Mench M.M. Fuel Cell Engines

Подождите немного. Документ загружается.

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

390 Other Fuel Cells

Anode

Electrolyte

Cathode

Interconnect

Fuel

Oxidant

(b)(a)

Figure 7.11 Schematic of segmented cell-in-series design: (a) banded and (b) bell-and-spigot

conﬁguration. (Image courtesy of Siemens Power Generation.)

Other cell designs, such as radial conﬁgurations and more recently microtubular designs,

have also been developed and demonstrated [21]. The microtubular designs can achieve

lower operating temperature, since it has long been recognized that thinner electrolyte and

electrode layers promote operation at lower temperature [22].

Durability Most of the durability issues of the SOFC are related to the brittle nature of

the ceramic structures that make up the stack components. Thermal expansion mismatches

between the various components lead to high thermal stresses during both manufacture and

operation. In fact, a planar fuel cell electrolyte/electrode plate will typically be nonplanar

and slighty curved at room temperature because it is designed to be planar at opera-

tional temperature. During operation, temperature gradients as large as 100

◦

C or larger

can persist across the stack location, depending on design, which can lead to failure. In

the tube-style bundle, individual tubes that fail can be replaced without massive disruption

to the remaining fuel cell stack, as discussed. Because of the series/parallel stack bundle

design, individual cells that fail can be tolerated, an important advantage of this approach.

The thermal mismatch during operation is not the only concern, however. The manufac-

turing approach is tremendously important in SOFC operation. A high rate of wasted

materials due to breakage is typical during the manufacture and deployment of the fuel

cells.

Durability of SOFCs is not solely related to thermal mismatch issues. The electrodes

suffer a strong poisoning effect from sulfur (in the parts-per-billion range), requiring the

use of sulfur-free fuels. Additionally, anode oxidation of nickel catalysts can decrease

performance and will do so rapidly if the SOFC is operated below 0.5 V. This can be a

problem in stack operation if not every cell is monitored, since individual cells in series

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

7.1 Solid Oxide Fuel Cells 391

can easily fall below 0.5 V due to ﬂow maldistribution, degradation, cold temperature, or

other effects. Therefore, operation above 0.5 V in all cells is required.

As SOFC reduced operating temperature targets are achieved and use of inexpensive

metallic interconnects become feasible, accelerated degradation from metallic interaction

becomes a problem. Metals utilizing chromium (e.g., stainless steels) have shown limited

lifetimes in SOFC and can degrade the cathode catalyst through various physicochemical

pathways [23]. As more data are available and new materials are used, there will always be

durability issues. Durability extension is a continual improvement process for any system,

and SOFCs are no exception. However, compared to PEFCs, longterm durability of SOFCs

in large stationary systems is slightly more advanced.

7.1.2 Summary of Advantages and Disadvantages

The SOFC has many technical advantages compared to other fuel cell systems. Like

all fuel cell systems, however, it is not a perfect ﬁt for all applications and has some

technical challenges that must be overcome. Despite the advantages of the high temperature,

development of the SOFC system is now based on achieving a lower temperature operation

to enable metal components and interconnects and better sealing.

The main advantages of the SOFC system are as follows:

1. Non–noble metal catalysts are used, so that the raw material costs are low.

2. High operating temperature and water generation at the anode greatly reduce

activation polarization and eliminate the need for expensive catalysts. This also

provides a tolerance to a variety of fuel stocks and enables internal reformation of

complex fuels.

3. High-quality waste heat combined with a bottoming or cogeneration system enables

a potential for high overall system efﬁciencies (∼80%) utilizing a bottoming or

cogeneration cycle [19].

4. Tolerance to carbon monoxide (CO) is a major poison to Pt-based low-temperature

PEFCs.

5. Solid ceramic electrolyte used does not suffer electrolyte vaporization loss or ex-

cessive corrosion seen in high-temperature liquid electrolyte systems.

The major disadvantages of the SOFC include the following:

1. The system power density is lower than PEFC systems, especially for tubular SOFC

designs.

2. The high temperatures involved preclude the use of larger SOFC systems for tran-

sient applications. Also, the electrolyte is not conductive until an elevated tem-

perature is reached, so that some kind of preburner heating system is needed for

startup.

3. Damage to the components due to thermal stresses in manufacturing and operation

is difﬁcult to control, leading to a high rate of damage to the highly brittle thin

ceramics involved, even before installation and operation. The manufacturing waste

residual is high, leading to increased manufacturing costs.

