-Phase Zone in Cell Structure41

Reactant Utilization In general, increasing reactant gas utilization or decreasing inlet concentrations result in decreased cell performance due to increased concentration polarization and Nernst losses42. This phenomena is closely related to the compositions of the fuel and oxidant employed in fuel cells. As noted above, the oxidant composition and utilization affect the cathode performance. The composition performance dependence is clearly seen when the PAFC is operated on air versus pure oxygen. The use of ambient air (21% O2) over that of pure oxygen results in a decrease in current density by a factor of three43. In terms of utilization, Figure 2-8 depicts the magnitude of the voltage loss with increasing oxidant utilization.

Page 42: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-11

Figure 2-8 Voltage Losses in a PAFC with Increasing Oxygen Utilization44

As fuel cells will typically be fueled on a reformed hydrocarbon, low levels of CO2, CO, and residual (unreformed) hydrocarbons, such as CH4, C2H6, etc are present in the anode gas feed stream. Except for CO, these constituents are electrochemically inert and act as diluents as previously shown in Table 1-3. Because the anode reaction is nearly reversible, the fuel composition and hydrogen utilization do not strongly influence cell performance45. This characteristic has been confirmed in experiments in which for operation with pure hydrogen, the cell voltage remains nearly constant at hydrogen utilizations up to 90%46.

Fuels & Poisons PAFCs have been operated on many fuels including natural gas, propane, butane, landfill gas, hydrogen, and anaerobic digester gases. In general, the restriction on fuels stems from the use of the electrolyte material, H3PO4, which allows hydrogen ions to migrate freely in the fuel cell. The production of H+ ions at the anode is not accomplished with hydrocarbon gases. Table 2-2 shows the fuel composition limitation of natural gas for a PAFC.

Fuel cell contaminants and impurities are delivered to the cell stack either from productions within the fuel processor (reformer and shift converter) or originate from the fuel gas source at the point of entry into the plant47 As previously shown Table 1-3, the contaminants that are poisonous to PAFCs are primarily CO, S in the form of H2S and COS, NH3, and Cl. CO poisons the electrocatalytic activity of Pt electrodes due to its tenacious adsorption characteristics. Since CO accompanies the incoming fuel, only the anode suffers reduced performance by the blocking of H2 oxidation on Pt. H2S behaves similarly to CO in that it adsorbs on Pt and blocks active reaction sites. Interestingly, it has been reported that there is some “synergistic” effects between polarizations and the presence of both H2S and CO. That is, the voltage losses are much greater when both constituents are in the fuel stream than when either of the contaminants are present by themselves48. In general, the limitations on poisonous gas species before any cell performance degradation are 1-3% (by vol.) CO, <50 ppm H2S+COS, and < 20 ppm H2S49.

Page 43: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-12

Table 2-2 Fuel Composition Limitations of Natural Gas on the PC25

Composition Max Allowable (% volume) Methane 100 Ethane 10 Propane 36 Butane 1.25 Pentane 0.5 Hexane 0.5

Propylene 3.5 Carbon Dioxide 3

Oxygen 5 Nitrogen 18

Total Sulfur 30 ppm Ammonia 1 ppm Chlorine 0.05 ppm

Supply Pressure 4-14 inches H2O

Thermal Management Since the electrochemical reactions are exothermic and ohmic resistances are present, the heat generated must be removed to try to maintain an isothermal operating behavior. At least three methods of stack cooling have been developed (i) air cooling, (ii) water cooling, and (iii) dielectric liquid cooling (or oil cooling)50. Air and water cooling are the most commonly employed methods. The water-cooled design offers better heat removal characteristics but increases the design complexity and is typically designed for larger installations. Water cooling can employ both boiling-water cooling and pressurized water cooling51. Cooling plates can be inserted every 5-7 cells depending on the size of the stack52. The plate consists of insulated cooling pipes embedded in graphite plates as shown in Figure 2-9.

Air cooling is simple and is utilized in smaller applications. Forced excess air is employed to cool the stack. This method while simple, easy to control, and reliable is less efficient due to lower heat removal rates and larger auxiliary power requirements53. In the case of air cooling, a cooling plate is installed about every six cells54.

Page 44: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-13

Figure 2-9 PAFC Water Cooling Stack Configuration55

Cell Life The definition of fuel cell lifetime has not been well established, but to date has been defined as the duration at which the output voltage (at rated current) has decreased by 10%56. For many years, fuel cell developers (of all types) have targeted a cell life of 40,000 hours; a goal that can significantly affect plant economics. This target has been demonstrated by several IFC PAFC field units, however, in some cases only 35,000 hours were achieved57. Inspection of cell components after 35,000 hours indicated many were still in good condition and the stack could be overhauled to extend its life without full replacement to beyond 40,000 hours58. In general, the PAFC is estimated to decrease its voltage by 3mV every 1,000 hours or about 3% voltage drop per year for (0.75 V operation at beginning-of-life).

The cause of reduced cell life in PAFCs, as in MCFCs, is attributed to corrosion. In PAFCs the corrosion issue is centered on corrosion of carbon components, the chief of which are the carbon support for the catalyst layer and the separator (or bipolar) plate. During operation, product water from cell reactions is removed by natural evaporation in the form of steam that is carried away with the process air flow. While the vapor pressure of phosphoric acid is lower than that of water, the acid is entrained and carried out of the cell with the product water vapor and into the stack manifold. Additionally, for operation at voltages above 0.8 V carbon corrosion and Pt dissolution rates in acid are accelerated. Therefore, low current and hot idle open circuit operation are avoided59. Other factors that affect PAFC performance decay are sintering of Pt particles and electrolyte flooding.

SYSTEM OPERATION System flowsheets are generally proprietary, and difficult to obtain in the open literature. However, the 200 kW ONSI PAFC plant, being in an advanced state of development, has had some exposure and a presentation of the design becomes possible. This section will give a brief

Page 45: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-14

overview of the operation of the commercially available PAFC cogeneration system as documented by Braun, et al (Ref 9). Figure 2-10 displays a schematic diagram of the ONSI 200 kWe PC25 system. Pipeline natural gas enters the system at station 1 and is preheated in the integrated heat exchanger to a temperature necessary for the catalytic removal of sulfur in the hydrodesulphurization (HDS) unit. Superheated steam is then injected into the natural gas stream (station 5) at a steam-to-carbon ratio of approximately 3:1 (3 H2O molecules for every 1 Carbon atom). At 315°C the fuel-steam mixture enters the reforming unit where most of the methane and higher hydrocarbons catalytically react with the steam, producing hydrogen and carbon monoxide (e.g., CH4 + 2H2O = 4H2 + CO).

The reformed gas exits the reformer at high temperature (near 460°C) and enters the integrated heat exchanger, serving as the heat source to preheat incoming fresh fuel and to superheat steam for the purpose of steam-injection downstream (station 4). The reformed gas exit temperature is controlled by a hot water circulation loop in the integrated heat exchanger (entering at station 25).

The reformed gas then enters the low temperature shift converter (thermally integrated with the HDS unit) where, by further reacting with H2O, additional hydrogen is produced via the water-gas shift reaction (CO + H2O = H2 + CO2). At a temperature near 200°C, the prepared fuel enters the anode compartment of the fuel cell at station 8. The fuel electrochemically reacts with oxygen supplied by ambient air entering the plant at station 14. In the fuel cell unit, both electrical and thermal energy are produced. Electrical energy is sent to the inverter for DC/AC conversion and thermal energy is transferred to a pressurized water stream which jackets the fuel cell.

The depleted fuel exits the fuel cell and is delivered to the reformer burner at station 9 in order to complete the combustion of the unreacted fuel constituents, while providing heat for the endothermic reforming reaction. The product gas is then used for preheating combustion air for the reformer burner (station 12). The O2-depleted oxidant exiting the cathode compartment of the fuel cell at station 15, is mixed with the product gas (station 13) before being sent to the stack gas condenser for recovery of water from stream 16, necessary in order to meet the steam needs of the reformer without make-up. Here the H2O required (125 kg/h) for the reformer is cooled by a glycol-water mixture entering at station 35, condensed out of the stack gas and recycled.

The glycol-water mixture serves as the working fluid in a closed loop cycle (31-32-33-34-35-31) for the purpose of heat recovery. Low temperature heat is transferred to the glycol-water mixture in both the condenser and the thermal control heat exchanger. In the thermal control heat exchanger, heat is transferred to the glycol-water stream to meet the necessary operating parameters (station 27) required by the fuel cell water cooling jacket. Heat from the glycol is delivered to the customer in the heat recovery unit or, if the load is insufficient, dumped as waste heat in the cooling module. The 100°C glycol-water mixture (station 32) must be cooled to 50°C (a duty of about 245 kW) before admittance into the stack gas condenser, in order to assure adequate recapture of H2O for reforming. This is ideally accomplished if the hot water user returns are lower than 50°C.

Page 46: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-15

Figu

re 2

-10:

ON

SI 2

00 k

We

Phos

phor

ic A

cid

Cog

ener

atio

n U

nit (

PC25

Sys

tem

) (9)

Page 47: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-16

Thermodynamic Analysis At full load system operation, the energy input of the fuel is 558 kW (HHV), and 200 kW of electricity and a 245 kW maximum of low-temperature thermal energy in the form of hot water are produced. The corresponding fuel cell electric efficiency, defined as the ratio of electrical (AC) power output to the energy input with the fuel, is 36%. Maximum overall system efficiency (ηco ) is near 80% .

Of the 20% losses incurred during operation, 10% are expelled with the stack gas, 7% are due to heat losses in equipment, and 3% are lost in the DC/AC inverter. Strictly on the basis of an 80% First Law efficiency, the 200 kWe PAFC design operates as (or nearly as) efficient as most conventional thermal and cogeneration systems. However, the operating conditions for heat recovery are severely limited by strong dependence upon the water conditions returning to the heat recovery unit.

Figure 2-11 displays the amount of heat recovery and corresponding system efficiency for various water return and delivery conditions. To recover the maximum amount of thermal energy available (245 kW) and achieve system efficiencies of 80%, low water return temperatures (27 - 49°C, or 80 - 120°F) are required. Additionally, performance is very dependent on hot water delivery temperature requirements.

50 60 70 80 90 100 110

Water Delivery Temperature (°C)

0

50

100

150

200

250

Hea

t Rec

over

ed (k

W) a

t Ful

l Loa

d (2

00 k

We)

Ope

ratio

n

Estimated Heat Recovery and System Efficiency

35.8

44.8

53.7

62.7

71.6

80.6

Syst

em (1

st L

aw -

HH

V) E

ffic

ienc

y (%

)

273849

60

71

82

ReturnTemps. ( C)

Figure 5-3 Figure 2-11 Efficiency Performance vs. Water Return Conditions (9)

Page 48: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-17

DEVELOPMENT STATUS AND TECHNICAL ISSUES Of all the fuel cell types, phosphoric acid fuel cell technology is clearly in the highest state of development. With the commercial release in 1993 of the PC25 by ONSI Corporation, a subsidiary of International Fuel Cells (IFC), PAFC technology in the U.S. has been in an advanced development status for over five years. PAFC technology has moved from the demonstration of stationary power plants up to 11 MW to production and cost reduction of 200 kW commercial cogenerators. IFC is the technical leader in the technology with close competition coming from Japanese companies, primarily Fuji Electric. Over one hundred PC25 units, each the size of a minivan and generating 200 kWe, are now in operation, often in locations such as hospitals and remote hotels where grid power is expensive and reliability is worth a premium60. ONSI has several installations in which the PAFC unit is operating on alternative fuels, such as methane from landfills, anaerobic digester gases, propane, naptha, and butane (expected). Good transient capabilities have also been reported with 10 kW/sec ramping in grid-connected mode and 80 kW/sec in stand-alone configurations61. ONSI states that in over 2 million hours of total operation, their PAFCs have demonstrated better than 95% reliability and a mean time between forced outage of 2200 hours-- a figure that bests on-site, diesel-powered generators62. The current performance specifications for the ONSI PC25 unit are listed in the Table 2-3 below:

There are no technical hurdles to a viable PAFC product since the technology has been “commercial” for about five years. However, costs are still 2-3 times higher than the commercial market will sustain. Thus, for PAFCs, the major hurdle in sustained development and product penetration into commercial markets is solely cost reduction. To achieve 50-65% cost reductions, not only higher sales volumes will be required but design improvements in the power plant itself. It has been stated that in order to realize significant cost reduction to levels which will permit widespread commercial entry, costs associated with every element of the power plant, including fuel processor, cell-stack design, power conditioning and control, and ancillary components plus assembly, must be reduced63.

Design improvements of the cell and stack surround improved electrodes, higher power density, reduced materials costs, and life of the stack. The primary loss in performance (voltage and, hence, efficiency) in PAFCs is associated with activation polarization which is a result of the reaction pathway and catalyst activity of the electrodes64. Also, higher power density would improve the kW/kg of stack and hence reduce cost. In terms of life, the endurance of cell stacks is severely affected by the cell operating point. As previously stated, cell life deteriorates at low current (high voltage) operation. This is a relatively minor issue; at the other end of the operating envelope, the cell performance stability at high current densities has been recently stated as the major factor affecting stack overhaul periods65.

Page 49: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-18

Table 2-3 Performance Specifications for the ONSI PC25 200 kW Unit (61,62)

Parameter Specification Comments

Power Capacity 0 to 200 kW with natural gas fuel

Thermal Energy (Cogeneration)

740,000 kJ/hour at 60�C; or 369,000 kJ/hour at 120�C or 60�C

700,000 Btu/hour at 140�F; 350,000 Btu/hour at 250�F

Efficiency (HHV) 36-38% Electric 80% max total w/ thermal

Ambient Temperature Range -30 to 45�C, up to 1500 m elev.

Voltage and Phasing 480/277 volts at 60 Hz 400/230 volts at 50 Hz

Electric Connection Grid-connected for on-line service and grid-independent for on-site premium service

Power Factor Adjustable between 0.85 to 1.0

Transient Overload None

Grid Voltage Unbalance 1%

Grid Frequency Range +/-3%

Voltage Harmonic Limits <3%

Plant Dimensions 3 m (10 ft) wide by 3 m (10 ft) high by 5.5 m (18 ft) long

Not including a small fan cooling module (5)

Plant Weight 18,200 kg (40,000 lbs)

MARKETS AND MANUFACTURERS PAFCs are targeting the commercial sector where waste heat can be utilized. These applications are typically hospitals, hotels, schools, and high value commercial buildings requiring high power quality, premium power services. For installed costs between $1500 – $2000/kW, the U.S. market size has been estimated at 10-125 MW/yr in high value commercial buildings and increasing to 250 MW/yr for $1000/kW cost66. If the goal of $1500/kW (installed) is achieved, the ONSI 200 kW units are expected to be competitive with approximately 50% of the retail market (1996 data)67.

PAFC development historically included transportation applications, such as transit buses. However, due to the rapid advancements of proton exchange membrane fuel cells, PAFCs are not likely to compete in light and medium duty vehicular transportation. Future applications for PAFCs may be found in marine or locomotive transportation markets and possibly in space applications.

Page 50: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-19

The ONSI Corporation of Windsor, Ct, a subsidiary of International Fuel Cells Corporation, is the only U.S. PAFC developer. In the past, both Westinghouse and Energy Research Corporation (ERC) were involved in PAFC development. While ERC still retains patents and PAFC designs, they are focused on MCFC development. Westinghouse sold their PAFC technology to the Fuel Cell Corporation of America of Large, PA in the early 1990s. However, the Fuel Cell Corporation of America as of 1996, had yet to develop a complete, integrated system and had no operating experience or demonstration units in the field68. Today, no literature can be found on them and they are assumed to no longer be in the running.

Foreign PAFC developers include Fuji Corporation, Mitsubishi Electric Co., and Toshiba Corporation in Japan. Ansaldo in Europe has obtained licensing rights from ONSI to sell and eventually build 200 kW units69.

CONCLUSIONS Of all fuel cell types, the phosphoric acid fuel cell is the most advanced, with no significant technical hurdles. It has been successfully demonstrated and presently has over 100 field units operational throughout the world. While the goal of PAFC developers has been to manufacture economically feasible systems for on-site applications, this has not yet been accomplished due to high system costs and low sales volumes. To achieve further reductions in systems costs, improvements in system and component designs and manufacturing methods must be achieved. Braun, et al. (70) have shown that the removal of the waste heat recovery system and the integration of the fuel cell and fuel preparation equipment with onsite boilers could result in a 10% plant cost reduction of the ONSI 200 kW power plant70.

Excitement surrounding PAFCs has been minimal (relative to PEM technology) due to both cost considerations, and relative anonymity of stationary fuel cell power plant market potentials. While performance is excellent, demonstrating good electric efficiency, fuel flexibility, and operating ruggedness, the biggest push for fuel cells may be a result of environmental considerations. In general, power plants account for approximately 25% of all CO2 and 40% of NOx, SOx emissions and other airborne pollutants71. The Kyoto Protocol and Clean Air Acts indicate increasing global environmental awareness that stimulates interest and investment in more efficient and clean alternative energy conversion technologies, such as fuel cells. The demonstrated emission output of ONSI’s PC25 units have a distinct advantage over other competing technologies. Some emission outputs are listed in Table 2-4 below. How this translates in terms of market penetration and the ultimate success of the PAFC remains to be seen. Nevertheless, the PAFC will assist in paving the way for next generation fuel cell technologies, such as PEMs, MCFCs, and SOFCs. Each fuel cell technology has its advantages and given the ever-increasing electric capacity demands, they may all find a market niche.

Page 51: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-20

Table 2-4 Comparison of 200 kWe PC25 Emissions vs. Conventional Plants72

Technology

Emissions

PC25 PAFC Cogenerator*

(lbs/MWh)

Pulverized Coal Fired Steam Plant

(lbs/MWh)

Combined Cycle Gas Turbine (lbs/MWh)

NOx 0.016 2.56 0.22

SOx 0 7.60 0

CO2 1122 1868 761

Particulate 0.00003 0.183 --

Total Hydrocarbons

0.00036 0.01 --

CO 0.023 -- --

Efficiency (HHV)

36% 37% 53%

*Operating on natural gas

Page 52: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-21

REFERENCES 1 E.W. Hall, et al., “Eighty Months of Commercial Experience with PC25 Fuel Cell Power Plant,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, November, p. 24. 2 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 1-15. 3 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, pp.285-288. 4 F.S. Kemp, “ On-Site Fuel Cell Power Plant Technology and Development Program Annual Report-1984,” GRI-85/0195, prepared by United Technologies Corp. for the Gas Research Instititute, Research Project FCR-7030, 1984. 5 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p.299. 6 Ibid., Figure 8.15, p. 300. 7 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 94. 8 Ibid. 9 R.J. Braun et al., "An Analysis of A Phosphoric Acid Fuel Cell Cogeneration System: I. Improvements via Ssytem Integration," ECOS '95 Proceedings, Istanbul, Turkey, 1995. 10 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, Figure 8.5, p.285. 11 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 3-3. 12 A.J. Appleby, in Proceedings of the Workshop on the Electrochemistry of Carbon, Edited by S. Sarangapani, J.R. Akridge, and B. Schumm, The Electrochemical Society, Inc., Pennington, NJ, p. 251, 1984. 13 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p.306. 14 E. Barendrecht, “Electrochemistry of Fuel Cells” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 94. 15 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 306. 16 Ibid., p. 309. 17 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 97. 18 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 303. 19 Ibid. 20 Ibid., p. 305. 21 Ibid. 22 Ibid. 23 K. Kinoshita, F.R. McLarnon, E.J. Cairns, Fuel Cells, A Handbook, prepared by Lawrence Berkeley Laboratory for the U.S. Department of Energy under Contract DE-AC03-76F00098, May 1988, pp. 37-53. 24 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 301. 25 Ibid., p. 309. 26 S.D. Moody and W.R. Dunbar, “Phosphoric Acid Fuel Cell Technical Report,” Center For Energy Studies, Marquette University, Milwaukee, WI, 1993. 27 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, Figure 8.17, p. 301. 28 Ibid., p. 315. 29 Ibid.