The SOFC has seen tremendous advancement in recent years, especially in terms of

cost per kilowatt. High system reliability with acceptable degradation has been observed in

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

392 Other Fuel Cells

prototype systems. Additional development will be toward a commercial system to compete

in stationary and distributed power markets, where the high efﬁciency of CHP systems using

the high-quality waste heat of the SOFC can be exploited.

7.2 MOLTEN CARBONATE FUEL CELLS

Development History The MCFC utilizes a mixture of alkali metal carbonates retained

in a solid ceramic porous matrix that become ionically conductive and molten at elevated

(>600

◦

C) temperatures. MCFCs were ﬁrst studied for application as a direct coal fuel cell in

the 1930s [24]. Development of MCFC systems has focused on large, kilowatt-to-megawatt

stationary distributed power applications, due to the fact that these high-temperature systems

are really only suited for steady power generation, and the efﬁciency can be maximized

by utilizing the waste heat. Theoretical electrical conversion and cogeneration efﬁciencies

of MCFC units are expected to reach 50 and 85%, respectively. Dutch scientists G. H. J.

Broers and J. A. A. Ketelaar developed MCFCs in the 1950s. In the United States, MCFC

research at Texas Instruments funded by the U.S. Army began in the 1950s and sought to

develop fuel cells for combat operations. Since operation on battleﬁeld fuel was desired,

a high-temperature fuel cell was sought that could run on reformed diesel fuel. Testing

of several prototype units was conducted, but further development abated. Development

in the 1980s continued in various countries, primarily in the United States, Japan, and

Europe.

Recently, development has accelerated and reached the commercial stage. In the United

States, Fuel Cell Energy (FCE) of Danbury, Connecticut, manufactures Direct FuelCell

units ranging from 300 to 2.4 MWe. These units are designed to run on a variety of

renewable and conventional fuel stocks using direct internal reformation technology. In

January 2007, FCE had over 150 million kilowatt-hours of generated power at customer

sites [25]. Recently FCE partnered with the Linde Group, a worldwide leader in industrial

gases, to sell and market the FCE Direct FuelCell products worldwide. Fuel Cell Energy has

other strategic alliances in the United States (Caterpillar), Europe (MTU Friedrichshafen),

and South East Asia (Marubeni), and now has a facility in Connecticut capable of building

up to two-hundred 200-kWe systems per year [26]. In Asia, Hitachi, Toshiba, and others

have designed and built prototype units and have commercial product development plans.

Europe and South Korea also have several active research organizations and industry in

this area. Reviews of the MCFC development and commercial prospects are given in refs.

[27, 28].

7.2.1 Operation and Conﬁgurations

The MCFC normal operation temperature is 650

◦

C, with nearly atmospheric pressure,

although elevated pressure systems have been utilized. At temperatures much above 650

◦

electrolyte loss and corrosion are too rapid. Much below this temperature, performance is

poor. Due to the presence of liquid electrolyte and mechanical damage concerns to the

brittle ceramic electrolyte retaining matrix, the differential pressure between the anode and

cathode must be very low.

At the elevated temperature, the Nernst voltage and the maximum thermodynamic

efﬁciency of the MCFC are lower than for PAFC or PEFC stacks. Also, the high temperatures

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

7.2 Molten Carbonate Fuel Cells 393

Hydrogen Oxidation Reaction (HOR)

+ CO

O + CO

+ 2e

Anode

Cathode

Matrix

Oxidizer gas

Fuel gas

+ 2CO

+ 4e

2CO

Separator Plate

Oxidizer Reduction Reaction (ORR)

Figure 7.12 Schematic of reactions occurring in MCFC. (Adapted from [28].)

require very long startup times on the order of tens of hours. There are also great advantages

to the higher temperature systems, including rapid kinetics so that less expensive catalysts

can be used. Additionally, the high-quality waste heat from the MCFC can be used in a

bottoming cycle or other purposes to increase the overall combined efﬁciency as in the

SOFC.

Operating Conditions The MCFC global electrochemical reaction is the same as for other

hydrogen air fuel cells, which is the same as a balanced combustion reaction

→ H

O (7.2)

A schematic of the fuel cell reactions and the main components of the MCFC is given in

Figure 7.12. The global anode hydrogen oxidation reaction (HOR) is

+ CO

2−

→ H

O + CO

+ 2e

−

(7.3)

The global oxidizer reduction reaction (ORR) on the cathode is

+ 2CO

+ 4e

−

→ 2CO

2−

(7.4)

Operation on a CO stream is possible as well, as with the SOFC. Thus, CO, a minor

species product of fuel reformation and a major poison to PEFCs, can actually be used as a

fuel. In practice, this means that use of reformed gas without any CO cleanup is perfectly

acceptable at the anode. On-anode reformation is also achievable, as with SOFCs, since

water is generated after anode, which is necessary for steam reformation. A wide variety of

fuels, including natural gas, coal gas, and biologically produced gases, can be successfully

used with the MCFC system [29, 30].

Unique to this fuel cell system, CO

is consumed in the electrochemical reaction

at the cathode, is converted into the carbonate anion, which carries the current through

Electrochemical redox reactions have been called “cold combustion” by some, since the overall reaction is

similar, yet it occurs at lower temperature.

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

394 Other Fuel Cells

Figure 7.13 Structure and material selection of typical MCFC. (Reproduced with permission from

[31].)

the electrolyte, and is converted back into CO

an the anode. Since the net exchange of

carbonate in the electrolyte is zero, a steady state can be achieved. However, CO

does need

to be supplied to the cathode. This additional complexity of the MCFC can be accomplished

by anode gas recycling or use of reformed gas products.

The general in-cell conﬁguration and most common cell materials are shown in Figure

7.13. In order to retain the molten carbonate electrolyte, which becomes mobile at the

operating temperature, a thin (0.5–1-mm) porous α-LiAlO

or γ -LiAlO

matrix is used.

The particle size in the matrix must be small and tightly controlled in manufacturing, so that

the overall electrolyte plate can be thin to reduce ohmic losses and hold the electrolyte in

by capillary forces. The pore sizes in the matrix and electrodes are tightly controlled so that

electrolyte leakage into channels and catalyst corrosion are minimized while optimizing the

three-phase boundary between electrolyte, catalyst, and reactant. The operating temperature

of the MCFC (650

◦

C) is actually a nice range. The high temperatures facilitate the use of

nickel-based alloys for catalysts, a major advantage in raw material cost compared to the

PAFC and PEFCs. However, since the operating temperature of the MCFC is considerably

lower than the SOFC (800–1000

◦

C), metal separator plates and interconnects can be used.

This is a tremendous advantage compared to conventional SOFC systems. Stackwise, both

internally and externally manifolded conﬁgurations have been used, depending on the

manufacturer [31].

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

7.2 Molten Carbonate Fuel Cells 395

Figure 7.14 Relative performance of MCFCs at 1 atm pressure. (Adapted from Ref. [26].)

The performance of the MCFC has improved greatly over the years, at the beginning

of life and during its life history. Figure 7.14 shows the historical development in the

performance of the MCFC. Improvements in the materials, electrolyte matrix, and other

aspects have resulted in steady improvement in the power density. Continued improvement

in durability is needed for stationary applications, however.

A model of MCFC performance has been developed based on the dominating losses

in the system, ohmic resistance, and anode and cathode kinetic losses. The fuel cell perfor-

mance model from ref. [32] is given in Eqs. (7.5)–(7.8):

cell

= E

(

T, P

)

− i

(

− R

)

(7.5)

This is similar to our basic model developed in Chapter 4, where the total anodic and

cathodic resistances are lumped into two terms, and the ohmic (IR) losses constitute the

other resistance:

= 9.84 × 10

−3

exp



23,800



(7.6)

The anode polarization is quite low, compared to the cathodic and ohmic losses, but it can

be signiﬁcant at high current:

= 9.50 × 10

−7

exp



27,900



−0.5

(7.7)

Finally, the cathodic activation losses are written as

= 6.91 × 10

−15

exp



179,200



−0.75

0.5

(7.8)

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

396 Other Fuel Cells

Of course, the choice of electrolyte thickness, type, time of service, and other factors not

accounted for in the above equations will affect performance signiﬁcantly. However, the

above model can be used for qualitative estimation of MCFC performance variation with

temperature and pressure.