Page 53: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-22

30 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 3-10. 31 Ibid. 32 Ibid., p. 3-7. 33 Ibid., p. 3-10. 34 Ibid, Figure 3-1, p. 3-5. 35 A.P. Fickett, in Proceedings of the Symposium on Electrode Materials and Processes for Energy Conversion and Storage, edited by J.E McIntyre et al., The Electrochemical Society, Inc., Pennington, NJ, 1977, p. 546. 36 A.J. Appleby, J. Electroanal, Chem., 118, 31, 1981. 37 J. Huff, “Status of Fuel Cell Technologies,” in Fuel Cell Seminar Abstracts, 1986 National Fuel Cell Seminar, Tucson, AZ, October, 1986. 38 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 3-11. 39 V. Jalan, et al., “Development of CO and H2S Tolerant PAFC Anode Catalysts,” in Proc. Of the Second Annual Fuel Cell Contractors Review Meeting, 1990. 40 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 301. 40 Ibid., p. 303. 41 J. Hiramoto and R. Anahara, Fuji Electric J., 9, 555, 1982. 42 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 3-12. 43 Ibid. 44 P.W.T. and L.L. France, in Proc. Of the Symposium on Transport Processes in Electrochemical Systems, R.S. Yeo, et al., Editors, The Electrochemical Society, Inc., Pennington, NJ, 1982, p. 77. 45 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 3-14. 46 “Advanced Water Cooled Phosphoric Acid Fuel Cell Development,” Final Report, Report No. DE/MC/24221-3130, International Fuel Cells Corporation for U.S. DOE, South Windsor, CT September, 1992. 47 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 3-15. 48 Ibid., p. 3-16. 49 Ibid. 50 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, Figure 8.17, p. 293. 51 Ibid. 52 J.M. King and G.W. Scheffler, “Phosphoric Acid Fuel Cell Subsystem Technology Final Report,” GRI-88/0215, prepared by International Fuel Cells Corporation for GRI, Research Project FCR-9594, September 1988, p. 3-6, 3-9. 53 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, Figure 8.17, p. 296. 54 J.M. King and G.W. Scheffler, “Phosphoric Acid Fuel Cell Subsystem Technology Final Report,” GRI-88/0215, prepared by International Fuel Cells Corporation for GRI, Research Project FCR-9594, September 1988, p. 3-6, 3-9. 55 T. Hirota, et al., Fuji Electric J., 61(2), pp. 133-187, (1981) 56 R. Anahara, “Research, Development, and Demonstration of Phosphoric Acid Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, Figure 8.17, p. 310. 57 E.W. Hall, et al., “Eighty Months of Commercial Experience with PC25 Fuel Cell Power Plant,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, November, p. 27. 58 Ibid., p. 28.

Page 54: Review of State-of-the-Art Fuel Cell Technologies for Distri

2-23

59 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 3-3. 60 Scientific American, December, 1996. 61 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 1-10. 62 E.W. Hall, et al., “Eighty Months of Commercial Experience with PC25 Fuel Cell Power Plant,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, November, p. 26. 63 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 310. 64 Ibid. 65 Ibid. 66. D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996, 3-9. 67 Ibid. 68 Ibid. 69 Ibid. 70 R.J. Braun et al., "An Analysis of A Phosphoric Acid Fuel Cell Cogeneration System: I. Improvements via Ssytem Integration," ECOS '95 Proceedings, Istanbul, Turkey, 1995. 71 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 181. 72 “The PC25 Fuel Cell Power Plant,” (Brochure), prepared by International Fuel Cells Corp., South Windsor, CT, August, 1986.

Page 55: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-1

3. Proton Exchange Membrane Fuel Cells (PEMFCs)

The proton exchange membrane fuel cell (PEMFC) is receiving increasing attention as major technical milestones have been reached. In the last several years, cell power density (power out per unit weight or volume) has increased by an order of magnitude and substantial cost reductions have been achieved. These advances are largely the result of a technology push to market PEMFCs in the transportation sector. In 1997, over $750 million dollars from automakers, such as Daimler-Benz and Ford Motor Company, was committed for investment into Ballard Power (the leader in PEMFC development) to develop cost-competitive light-duty automobile PEMFC power plants by 2004. Additional collaborations of automakers with other fuel cell manufacturers, United Technologies, for example, that will rival the Ballard-Ford-Benz deal are also expected1.

The excitement surrounding PEMFCs is tempered by the reality that many technological barriers stand between failure and success in the ultra competitive automotive market. For this reason, it is anticipated that PEMFCs will first be marketed in stationary applications. The low-operating temperature, rapid startup, high power density, and simplicity characteristics that make the PEM fuel cells attractive for transportation, also make them attractive in remote, standby, and premium power onsite markets.

This section will outline the technical aspects of PEMFCs, such as electrochemistry, cell performance, and fuel processing. Cell construction materials and anticipated costs are then presented followed by an overview of development status and manufacturers. The section concludes with a summary of PEMFC advantages, and technical and economic issues that require resolution.

ELECTROCHEMISTRY Proton exchange membrane fuel cells operate at the lowest temperature (80°-100°C) of all the fuel cell types. The electrolyte consists of a solid polymeric membrane (fluorinated sulfonic acid) fitted between two platinum impregnated porous electrodes. The polymeric membrane has two primary functions: (i) as electrolyte, possessing excellent hydrogen proton conductivity characteristics, and (ii) as reactant gas barrier which separates the fuel gas (hydrogen or hydrogen rich mixture) from the oxidant gas (O2 or air). Unlike liquid electrolytes, solid electrolytes generally provide better resistance to gas crossover from the fuel and air electrodes. This is significant in that loss of fuel in the depleted oxidant stream can be averted and opportunities for developing an explosive mixture of hydrogen and oxygen are reduced. The electrodes are thin gas-diffusion electrodes and are comprised mainly of a carbon cloth backing and micro-platinum particles deposited on a carbon support. The platinum loading acts as an electrocatalyst ensuring the electrode reactions take place at a sufficient rate.

The cell assembly shown in Figure 3-1 is the heart of the electrochemical system. It consists of a solid polymeric membrane, anode, and cathode. The components are pressed together under

Page 56: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-2

elevated temperature and pressure to form the membrane electrode assembly (MEA)2. Graphite bipolar separator plates are then pressed against the MEA to provide current collection, structural support, and reactant gas pathways. The MEAs and separator plates are stacked to meet the design power requirements (see Figure 3-2). The entire stack assembly can be held together by mechanical screws or pneumatic pressure.

During operation, hydrogen (typically reformed from natural gas, propane, or methanol) and air (or pure O2) flow down channels in the bipolar plates where on one side, each electrode face is exposed to the reactant gases. Figure 3-3 depicts a schematic of the electrochemical reacting system. Because of its chemical affinity for oxygen, hydrogen entering the anode compartment is adsorbed to the anode, where it diffuses through the electrode to the electrocatalyst sites and is oxidized (i.e., a loss of electrons occurs). The hydrogen ions then migrate through the electrolyte to the cathode, carrying the positive charge and thus accomplishing useful work. At the cathode, the gas diffuses to the electrocatalyst for oxygen reduction. Similarly, oxygen entering the cathode compartment is adsorbed to the cathode, diffuses to the electrocatalyst sites, and is reduced by hydrogen protons that have migrated through the polymer membrane. The electrode reactions that occur in the presence of the platinum electrocatalysts are summarized as follows:

Anode: H2 2H+ + 2 e-

Cathode: O2 + 4H+ + 4e- 2H2O

Overall: 2H2 + O2 2H2O

Typically, the electrode reactions are rate limiting. For PEMFCs, the cathode reduction reaction is the slowest reaction and therefore, much research has been carried out to characterize and improve the electrocatalytic performance of this electrode3,4,5,6,7,8,9. In general, because of its good oxygen solubility characteristics, the electrocatalysis of the cathode reaction is superior to that of the phosphoric acid fuel cell10.

OPERATION AND CELL PERFORMANCE The proton exchange membrane fuel cell typically operates at about 85°C. This temperature is lower than the PAFC, MCFC, and SOFC which are in a more advanced state of development. The operating temperature is set by both the thermal stability and the ionic conductivity characteristics of the polymeric membrane11. To get sufficient ionic conductivity, the proton-conducting polymer electrolyte requires liquid water. Thus, temperatures are limited to less than 100°C. The low-operating temperature allows the PEMFC to be brought up to steady-state operation rapidly. This characteristic, coupled with the lightweight, high power density features makes PEMFCs attractive for transportation applications. However, the low-temperature operation results in low-grade waste heat which is not suitable for most cogeneration applications and makes thermal integration with fuel processing equipment difficult.

The PEMFC can operate at elevated pressures and it is often beneficial to do so in order to obtain high power densities. Air pressures up to 8 atm have been used12. Also, the solid polymer membrane can support substantial differential reactant pressures which enables some flexibility

Page 57: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-3

in the PEFC system design and opportunities for maintaining water balance in the membrane (a critical issue which will be discussed later in this section)13,14.

Fuels & Poisons As with many of the lower temperature fuel cells, the PEMFC requires a pure H2 source for operation. Since hydrogen is not readily available, it is typically obtained by reforming a hydrocarbon fuel, such as methanol or natural gas. The reformate (the product fuel stream from the reformer) often contains significant amounts of residual CO and CO2. A typical composition of steam-reformed methanol is 75% H2, 24% CO2, and 0.3 to 1 % CO by volume15 . CO ppm levels of 50 or greater poison the catalyst, causing severe degradation in cell performance. Poisoning of the platinum catalyst occurs because of “tenacious adsorption” of CO on the catalyst sites16,17. Since the electrode reactions are surface phenomena, the greater the poisoning, the less area available for the electrochemical cell reactions. Additionally, it has been recently observed that the sustained use of CO2 in PEMFCs gradually degrades cell performance (by as much as 30% in just 1000 hours), effectively limiting cell life18. Thus, if natural gas or methanol fuels are to be used, additional processes or measures must take place to limit oxides of carbon concentration.

Methods to reduce CO in the reformate include external selective oxidation using alloy catalysts and bleeding of O2 into the reformate. External selective oxidation is a process to convert CO to CO2 by mixing reformate with a small amount of air then passing it through a column packed with 0.5% platinum on alumina19. This can limit CO levels to about 100 ppm. To obtain further reductions to sub-10ppm levels, bleeding of about 2% O2 into the CO containing reformate is employed20. Thus, selective oxidation is used to convert the bulk of the CO to CO2, and O2 bleeding can be used to convert the residual low-level CO content. Use of these two methods combined can enable cell performance on par with that of pure hydrogen21. However, partial oxidation of CO to CO2 may be ineffective due to the potential for reverse water-gas shift of the reformate, (i.e., formation of CO and H2O from H2 and CO2)22. The reverse water-gas shift reaction may indeed be contributing to poor life performance. In any case, the current strategy favors long-term CO2 performance degradation over shorter term CO electrocatalyst poisoning. One developer has indicated that no performance degradation on CO2 in their units has been observed23. Additional surveying of the literature is required to further confirm and assess CO2 performance degradation effects and preventive measures.

Fuel type and processing represents the single greatest challenge facing fuel cells in general. This is particularly true for PEMFCs considering their susceptibility to electrocatalyst poisoning from low-level CO and more recently, CO2 in life testing. Given sufficient fuel processing, PEMFCs are expected to operate using hydrogen, methanol, propane, and natural gas fuels.

As Table 1-3 (Section 1) shows ammonia, chlorine, and sulfur are also poisons to PEMFCs.

Efficiencies PEMFC stacks have an electrical efficiency (dc power to LHV of the anode feed gas) of nearly 50%24. However, due to the poor thermal integration of fuel cell stack with conventional steam-reforming equipment, the overall system efficiencies have been limited to 42%25. In general,

Page 58: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-4

PEMFC systems have the lowest electrical efficiencies of all fuel cell systems. A comparison of fuel cell efficiencies is displayed in Tables 1-1 and Table 1-3 in Section 1 of this report.

Current-Voltage and Power-Current Performance Curves As previously discussed in section II of this report, the theoretical reversible cell voltage (directly related to fuel cell efficiency) is influenced by temperature and pressure according to the Nernst equation. Thus, an increase in operating pressure will increase the voltage output of the cell as the partial pressure of the reactants (most significantly O2) is higher. Figure 3-4 depicts a voltage-current density plot for a single PEMFC with a Dow membrane. The performance curve translates into a power density of 520 watt per square foot of cell area (at an operating current density of 1 A/cm2). In general, membrane performance has been increased by an order of magnitude in the last ten years; some have registered power densities up to 2500 W/ft2 (26). Thus, considering only a cell stack with 1kW/ft2, some 75 kW (100hp) of power might be developed using only 75 - 1 ft2 cells (roughly the same volume occupied by a large suitcase.)

Water Management There are several water transport mechanisms in PEM fuel cells and their effective management is critical. As the cell reaction indicates, water is produced at the cathode. The product water does not dissolve in the electrode or the electrolyte (both are insoluble and cannot be diluted by water). Instead, water produced from the reduction reaction is typically ejected into the cathode compartment where it is carried away by the oxidant stream27. The production of water at the cathode can cause flooding of the electrode if it is not removed at a sufficient rate. The presence of excess water in the cathode inhibits reactant gas diffusion into the electrode. Thus, hydrogen protons arriving at the cathode find reduced amounts of O2 available for reduction reaction. Additionally, because the membrane ionic conductivity is directly proportional to water content, too much removal of water can dehydrate the membrane and severely impair the cell performance. This problem is exacerbated by the electro-osmotic pumping of water from anode to cathode (see Figure 3-5). This effect is characterized by the hydration of hydrogen protons by membrane water content (or humidified fuel gases). Movement of the hydrogen protons as they migrate through the electrolyte is then accompanied by the drag of water molecules to the cathode side. The result can be drying of the membrane and clogging of the cathode’s porous electrocatalytic sites which ultimately inhibits oxygen mass transfer rates. An additional water transport mechanism occurs by back diffusion of water along the concentration gradient from cathode to anode. In general, to assist in maintaining sufficient water content in the membrane, the fuel gases are humidified. This is accomplished using the product water expelled from the stack in the cathode effluent.

The mechanisms for water transport are then summarized as: (i) water production and removal at the cathode, (ii) water drag from anode to cathode, (iii) water transport from fuel gas to anode, and (iv) back diffusion from cathode to anode. A water imbalance can cause either flooding of the electrodes or dehydration of the membrane, both of which degrade cell performance. It is therefore crucial for cell performance to utilize effective control schemes for the addition of water at the anode and removal of it from the cathode

Membrane water management is complex and several developers employ different strategies which generally fall into the categories of passive or dynamic water removal control. Passive

Page 59: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-5

water removal control methods traditionally use a porous hydrophilic collector with wicking pathways that facilitate product water transport to the collector and then removal28. The collected liquid-vapor mixture is separated and used for reactant gas humidification. Another measure employed is to maintain a pressure differential between cathode and anode compartments which inhibits the ingress of water from anode to cathode. Ballard Power Systems and others often operate at a cathode-to-anode pressure imbalance of 5:3 atm. In dynamic water removal control schemes, the excess air in the cathode pushes the water out of the cell where the liquid is then separated in a drum29. It is “dynamic” in the sense that the amount of air flow is dependent on the stoichiometry and operating point (load) of the cell.

More recently, Ballard has developed a fuel cell stack which makes use of the back diffusion of water along the concentration gradient from cathode to anode, effectively enabling anode product water removal30. Testing of this stack design has also resulted in a substantial reduction of parasitic loads due to reduced excess air flow rate requirements. Air flow rate can now be operated near the stoichiometric level of 1.0, down from the more typical level of 2.0 (or 100% excess air)31.

Thermal Management As Table 1-3 shows, the medium used to cool the PEM fuel cell stacks is typically accomplished with water. Water or excess air is circulated through the cell stacks via cooling plates on either side of the MEA as shown in Figure 3-6. In transportation applications, the liquid coolant is circulated through a heat exchanger where the thermal energy is dumped to the ambient air via convection from vehicle movement or a fan, in much the same manner as automobile radiators.

CELL CONSTRUCTION AND MATERIALS The membrane-electrode assembly (MEA) makes up the heart of the PEMFC stack. In addition to the MEA, the cell separator plate is a crucial stack component. As with the MEA, it is a repeating unit that provides reactant gas flow pathways, current collection, gas separation from adjacent cell components, and structural strength. These component materials and costs are discussed below.

Membrane The PEMFC uses a solid polymer membrane (fluorinated sulfonic acid). The membrane is composed of a fluorocarbon polymer “backbone structure, similar to Teflon,” to which sulfonic acid groups (SO3H) have been chemically bonded32. The membranes used are very thin (.05 to 0.18 mm) and are safely and easily handled33. Properties of the membrane that affect fuel cell performance include water content, oxygen solubility, conductivity, and thermal stability34. These properties are intimately connected with the equivalent weight of the membrane (ratio of weight of the polymer to the number sulfonic acid groups)35. The lower the equivalent weight, the better the ionic conductivity.

Several manufacturers have developed membranes. The most visible of these have been DuPont’s Nafion and more recently, a Dow Chemical membrane. The current cost of these membranes is prohibitive at $500/m2, but it is anticipated that the costs will come down with

Page 60: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-6

high volumes and competition. The transportation commercial cost goal for membranes is $15/m2 of material36. Table 3-1 lists the PEMFC cell materials and cost estimates.

Electrodes Porous, gas-diffusion electrodes, very similar to PAFCs are employed in solid polymer membrane fuel cells. They are comprised of a platinum electrocatalyst supported on porous carbon cloth with a hydrophobic coating. Previously, the high cost of electrodes was associated with the high platinum loading requirements for electrocatalysis. Cost reduction without detriment to performance was accomplished by employing supported platinum catalysts on the carbon surface in much the same manner as PAFCs. This reduced the platinum requirements by as much as a factor of 80 (4mg/cm2 to 0.05 mg/cm2)37,38.

With a reduction in the noble metal (Pt) requirements, electrode costs have been reduced substantially. One study shows that the platinum cost for an 80 kW PEMFC plant has been reduced from $30,000 to $500 by using loadings of 0.25mg/cm2 or less and low-cost manufacturing techniques39. Automakers have recently indicated that only $225 of platinum will be necessary for a mid-sized automobile40.

Separator plate The separator plate must be impervious to gas diffusion. Traditionally, this component is constructed from flat graphite with machined channels to accommodate reactant gas flow41. Both impervious graphite plate stock and milling are of high cost. Cost reduction efforts in these areas are focused on the use of alternative materials, such as molded composites and coated metals.

Cell Stack A cost breakdown for the stack components in a pilot plant scenario is shown in Table 2. Cost reductions are anticipated for production units. The projected total stack cost for the various markets are summarized below42:

Utility Applications $350 - 500/kW

Heavy Duty Transportation $100 - 300/kW

Automotive $20 - 50/kW

STATUS OF DEVELOPMENT AND COMMERCIALIZATION Ironically, while much of the press attention is on PEMFCs for transportation, they are first targeted in stationary power applications. Although the state of PEMFC development is lagging other fuel cells, Ballard Power Systems, the leader in PEM development, is quietly and aggressively pursuing the commercial market sector in the 250 kW range43. In fact, stationary power development at Ballard is the largest business activity and at least one-third of the company’s resources are devoted to it44. Presently, fuel cell developers and utility experts are not in agreement on which of the numerous power markets PEMFCs will find the greatest success in45.

Page 61: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-7

Due to their lightweight, power density, and load-following characteristics, PEMFCs are also being aggressively developed for transportation applications. Automotive giants Daimler-Benz and Ford Motor Company have committed $750 million for PEMFC into research, development, and cost reduction. With this capital, PEMFCs are expected to overcome the numerous fuel processing, system, and cost reduction hurdles within the next five years. The potential dual market for PEMFCs has given rise to low capital cost expectations that may make them competitive in small-scale retail power markets.

PEMFCs are currently being developed primarily for sizes less than 500 kW. The applications included are listed in the table below:

PEMFC APPLICATIONS

Transportation Stationary Military/Portable

Light Duty Vehicles (50 – 100 kW) Residential (2 – 10 kW) Small (20 – 100 W)

Medium Duty Vehicles (200 kW) Commercial (250–500 kW) Field Generators (5 kW)

Marine Onboard Power & Propulsion Battery Replacements (100 kW)

Submarine (40 – 400 kW)

Existing plants To date, several PEMFC power plants have been developed for both transportation and stationary applications. The most visible plants are those prototypes developed for transportation. Ballard Power Systems is pacing the commercialization efforts with PEMFCs in several demonstration transit buses in the Chicago Transit Authority and British Columbia transit system, and the NECAR I and II with Daimler Benz. In stationary power, they have demonstrated proof of concept prototypes of 10 and 30 kW systems and are now focusing on a market entry 250 kW distributed generation unit for 200046.

H-Power has developed 10 kW stacks for Ford and is currently working on 3 kW residential units which they expect to sell for about $5,000 in moderate volume47. Additionally, H-Power has developed the first wholly unsubsidized, fully commercial fuel cell unit for a trailer-mounted, electric powered highway construction sign48.

Economics To be successful in the transportation sector, PEMs will need to be available for at least $150/kW49. In fact, automakers believe that in light duty applications they will need to be around $25-50/kW50. This means that the current cost of $500/kW will need to be reduced by another order of magnitude. To accomplish this, production volumes on the order of one million units per year may be necessary.

Page 62: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-8

Most generally agree that stationary onsite fuel cell systems will need to be on the order of $1500/kW to compete. Due to PEMFCs lower system efficiencies, utility experts believe they will need to be of lower cost or will have to offer other benefits to potential customers in order to compete against PAFCs in commercial markets51. The stack costs are broken down in Table 3-1. Balance-of-plant cost estimates are not available at this time.

Technical & Economic Issues For stationary power applications, the technical issues that are barriers to PEMFC commercialization are summarized as:

1. Electrocatalyst tolerance to low-level CO concentrations in the fuel feed gas 2. The long term effects of CO2 concentration on cell life. 3. Water management and membrane operating temperature limits. 4. Optimal system integration (e.g., current thermal integration difficulties with fuel

reformer.) 5. Membrane and balance-of-plant (BOP) costs. 6. Manufacturing costs (stack and cell components).

Consistent with the main technical issues, the recent research areas have focused on:

1. CO conversion and CO-tolerant catalysts to allow operation on reformed fuels, 2. Polymer membranes for high power density and low cost, 3. Low Pt catalyst loadings for low cost and alternative (alloy) catalysts.

One less severe requirement imposed by transportation applications is that of cell life at 3,000-5,000 hours. For stationary applications, the economical target is 40,000 hours. However, transient operation poses severe problems as current reformer start-up times are approximately 10 minutes in duration52. Little work has been published in the area of systems integration, but it is believed that this aspect is most carefully guarded by prospective developers. In general, the technology is considered viable but costly. This is evidenced by the current focus of many developers on cost reduction methods.

MANUFACTURERS Ballard Systems is the leader in PEMFC development. They are currently developing 250 kW distributed power systems for commercial markets and are very involved in transportation development for both light and medium duty applications. Relative to Ballard, Plug Power, H-power, Mechanical Technology Inc., Energy Partners, and International Fuel Cells (IFC) appear to be next in development status, with Allied Signal closing. H-Power is focused on battery replacement, transportation market, and small (sub 10 kW) residential and dispersed generators. Mechanical Technology Inc. is focused on both transportation and stationary power. IFC has been involved in PEMs for about ten years and is currently working on 50 kW stacks for Ford and smaller unmanned undersea vehicles53.

A list of PEMFC manufacturers is presented in Table 3-2. The rest of the companies listed are in a relatively equal state of development and commercialization.

Page 63: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-9

CONCLUSIONS PEMFCs have received increased attention recently as significant technical issues have been hurdled. Of note is the significant investment of automakers ($750 million plus) into PEMFCs for transportation use. While penetration into the transportation market by 2004 represents the most difficult challenges, PEMFCs are anticipated to enter into the stationary power residential and commercial markets in a much nearer timeframe (~2000).

PEMFCs are the lowest operating temperature fuel cell at 80 – 90°C and have the lowest electrical efficiency of all fuel cell systems. The solid polymer membrane provides excellent hydrogen proton conductivity and has the highest power density of any type of cell. However, the low operating temperature along with the various water transport mechanisms, make water management in PEMFCs critical to sustained performance. Additionally, because of the lower temperature operation and the platinum electrocatalyst requirements, CO and CO2 concentration in the anode feed gas (fuel) must be rigorously limited to avoid poisoning and voltage degradation.

The specific characteristics that PEMFCs have that may make them advantageous over other, higher temperature fuel cells is summarized in the table below54.

PEM Characteristic Possible Advantages Low Temperature Operation Lower cost Rapid startup and shut-down Material corrosion problems are reduced

Solid Electrolyte Simpler cell design No corrosive liquid in cell and no electrolyte redistribution Able to withstand large pressure differentials and control is therefore simplified

High Power Density Lower unit cost and raised fuel cell efficiency Low weight and volume requirements

These traits indicate that PEMFCs are suitable for transportation and some stationary applications. Presently, the best stationary applications for PEMs are uncertain as developers and utility experts are in disagreement. Utility experts believe that PEMFCs are best suited for standby, peaking, and premium power markets. Ballard Power is aggressively pursuing distributed applications aimed at the commercial sector with large systems of 250 kW. In commercial applications, the brass ring for capital cost is believed to be about $1500/kW, while the transportation sector requirements are much more severe at $25-50/kW.

The infusion of R&D capital by the automakers is expected to accelerate PEMFC development schedules and assist in meeting cost goals as the potential dual market will improve manufacturing economies of scale.

Page 64: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-10

TABLE 3-1

PEM Fuel Cell Materials and Costs

Component Material Dimension Cost Current Projected

Membrane Fluorinated sulfonic acid polymer (Nafion by DuPont)

Perfluorinated sulfonic acid polymer (by DOW)

(50 – 170 µm thick)

$500/m2

--

$15/m2

--

Electrodes Porous carbon cloth with Pt catalysts

~0.1 mg Pt/cm2

Separator Plate Milled Graphite

Membrane Electrode

Assembly (MEA)

Membrane, anode, and cathode $400 - 500/kW

Cell Separators Milled graphite $200 – 300/kW

Balance of Stack Components

Screws, cooling plates, manifolding, etc.

$50 - 100/kW

TOTAL $650 – 900/kW $350-500/kW*

* Adapted from Refs 13, 15, and 18.

Page 65: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-11

Table 3-255 PEMFC Companies Engaged in Development & Commercialization

Manufacturer / Vendor Target Market

1. Ballard Power Systems Transportation Light Duty (Automobiles) Medium Duty (Transit buses) Stationary Commercial (250kW initially)

2. Mechanical Technology Inc. Transportation

Stationary Distributed Power

3. Energy Partners Transportation

Stationary Distributed Power

4. H-Power Transportation

Stationary - 2-5kW Residential & Dispersed -Battery Replacement

5. International Fuel Cells Transportation

6. Allied Signal Transportation

Stationary Distributed Power

7. Analytic Power Transportation

Stationary

8. ElectroChem - OEM component supplier - Small Distributed Power

10. Delphi (Div. of GM) Transportation

11. Northwest Power Systems Stationary Remote Power

Page 66: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-12

Figure 3-1 Schematic of a Membrane Electrode Assembly (MEA)

Fuel Cell Stack Components

Figure 3-2

Page 67: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-13

Figure 3-3

Chemical Reacting System within a PEMFC

(from reforming)

Anode: H2 2H+ + 2e- Cathode: 1/2 O2 + 2H+ + 2e- H2O

Page 68: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-14

Figure 3-4 PEM Cell Potential vs. Current Density56

0 200 400 600 800 1000 1200 1400 1600

Current Density (mA/cm2)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

1.10

1.20

Cell

Pote

ntia

l (Vo

lts)

(for a single DOW membrane cell with 0.05 mg/cm^2 platinumloaded gas diffusion electrode at 70°C and 1 atm)

Page 69: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-15

Ano

deC

atho

de(2

) H

2O -

H+

Wat

er D

rag

(3)

Bac

k-D

iffus

ion

of H

2O

(1)

H2O

Mas

s Tra

nsfe

r

(4) H

2O

Eva

pora

tion

H2,

H2O

O2,

N2

Mec

hani

sms:

1) W

ater

tran

spor

t fro

m fu

el g

as to

ano

de

2) W

ater

dra

g fr

om a

node

to c

atho

de

3) B

ack

diff

usio

n fr

om c

atho

de to

ano

de

4) W

ater

pro

duct

ion

and

rem

oval

at t

he c

atho

de

Inle

t Fue

lIn

let A

ir

2 H

2

4H

+ + 4

e-

H2

O2 +

4H

+ + 4

e-

2 H

2O

Asp

ects

of W

ater

Mgm

t1)

Ion

ic c

ondu

ctiv

ity d

irect

ly p

ropo

rtion

al to

wat

er c

onte

nt

2) L

ife /

dura

bilit

y of

mem

bran

e ch

alle

nged

by

cycl

ing

3) O

pera

ting

curr

ent i

nflu

ence

s wat

er c

onte

nt

Figu

re 3

-5 W

ater

Tra

nspo

rt in

a P

EM

FC

Page 70: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-16

Cat

alys

t (Pt

or

Ru)

- poi

sone

d by

CO

, S- E

xpen

sive

Car

bon

Flow

Fie

ld

Car

bon

Flow

Fie

ld

Mem

bran

e- h

igh

ioni

c co

nduc

tivity

- sen

sitiv

e to

wat

er c

onte

nt

Ele

ctro

de

- Ele

ctri

cally

Con

duct

ive

- Por

ous C

arbo

n Su

ppor

t- H

ydro

phob

ic (w

etpr

oof)

H2

/H

2O O2/N

2H

2OO

2/N2

H2O

O2/N

2H

2O

H2O

Coo

lant

Pas

sage

- ∆T

= 10

°C

H2O

Coo

lant

Pass

age

H2

/H

2OH

2 /

H2O

ME

A =

Mem

bran

e E

lect

rode

Ass

embl

y

Figu

re 3

-6 P

EM

Sta

ck C

ross

-Sec

tion

Page 71: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-17

REFERENCES 1 Barron’s, May 11, 1998. 2 K.B. Prater, “Polymer electrolyte fuel cells: a review of recent developments,” J. Power Sources, 51, 1994, p. 133. 3 K. Stasser, Proc. Grove Anniversary Fuel Cell Symposium (London), J. Power Sources, 29(1/2), 1989. 4 C. Seymour et al., Program and Abstracts 1990 Fuel Cell Seminar, Phoenix, AZ. 5 D. Bloomfield et al., Program and Abstracts 1990 Fuel Cell Seminar, Phoenix, AZ. 6 P. Patil et al., Symp. Proc. 10th Int. Electric Vehicle Symp. (Hong Kong), 1990, p. 657. 7 H. Creveling, Program and Abstracts 1990 Fuel Cell Seminar, Phoenix, AZ. 8 S. Srinivasan et al., Proc. 1st Int. Fuel Cell Workshop on Fuel Cell Technology Research and Development (Ohokayama, Tokyo), 1989, pp. 119-50. 9 S. Srinivasan, B. Dave, K. Murugesamoorthi, A. Parthasarathy, and A. Appleby, “Overview of Fuel Cell Technology,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993. 10 E. Barendrecht, “Electrochemistry of Fuel Cells,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 111. 11 S. Srinivasan, B. Dave, K. Murugesamoorthi, A. Parthasarathy, and A. Appleby, “Overview of Fuel Cell Technology,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p. 65. 12 Ibid. 13 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995. 14 K.B. Prater, “Polymer electrolyte fuel cells: a review of recent developments,” J. Power Sources, 51, 1994, p. 133. 15 D. Watkins, “Research, Development, and Demonstration of Solid Polymer Fuel Cell Systems,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p. 501. 16 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p 430. 17 K.B. Prater, “Polymer electrolyte fuel cells: a review of recent developments,” J. Power Sources, 51, 1994, p. 130. 18 D. Rastler et al., “State-of-the-Art Assessment of Polymer Electrolyte Membrane Fuel Cells for Distributed Power Applications,” prepared for the Electric Power Research Institute, Interim Report, EPRI TR-107064, November, 1996, p. 5-5. 19 D. Watkins, “Research, Development, and Demonstration of Solid Polymer Fuel Cell Systems,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p. 502. 20 Ibid. 21 Ibid. 22 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 90. 23 J. McClellen, ONSI Corporation, private communciation, October, 1998. 24 Ibid., p. 88. 25 D. Rastler et al., “State-of-the-Art Assessment of Polymer Electrolyte Membrane Fuel Cells for Distributed Power Applications,” prepared for the Electric Power Research Institute, Interim Report, EPRI TR-107064, November, 1996, p. 4-5. 26 D. Watkins, “Research, Development, and Demonstration of Solid Polymer Fuel Cell Systems,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p. 500. 27 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 73. 28 D. Watkins, “Research, Development, and Demonstration of Solid Polymer Fuel Cell Systems,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p. 509.

Page 72: Review of State-of-the-Art Fuel Cell Technologies for Distri

3-18

29 Ibid. 30 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 85. 31 Ibid. 32 K.B. Prater, “Polymer electrolyte fuel cells: a review of recent developments,” J. Power Sources, 51, 1994, p. 130. 33 Ibid. 34 S. Srinivasan, B. Dave, K. Murugesamoorthi, A. Parthasarathy, and A. Appleby, “Overview of Fuel Cell Technology,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p. 66. 35 Ibid. 36 D. Rastler et al., “State-of-the-Art Assessment of Polymer Electrolyte Membrane Fuel Cells for Distributed Power Applications,” prepared for the Electric Power Research Institute, Interim Report, EPRI TR-107064, November, 1996, p. 5-8. 37 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 78. 38 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 430. 39 Ibid., p. 434. 40 “The Automaker’s Big Time Bet on Fuel Cells,” Fortune, March, 1998. 41 D. Rastler et al., “State-of-the-Art Assessment of Polymer Electrolyte Membrane Fuel Cells for Distributed Power Applications,” prepared for the Electric Power Research Institute, Interim Report, EPRI TR-107064, November, 1996, p. 5-4. 42 Ibid. 43 D.Dunnison, “PEM Fuel Cell Power Plant Development at Ballard Power Systems,” Proc. 2nd Annual Distributed Resources Conference, EPRI TR-107585, November, 1996, p. 6-11. 44 Ibid., p. 6-9. 45 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996, 3-70. 46 D.Dunnison, “PEM Fuel Cell Power Plant Development at Ballard Power Systems,” Proc. 2nd Annual Distributed Resources Conference, EPRI TR-107585, November, 1996, p. 6-11. 47 New York Times, June 17, 1998. 48 Ibid. 49 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 338. 50 “The Automaker’s Big Time Bet on Fuel Cells,” Fortune, March, 1998. 51 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996. 52 W.L. Mitchell, “Development of an Integrated Reformer/Fuel Cell Power System for Petroleum and Alcohol Fueled Vehicles, Part I: Brassboard Demonstration,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p.232. 53 D. Rastler et al., “State-of-the-Art Assessment of Polymer Electrolyte Membrane Fuel Cells for Distributed Power Applications,” prepared for the Electric Power Research Institute, Interim Report, EPRI TR-107064, November, 1996. 54 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 338. 55 Adapted from, “State-of-the-Art Assessment of Polymer Membrane Fuel Cells for Distributed Power Applications,” EPRI TR-107064, November 1996. 56 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995.

Page 73: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-1

4. Molten Carbonate Fuel Cells (MCFCs) Molten carbonate fuel cells (MCFCs) are often considered a 2nd generation fuel cell as they are anticipated to be commercialized after phosphoric acid fuel cells. Like PAFCs, MCFCs are a liquid electrolyte-based fuel cell that makes use of flat, planar configured fuel cell stacks. However, at 650°C, the operating temperature is substantially higher and the molten carbonate electrolyte makes for a more corrosive environment than that encountered in the PAFC. The higher operating temperature enables internal reforming of hydrocarbon fuels, improving system design and efficiency. Additionally, the elevated operating temperature combined with fast electrode kinetics eliminates the need for expensive noble metal electrocatalysts and results in the highest electric efficiency of all fuel cell types. These same features also allow fuel flexibility, as carbon monoxide is not poisonous to the MCFC. These advantages are somewhat offset by the more severe material requirements for high temperature operation in the corrosive and oxidizing environments. Cell materials that demonstrate the necessary corrosion stability at reasonable costs have been primarily stainless steel alloys, ceramic composites, and semiconducting oxides. Long cell life (~40,000 hours) with the current state-of-the-art cell components represents a significant and ongoing technical challenge to the development of a commercial MCFC product.

Two U.S. companies are presently pursuing the technology, Energy Research Corporation (ERC) and MC-Power. ERC, in Danbury Connecticut, has been developing MCFC technology for 30 years and is the technical leader. MC-Power is a more recent collaboration between the Institute of Gas Technology (IGT), Bechtel Corporation, and Ishikawajima-Harima Heavy Industries. Both U.S. companies anticipate commercial products within the next 3 years.

ELECTROCHEMISTRY The cell components in molten carbonate fuel cells are relatively thick (~1mm) with respect to other fuel cell types. When the individual cell components are assembled, the thickness of a single cell is on the order of 1 cm. The electrolyte is typically made from a mixture of lithium and potassium carbonate salts (e.g., 62 mol% Li2CO3 – 38 mol% K2CO3) that has a thickness of about 0.5 mm1. A unique feature of the MCFCs is that the electrolyte is immobilized using a ceramic (lithium aluminate, LiAlO2) structure that also provides enough plasticity to act as a gas tight seal which can prevent the escape of gas to the ambient2. The anode is nearly all nickel doped with about 10 wt% chromium to inhibit mechanical creep. The nickel oxide cathode is made to be about 1 mm thick and is a chief target for the improvement of cell endurance. Finally, each cell is separated with a separator (or bipolar) plate often constructed from stainless steel on Incoloy sheets3. The cell stack structure is shown in Figure 4-1.

As Figure 4-1 shows, molten carbonate fuel cell stacks consist of repeating fuel cell units, each comprised of an anode, cathode, electrolyte matrix and a bipolar separator plate between cells. Reactant gases, which typically consist of desulphurized and reformed natural gas and ambient air, flow over the electrode faces (anode and cathode) in channels through the bipolar separator

Page 74: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-2

plates generating electrical and thermal energy by the simultaneous electrochemical oxidation of fuel and reduction of oxygen.

Interestingly, the mobile ionic species of the MCFC is the carbonate ion (CO32-). This ion is

formed in the cathode reactions from oxygen and carbon dioxide. The cell depicted in Figure 4-2 illustrates the electrochemical processes at work. The oxidant entering the cathode is an air-CO2 mixture. Oxygen, carbon dioxide, and electrons react at the cathode to form carbonate ions CO3

2-. These ions transport charge through the molten carbonate electrolyte to the fuel electrode (anode) where they react primarily with hydrogen. Hydrogen entering the anode channels is adsorbed into the anode and oxidized by reacting with the carbonate ions to form water and carbon dioxide, and releasing electrons at an electrochemical potential greater than that in the cathode4. These reactions are written as, Anode: H2 + CO3

2- H2O + CO2 + 2e- Cathode: ½ O2 + CO2 + 2e- CO3

2- Overall: H2 + ½ O2 H2O Water produced from the oxidation of hydrogen reacts with carbon monoxide within the cell to produce additional hydrogen for electrochemical use. This "water-gas shift reaction" (CO + H2O = H2 + CO2) is possible because the reaction rapidly reaches equilibrium at the high operating temperature of the fuel cell5. Since carbon dioxide is consumed at the cathode and produced at the anode, anode exhaust gas is typically sent to a combustor for "afterburn" then looped back to the cathode where CO2 is required.

CELL MATERIALS AND CONSTRUCTION In general, the materials employed in MCFCs have not changed in the past 25 years. The cell material requirements vary depending on the component, but all must withstand both high temperature operation and a corrosive environment. The MCFC utilizes ceramics, metals, and metal-alloys. The electrolyte structure, anode, and cathode are constructed from LiAlO2 and alkali carbonates (K2CO3 and Li2CO3), nickel, and nickel oxide, respectively. While optimal electrolyte composition is still being studied, the materials utilized have remained essentially unchanged for some time. The most significant MCFC advancements have been in fabrication processes. In the early 1980s, electrolyte structures were formed from hot pressing mixtures of LiAlO2 and alkali carbonates at 5000 psi and 440°C conditions6. This manufacturing method resulted in poor quality and high cost. The application of tape casting to electrolyte structure fabrication to form uniformly thin electrolytes in a repeatable manner has been the most visible of the manufacturing process advancements. Table 4-1 lists the history and current state-of-the-art materials employed in MCFCs. A brief discussion of each of the cell components follows.

Page 75: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-3

O2, CO2

Figure 1 Fuel Cell Stack Components

Figure 2 Chemical Reacting System within an MCFC

3

C O

2

H2 + CO3= H2O + CO2 + 2e-

(Nat. Gas, coal gas, or other)

(Air, CO2)

Figure 4-1 Fuel Cell Stack Components

Figure 4-2 Chemical Reacting System Within an MCFC

Page 76: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-4

Table 4-1 History of State-of-the-art Component Materials in an MCFC1

Electrolyte / Electrolyte Structure

The electrolyte must exhibit good ionic conductivity (for transport of CO32- ions) and resistance

to gas crossover. The ionically conductive electrolyte material is based on a mixture of lithium and potassium carbonates that are in liquid phase above about 500°C. The electrolyte composition is often the eutectic composition of 62 mol% Li2CO3 – 38 mol% K2CO3

7. A unique feature of MCFCs is the use of an electrolyte matrix to immobilize the liquid carbonate. The matrix is composed of a mixture of ceramic powder (lithium aluminate, LiAlO2) and carbonate electrolyte that forms a semisolid structure. The fabrication of the electrolyte structure is accomplished by tape-casting, a process commonly used in ceramic and microelectronic technology. The tape-casting process involves dispersing the ceramic (LiAlO2) or metal (Ni)

1 Adapted from Selman, 1993 and 1998 Fuel Cell Handbook, DOE/FETC-99/1076.

Page 77: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-5

powder in a solvent containing organic binders, plasticizers, and additives. The resultant mix, termed a ”slip”, is cast over a smooth, moving substrate in a thin film, whose thickness is maintained by a stationary, knife edge device (“doctor blade”)8. The slip is dried, becoming a semi-stiff structure which is assembled into the fuel cell. The organic binder is burned out by heating to 250-300°C and the carbonate is melted and subsequently absorbed into the ceramic structure during cell start-up.