Reforming As discussed, since the MCFC produces waste heat and steam at the anode like

the SOFC and can use CO as fuel, an excellent opportunity exists for internal reformation

of the fuel gas. There are two types of internal reformation practiced, indirect internal

reformation (IRR) and direct internal reformation (DIR). In IRR, the fuel gas is mixed with

water vapor, heated inside the stack with waste heat, and reformed over a catalyst bed into a

hydrogen–CO mixture. The reformed mixture then enters the active fuel cell area. In DIR,

the fuel gas is reformed directly in the active area ﬂow ﬁelds. Although the DIR approach

is more compact and can theoretically be used to remove a signiﬁcant portion of the waste

heat from the stack, IRR allows the use of different catalysts speciﬁcally for reformation,

prolonging system life [32].

Example 7.1 Nernst Voltage Effect of CO

The MCFC is a little different than other

fuel cells in that carbon dioxide is a reactant at the cathode and a product at the cathode.

Write a symbolic expression for the expected change in thermodynamic voltage with partial

pressure of all reacting species.

SOLUTION From the deﬁnition of Nernst voltage given in Chapter 3 and the governing

anodic and cathodic reactions given, we can show that

E =−

G







1/2

2,c

2,a



− E



−

G







1/2

2,c

2,a





−



−

G







1/2

2,c

2,a





− E



)

1/2

2,c

/(P

2,a

)





)

1/2

2,c

/(P

2,a

)



where the subscripts a and c on the partial pressure of the CO

represent anode and cathode,

respectively.

COMMENTS: In this fuel cell, the CO

pressure on the anode and cathode enter the

Nernst equation. The effect of a concentration difference is only a few millivolts, though,

and the MCFC cannot withstand high pressure differentials between the anode and cathode

due to the liquid electrolyte and material strength limits of the electrolyte matrix.

Durability The lifetime durability of MCFCs has been greatly improved during the most

recent period of development. The MCFC system still suffers from a variety of durability

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

7.2 Molten Carbonate Fuel Cells 397

issues, as discussed below [31, 32]:

1. Electrolyte Loss Loss of the electrolyte will cause openings in the electrolyte

matrix, resulting in gas crossover. Electrolyte loss can come from vaporization

of the liquid electrolyte or reaction with the matrix, catalyst, or support materials.

Electrolyte vaporization was thought to be a lifetime limiting issue, but stack testing

has shown the evaporative losses to be much less than equilibrium thermodynamics

would predict.

2. Nickel Shorting The lithiated porous nickel oxide cathode catalyst is slightly solu-

ble in the electrolyte, which can result in dissolution of the catalyst in the electrolyte.

This reduces the cathode activity but also can lead to a nickel metal bridge between

the anode and cathode. As the nickel dissolves into Ni

ion, it migrates under the

voltage gradient to the anode and is eventually reduced with dissolved hydrogen,

precipitating in the electrolyte [33]. Over time, this bridge works back to the cathode,

causing shorting. This issue is more serious in high-pressure systems. Approaches

to mitigate this serious problem include thicker electrolyte bridges, the use of alloy-

ing agents to the NiO cathode to reduce solubility, and the use of lithium–sodium

carbonate mixtures (rather than the Li–K mixture normally used) to reduce NiO

solubility, or introduction of a ferrite layer [34].

3. Separator Plate Corrosion The stainless steel separator plate used can suffer se-

vere corrosion in the high-temperature environment at the anode and the cathode.

Corrosion with the electrolyte reduces contact efﬁciency and also consumes elec-

trolyte. This problem can be mitigated with a nickel or alternative metal coating on

the anode [35] and reduction of the separator plate surface area on the cathode.

4. γ -Li–Al

Matrix Phase Transformation The γ -Li–Al

used for the elec-

trolyte matrix can undergo phase transformation to α-Li–Al

over time. This

results in particle growth in the matrix, which reduces capillary suction and results

in lost electrolyte. Switching to an α-Li–Al

matrix material as the initial struc-

ture may avoid this problem. However, sintering and other changes with this phase

have also been observed [36].

There are of course other degradation modes, but they tend to be more minor. There has

been a resurgent interst in these systems, and progress has reduced the decay rate to nearly

acceptable levels. Measured voltage decay rates in MCFC stacks manufactured by several

companies show losses approaching the lifetime goal of 0.25% voltage loss per 1000 h, or

4% over the 40,000-h lifetime between overhaul for stationary applications [32].