In all fuel cells, the three phase interface of electrolyte/solid electrode/reactant gas is critical for cell performance, i.e., the interface is crucial to enable the electrochemical cell reactions at the electrodes to occur at a sufficient rate and with maximum surface area. Restated, in order for electrode reactions to occur, the liquid electrolyte, solid electrode, and reactant gas must all be in physical contact with one another. The greater the contact area, the greater the conduction pathways, and the higher the current density. Thus the requirement of substantial and intimate physical contact area of all three phases, necessarily involves impregnation of the molten carbonate into the porous electrode structure. The methods employed to control electrolyte distribution in cell components of phosphoric acid and alkaline fuel cells are not available to the MCFC. The approach chosen to control the electrolyte distribution for MCFCs is that of capillary equilibrium. This method, depicted in Figure 4-3, relies on a balance in capillary pressures to establish the relative amounts of electrolyte in each component9. By properly coordinating the pore diameters in the electrodes with those of the electrolyte matrix (which contains the smallest pores), the electrolyte distribution is determined10. The pores sizes of the electrolyte structure are below 1 µm, the anode pore sizes typically range between 3- 6 µm, and the cathode pores are between 7 –15 µm. In this arrangement, the electrodes are only partially filled enabling reactant gas access, and the electrolyte matrix is entirely filled, providing ion conduction and a barrier to gas crossover.

The control over the optimum distribution of molten carbonate electrolyte in the different cell components is termed “electrolyte management,” and is critical for attaining sustained high performance and long life11. Numerous mechanisms affect the distribution of electrolyte in the cell components, including consumption by corrosive reactions, potential driven migration, and creepage of salt and salt vaporization. In general, no matter how the electrolyte is managed, MCFCs experience a gradual, but sustained loss of electrolyte which results in performance decay12. Thus, electrolyte management is a key aspect to achieving long-term endurance performance and is a dominant research area.

Anode

As Table 4-1 shows, the anode is made from a porous sintered nickel with small percentages of chromium or aluminum. The chromium (or aluminum) additive is oxidized in situ and is lithiated by the electrolyte to form LiCrO2 particles on the nickel surface which form the sinter13. (Recall that “sintering” is a process whereby synthesized powder particles are heated at increasing temperatures until bonds between particles are formed and diffusion of chemical species into the powder microstructure takes place.) The presence of oxides is to reduce further sintering of nickel particles and stabilize the sinter against mechanical creep (i.e., deformation under compressive loads) which can decrease anode porosity, increase contact resistance, and increase risk of gas crossover14. The anode porosity ranges between 45-70% (by volume) with median pore sizes of about 5 µm, and a thickness of 1mm. The anode is typically filled to 50-

Page 78: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-6

60% of its pore volume with electrolyte15. The present cost of nickel anodes is relatively high, and ongoing research is investigating whether other materials (e.g., copper) can be substituted yet maintain the stable, non-sintering, creep resistance characteristics of nickel16.

Figure 4-3 Capillary Equilibrium in a MCFC2

While the nickel-based anode adequately meets the majority of the performance requirements, there is a need to provide better resistance to fuel gases containing sulfur17. Sulfur poisons the electrode by chemisorption on the nickel surfaces, blocking of the electrochemical reaction sites which also poisons the reaction sites for the water-gas shift reaction18. Alternative materials, such as ceramic anodes (LiFeO2), are being tested.

Cathode Cathode material functionality requires high electrical conductivity, structural strength to withstand large, mechanical compressive loads, and a low dissolution rate in the acidic molten carbonate electrolyte. The present cathode material is made of lithiated nickel oxide (LiNiO). As Table 4-1 shows, the electrode is about 60% porous and 0.5-1mm in thickness. The average pore sizes range between 5-7µm. The cathode is typically filled to 20-25% of its pore volume with the molten carbonate electrolyte, extending the reaction surface area19. While the NiO is relatively stable in the severe oxidizing environment of the cathode compartment, the nickel dissolves in the electrolyte, migrates towards the anode, precipitates, and eventually shorts out the cell. The dissolution of the cathode occurs according to the following mechanism:

2 1998 Fuel Cell Handbook, DOE/FETC-99/1076.

H2 + CO3= H2O + CO2 + 2e- ½ O2 + CO2 + 2e- CO3

=

Pores

Page 79: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-7

NiO + CO2 Ni2+ + CO32-

The mechanism involved in the rate of dissolution is complicated as it demonstrates dependence on a variety of variables, such as electrolyte composition, CO2 partial pressure, H2O partial pressure, and temperature20. The effect of CO2 partial pressure and electrolyte composition are, perhaps, more significant. In general, the loss of cathode material is often considered acceptable. For example, developers at Energy Research Corporation have reported 30-40% loss in NiO material (or only 10% of the cathode thickness) over 40,000 hours of operation21. However, the loss of active material is not the major concern, but the dissolving of nickel followed by the precipitation of nickel particles in the electrolyte matrix and anode-to-cathode short circuiting22. This has been the primary life limiting constraint on MCFCs, particularly in pressurized operation where CO2 partial pressures increase, resulting in increased dissolution rates.

Stack Construction The thin cells are assembled one on top of the other to form a stack, as Figure 4-1 illustrates. The actual stack structures employed by developers are more complicated than depicted in Figure 4-1. Depending on design philosophy, they may also consist of perforated anode and cathode current collectors, extended length electrolyte tiles (matrix) which butt up against the gas manifold, forming a wet seal between anode and cathode gases, and cell separators for internal manifolding designs. However, in stacking of the thin cell components the stack still maintains many of the two dimensional characteristics of single cells23. The most salient variance in design features of cell stacks is the manner in which reactant gases are fed to the cells. The two stack designs presently being developed are external and internal manifolding. These are schematically illustrated in Figure 4-4.

In external manifolding, the oxidant and fuel gases are fed to separate plenums where they admitted to the cell inlets. The fuel and oxidant gases flow perpendicular to each other in a “cross-flow” configuration, and exit on the opposite side of the stack. External manifolding has traditionally been the simplest design to employ. Its simple design allows large manifold cross-sections with a minimum amount of material and has low projected manufacturing costs24. An additional advantage of external manifolding is that it enables internal reforming of the fuel gas, increasing system efficiency. The disadvantages to external manifolding are related to the manifold gasket which demonstrates gas leakage and facilitates electrolyte migration (“ion pumping”) from the positive to the negative end of the stack25. Internal manifolding was developed to specifically address these issues. Energy Research Corporation (ERC), in Danbury, CT is the leader and sole U.S. developer of externally manifolded MCFC designs.

In internal manifolding the reactant gases are fed via pipe connections to the cell stack. When properly aligned, the holes in each of the numerous stacked separator plates serve as an internal gas pathway (or duct), distributing reactants to each cell. The method of internal manifolding gives flexibility in reactant gas flow patterns, allowing cross-flow, co-flow, or counter-flow configurations. Each flow pattern has its advantages, but it appears that most developers have chosen either cross-flow or coflow operation (coflow has the most uniform temperature distribution)26. Although internal manifolding prohibits internal reforming, strategic placement of the external reformer directly next to the cell stack enables utilization of the exothermic heat generated from the cell reactions. Additionally, use of an external reformer isolates the

Page 80: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-8

reforming operation from cell stack operations, and prevents the exposure of reforming catalyst to electrolyte attack27. While internal manifolding does manage to hurdle the problems of gas leakage and electrolyte migration by the elimination of manifold gasketing, its disadvantages lie in manufacturing costs and operation. The cell structure in internal manifolding is more complicated than external manifolding by virtue of the addition of a cell separator plate which functions as both flow guide, separating fuel gas on one side from oxidant on its other side, and current collector. In addition to the high manufacturing costs of this plate, the exposure to both fuel and oxidant gases make corrosion of the plate a concern. MC-Power in Burr Ridge, IL is the sole U.S. manufacturer pursuing this technology.

Figure 4-4 Internal and External Manifolded Fuel Cell Stacks

OPERATION AND PERFORMANCE Molten carbonate fuel cells typically operate at about 650°C. At this temperature, the electrode kinetics are fast enough to enable the use of lower cost catalyst materials and reduce activation polarizations. The cathode reactions are generally rate limiting28, and cathode polarizations (activation, concentration, etc.) account for about 40% of the voltage loss in MCFCs29. Ohmic (resistive) losses represent only 1/6 of the total voltage loss. The fast electrode kinetics (arising

Page 81: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-9

from elevated operating temperature) and the carbonate electrolyte facilitate the use carbonaceous fuels, such as carbon monoxide. Carbon monoxide can be oxidized directly at the anode, according to

CO + CO3= 2 CO2 + 2e- (Anode Reaction involving CO)

but this reaction is not thermodynamically favorable. Instead, the carbon monoxide participates in the mildly exothermic water-gas shift reaction,

CO + H2O CO2 + H2 (Water-gas shift)

consuming one mole of water and producing one mole of hydrogen which can be utilized in the anode oxidation reaction. The water gas shift reaction achieves rapid equilibrium at the anode and serves as an indirect source of hydrogen30. The high operating temperature (650°C) and the heat released from the exothermic electrode reactions are sufficient to enable the use of internal reforming. The reforming of methane is most common and is accomplished via the endothermic steam-methane reforming reaction,

CH4 + H2O CO + 3 H2 (Reforming Reaction)

This reaction occurs simultaneously with the electrochemical oxidation of hydrogen at the anode. In internal reforming MCFCs, the heat for the endothermic reforming reaction is supplied by the heat liberated from fuel cell reactions, eliminating the need for an external fuel reformer and simplifying the process design. Two types of internal reforming have been developed: direct internal reforming (DIR) and indirect internal reforming (IIR). Figure 4-5 depicts the conceptual differences. IIR stacks have separate reformer plates at regular intervals in the stack, typically one for every 5 or 6 cells31. In DIR stacks, reforming and anodic reactions occur in each cell, although the reforming catalyst and its support are spaced away from the anode so that they are not wetted by the corrosive molten carbonate32.

High temperature and low pressure favor the forward direction of the reforming reaction, and therefore, internal reforming MCFCs are likely to operate at atmospheric pressure. The reforming reactions are sustained by a supported nickel catalyst which provides adequate catalytic activity to meet the H2 needs of the fuel cell33. The relationship between the methane converted to hydrogen (production) and fuel utilization (hydrogen consumption) is depicted in Figure 4-6. At open circuit conditions, about 83% of the methane is converted to hydrogen. As current is increased (and hence, fuel utilization), the amount of methane converted increases, reaching a maximum of about 98.5% for utilizations higher than 55%. Internal reforming has been successfully demonstrated in fuel cell stacks and is expected to be employed in ERC’s commercial units34.

Page 82: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-10

Figure 4-5 Internal Reforming Configurations35

Figure 4-6 Methane Conversion in an Internal Reforming MCFC36

Page 83: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-11

Reactant Utilization Fuel utilization is defined as the mass rate of hydrogen consumption over the mass flow of hydrogen supplied to the cell. In order to maintain a positive partial pressure in anode and cathode compartments (and therefore an electrochemical driving force), not all the fuel and oxidant may be utilized in a cell. In general, reactant utilization and current density have similar impacts on cell voltage; the lower the reactant utilization, the higher the cell (or stack) voltage. However, operating at reduced reactant utilization means lower system efficiency due to the inefficient fuel use. Fuel utilization between 75-85% and oxidant utilization of 50% is typically the compromise made in operating point selection to maximize overall performance.

The residual fuel exiting the anode compartment is often combusted in an afterburner that requires a catalyst for oxidation of the low-Btu gas. Additionally, as CO2 is produced at the anode and is required at the cathode, the exhaust is recirculated from anode exit to cathode inlet. Figure 4-7 shows a simple process flow schematic for an MCFC system.

Figure 4-7 Simple Process Flowsheet for a MCFC Power Plant37

Page 84: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-12

Fuels & Poisons The molten carbonate fuel cell is capable of operating on several fuel types. To date, MCFCs have operated on reformed or simulated natural gas and simulated coal gas. It is expected that MCFCs will also be capable of operating on ethanol, landfill gas, and military logistic fuels (JP-8 and diesel). Given the expected markets of MCFCs, early commercial units will be operated on natural gas and gasified coal is eventually expected to be the major source of fuel for MCFCs in the 21st Century38.

Contaminant studies have identified H2S, HCl, H2Se, and As as four species that significantly degrade carbonate fuel cell performance39. While the tolerance to each contaminant may depend on the cell operating conditions (temperature and pressure), inlet gas composition, component materials, and system operation (e.g., anode gas recycle, venting, gas cleanup), the acceptable tolerance levels for all types is presently judged to be less than 1 ppm (parts per million)40.

The adverse effects of H2S have been the most extensively studied and have been found to occur from41,42,43,44:

• Chemisorption on Ni surfaces to block active electrochemical sites

• Poisoning of catalytic reaction sites for the water gas shift reaction

• Oxidation to SO2 in a combustion reaction, and subsequent reaction with carbonate ions in the electrolyte.

Poisoning of the catalyst sites for the water gas shift appears to be minor. If the anode gas is recirculated via a combustion chamber, then any contaminants delivered to the anode compartment are carried to the cathode side of the cell. In this circumstance, any SO2 present reacts with carbonate ions to produce alkali sulfates which migrate through the electrolyte to the anode where the concentration of sulfur is increased45. It appears that sulfur concentrations of less than 0.5 ppm will be necessary for sustained long-term performance.

Voltage-Current Performance The cell performance (i.e., voltage-current performance) of MCFCs has improved substantially since 1980. Early designs could demonstrate high voltage but only at low current densities. While MCFC performance advancements have not improved by the orders of magnitude such as in PEMs, they have increased current density by about one order of magnitude since 1967. Figure 4-8 depicts the progress of MCFCs over about 25 years.

Page 85: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-13

Figure 4-8 MCFC Cell Voltage Performance Advancements46

The voltage-current characteristic for MCFCs, as with other fuel cells, demonstrates dependency on such operating parameters as cell temperature, pressure, reactant (fuel and oxidant) utilization, and reactant gas composition. Like the PAFC, the factors involved in choosing the operating point for MCFCs include stack size, heat transfer rate, voltage level, load requirements, and cost47. Typical cell operating parameters are 650°C cell temperature, atmospheric pressure, and 75-80% fuel utilization. State-of-the-art average cell voltage ranges between 0.75 – 0.85 V with a current density of near 150-200 mA/cm2. As previously stated, ohmic resistance has a large influence on the operating voltage of the fuel cell. ERC has determined that the electrolyte matrix accounts for 70% of the total ohmic loss of the cell48. Producing thinner electrolyte matrices would reduce the ohmic losses, but thickness is limited by gas crossover and cathode (nickel oxide) dissolution considerations. For example, it has been observed that an increase in electrolyte thickness has decreased the rate of NiO dissolution in carbonate (effectively increasing cell life) by increasing the Ni2+ diffusion path length and thus lowering the transport rate. The importance of cell life is a dominant consideration.

As with other fuel cell types, temperature elevation improves the voltage performance in spite of theoretical considerations. While theoretical voltage is reduced with temperature increase, faster electrode kinetics are achieved by reducing the activation polarization and producing a net

Page 86: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-14

voltage gain. In MCFCs, the average temperature is about 650°C. In the temperature range between 575° to 650°C, about 1/3 of the voltage improvement is attributable to ohmic polarization reduction and the remainder to anode and cathode polarization reductions49. Operation above 650°C produces further gains in voltage performance at the expense of cell life. Temperature related effects on cell life include increase in electrolyte loss from evaporation and material corrosion.

Increases in operating pressure will improve cell voltages due to increases in the partial pressure of the reactants, gas solubilities, and mass transport rates. Increase in mass transport enables higher current densities which are imperative for stack cost reduction efforts (i.e., more power per unit volume means using less material). Studies by Japanese researchers have shown that the performance gain with an increase in pressure is largest in the 1-3 atm range, mostly due to decrease in cathode polarization50. However, increase in operating pressure will also result in increased nickel oxide dissolution (CO2 partial pressure increase) and undesirable side reactions, such as carbon deposition from carbon monoxide and methane decomposition, and methane formation (methanation). ERC is developing atmospheric operating MCFCs and MC-Power is pursuing 1-3 atmosphere operation.

Efficiencies MCFCs are the most efficient of all fuel cell types. Efficiencies are expected to be near 50% -60% (LHV) and have been verified up to about 44%. The efficiency demonstrated in the 1.8 MWe MCFC demonstration plant in Santa Clara, California was 43% and given the conservative operating mode, it was judged that operational plant target of 49% could easily be achieved51. In general, thermal integration of high temperature equipment such as through the use of internal reforming enables higher operating efficiencies.

Thermal Management Like SOFCs, MCFCs are maintained at 650°C through the use of air-cooling accomplished by operating with high excess air amounts. In the Santa Clara Demonstration Project, ERC’s carbonate plant operated with an oxidant mass flow of over 10 times that of the fuel52. Water cooling presents the opportunity for large thermal gradients and is not employed. The reactant gas flow configuration has some effect on the temperature distribution within MCFCs. As previously stated, the nature of cross-flow configuration result in more non-uniform temperature distributions than other flow patterns. This is a result of the reactant consumption characteristics and subsequent current distribution over the planar area of the cell. Additionally, cell size impacts how heat generated within the stack is managed. The larger the cell, the more difficult it is to ensure that the heat generated is convected away to prevent cell hot spots and excessive temperature rise within the components53. ERC has performed thermal cycling tests on a 10 kW stack and indicated no deleterious effects after 10 cycles54.

Most significant on temperature control is the effect of internal reforming. When employed within a cell stack, the heat from the cell reactions may be transferred directly to the highly endothermic internal reforming reaction. Thus, the reforming reaction acts as a heat sink which can reduce the cooling load by as much as 50%55. Various internal reforming configurations have been tested by ERC and it appears that the IIR-DIR method (see Figure 4-5) results in the most uniform temperature distribution over the planar cell area56.

Page 87: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-15

Life Since its inception in about 1950, MCFC development has continuously battled the corrosive effects of carbonate salts. Today, it remains a technical challenge. The most significant life limiting sources have been related to cathode dissolution (nickel oxide) and electrolyte evaporation. Current life estimates by the various developers agree that 40,000 hours is feasible, however, endurance testing of full-scale stacks remains to be performed to resolve life issues. To obtain a 40,000 hour life, a 2 mV/1000hr degradation is estimated to be necessary. Present performance is closer to -5mV/1000 hr57. European MCFC developers have been more open about their expectations. One developer using a licensed IMHEX design (internally manifolded heat exchange design developed by the Institute of Gas Technology) judges that the present lifetime is about 25,000 hours58.

MANUFACTURERS & MARKETS There are currently two U.S. companies pursuing molten carbonate fuel cells, Energy Research Corporation (ERC) based in Danbury, Connecticut, and MC-Power, in Burr Ridge , Illinois. International Fuel Cells Corporation (IFC) also has been active in MCFC development, but has refocused company resources on both PEM and PAFC development. Future MCFC development by IFC is uncertain. ERC has been developing carbonate fuel cell technology since 1969 and is considered the technology leader. They are currently pursuing externally manifolded, atmospheric pressure operating carbonate power plants in the 300 kW – 3 MW range. ERC has demonstrated MCFC technology at large scales with the construction and test of a 1.8 MW power plant (the largest fuel cell power plant in North America) in Santa Clara, California. The $46 million dollar project was funded by a consortium of gas and electric utilities, the Department of Energy, the Electric Power Research Institute, the California Energy Commission, and ERC. Successful operation occurred for about 4,000 hours before a stack short, due to use of a conducting insulation material in the stack modules, forced a premature shutdown. ERC has chosen the Indirect Internal Reforming (IIR)-Direct Internal Reforming (DIR) combination as its baseline MCFC design. While ERC will focus on their 2.5 MW product for market entry, they also plan on offering 300 kW and multi-MW scale skid-packaged power plants. Figure 4-9 depicts a rendering of the proposed commercial 2.5 MW power plant. All plants will operate off of natural gas with diesel fuel processing options. Once experience has been obtained with the early units, ERC plans on producing larger power plants operating on natural gas and/or coal gas59.

MC-Power has been developing carbonate fuel cells since 1987 and is slightly behind ERC with respect to the expected commercial product release timeframe. They originally licensed an internally manifolded MCFC design from IGT and are now partnered with several firms to commercialize a carbonate fuel cell product by 2002. Recently, MC-Power completed demonstration testing of a 250 kW power plant at Miramar Naval Air Station in California. Further demonstration tests for stack sizes of 75kW and 175kW are planned for late-1999 through 2001 before commercial product release in 200260. MC-Power plans to enter the power market with 250 kW and 1 MW units61. Figure 4-10 shows a layout of a proposed 500 kW commercial plant. Table 4-2 summarizes the manufacturers, anticipated product sizes, and market segmentation.

Page 88: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-16

Figu

re 4

-9 E

RC

Com

mer

cial

2.5

MW

e M

CFC

Pow

er P

lant

Page 89: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-17

Figu

re 4

-10

M

C-P

ower

500

kW

e C

omm

erci

al M

CFC

Pow

er

Plt

Page 90: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-18

Table 4-2 North American Manufacturers and Product Entry

Manufacturer Technology Product Size Markets

Energy Research Corporation (ERC)

External Manifolded (1 atm operation)

300 kW – 2.5 MW Commercial Light Industrial

Distributed power Niche

Transportation (Marine)

MC-Power Internally Manifolded

(1 –3 atm operation)

250 – 1 MW Commercial Light Industrial

Distributed power Niche

Examples of the commercial market segment includes hotels, schools, small to medium sized hospitals, and shopping malls. It is anticipated that MCFC unit will operate in a cogeneration mode, supplying heat and electricity for cooling and lighting loads. The light industrial market includes the chemical, paper, metal, food, and plastic industries. The traditional utilities, unregulated subsidiaries, municipal utilities, and new, energy service providers are considered customers for the distributed power segment. Special niche market applications are computer centers and other customers who require premium power quality and high reliability. Additionally, producers of renewable or “opportunity” fuels such as landfills, waste water treatment plants, and refineries62.

In order to penetrate these markets, however, commercialization barriers, such as high volume productions (on the order of 200-400 MW/yr) to drive down stack manufacturing costs to $200-$400/kW and plant footprint will have to be overcome. If cost targets are achieved, the potential markets are large enough to sustain and continue MCFC development and commercialization. According to an EPRI study, if current industrial electric rates hold and ERC and MC-Power attain their cost targets, their MCFC cogeneration products will be competitive with about 20% of the current industrial market rates63.

Foreign MCFC developers include Hitachi Ltd., Ishikawajima-Harima Heavy Industries (IHI), and Mitsubishi Electric Co. in Japan and ECN in the Netherlands.

STATUS OF DEVELOPMENT AND COMMERCIALIZATION Molten carbonate fuel cells are poised to enter the commercial market as early as 2000. ERC leads the commercialization effort world-wide, but not substantially. Several demonstration plants are planned by both ERC and MC-Power for the remainder of this year and into the next. U.S. manufacturers are focusing development activities on improvement of component performance, endurance tests, cost reduction, and high volume manufacturing methods. To

Page 91: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-19

obtain competitive capital costs ($1000/kW), a manufacturing volume of 400 MW/yr will be required64. ERC currently has a 17 MW/yr manufacturing capability which could be expanded to 50 MW/yr. MC-Power has a 2MW/yr capability that could be expanded to 15 MW/yr65.

Survey of the open literature reveals the following development issues:

1.) High power density (and small plant footprint) is a must for commercialization. Recent stack designs have registered power densities of about 0.125 W/cm2 (0.75 V @ 165 mA/cm2) on natural gas. This compares unfavorably with other fuel cell types. Additionally, increasing power density can help achieve stack cost reduction goals. To reduce cost and plant footprint developers have set power density goals of 0.18 to 0.225 W/cm2 (66).

2.) Life. NiO (cathode) dissolution in the electrolyte, electrolyte management, and hardware corrosion protection are the three major factors in establishing long life characteristics in MCFCs. Alternative cathode materials, such as LiCoO2 or LiFeO2 are being investigated to improve the current NiO design. Additionally, optimizing the electrolyte composition, for example by adding lithium sodium instead of lithium potassium, to the melt could reduce NiO solubility in the carbonate. Corrosion protection is currently insufficient and costly, especially with regard to the reforming catalyst and bipolar and separator plates. If internal reforming is used, the catalyst (nickel on magnesium oxide or lithium aluminum oxide) may degrade over a few thousand hours67. The bipolar plate currently uses stainless steel cladded with nickel. The plate is costly, representing over 50% of the material costs68.

3.) Cost Reduction. Cost reduction is a major issue, especially surrounding the balance of plant which ERC estimates at about 70% of the total power plant cost69. ERC indicates that materials cost represent about 65% of the stack cost and have targeted increasing this number to 80%.

4.) Systems Integration and Thermal Management. A fair amount of work is being performed in this area and is likely to continue in assisting plant and stack design. However, published application studies are few and far between.

5.) Reliability and durability of stacks.

Table 4-3 summarizes the status of MCFC technology.

Page 92: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-20

Table 4-3 Status Summary of MCFC Technology

Company ERC MC-Power

Development Targets

Current Projected Current Projected

Manifold Design External Manifold Internal Manifold

Operating Temperature

650°C 650°C 650°C 650°C

Operating Pressure 1 atm 1 atm 1 atm 1 - 3 atm

Reforming Direct Internal (DIR)

Combination (IIR-DIR)

External External

Power Density 0.125 W/cm2 ~0.2 W/cm2 0.125 W/cm2 ~0.2 W/cm2

Efficiency (HHV) 47% 58% 54%

Life (hours) ~15,000* 40,000 ~17,000* 40,000

Size 1.8 MW 300 kW – 2.5 MW 250 kW 250 kW – 1 MW

Footprint -- 4500 ft2 (2.5 MW) -- 4,000 ft2 (1MW)

Cost N/A

$1650/kW (early units)

$1250–1300/kW (mature)

N/A

$1750/kW (early units)

$1400 – 1550/kW (mature)

Timeframe -- 2001 -- 2002

* Demonstrated Life. Present Life expectancies are extrapolated from short-stack tests of 3-4,000 hours.

Page 93: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-21

CONCLUSIONS Molten carbonate fuel cells still have technical and economic issues but are advancing rapidly and are poised to enter the commercial market as early as 2001. The high operating temperature offers the highest efficiency of all fuel cell types and allows the use of internal reforming while producing sufficiently high grade waste heat for utilization in a bottoming cycle or cogeneration. In a bottoming cycle or cogeneration configuration, Braun (70) and Richter (71) have independently shown system efficiencies of 65-70% (HHV) are possible. The molten carbonate electrolyte does present some significant technical challenges to life performance, due to the solubility of the present cathode material and the corrosive nature of the molten salts which evaporate and migrate out of the immobilizing matrix, coating cell components. Also, the high balance-of-plant costs represent the most significant fraction of the power plant capital cost. Increases in power density and reduced component material and fabrication costs are expected to alleviate BOP costs. While costs are relatively high, MCFCs offer an advantage over other types in that their end-of-life metal components have high salvage value enabling recovery of most of the stack material cost72. In contrast, PAFC and PEM stacks are largely carbon-based materials with high initial costs, but little or no salvage value. If cost targets are obtained, EPRI estimates MCFCs will be competitive with about 20% of the industrial cogeneration markets (using constant 1996 electric rates)73. Product sizes will range from 250 kW – 2.5 MW and the markets targeted are commercial, light industrial, distributed power, and special niche markets. The commercial segment is expected to include hospitals, schools, and hotels. However, plant footprint will have to be reduced to be attractive with other fuel cell types in these applications and avoid siting restrictions. MCFCs are expected to find their largest market in industrial cogeneration applications, where the available high-grade heat can be effectively capitalized on.

Page 94: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-22

REFERENCES 1 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.4-5. 2 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 349. 3 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.4-7. 4 K.R. Williams, An Introduction to Fuel Cells, Elsevier Publishing Company, New York, 1966. 5 K. Kinoshita, F.R. McLarnon, E.J. Cairns, Fuel Cells, A Handbook, prepared by Lawrence Berkeley Laboratory for the U.S. Department of Energy under Contract DE-AC03-76F00098, May 1988. 6 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 368. 7 Ibid, p. 368. 8 Ibid, p. 369. 9 Ibid, p. 352. 10 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.4-2. 11 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 354. 12 Ibid, p. 354. 13 Ibid, p. 374. 14 Ibid, p. 374. 15 Ibid, p. 374. 16 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.4-9. 17 Ibid, p. 4-9. 18 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 389. 19 Ibid, p. 385. 20 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.4-5. 21 M. Farooque, ERC, "Development on Internal Reforming Carbonate Fuel Cell Technology, Final Report," prepared for U.S. DOE/METC, DOE/MC/23274-2941, pp. 3-18, October, 1990. 22 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 385. 23 Ibid, p. 398. 24 Ibid, p. 399. 25 Ibid, p. 399. 26 Ibid, p. 402. 27 Ibid, pp. 414-420. 28 E. Barendrecht, “Electrochemistry of Fuel Cells,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 101. 29 “Commercialization of Fuel Cells,” S.S. Penner, Editor, A report of the U.S. Department of Energy Advanced Fuel-cell Commercialization Working Group, Energy-The International Journal, Pergamon Press, New York, 1995, p. 101. 30 K. Kinoshita, F.R. McLarnon, E.J. Cairns, Fuel Cells, A Handbook, prepared by Lawrence Berkeley Laboratory for the U.S. Department of Energy under Contract DE-AC03-76F00098, May 1988.

Page 95: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-23

31 T. Tanaka, et al., Proc. 3rd Molten Carboante Fuel Cell Symposium, PV93-3, (The Electrochemical Society, Pennington, NJ, 1993), p. 37. 32 J. Ohtsuki, et al., Proc. 3rd Molten Carboante Fuel Cell Symposium, PV93-3, (The Electrochemical Society, Pennington, NJ, 1993), p. 48. 33 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.4-28. 34 H. Maru, et al., “ERC Direct Carbonate Fuel Cell Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054. 35 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 419. 36 J.H. Hirschenhofer, D.B. Stauffer, and R.R. Engleman, Fuel Cell Handbook, 3rd Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, 1994. 37 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998. 38 H. Maru, et al., “ERC Direct Carbonate Fuel Cell Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054. 39 A. Pigeaud, ERC, and G. Wilemski, Physical Sciences, "Effects of Coal-Derived Trace Species on the Performance of Carbonate Fuel Cells," in Proc. of the Fourth Annual Fuel Cells Contractors Review Meeting, U.S. DOE/METC, pp. 42-45, July 1992. 40 1994 Fuel Cell Seminar Abstracts, pp. 539-542. 41 W.V. Vogel and S.W. Smith, J. Electrochem. Soc., 129, 1441, 1982. 42 S.W. Smith, H.R. Kunz, W.M. Vogel and S.J. Szymanski, in Proc. of the Symposium on Molten Carbonate Fuel Cell Technology, edited by R.J. Selman and T.D. Claar, The Electrochemical Society, Inc., Pennington, NJ, p. 246, 1984. 43 R.J. Remick, E.H. Camara, paper presented at the Fall Meeting for The Electrochemical Society, Inc., New Orleans, LA, October 7-12, 1984. 44 R.J. Remick, in Proceedings of the Fourth Annual Contractors Meeting on Contaminant Control in Hot Coal-Derived Gas Streams, DOE/METC-85/3, edited by K. E. Markel, U.S. Department of Energy, Morgantown, WV, p. 440, May 1984. 45 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 4-28. 46 Ibid, p. 4-4. 47 Ibid, p. 4-11. 48 A. Pigeaud, ERC, and G. Wilemski, Physical Sciences, "Effects of Coal-Derived Trace Species on the Performance of Carbonate Fuel Cells," in Proc. of the Fourth Annual Fuel Cells Contractors Review Meeting, U.S. DOE/METC, pp. 42-45, July 1992. 49 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 4-17. 50 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993. 51 M. Farooque, A. Kush, A. Leo, H. Maru, and A. Skok, “Direct Fuel Cell Development and Demonstration Activities at Energy Research Corporation,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, November, p. 15. 52 E. Gillis, “Molten Carbonate Fuel Cell Technology,” EPRI Journal, April/May 1994, pp. 34-36. 53 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 398. 54 H. Maru, et al., “ERC Direct Carbonate Fuel Cell Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054.

Page 96: Review of State-of-the-Art Fuel Cell Technologies for Distri

4-24

55 J.R. Selman, “Research, Development, and Demonstration of Molten Carbonate Fuel Cell Systems,” in Fuel Cell Systems, Edited by L.J. M. Blomen, M.N. Mugerwa, Plenum Press, New York, NY, 1993, p. 412. 56 H. Maru, et al., “ERC Direct Carbonate Fuel Cell Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054. 57 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 4-29. 58 Dr. Peter Kortbeek, BCN Dutch Fuel Cell Corporation, private communication, November, 1998. 59 H. Maru, et al., “ERC Direct Carbonate Fuel Cell Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054. 60 R.O. Petkus, “MC-Power’s MCFC Generator Verification Test Program,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, November. 61 H. Maru, et al., “ERC Direct Carbonate Fuel Cell Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054. 62 E. H. Camara, “M-C Power Commercialization Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054. 63 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996. 64 Ibid. 65 H. Maru, et al., “ERC Direct Carbonate Fuel Cell Program Overview,” Proc. Of the Fuel Cells ’97 Review Meeting, DOE/FETC-98/1054. 66 “Commercialization of Fuel Cells,” S.S. Penner, Editor, A report of the U.S. Department of Energy Advanced Fuel-cell Commercialization Working Group, Energy-The International Journal, Pergamon Press, New York, 1995, p. 399. 67 “Technical and Economic Assessment of Molten Carbonate Fuel Cell Manufacturing Costs,” prepared for the Electric Power Research Institute,EPRI TR-101525, Project 1677-22, Final Report, October, 1992. 68 Ibid. 69 “Commercialization of Fuel Cells,” S.S. Penner, Editor, A report of the U.S. Department of Energy Advanced Fuel-cell Commercialization Working Group, Energy-The International Journal, Pergamon Press, New York, 1995, p. 413. 70 R.J. Braun, et al., “Improvements of a Molten Carbonate Fuel Cell Power Plant via Exergy Analysis,” Proc. Of the Symposium on Second Law Analysis, Costing, and Design Optimization, Advanced Energy Systems Div., ASME Congress & Expo, Atlanta, 1996. 71 K.V. Lobachyov, and H.L. Richter, “Addition of Highly Efficient Bottoming Cycles for the Nth Generation Molten Carbonate Fuel Cell Power Plant,” J. Energy and Resources Tech., 1997. 72 “Technical and Economic Assessment of Molten Carbonate Fuel Cell Manufacturing Costs,” prepared for the Electric Power Research Institute,EPRI TR-101525, Project 1677-22, Final Report, October, 1992. 73 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996.

Page 97: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-1

5. Solid Oxide Fuel Cells (SOFCs) Solid oxide fuel cells (SOFCs) are rapidly approaching commercialization in both commercial and distributed generation market sectors. Two different solid oxide fuel cell geometries are being developed, tubular and planar. The tubular design is the most advanced and, under the development of Siemens Westinghouse, the world leader in solid oxide fuel cell technology, has reached the field unit demonstration phase of commercialization. To date, Siemens Westinghouse has demonstrated 1 kW, 25 kW, 100 kW, and 250 kW (in 1999) solid oxide fuel cell stack technology and is likely to compete with MCFCs in distributed power generation. Planar SOFC technology lags tubular development, but developers are making good progress towards hurdling performance and cost issues, and some may be situated to offer a commercial product as early as 2002.

Solid oxide fuel cells employ a solid state electrolyte and operate at the highest temperature (1000°C) of all fuel cell types. The high operating temperature necessitates the use of exotic ceramics and metal-ceramic composites. While ceramic materials can be costly, the larger fraction of cost is associated with the labor and cost-intensive manufacturing techniques to fabricate stack components. If manufacturing cost targets are achieved, SOFC systems are likely to be one of the cheapest fuel cell technologies available at $800-$1000/kW. An additional driving force for pursuing this technology is the potentially high system efficiencies that SOFCs offer. When integrated with a gas turbine (SOFC-GTs), SOFCs are expected to achieve 70-75% (LHV) system efficiencies, representing a significant leap over current power generating technologies. More unconventional applications of SOFCs are also receiving attention. BMW of North America recently announced that they will employ SOFCs operating on gasoline to meet the on-board electrical requirements of its passenger vehicles while keeping the IC engine for drive power1. Renewable applications in which the SOFC is operated as both electrolyzer and converter have recently been put forth2.

This section will outline the technical aspects of SOFCs, such as electrochemistry, cell materials and component manufacturing. Cell operation and performance are then presented followed by an overview of development status and manufacturers. The section concludes with a summary of SOFC advantages, and technical and economic issues that require resolution.

ELECTROCHEMISTRY Solid oxide fuel cells operate at the highest temperature (1000°C) of all the fuel cell types. Unlike the phosphoric acid and molten carbonate fuel cells, the solid oxide fuel cell is an all-solid-state device. As solid polymer membrane fuel cells rely on a hydrated electrolyte to ensure ionic conductivity, the SOFC is technically the only true solid-state device. In general, the solid phase electrolyte is simpler in design than PAFCs or MCFCs as it requires only two phases (gas-solid) for the charge transfer reactions at the electrolyte-electrode interface. The two-phase contact simplifies the design because it eliminates corrosion and electrolyte management concerns commonly associated with the liquid electrolyte fuel cells3. Additionally, the solid electrolyte enables flexibility in cell design as many geometries can be produced. However, due

Page 98: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-2

to the high operating temperature and oxidizing environment, material requirements for SOFCs are extremely stringent. The high operating cell temperature is required to maximize the ionic conductivity of the electrolyte and ensure good electronic conductivity of the electrodes and interconnect. As a result, the critical cell components are made from various ceramics, metal-ceramic composites, and high temperature alloys.

The cell components are usually thin (10 – 1000µm) layers typically deposited one on top of the other by sophisticated (and often costly) manufacturing techniques to form a cell structure. Depending on the cell design, the thermal expansion coefficients of as many as four different ceramic layers must be well matched. The ceramic electrolyte consists of an yttria-stabilized zirconia (YSZ) which exhibits good oxygen ionic conductivity and permits little electronic conductivity at 1000°C. Like the other fuel cell types, the electrodes are porous, gas-diffusion electrodes. The anode is about 20-40% porous and is formed from metallic nickel and an yttria-stabilized zirconia skeleton (for thermal compatibility)4. The nickel serves as both an electrocatalyst and electronic conductor. The cathode is made from strontium-doped lanthanum manganite with about the same porosity as the anode5. The cells are electrically connected to each other through the cell interconnect or bipolar plate which is made from strontium-doped lanthanum chromite or a high temperature metal alloy.6,7,8,9 The materials used in the stack or tube structure will be discussed in greater detail shortly.

Depending on the cell configuration (tubular or planar), the reactant gases may flow in annular, radial, or parallelepiped spaces as shown in Figure 5-1. In radial SOFC designs, reactant gases do not flow in an open passage along an electrode surface, but diffuse through the porous electrode microstructure from center to periphery of the disk. However, no matter which design is employed, the fundamental electrochemical processes of cell operation remain the same. During operation in a planar stack as in Figure 5-1(b), hydrogen (typically reformed from natural gas, propane, diesel, methanol, or jet fuel) and air (or pure O2) flow down channels in the bipolar plates where on one side, each electrode face is exposed to the reactant gases. Figure 5-2 depicts a schematic of the electrochemical reacting system. Oxygen entering the cathode compartment is adsorbed to the cathode, diffuses to the electrode-electrolyte interface and is reduced (i.e., gain of electrons) by the incoming electronic charge. Whereas hydrogen is the charge carrier in PEMs or PAFCs, the mobile ionic species in SOFCs are negatively charged oxygen ions. The oxygen anions migrate across the electrolyte through vacant lattice sites within the solid oxide, carrying the negative charge to the electrolyte-anode interface where hydrogen is oxidized, thus accomplishing useful work. In a similar manner, because of its chemical affinity for oxygen, hydrogen entering the anode compartment is adsorbed to the anode, where it diffuses through the porous electrode to the electrode-electrolyte interface and is oxidized (i.e., loss of electrons). The high cell operating temperature enables high reactant activity and therefore facilitates fast electrode kinetics. This is especially advantageous as precious platinum electrocatalysts are not required and the electrodes cannot be poisoned by carbon monoxide. As a result, carbon monoxide is a potential fuel in SOFCs (and MCFCs). However, because of the chemical kinetics of the carbon monoxide reaction, this reaction is not favored for oxidation but for water-gas shift.

Page 99: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-3

Figure 5-1 Tubular and Planar SOFC Designs

(c) Radial Design10

Air Flow Air Electrode

Interconnection

Fuel Electrode

(a) Tubular Design

(b) Flat Plate (Bipolar) Design

Page 100: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-4

The electrode reactions are:

Anode: 2 H2 + 2 O= 2 H2O + 4 e- 2 CO + 2 O= 2 CO2 + 4 e- Cathode: O2 + 4 e- 2 O=

Overall: H2 + ½ O2 H2O CO + ½ O2 CO2 As with other fuel cell types, the electrode reactions are typically rate limiting. However, in SOFCs the operating temperatures are sufficiently elevated that performance issues are not related to kinetics but to Ohmic losses across components and component interfaces.

(Y ttria-stab ilizedZ ircon ia)

C O

(Y O - ZrO )2 3 2

(H , CH , CO, H O, CO )2 4 2 2

FIGURE 5-2 Chemical Reacting System within an SOFC11

CELL COMPONENTS AND CONFIGURATIONS Figure 5-1 depicts tubular and planar SOFC cell components for the various designs being developed. State-of-the-art SOFC designs make use of thin film concepts where films of electrode, electrolyte, and interconnect material are applied one over the other and sintered* at high temperature to form a single cell12. Presently, the manufacturing processes employed to fabricate the cells differ according to the cell configuration and manufacturer. These differences, coupled with the effort by some developers to produce a lower temperature (650-800°C) solid oxide fuel cell has generated a stratified fuel cell technology; larger demonstration units of the

* Sintering is a process whereby synthesized powder particles are heated at increasing temperatures until bonds between particles are formed (see Ref 15, p. 150.) At elevated temperature, diffusion of chemical species into the powder microstructure takes place altering the properties. Temperature and powder grain size are important variables in sintering.

Page 101: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-5

tubular type are underway while planar SOFC development is currently between small demonstration and stack development. Relative to the other fuel cell types, SOFC development is especially dependent on materials research and manufacturing processes. The development of suitable low cost cell component materials and low cost fabrication techniques represents a significant challenge for SOFCs13. Additionally, due to the many geometric configurations that can be developed from solid components, many developers are proceeding in different directions. This is most clearly evidenced by the ongoing planar stack development. As planar SOFC technology matures, it may be expected that some design conformity will eventually result (as previously occurred with PAFC and MCFC development)14. While rigorous treatment of SOFC materials and stack development is outside the scope of this report, the following attempts to consolidate the varied design efforts and materials issues.

CELL COMPONENTS The solid oxide fuel cell had its early beginnings in steam electrolysis for producing hydrogen and as an oxygen sensor.15,16 In both applications, the principle operation was based on the use of yttria-stabilized zirconia (YSZ). Presently, the electrolyte is still zirconia-based and the critical cell components are made from relatively exotic ceramics, metal-ceramic composites, and high temperature super alloys. The tubular design configuration of Siemens Westinghouse is the most advanced of all types. Table 5-1 illustrates the material history and current status of tubular development.

Table 5-1 History of Tubular Solid Oxide Fuel Cell Component Technology17

Page 102: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-6

Electrolyte The present electrolyte is essentially unchanged from that used in early SOFC development, consisting of zirconia (ZrO2) doped with 8 mol% of yttria (Y2O3). Yttria is mixed with zirconia to enable vacant oxygen-ion sites in the crystalline lattice for O= conduction18. In order for the ceramic electrolyte to demonstrate an ionic conductivity within an order of magnitude of aqueous electrolytes, the material temperature must be near 1000°C. At this temperature, the electrolyte exhibits good conductivity for the transport of oxygen anions and low electronic conductivity over a wide range of oxygen partial pressures (activities)19. At temperatures below 1000°C, the ionic conductivity decreases (ohmic resistance increases) causing a dramatic decrease in cell performance. This is shown in the voltage-current characteristic of Figure 5-3.

There are many requirements for a viable solid electrolyte material. The solid electrolyte has two primary functions: (i) as electrolyte, possessing good oxygen anion conductivity characteristics, and (ii) as reactant gas barrier which separates the fuel gas (hydrogen or hydrogen rich mixture) from the oxidant gas (O2 or air). An electrolyte exhibits good gas barrier qualities when it is of high density with no porosity. Unlike liquid electrolytes, solid ones generally provide better resistance to gas crossover from the fuel and air electrodes.20 This is significant in that loss of fuel in the depleted oxidant stream can be averted increasing thermodynamic efficiency and opportunities for developing an explosive mixture of hydrogen and oxygen are reduced. The electrolyte must also exhibit good chemical stability with respect to (i) other cell components (i.e., electrodes), and (ii) in both the oxidizing and reducing environments of the cathode and anode compartments, respectively. In addition to these requirements, it is important that the electrolyte layer be thin to minimize ohmic loss and that it maintain thermomechanical compatibility (i.e., matching of thermal expansion coefficients) with adjacent components. This is particularly important in planar SOFC stacks to minimize interfacial stresses and stress concentrations at geometric discontinuities.21

Figure 5-3: SOFC Performance with Varying Temperature (planar stack) 22

Two Cell Planar Stack Performance with 67% H2 + 22% CO + 11% H2O/Air

Page 103: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-7

While yttria-stabilized zirconia (YSZ) is most widely used, the high operating temperature creates problems in life, thermal cycling, gas sealing (planar designs), fabrication techniques, mating component interconnects, and material and manufacturing costs. In an effort to reduce these problems, two different approaches have been taken: (1) improve material and cell design to allow sustained operation within the 1000°C environment and (2) reduce the operating temperature of the SOFC while maintaining performance. Efforts to improve material and cell design have focused on increasing the fracture toughness of components, use of alternative materials or adding constituents to alter the existing electrode properties, developing high temperature gas-tight seals, reducing component polarizations, and investigating lower cost, fabrication processes.23,24 Some examples of the above include: increasing the mechanical strength of the electrolyte structure by adding alumina powder; substitution of jet vapor deposition for the capital intensive electrochemical vapor deposition process currently employed; and the use of mixed conduction (i.e., exhibit both ionic and electrical conductivity) materials, such as ceria (CeO2) in the fuel electrode.

It is anticipated that an operating temperature reduction (600 to 800°C) would enable metals to be substituted for ceramics in the electrodes and cell interconnect and thereby achieve a reduction in cost.25 The chief hurdle in lower temperature operation (600–800°C) has been the increased ionic resistance of the YSZ electrolyte. Even at 1000°C, the ohmic loss across the electrolyte is significant with respect to the rest of the cell components. This inherent resistance is particularly true for planar SOFCs because they have typically employed thick (100-200µm) electrolytes. To counter this effect, significant effort has been expended to develop ultra-thin YSZ electrolytes to keep the product of ionic conductivity and electrolyte thickness constant, and thereby maintain cell performance for intermediate temperature (600-800°C) operation. Siemens Westinghouse, SOFCo, and AlliedSignal are pursuing thin film zirconia. An electrochemical vapor deposition process produces the thin electrolytes (20-40µm) used by Siemens Westinghouse (tubular designs).26 AlliedSignal is producing 5-10µm electrolytes by a tape-calandering process for their planar SOFC designs.27

Alternatively, other research efforts have focused on search of a lower-temperature electrolyte material. One potential candidate has been ceria oxide (CeO2) based electrolytes. These have shown promise, but unfortunately also demonstrate unacceptably high electrical conductivities.28,29 Most recently, doped lanthanum gallate (La0.9Sr0.1Mg0.2Ga0.8O3) has proven to be an attractive alternative.30,31,32 This electrolyte has demonstrated good ionic conductivity and electronic resistivity at operating temperatures between 700-800°C. SOFCo and TMI in cooperation with EPRI will perform 1 kW stack development employing this electrolyte for the year 200033.

Anode Expanding on the cell component requirements summarized in Section 1 of this report, a suitable anode material should exhibit (i) effective catalysis for oxidation reactions, (ii) high electronic conductivity, (iii) thermal expansion compatibility with mating components*, (iv) sufficient porosity to allow rapid reactant gas diffusion, (v) chemical and mechanical stability in the high * This requirement has seen some relaxation in planar development. Due to the nature of the planar, radial SOFC design developed by TMI, the use of incompatible thermal expansion coefficients between electrode, electrolyte, and interconnect materials is possible and has been test demonstrated (see Ref 9, p. 125.)

Page 104: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-8

temperature environment, and (vi) resistance to poisoning by fuel gas contaminants.34 The present anode material is formed from a porous cermet† of nickel and an inert yttria-stabilized zirconia skeleton35. The cermet is formed in situ within the cell and becomes an electronic conductor at nickel contents great than 30% by volume36. The YSZ skeleton serves to better match the thermal expansion coefficients between electrolyte and electrode, and inhibit sintering of the metal particles37. The nickel supplies the electrical conductivity, but the amount of nickel in the compound is limited by thermal expansion coefficient matching with the electrolyte. As Table 5-1 shows, the fuel electrode is typically manufactured with a porosity between 20 to 40%. Depending on the type of cell design, the fuel electrode can be made from slurry-EVD or slurry-sintering processes. Manufacturing of the anode for tubular and planar designs will be described in the next section.

It is important that the anode be fuel contaminant tolerant, especially with regard to sulfur levels. To date, testing has demonstrated that SOFC tolerance to sulfur is of at least an order of magnitude higher than other fuel cell types. When sulfur tolerant electrode materials are employed, as in radial planar SOFCs, sulfur tolerance can be three or more orders of magnitude higher (~2000 ppm)38.

Cathode Material requirements for the cathode include (i) high electrocatalytic activity for oxygen reduction reactions, (ii) chemical and mechanical stability in the high temperature oxidizing environment, (iii) thermal expansion compatibility with mating components, (iv) sufficient porosity to allow rapid reactant gas diffusion, and (v) high electronic conductivity for current collection39. The high temperature, oxidizing environment of the cathode compartment makes for severe material requirements. The present cathode material is made from strontium-doped lanthanum manganite (LaMnO3) with about the same porosity as the anode40. This material exhibits good electronic conductivity if somewhat poorer thermal compatibility.

Material cost and weight for cathodes are significant due to the use of high purity, raw earth content materials like lanthanum. The doped lanthanum manganite air electrode tube represents over 90% of the weight of a tubular SOFC41. Siemens Westinghouse has demonstrated use of lower purity raw materials to synthesize the air electrode powder for cathode fabrication.

Cell Interconnect The requirements for cell interconnect materials are (i) high electronic conductivity with no electronic conductivity, (ii) chemical and mechanical stability in both the anode and cathode environments, (iii) impervious to gas phase diffusion from either fuel or oxidant species, and (iv) thermal expansion compatibility* with mating components42. In tubular and some planar designs, the present interconnect material of choice is doped lanthanum chromite (LaCrO3). In radial planar SOFC designs, high temperature metal alloys are also being used.

CELL CONFIGURATIONS Since the electrolyte is a solid, it can be cast into many different shapes affording many different cell assembly designs. The different SOFC designs being developed by manufacturers vary † A cermet is a mixture formed from ceramic and metal particles. * see footnote for anode.

Page 105: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-9

depending on the design philosophy incorporated to address cost and application issues. To date, effort has been expended at producing the tubular and planar (both flat and radial) configurations shown in Figure 5-1. Siemens Westinghouse, the leader in SOFC technology, is pursuing the tubular design. However, the currently high manufacturing costs of fabricating tubular components has limited Siemens Westinghouse to manufacturing economies of scale of 1 MWe or greater43. The high costs anticipated for tubular designs have helped to stimulate research interest in SOFC planar technology. SOFCo (a limited partnership between Ceramatec and McDermott Technology), AlliedSignal, Ztek, and TMI are the North American manufacturers pursuing this technology. The following summarizes the design aspects of the respective SOFC stack technologies.

Tubular Configuration The Siemens Westinghouse tubular cell design is the most advanced solid oxide fuel cell. It has been in development since the late 1950s, and is currently in the field demonstration and design scale-up phase before commercialization. Figure 5-4 depicts a tube cross-section for the largest tubular design produced to date. The cell diameter is currently 2.2 cm and the cell tube is 150 cm in length. The tube diameter is set (or limited) by the relatively large ohmic loss suffered from the long, circumferential current path44.

2.2 cm

Current flow

Figure 5-4 Cross-Section of Siemens Westinghouse Tube Design45

Tubular cell designs are fabricated by sequential processes beginning with the air electrode support tube followed by deposition of electrolyte, cell interconnect, and finally the anode. The air electrode support tube is fabricated by extruding a cylindrical section to the desired length, inserting a plug to close one end, and then sintering at high temperature46. The electrolyte is then applied over the porous air electrode in a uniformly thick (20-40µm) gas-tight film by electrochemical vapor deposition. This process is generally considered costly, but Siemens Westinghouse believes it preferable due to its ability to deposit uniformly dense layers reliably and in acceptable cycle times47. The cell interconnect is then applied through a plasma-spraying step. The anode is sequentially formed on the electrolyte by the deposition of a Ni-YSZ slurry

Page 106: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-10

on the electrolyte layer and completed by impregnation of YSZ by electrochemical vapor deposition process (EVD). More recently, the EVD process has been replaced by the lower cost sintering process48.

Figure 5-5 Siemens Westinghouse Tubular Stack Design49

ProcessAir 630°C

Exhaust850°C

DesulfurizedNatural Gas

Tubular Cell 1000°C

Prereformer

Combustion Plenum

Depleted FuelRecirculation Plenum

Stack ReformerFuel Ejector (jet pump)

Figure 5-5 depicts the manifolding of fuel and oxidant gases for an array of vertical tubes. The air electrode tube is closed at one end. The tubular approach of one closed end eliminates the need for gas seals between cells. In this design, air is admitted into the alumina injection tube from the top of the header and flows down the center of the tube increasing in temperature. In this arrangement, the alumina injection tube sits inside the tubular cell extending to the proximity of the closed end of the air electrode. The air exits the injection tube and flows upward in the annular cathode space in a “co-flow” arrangement with the fuel. Fuel is delivered at the bottom of the tube bundle through small injection holes and flows upward along the exterior of the tube in a parallel (co-flow) direction with the oxidant and is reformed before it undergoes electrochemical oxidation at the anode. Typically 85% of the fuel has been consumed electrochemically before the depleted gas reaches the combustion plenum50. The depleted gases exit to a common plenum where they are combusted, liberating additional heat for preheat of the incoming air stream, and are then exhausted out of the stack where the available high grade heat

Page 107: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-11

may be used to drive a power cycle or to serve thermal loads. Not all of the reaction products are immediately exhausted in a bundled tubular stack. As Figure 5-5 illustrates, a portion of the gases are recirculated and mixed with fresh, incoming fuel. The recirculated exhaust gases serve to preheat the fuel and supply the necessary H2O for fuel reforming. The amount of recirculated gas is controlled by the jet pump, in which the high momentum fuel gas entrains the necessary fraction of exhaust gases for the reforming reactions. The fuel mixture is preheated to about 700°C before it reaches the prereformer where higher hydrocarbons and some methane are reformed to hydrogen and carbon monoxide51. Inside the stack, heat for reforming is supplied by radiant heat from the cells that dissipate the heat released during the electrochemical reactions52.

Figure 5-6 illustrates the cell-to-cell connection in the tubular bundle configuration. The series-parallel design is to protect the system against complete stack failure in the event of individual cell failure53. In a large (100 kW) demonstration unit, one tube bundle consisted of 24 cells arranged in a rectangular array which configured 3 cells in electrical parallel and 8 in electrical series54.

The main advantage that the tubular cell design offers is that it eliminates the need for high temperature, leak-free gas manifolding of the fuel and oxidant streams. However, the long current path around the circumference of the tube results in large resistive losses. This effectively places an upper design limit on tube diameter. The tubular geometry also results in lower power densities than the planar SOFC designs. The next generation tubular design by Siemens Westinghouse targeted to address these issues is depicted in Figure 5-14.

Figure 5-6 Cell-to-Cell Connections in a Tube Bundle55

Flat (Bipolar) Planar Configuration

The bipolar or flat plate cell structure of Figure 5-1 is also the common configuration for PEMs, PAFCs, and MCFCs. The high manufacturing cost anticipated for the tubular design was a

Page 108: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-12

major driving force in SOFC planar development. Some of the recognized advantages of planar SOFCs are (1) lower manufacturing costs due to multiple fabrication options, (2) high power density (and lower ohmic loss), (3) ease of heat removal, (4) structural ruggedness, and (5) stack design flexibility56,57. While there are many common elements to tubular and planar components (e.g., materials), the cell and stack geometries are significantly different (see Figure 5-1). As Figure 5-6 illustrates, the cells are configured in simple electrical series connection with perpendicular current flow and no use of external cell interconnections. Additionally, two more important differences are the operating temperature and the use of gas seals at plate edges. As previously stated, reduction of the operating temperature is desirable to relax design constraints. For example, reduction from 1000°C to 700°C would reduce thermal stresses, improve cell life, and enable the use of inexpensive stainless steels lowering costs. However, the manifolding of process gases to and from the cell stack requires sealing to separate fuel and oxidant between cells. Compressive seals have been suggested but there is concern with cracking of ceramic components from the imposed non-uniform stress distribution58. In recent presentations at the 1998 Fuel Cell Seminar, planar SOFC developers state that high temperature gas sealing issues have been successfully overcome. Hence, although gas sealing has been identified as a technical hurdle to commercialization, there is some question as to whether it remains as one.

Unlike tubular designs, fabrication of individual bipolar stack components can be performed separately59. The electrolyte and interconnect layers can be formed by tape casting (or calandering). Processes commonly found in the electronics industry, such as slurry method, screen-printing, or plasma spraying apply the electrodes60. The fuel cell stacks are assembled by stacking the layers much the same as in PAFC or MCFC technology61.

Radial Planar Configuration Technology Management, Inc. (TMI) and AlliedSignal are developing the radial planar SOFC design. This design has been suggested to offer low cost and design simplicity62. The conceptual design conceived by TMI was previously depicted in Figure 1. In this design, reactant gases are supplied through holes at the center of the cell. The fuel and oxidant, delivered to each cell by internal manifolding, flow radially outward in parallel directions (“coflow” arrangement)63. Interestingly, in this design the reactant gases are not distributed in traditional gas pathways, but flow radially through the porous electrodes from center to periphery. The use of internal manifolding eliminates the need for external manifolds as in the bipolar design.

The electrolyte, anode, and cathode cell components are made from the conventional SOFC materials previously stated, however the separator plate is made from a high-temperature metal alloy64. One advantage of the radial design is that the unrestrained perimeter of the circular cell components relaxes thermal expansion matching requirements. This may allow the use of new or modified materials, such as sulfur tolerant electrodes65.

OPERATION AND PERFORMANCE The solid oxide fuel cell operates at the highest temperature of all the fuel cell types (1000°C). A penalty for higher temperature operation occurs because the increased Gibbs energy of formation (less negative) lowers the theoretical reversible cell voltage as seen by,

Page 109: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-13

VG

nF=

− ∆

where, G = Gibbs free energy of formation for the overall reaction n = the number of moles of charge transferred in the reaction F = Faraday’s Constant This disadvantage is largely offset by the lowered activation polarization during cell operation66. That is, the high operating temperature facilitates fast electrode kinetics, effectively lowering the cell voltage losses associated with activation overpotentials so that theoretical penalty for higher temperature operation is nearly eliminated. Figure 5-3 illustrates the trend of increased cell voltage as a function of temperature.

Fuel cell operation at elevated pressures is of importance because it is expected to improve cell performance (thereby reducing plant footprint) and reduce balance-of-plant costs. Recently, tubular SOFC designs have demonstrated enhanced performance operating at elevated pressures. A Siemens Westinghouse SOFC integrated gas turbine system recently recorded a 20% increase in cell power output operating at 15 atmospheres and 33% improvement at 10 atmospheres67,68. Figure 5-8 depicts the performance gains for pressurized operation in an air electrode-supported tubular SOFC operating at 1000°C. Unlike liquid electrolyte fuel cells, the SOFC demonstrated no deleterious effects from several thousand hours of sustained pressurized operation. In a SOFC-gas turbine system, the fuel cell is typically configured in a topping “cycle” or replaces a pressurized combustor as shown in Figure 5-9. Pressurized operation is likely to be applied in SOFC-gas turbine configurations only and is expected to significantly enhance power plant efficiencies to nearly 80% (LHV)69.

Air Electrode Supported Cell Performance at 1000°C

[2.2 cm diameter, 150 cm active length]

Figure 5-8: SOFC Performance with Varying Pressure (tubular stack)70

Page 110: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-14

Exhaust

Process Steam

Rankine BottomingCycle

HRSG- Process Heat

- Lower PressureSOFC-GT

kWe

SOFC(1000°C)

GasTurbine

AirCompressor

Air

NaturalGas

Figure 5-9 Simple SOFC-Gas Turbine Cycle

Fuels & Poisons Unlike any of the lower temperature fuel cells, the SOFC is more tolerant of fuel impurities and can operate using hydrogen, carbon monoxide, and methane fuels directly at the anode (see Table 1-3). The direct oxidation of methane at the anode is of thermodynamic interest since above operating temperatures of 500°C, it offers higher theoretical cell voltages than operation on hydrogen (and carbon monoxide above 600°C) thereby increasing the fuel-to-electric efficiency. This is illustrated in Figure 5-10. Additionally, the endothermic reforming reaction could be eliminated offering additional advantages in terms of fuel processing. Gardner discusses the thermodynamic advantages of the electrochemical oxidation of methane in more detail71. Due to these potential advantages, the direct oxidation of methane at the anode has recently gained interest in both conventional and renewable applications (cf. Mogensen, Ref 2). However, preliminary studies of this capability currently indicate that thermal decomposition of CH4 and subsequent carbon deposition on the anode results in decreasing performance72,73,74. The current strategy is to produce hydrogen within the cell. The high operating temperature enables internal reforming to take place within the anode compartment of the cell. Additionally, the necessary H2O for the reforming reaction can be supplied through the use of anode gas recirculation75,76. Thus, no costly external reformers or catalysts are required to produce hydrogen.

Another important aspect of SOFC operation is that no recycle of CO2 from the anode exhaust to the cathode inlet, as required by MCFCs, is necessary because O2 alone is consumed in the cathode reactions77. This enables simpler system process design and can reduce blower power requirements.

The relative insensitivity of SOFCs to gas contaminants normally considered “poisons” to lower temperature fuel cells makes it especially attractive for unconventional fuels, such as biomass or

Page 111: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-15

coal gasification. Typical contaminants found in coal gas include hydrogen sulfide (H2S), hydrogen chloride (HCl), and ammonia (NH3). Test results from experiments of a tubular SOFC operating on a simulated air-blown coal gas containing 5000 ppm NH3, 1 ppm HCl, and 1 ppm H2S favorably demonstrate no loss in performance due to ammonia or hydrogen chloride. The addition of H2S registered an immediate voltage loss followed by a linear voltage degradation with time. It is of note that there was a reversible recovery of voltage when the H2S contaminant was removed from the feed gas illustrating no permanent poisoning takes place78. Planar SOFCs have exhibited superior sulfur resistance than tubular designs; successful tests with resistance to sulfur levels over 10 ppm have been demonstrated79 using thin film technology and over 2000 ppm using radial, thick zirconia electrolytes and sulfur tolerant electrodes80. In the absence of impurities, Siemens Westinghouse cells are demonstrating lifetime capabilities greater than 69,000 hours81. While hydrogen sulfide was the largest source of voltage loss due to a contaminant in this design, the use of H2S as a potential fuel is being studied in SOFC designs employing ceria-based electrolytes82. A recent listing of poisons to solid oxide fuel cells is shown in Table 1-3.

Figure 5-10 Theoretical Cell Voltage for CH4, H2 and CO Fuels

300 500 700 900 1100 13000.60

0.70

0.80

0.90

1.00

1.10

1.20

1.30

1.40

Cell Temperature [°K]

Theo

retic

al C

ell V

olta

ge [

Volts

]

CH4

CO

H2

Fuel Utilization Fuel utilization is defined as the mass rate of hydrogen consumption over the mass flow of hydrogen supplied to the cell. All the fuel delivered to a fuel cell cannot be oxidized in the cell because the cell voltage adjusts to the lowest chemical potential for the gas mixture composition at the anode and cathode outlets83. Since electrodes act as isopotential surfaces, the cell voltage

Page 112: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-16

cannot exceed the minimum value of the Nernst potential84. Thus, in order to maintain the reaction driving force (and positive cell voltage), some fuel and oxidant content must remain in the cell gas outlet. The effect of fuel utilization on cell voltage is shown in Figure 5-11.

In SOFCs, stacks are often operated at 85% fuel utilization85. As Figure 5-11 indicated, fuel utilization has an impact on the fuel cell efficiency and moreover, a significant impact on the overall system efficiency. The remaining 15% of the fuel in the exhaust outlet is either recirculated back to the fuel cell inlet or combusted downstream of the fuel cell stack. The impact of fuel utilization on system design and efficiency is a potential area for further study.

Oxidant gas composition also impacts SOFC performance. As with other cell types, SOFCs operating efficiency is improved when operating on pure oxygen versus air. This trend is depicted in Figure 5-12.

Oxidant Utilization is 25%; (o - Pure O2; ∆ - Air)

Figure 5-11: Effect of Fuel Utilization on Cell Voltage at Varying Temperatures86

Efficiencies As previously discussed, the high operating temperature of SOFCs reduces the theoretical cell potential. In practice, the electrode kinetics are fast at elevated temperatures effectively minimizing the activation overpotential and enabling cell efficiencies of 50% or greater on a lower heating value basis. This efficiency is lower than molten carbonate fuel cells, but when combined with bottoming cycles, the overall system efficiency can exceed 75% (LHV).

Voltage- Current and Power Density Performance The voltage level of any fuel cell is determined by the magnitude of the activation, ohmic, and concentration polarizations. In SOFCs, the activation polarizations are nearly eliminated due to

Page 113: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-17

the high reactant activity at 1000°C operation. Figure 5-13 depicts the voltage-current performance of a Siemens Westinghouse tubular cell operating between 900° and 1000°C with 400% theoretical air.

1000°C Operation; Oxidant Utilization is 25%

Fuel (67% H2/22% CO/11%H2O) Utilization is 85%

Figure 5-12: Effect of SOFC Performance Operating with Pure Oxygen (o) and Air (∆)87

In a SOFC, the largest polarization (or voltage drop) is associated with ohmic resistance in the cell components. A breakdown of the contribution to ohmic polarization in tubular cells is listed below88:

Component Contribution to Polarization Cathode 45% Interconnnect 25%

Anode 18% Electrolyte 12% The contribution to the total ohmic polarization by the cathode is clearly the largest. This is due to the long, circumferential electrical current path required as depicted previously in Figure 5-3. Research efforts to reduce the ohmic loss in tubular cells (and reduce physical size and cost) have culminated in an advanced geometry tubular design shown in Figure 5-14(a). The vertical ribs shown in this design shorten the current path length, reducing the internal cell resistances and improving the power output. This design has been tested and demonstrated a 33% increase in stack power density over the current tubular design89. Development of this cell design is currently underway. The present power densities of the two respective designs are shown in Figure 5-14(b). SOFC power density is superior to MCFCs and is expected to improve with continued development (PEMs have the highest power densities, followed by PAFCs).

Page 114: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-18

(1.56 cm diameter, 50 cm length)

Figure 5-13: Voltage-Current Characteristic of a Westinghouse AES Cell90

Thermal Management The cell operating temperature is typically maintained with excess process airflow of about 4-6 times that of stoichiometric requirements. The sensible heat gain of the process gases (from the exothermic electrochemical reactions and heat generated from electronic resistances), combined with the combustion of the depleted anode tail-gas is utilized to preheat the inlet air entering the stack. However, the use of excess air affects both the capital cost and system efficiency of the plant. The process air must be preheated in a large, expensive, high alloy heat exchanger and pumped through the system by a blower that consumes electrical energy which could otherwise be sold91. Recent research efforts have focused on ways of reducing the amount of excess air. Some of the methods include the use of internal reforming, increasing allowable air temperature rise in the stack, and heat removal by conduction out of the stack92. One recent study indicates that the current operating stoichiometric numbers of a Westinghouse tubular design are excessive93. In general, this is an area of continued research.

Page 115: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-19

Figure 5-14: Advanced Geometry SOFC and Power Density Comparison94

Life Early Siemens Westinghouse porous support tube designs demonstrated over 69,000 hours of operation with less than 0.5% voltage degradation95 in single-cell laboratory tests. The commercial tubular design will be the air electrode supported (AES) design previously depicted in Figure 5-3. This design has been tested for over 20,000 hours with less than 0.1% performance degradation96. Current cell life of tubular designs are anticipated at 10 years with nearly 8 years (69,000+ hours) demonstrated. While single and two cell life tests have demonstrated long life, 25 kW stack tests have only about 15,000 hours operation, but appear to

(a) Current Flow Path for Advanced SOFC Geometry

(b) Power Density Comparison of SOFC Geometry Designs

Page 116: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-20

promise acceptable component life to this point. Siemens Westinghouse expects commercial SOFCs to have lifetimes between 10 – 20 years (2-4 times greater than other fuel cell types).

MANUFACTURERS AND MARKETS Solid oxide fuel cells are the only fuel cell technology expected to span all traditional power generating markets (Residential, Commercial, Industrial/Distributed Generation, Utility). SOFCs are also likely to penetrate niche markets, such as small portable generators, and remote and premium power applications. As previously discussed, the SOFC planar technology arose to address lower cost manufacturability of high temperature fuel cells. Hence, development has primarily emphasized “cost-optimized” over “performance-optimized” designs. Additionally, most of the planar technology companies are relatively small outfits with limited resources. For these reasons, it is presently anticipated that the smaller (< 300 kW) markets will be served by planar SOFC technology where the capital outlays are significantly less. In these markets, tubular designs are not expected to compete in the near future (10-15 years) due to the inherently higher manufacturing costs. The table below lists the major North American Manufacturers and their expected market entry product lines.

Table 5-2 North American SOFC Manufacturers & Market Entry

Manufacturer Technology Product Size Market(s)

Siemens Westinghouse

Tubular

(1000°C)

1 –5 MWe (initial)1

< 50 MWe (longterm)

Large Commercial & Industrial Cogeneration

Distributed Generation

SOFCo Flat Planar (700-800°, 1000°C)

10 – 50 kWe Commercial – HVAC

Ztek Radial Planar (1000°C)

25 – 50 kWe

250 – 300 kWe

Commercial – HVAC

Commercial Cogen2

AlliedSignal Flat & Radial Planar (600 - 800°C)

Small portable (500 W)

(Large Commercial?)

Portable Power

Commercial SOFC-GT

TMI Radial Planar (700-800°, 1000°C)

20 – 100 kWe Commercial - HVAC

1 Siemens Westinghouse SureCell SOFC initial entry will be a 1.3 MW unit, but a smaller 300 kWe may also be sold in international markets‡. 2 Ztek’s 250 kW generator unit is expected to be comprised of an SOFC integrated with a gas turbine.

‡ S. Veyo, Siemens Westinghouse, private communication, November, 1998.

Page 117: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-21

Siemens Westinghouse is the world leader in SOFC technology. The company is the only developer to have successfully demonstrated sustained operation of large (25 –100 kW) stacks in the field. Also, they have commissioned an $18 million dollar pre-pilot manufacturing facility capable of 4 MW/yr to validate lower cost manufacturing processes. Due to the current capital and labor intensive manufacturing processes of the tubular based SOFC design, Siemens Westinghouse has indicated that cost competitive generators cannot be achieved with capacities less than 1 MW97. Thus, they have focused their efforts to develop 1-5 MW pressurized SOFC-gas turbine units for high system efficiency to reduce capital and product electricity costs98. At these sizes, one study has assessed that Siemens Westinghouse could produce units for under $900/kW, assuming high-volume manufacturing of several hundred MW per year99. These units have been given the trademark name of SureCell and a plant layout is depicted in Figure 5-15.

Planar developers are initially expected to enter the commercial market sectors at 2-50 kWe. SOFCo will initially enter the market within the 10-50 kW range100. Their stack technology is centered on their CPnTM design which consists of a multi-stack configuration that allows multi-staged oxidation of the fuel101. SOFCo’s (Salt Lake, Utah) business plan is aimed at market entry SOFC – HVAC products for small commercial retail or multi-family residences. Ztek (Waltham, MA) has developed 1 and 25 kW stacks to date. Their 25-kWe stack module will be the building block for larger applications. Ztek is planning on two main product lines. The first combines a 15-100 kWe SOFC module with an HVAC system. The second is a 250 kWe packaged system that would integrate seven 25 kW SOFC stacks with a small (60-100kW) micro-turbine102.

AlliedSignal is also engaged in planar SOFC development. To date, AlliedSignal has performed significant research and development for monolithic (an SOFC type currently on hold), flat planar, and radial geometry designs. Information regarding their market product focus is scarce. Recently, they have been developing a 500 W portable SOFC electric generator weighing approximately 7 kg in a 1 cubic foot space for DARPA103. They also have ongoing work in radial SOFC design. Interestingly, since AlliedSignal is considered a major developer of 75 kWe gas turbines and it is in their strategic interest, it is suspected that they may also be developing SOFC-GT generators.

Market analyses for SOFC based systems are favorable. EPRI estimates 500-600 MW a year for the next 10 years for SOFC-GT systems in the U.S.104 Internationally, the markets are easily equivalent to that of the U.S. Potential early-market U.S. customers include, rural electric generating and transmission utilities, remote power applications where cost of T&D installation is exorbitant, low emission regions such as southern California, and regions where the current cost of electricity makes fuel cell economics more competitive, such as premium power users.

If cost targets can be obtained, the potential commercial market sector for SOFCs is also large. A market analysis of commercial building applications indicated that there were over 500,000 potential sites for SOFCs with capacities ranging from 10 – 500 kW and operating in either stand-alone or cogeneration modes105. The likely early customers are hospitals, health care facilities, hotels, educational and office buildings that require premium power service.

Page 118: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-22

STATUS OF DEVELOPMENT AND COMMERCIALIZATION Current state-of-the art SOFC technology has demonstrated satisfactory efficiency and life performance. Significant effort is being poured into cost reduction, especially that associated with fabrication techniques. For example, sintering is a high temperature process which adds production complexity and cost. The materials may cost $7 - $15/kW, but manufacturing can drive this to $700/kW for the stack106,107,108. While manufacturing and materials costs are significant, the technology does not appear to have any “showstoppers.” Tubular SOFC technology is clearly in the lead over planar designs, however, planar SOFC developers are targeting similar timeframes for commercial release of their first product lines. Table 5-3 summarizes the status of SOFC technology.

TABLE 5-3 Status Summary of SOFC Technology

TUBULAR PLANAR

Development Targets

Current Projected Current Projected

Operating Temp 1000°C 1050°C 800-1000°C 750°C (10yrs)

Operating Pressure 10 atm 1 – 15 atm 1 atm 5- 10 ? atm

Life 69,000 hrs (cell)

10 – 20 years (stack)

15,000 hrs 40,000 hrs

Size 25,100,250 kW (demos)

1 – 5 MW (commercial)

1 kW, 25 kW 3, 50, 250 kW

Cost $25- 50k /kW $900-1600/kW $50k /kW $800/kW

Commercial Timeframe 2001 2002

Some of the development issues common to both tubular and planar SOFC technology are listed below109,110:

Development Issues (1) Lower, reliable manufacturing techniques for cell components

(2) Thermal management of stack heat flows (air-cooling, internal reforming, etc.)

(3) Systems application studies to best integrate and take advantage of the new technology

(4) Quality Assurance (non-destructive evaluation techniques to detect manufacturing flaws in cell and stack components)

(5) New and/or improved materials, including

• Further development of contaminant tolerant fuel electrodes

Page 119: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-23

• Improved interconnect materials (stability and conductivity over range of O2 partial pressures)

• Establishment of physical and mechanical properties of the cell and stack components versus temperature for design and modeling of stack performance

Tubular SOFC designs also require further development on castable insulation designs, air electrode fabrication and weight reduction, and higher power density designs. Similarly, planar SOFC technology requires continued effort on developing a suitable electrolyte material for lower temperature operation, stack scale-up, and perhaps, gas seal materials and designs.

CONCLUSIONS Presentations at the recent 1998 Fuel Cell Seminar in Palm Springs, CA showed significant progress in both performance improvements and cost reductions have been made. As with all other fuel cell types, cost is the primary barrier for commercialization of SOFC technology. If cost targets are obtained, SOFCs are likely to be very successful in a wide variety of markets. While it has not been discussed to this point, SOFCs may also find some use in the heavy transportation markets, such as marine and locomotive, where the number of cycles is reduced. Although the author deems it an unlikely application, some SOFC studies have even been performed on light duty, gasoline powered electric vehicles111. In any of the numerous potential markets the SOFC finds itself in, it does offer some advantages over other fuel cell types as shown below:

SOFC Characteristic Possible Advantages

High temperature operation Allows the direct use of hydrocarbon fuel (no external reforming)

High grade waste heat available for process needs or bottoming cycle

High Power Density Reduces materials cost per unit power output

Smaller plant footprints

Ceramic materials Improved resistance to gas crossover

No electrolyte management (no corrosive liquids)

Potentially longest life of all fuel cells

Fuel Flexibility / High Contaminant No significant fuel processing requirements enable Resistance ease in use of wide range of fuel types

High SOFC-GT System Efficiencies Integration of gas turbines for ultra-high system efficiencies (>70%) is easily accomplished

Page 120: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-24

The high operating temperature makes the SOFC attractive for commercial and industrial cogeneration, but it is expected that SOFCs will be primarily baseloaded to reduce thermal cycling of the ceramic components. Perhaps most singular among fuel cell types, is the nearly perfect match of SOFCs with small gas turbines. Solid oxide fuel cells integrated with small gas turbines can potentially obtain efficiencies of 70% (LHV) or greater and offer the additional benefit of small footprint. These performance and size characteristics give SOFC-GT systems a large market potential if cost targets can be obtained.

Successful market penetration traditionally requires low cost products. Unfortunately, manufacturing low-cost products typically requires high sales volume. This apparent paradox may be offset by non-traditional influences, such as utility deregulation, “green power” (biomass fueled SOFCs), government rebates on fuel cell systems, environmental policy, and the smaller domestic and international niche power markets. Many believe that widespread penetration of fuel cell systems is not necessary for successful commercialization of the technology. Light market penetration may be enough to give fuel cells the necessary toehold to elevate performance visibility and gradually increase market share.

PowerConditioning

SOFCModule

Turbo-Generator(1/4 of total

Process Air InletFiltering

Stack

Controls &Acquisition

Figure 5-15 Westinghouse SureCell (1 - 5 MW) Power Plant Layout112

Page 121: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-25

REFERENCES 1 BMW of North America, Inc., Company Press Release, “BMW Developing Gasoline Fuel Cell in Cooperation with DELPHI Automotive Systems,” PRNewswire, Munich Germany, April 29,1999, http://biz.yahoo.com/prnews/990429. 2 M. Mogensen and C. Bagger, “SOFC: The Key to Make Renewable Energy Profitable,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, pp. 96-99. 3 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.465. 4 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-4. 5 K. Kinoshita, F.R. McLarnon, E.J. Cairns, Fuel Cells, A Handbook, prepared by Lawrence Berkeley Laboratory for the U.S. Department of Energy under Contract DE-AC03-76F00098, May 1988. 6 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-4. 7 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.484. 8 S. Singhal, “Advancements in Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 266. 9 M. Petrik, C. Milliken, R. Ruhl, and B.Lee, “Status of TMI Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Abstracts, Palm Springs, CA, p. 124. 10 Ibid., p. 124. 11 Adapted from: W.R. Dunbar, et al., “Combining Fuel Cells with Fuel-Fired Power Plants for Improved Exergy Efficiency,” Energy, 16, No. 10, pp 1260, 1991. 12 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-4. 13 Ibid., p. 5-1. 14 Ibid., p. 5-15. 15 E. Barendrecht, “Electrochemistry of Fuel Cells,” in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.105. 16 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 416. 17 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-4. 18 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 135. 19 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 416. 20 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 133. 21 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 416. 22 Data from Allied Signal, 1992 via J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998. 23 Ibid., p. 5-11. 24 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 420.

Page 122: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-26

25 K. Frist, M. Karpuk, and J. Wright, “GRI’s Fundamental Research on Intermediate-Temperature Planar Solid Oxide Fuel Cells,” 1992 Fuel Cell Seminar Program and Abstracts, Tucson, AZ, p. 624. 26 S. Singhal, “Advancements in Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 267. 27 N. Minh, A. Anumakonda, B. Chung, R. Doshi, J. Ferrall, G. Lear, K. Montgomery, E. Ong, L. Schipper, and J. Yamanis, “High Performance Reduced-Temperature Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 262. 28 C. Milliken, S. Elangovan, J. Hartvigsen, and A. Khandkar, “Ceria Electrolyte for Solid Oxide Fuel Cell Applications,” prepared for the Electric Power Research Institute, Final Report, EPRI TR-109199, 1997. 29 W. Bakker, K. Huang, J. Goodenough, A. Khandkar, S. Elangovan, and C. Milliken, “Doped Lanthanum Gallate, A Superior Electrolyte for Low Temperature Solid Oxide Fuel Cells,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p 251. 30 Ibid., p. 250. 31 T. Ishihara et al., “Doped LaGaO3 Perowskite Type Oxide as a New Oxide Ionic Conductor,” J. AmChem. Soc., 116, 3801, 1994. 32 M. Feng and J. Goodenough, “A Superior Oxide Ion Electrolyte,” Eur. J. Solid State Inorg. Chem, T31, pp. 663-72, 1994. 33 W. Bakker, K. Huang, J. Goodenough, A. Khandkar, S. Elangovan, and C. Milliken, “Doped Lanthanum Gallate, A Superior Electrolyte for Low Temperature Solid Oxide Fuel Cells,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p 251. 34 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.472. 35 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996, p. 140. 36 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.472. 37 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-5. 38 M. Petrik, C. Milliken, R. Ruhl, and B.Lee, “Status of TMI Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Abstracts, Palm Springs, CA, p. 125. 39 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.467. 40 K. Kinoshita, F.R. McLarnon, E.J. Cairns, Fuel Cells, A Handbook, prepared by Lawrence Berkeley Laboratory for the U.S. Department of Energy under Contract DE-AC03-76F00098, May 1988. 41 S. Singhal, “Advancements in Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 267. 42 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.474. 43 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996. 44 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-6. 45 Ibid., p5-9. 46 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.477.

Page 123: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-27

47 S. Singhal, “Advancements in Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 267. 48 Ibid. 49 “Solid Oxide Fuel Cells | The New Generation of Power,” Siemens Westinghouse brochure, Pittsburgh, PA. 50 S. Veyo, “The Westinghouse Solid Oxide Fuel Cell Program: A Status Report,” Proc. Of IEEE, 1996, pp.1138-43. 51 Ibid. 52 S. Veyo, “Tubular SOFC Power System Operational Experience,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 458. 53 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.476. 54 S. Veyo, “Tubular SOFC Power System Operational Experience,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 458. 55 T.J. George and M.J. Mayfield, DOE/METC-90/0268, 1990. 56 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-8. 57 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 419. 58 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-10. 59 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.483. 60 K. Kordesch and G. Simader, Fuel Cells and Their Applications, VCH Publishers, New York, NY, 1996. 61 N. Minh, “High-Temperature Fuel Cells, Part 2: The Solid Oxide Fuel Cell”, ChemTech, 21, February 1991. 62 M. Petrik, C. Milliken, R. Ruhl, and B.Lee, “Status of TMI Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Abstracts, Palm Springs, CA, p. 124. 63 Ibid. 64 Ibid. 65 Ibid., p. 125. 66 K. Murugesamoorthi, S. Srinivasan, and A. Appleby, “Research, Development, and Demonstration of Solid Oxide Fuel Cells”, in Fuel Cell Systems, Leo J. Blomen and Michael N. Mugerwa, Editors, Plenum Press, New York, 1993, p.465. 67 M. Binder and M. Williams, “Status of U.S. Stationary Power Plant Development,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 7. 68 S. Veyo, “The Westinghouse Solid Oxide Fuel Cell Program- A Status Report,” Proc. Of IEEE, 1996, pp. 1138-43. 69 T. George, K. Lyons, and R. James, “Multi-Staged Fuel Cell Power Plants Targeting 80% LHV Efficiency,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, pp. 132-135. 70 S. Singhal, “Recent Progress in Tubular Solid Oxide Fuel Cell Technology,” Proc. Of the Fifth International Symposium on Solid Oxide Fuel Cells (SOFC-V), The Electrochemical Society, 1997. 71 F.J. Gardner, “Thermodynamic Processes in Solid Oxide and Other Fuel Cells,” J. of Power and Energy, 211, No. A5, Proceedings of the Institution of Mechanical Engineers, London, UK, 1997. 72 R. Baker and I. Metcalfe, “Study of the Activity and Deactivation of LaCaCrO3 in Dry CH4 Using Temperature Programmed Techniques,” Proc. Of the 4th International Symposium on Solid Oxide Fuel Cells, Th Electrochemical Society, 1995, pp.780-90. 73 T. Horita et al., “Oxidation of CH4 on Ni and Fe Anodes in Solid Oxide Fuel Cells,” Proc. Of the 4th International Symposium on Solid Oxide Fuel Cells, Th Electrochemical Society, 1995, pp.791-800.

Page 124: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-28

74 T. Aida et al., “Direct Oxidation of Methane on Anode of Solid Oxide Fuel Cell,” Proc. Of the 4th International Symposium on Solid Oxide Fuel Cells, Th Electrochemical Society, 1995, pp.801-9. 75 W. Lundberg, “Solid Oxide Fuel Cell Cogeneration System Conceptual Design,” Final Report, Gas Research Institute, Chicago, IL, 1989. 76 S. Veyo, “The Westinghouse Solid Oxide Fuel Cell Program- A Status Report,” Proc. Of IEEE, 1996, pp. 1138-43. 77 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-18. 78 N. Maskalick, “Contaminant Effects in Solid Oxide Fuel Cells,” Proc. Of the Fourth Annual Fuel Cells Contractor’s Review Meeting, U.S. DOE/METC, July, 1992, p. 127. 79 N. Minh, A. Anumakonda, B. Chung, R. Doshi, J. Ferrall, G. Lear, K. Montgomery, E. Ong, L. Schipper, and J. Yamanis, “High Performance Reduced-Temperature Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 263. 80 M. Petrik, C. Milliken, R. Ruhl, and B.Lee, “Status of TMI Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Abstracts, Palm Springs, CA, p. 125. 81 M. Binder and M. Williams, “Status of U.S. Stationary Power Plant Development,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 7. 82 J. Electrochemical Soc., v.145, number 5, pp. 1449-54. 83 K. Kinoshita, F.R. McLarnon, E.J. Cairns, Fuel Cells, A Handbook, prepared by Lawrence Berkeley Laboratory for the U.S. Department of Energy under Contract DE-AC03-76F00098, May 1988, p. 26. 84 Ibid. 85 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-24. 86 Ibid. 87 Ibid. 88 Ibid., p. 5-15. 89 S. Singhal, “Advancements in Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 268. 90 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p. 5-24. 91 K. Krist, K. Gleason, and J. Wright, “The Effect of SOFC Stack Heat Rejection on Electrical Generation Costs,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, pp. 519-22. 92 Ibid. 93 C. Haynes, “Reconsideration of Air/Fuel Ratios for Thermal Management of Tubular Solid Oxide Fuel Cell Stacks,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, pp. 406-9. 94 S. Singhal, “Advancements in Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 267. 95 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998, p.5-24. 96 S. Singhal, “Advancements in Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 266. 97 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996. 98 S. Veyo, “The Westinghouse Solid Oxide Fuel Cell Program- A Status Report,” Proc. Of IEEE, 1996, pp. 1138-43. 99 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996.

Page 125: Review of State-of-the-Art Fuel Cell Technologies for Distri

5-29

100 A. Khandkar, S. Elangovan, J. Hartvigsen, D. Rowley, and M.Tharp, “Recent Progress in SOFCo’s Planar SOFC Development,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, pp. 465-8. 101 D. Rastler, D. Herman, R. Goldstein, and J. O’Sullivan, “State-of-the-Art Fuel Cell Technologies for Distributed Power,” Final Report, EPRI TR-106620, July 1996. 102 EPRI Journal, May/June 1997. 103 N. Minh, A. Anumakonda, B. Chung, R. Doshi, J. Ferrall, G. Lear, K. Montgomery, E. Ong, L. Schipper, and J. Yamanis, “High Performance Reduced-Temperature Solid Oxide Fuel Cell Technology,” 1998 Fuel Cell Seminar Abstracts, Palm Springs, CA, p. 265. 104 EPRI Journal, May/June 1997. 105 P. Schafer, “Commercial Sector Solid Oxide Fuel Cell Business Assessment,” prepared for the Electric Power Research Institute, Interim Report, EPRI TR-106645, August, 1996. 106 N. Minh, “High-Temperature Fuel Cells, Part 2: The Solid Oxide Fuel Cell”, ChemTech, 21, February 1991. 107 K. Frist, M. Karpuk, and J. Wright, “GRI’s Fundamental Research on Intermediate-Temperature Planar Solid Oxide Fuel Cells,” 1992 Fuel Cell Seminar Program and Abstracts, Tucson, AZ. 108 B. Halpern et al., “Jet Vapor Deposition of Thin Films for Solid Oxide and Other Fuel Cell Applications,” Proc. Of the Fourth Annual Fuel Cells Contractor’s Review Meeting, U.S. DOE/METC, July, 1992. 109 J.H. Hirschenhofer, D.B Stauffer, R.R. Engleman, and M.G. Klett, Fuel Cell Handbook, 4th Edition, prepared for the U.S. Department of Energy under contract no. DE-AC21-94MC31166, November, 1998. 110 “Commercialization of Fuel Cells,” Penner, S.S., Editor, Energy-The International Journal, Vol. 20, Pergamon Press, New York, May, 1995, p. 418-22. 111 K. Yamada, et al., “Application of SOFC for Electric Vehicle,” Proc. Of the 4th International Symposium on Solid Oxide Fuel Cells, Th Electrochemical Society, 1995, pp. 33-41. 112 “Solid Oxide Fuel Cells | The New Generation of Power,” Siemens Westinghouse brochure, Pittsburgh, PA.

Page 126: Review of State-of-the-Art Fuel Cell Technologies for Distri

6-1

6. Conclusions

This report has presented an in-depth survey and study of the technical, economic, and commercial status of fuel cell technologies. In reviewing the developmental status of each of the technologies, it is apparent that phosphoric acid fuel cells are currently the leader, but it is not so obvious which technology will be ultimately the most successful. Thus, at this development stage, one is faced with the question of which fuel cell is the “best.” Or perhaps a more tractable question: which fuel cell is the best for a given application? At this time, the early PAFC market entry units (200 kW) have targeted commercial markets from about 500 kW to several MW for baseload operation (peak shaving). The commercial applications have been most frequently, hospitals and health care facilities, schools, hotels, and office buildings where premium power and heating are required. PEMs are targeting the same applications; however, due to their lower operating temperature they are not able to provide the same quality of heat for serving end-use thermal needs. MCFCs have not entered the market yet but developers are aiming for larger distributed generation and industrial applications where high-grade heat can be used and the manufacturing economies of scale are more cost effective. Like MCFCs, SOFCs are also targeting larger energy consumers but due to its solid electrolyte and fuel flexibility, some developers are pursuing smaller commercial and residential markets. The commercial release of SOFCs is expected to occur after MCFCs.

While fuel cells still offer many performance advantages over conventional power generating equipment, the gap may not be as large as previously expected. As fuel cell technology has improved and come closer to commercialization, other power generating technologies have not been idle in making performance gains. Diesel generators have continued to reduce NOx levels and improve efficiency. The gas turbine technology has broadened its market by the development and imminent release of microturbine units ranging from 30 – 75 kWe in size. Larger gas turbine technology has reached efficiency levels near 40% without combined cycle. First generation microturbines are likely to have efficiencies less than 30%, but this is compensated by the anticipated low capital costs (~$600/kW). In general, gas turbines (small and “micro”) are expected to give fuel cells the toughest competition. Fuel cell emissions are much lower than any its competitors and environmental concerns may ultimately give fuel cells the edge. If fuel cells can realize their target costs and continue to demonstrate their anticipated wide range applicability, they will eventually succeed in multiple markets (stationary, space, and transportation applications including locomotive and marine). Due to the many uncertainties, the relative success of each fuel cell type is not clear. In fact, each fuel cell may find a market niche by uniquely providing something another cannot. For example, if one requires fast start-up and load-following capabilities, the proton exchange membrane fuel cell is the best answer. If the specific application requires minimal transients with higher-grade process heat, then a solid oxide fuel cell system may be the best fit. However, some broad insight may be gained by looking beyond the current development focus of materials and manufacturability to the total fuel cell system and its design.

All fuel cell types currently suffer from high capital costs. On the one hand, this is to be expected for the technology is in relative infancy and low volumes and poor manufacturing economies of scale result in high capital costs. On the other hand, in many cases the cost of the

Page 127: Review of State-of-the-Art Fuel Cell Technologies for Distri

6-2

fuel cell stack alone makes up less than 30% of the total power plant cost. That is, for most stationary applications of fuel cells using hydrocarbon fuels, the balance-of-plant (BOP) including power conditioning (inverter) equipment represents 65-85% of the total fuel cell system cost1. Additionally, the lower the temperature of the fuel cell, the more complex and expensive the BOP becomes. Consequently, BOP cost reduction becomes the most important factor in achieving a cost competitive product. From the previous report sections, it is clear that for lower temperature fuel cells, thermal integration of fuel processing equipment (e.g., the reformer) with the power generating section becomes more complex and therefore, more cost-intensive. International Fuel Cells Corporation has proven in their 200 kW PAFC PC25 unit that complex process operations can be effectively managed and controlled, but after 5 years of manufacturing “commercial” product, capital cost goals of $1500/kWe have yet to be achieved.

Lower Temperature Fuel Cells (PEMs and PAFCs)

PEMs operate at 80°C and PAFCs near 200°C. The lower operating temperature results in rapid power up, but poorer heat rejection for cogeneration purposes. Unlike PAFCs, PEMs have a solid electrolyte that affords a simpler and more rugged fuel cell stack design without any of the corrosion issues facing liquid electrolytes types (PAFCs and MCFCs). However, the low operating temperature also results in the need for expensive catalysts that can be poisoned by low-level carbon monoxide content. In particular, due to low CO tolerance levels, PEMs require additional fuel processing steps than PAFCs, resulting in increased costs when operating on hydrocarbon-based fuels. PAFCs operate at the low end of the efficiency goal for fuel cell power plants and they also suffer from catalyst poisoning due to the presence of CO (at levels above 3-5% by volume) in the anode gas feed.

The complexity of the BOP is a significant factor in cost. Stationary PEM power plants require a large array of ancillary equipment aimed at integrating the high temperature fuel processing subsystem with the lower temperature power generating section. The potential cost advantage of PEMs over PAFCs is a factor which cannot be overlooked when considering low-temperature options. Stationary PEM power plants may ultimately draw benefit from the dual market scenario (transportation and stationary power) in which high volumes drive down costs.

While PAFC systems can be somewhat simpler in design and PEMs offer rapid start-up and high power density, both the reduced system efficiency and the fuel limitations are arguments for high temperature fuel cells.

Higher Temperature Fuel Cells (MCFCs and SOFCs)

The operating temperature of MCFCs is 650°C and SOFCs are currently near 1000°C. Some of the benefits from high temperature operation are improved reaction kinetics, which allow elimination of expensive catalysts and elimination of the reformer by internal reforming of the fuel within the fuel cell stack itself. These characteristics augment fuel cell efficiency. CO does not poison either of these fuel cells and is in fact, a useable fuel. Thus, simpler BOP designs result and product gas from the higher temperature operation is sufficient to drive a gas turbine or steam cycle for additional power output and higher system efficiency.

1 “Commercialization of Fuel Cells,” S.S. Penner, et al., editor, Energy-The International Journal, 20 (5), May, 1995, p. 413.

Page 128: Review of State-of-the-Art Fuel Cell Technologies for Distri

6-3

MCFCs currently have significant corrosion and electrolyte management issues. Additionally, they demonstrate low power densities and require CO2 recycle for the cathode. SOFCs, on the other hand, have a solid electrolyte with long life, flexible geometry, and no corrosive liquid electrolyte problems. The SOFC operating temperature enables excellent thermal integration with both reforming and turbines. While MCFCs are slightly ahead of SOFC technology, the tubular SOFC by Siemens Westinghouse is not far behind and other geometry SOFC designs are expected to enter the market at smaller scales enabling SOFC technology at sizes as low as 15 kW. Where MCFCs have corrosion issues, SOFCs have concerns revolving around the high operating temperature that places more severe requirements on materials, fabrication processes, and mating of assembly components. High technology manufacturing processes are developing rapidly and ultimately, mature SOFC capital costs are anticipated to be the lowest of all fuel cell types.

For low temperature applications with rapid transients, the solid electrolyte PEM fuel cell is an attractive option. At higher temperatures, the solid electrolyte SOFC is most attractive. Table 6-1 presents a comparison between these two types of fuel cells. The solid electrolyte was selected for future study in Phase 2 of this project.

Recommendations Phase 1 of the study assessed the present state-of-the-art performance, potential applications, anticipated costs, and advantages of each of the four competing fuel cell types. Using the information gathered in Phase 1, the solid oxide fuel cell (SOFC) was selected for further study in Phase 2 based on criteria, such as electric efficiency, operating parameters (e.g., temperature and pressure), fuel types and fuel contaminants, anticipated mature capital cost, and end-use application flexibility. In Phase 2 of the project, two end-use applications in the commercial-HVAC market, one characterized by predominately electric-only operation and another with more level cogeneration (heating and/or cooling) needs, have been broadly defined for fuel cell system integration and performance studies. The size (electrical) of the two applications are estimated at 5-50 kWe and 100-200 kWe, respectively.

Employing solid oxide fuel cells in these applications offers many advantages over other fuel cell types. In general, fuel cells operate on hydrogen and thus require fuel processing of conventional hydrocarbon fuels. The extent of the fuel processing system depends on the fuel cell type and operating temperature. Because of their relatively high operating temperature (1000°C), solid oxide fuel cells (SOFCs) do not need significant fuel preparation enabling modeling, analysis, and simulation efforts in the proposed research to be focused on optimal system integration strategies. While a SOFC can be deployed to meet base electrical loads, their characteristically high electric-to-heat ratio also makes them attractive to provide the thermal requirements of various end-use applications. The solid oxide fuel cell produces high-grade waste heat that can be recovered for system process heating, gas compression requirements, or exported for cogeneration (or trigeneration) purposes. The recovered waste heat can be utilized for space heating, process steam, and/or domestic hot water demands. As an on-site power generator, the fuel cell will operate in stand-alone or grid-connected configurations.

Phase 2 is presently underway and is focusing on the performance and cost of fuel cell-integrated building energy system configurations at the aforementioned electrical sizes.

Page 129: Review of State-of-the-Art Fuel Cell Technologies for Distri

6-4

Table 6-1 Solid Oxide Fuel Cells versus Proton Exchange Membranes

Fuel Cell Parameter SOFC PEM

Life + 10-20 years (Expected) + 8 yrs (70,000 hrs Demonstrated)

• 5 years (Expected) - 2-3 yrs (25,000 hrs Demonstrated)

Operating Temperature 650° – 1050°C

+ Fast electrode kinetics + No expensive catalysts + Internal Reforming possible + Excellent Thermal integration - Poor thermal cycling capabilities

60° - 95°C

- Slower reaction kinetics • Expensive catalysts (but low reqmt) - External reforming required - Difficult thermal Integration + Excellent thermal cycling capabilities (cold start & shut-down)

Pressurization + 1 – 15 atm (all Demonstrated) + 1 – 5 atm (3-5 atm demonstrated)

Electric Efficiency (LHV)

+ 45% (Demonstrated)

+ 50-55% (Expected)

35- 40% (Demonstrated)

40% (Expected)

System Efficiency (LHV)

+ 65% likely

+ 75% possible (SOFC-GT)

+ 40-50% (Expected)

Waste Heat High Temperature High grade enables cogeneration/trigeneration

High Pressure Steam/Hot water Absorption Refrigeration Bottoming cycles

Low-temperature Low grade only

Hot water

Fuels / Contaminants Fuel Flexible (not easily poisoned)

+ CH4, CO are directly useable + Successfully tested on PNG, Diesel #2, JP-8 fuels + Superior resistance to S2, Cl2, and NH3 than other fuel cells

Fuel Flexibility dependent on Processor (Rigorous fuel processing required)

- CO poisons at levels > 10 ppm

- S2, Cl2, and NH3 are poisonous

Power Density ~ 10 kg/kW ~ 1 kg/kW

Sizes / Markets 2 kW – 200 MW

Residential (2 – 10 kW) Commercial (15 – 250 kW) Heavy Transportation (1+ MW) Industrial (500 kW – 10 MW) Dis-Gen ( 1 – 25 MW) Utility (100 – 200 MW)

50 W – 250 kW

Residential (2- 10 kW) Commercial (15 – 250 kW) Transportation (25 – 250 kW)

Costs $800- $1600/kWe (System) ** $1500/kWe (System) Large reductions possible if lucrative dual market scenario occurs

Development Issues Materials & Fabrication costs Lower-Temp Op. Thermal Mgmt

Fuel Processing Systems Integration & costs Water Managemt Transient Op.

Page 130: Review of State-of-the-Art Fuel Cell Technologies for Distri

ENERGY CENTEROF WISCONSIN

595 Science Drive

Madison, WI 53711

Phone: 608.238.4601

Fax: 608.238.8733

Email: [email protected]

www.ecw.org

Printed on Plainfield Plus,

a recycled chlorine-free stock

containing 20% post-consumer waste.