7.2.2 Summary of Advantages and Disadvantages

The main advantages of the MCFC system include the following:

1. Fuel Flexibility The MCFC can internally reform a wide variety of fuel sources

and use carbon monoxide as a fuel, a major poison in PEFC applications. This

alleviates the need for a hydrogen infrastructure with this system.

2. Inexpensive Catalysts and Metal Materials High-temperature operation elimi-

nates the need for expensive noble metal catalysts found in PAFC and PEFC

applications. Non–noble metal catalysts are used, typically nickel–chromium or

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

398 Other Fuel Cells

nickel–aluminum on the anode and lithiated nickel oxide on the cathode. Since the

temperature is not too high, metal interconnects and current collectors can still be

used eliminating a major limitation of SOFC systems.

3. Quality Waste Heat The high-temperature waste heat can be used to power a

steam turbine in a bottoming cycle or for cogeneration of heat to increase overall

efﬁciency.

There are signiﬁcant drawbacks of the MCFC, however, including the following:

1. Long Startup Time The high temperature requires tens of hours for startup to avoid

damage. This eliminates the MCFC for all but continuous-power applications.

2. Low Power Density This is not a major issue in stationary and distributed power

applications, but the system size is still relatively large compared to other fuel cell

and competing technology systems.

3. Durability System longevity and durability still need to be improved. Systems

with liquid electrolyte all suffer from evaporative loss, and compared to the SOFC,

the MCFC has a highly corrosive electrolyte that, coupled with the high operating

temperature, accelerates corrosion and limits longevity of cell components, espe-

cially the cathode catalyst. The cathode catalyst (nickel oxide) has a signiﬁcant

dissolution rate into molten carbonate electrolyte, causing nickel sintering.

4. Carbon Dioxide Injection Carbon dioxide must be injected into the cathode

stream. This can be accomplished with recycling from the anode efﬂuent or in-

jection of combustion product but complicates the system design and dilutes the

oxygen mole fraction at the cathode, lowering cell voltage.

5. Electrolyte and Matrix Maintenance The liquid electrolyte interface between the

electrodes is maintained by a complex force balance involving gas-phase and elec-

trolyte liquid capillary pressure between the anode and cathode and refractory

electrolyte matrix. Signiﬁcant spillage of the electrolyte into the cathode can lead

to catalyst dissolution, and the vapor pressure of the electrolyte is nonnegligible,

leading to loss of electrolyte through reactant ﬂows.

Additionally, it is yet uncertain if the MCFC systems will be cost competitive with

other technologies. This issue ultimately relegated PAFCs to the premium power niche. The

MCFC is poised to become the second major commercially available fuel cell system, after

PAFCs discussed in the following section. Certainly, hundreds of kilowatts- and megawatts-

sized units will be in operation in the coming years from several manufacturers. A MCFC

plant is quiet with low emissions and the potential to reach 50% electrical conversion

efﬁciency and over 80% overall combined efﬁciency. Whether or not truly deep market

penetration into tens of thousands of units will occur will come down to the cost, durability,

and convenience comparisons with other technologies.

7.3 PHOSPHORIC ACID FUEL CELLS

Development History Phosphoric acid fuel cells were developed from the 1970s through

the 1990s by several companies in the United States, including ONSI, a division of In-

ternational Fuel Cells (IFC), which eventually became United Technologies Corporation

c07 JWPR067-Mench December 19, 2007 17:46 Char Count=

7.3 Phosphoric Acid Fuel Cells 399

Figure 7.15 Photograph of 100-kWe transit bus with PAFC system and methanol reformer.

(Reproduced with permission from Ref. [37].)

Fuel Cells (UTC Fuel Cells). Worldwide, Fuji Electric Company and Mitsubishi Electric

Company in Japan developed PAFC systems for residential and stationary power applica-

tions. The PAFC demonstration units have been developed for a wide variety of backup

power and even transportation applications. In the 1990s Georgetown University helped

operate a PAFC bus fueled by reformed methanol. The original stack was produced with a

Fuji Electric fuel cell stack, and a second system was installed with an IFC 100-kWe PAFC

stack, shown in Figure 7.15. This bus was operated successfully for a number of years and

then sent to the University of California–Davis. However, large relative system size and

rapid development of the PEFC have since limited development of the PAFC to stationary

power applications [37].

In April 1992, the ﬁrst commercially available fuel cell power unit, a 200-kWe PC25

PAFC system developed by ONSI, was delivered and installed [38]. Since the prototype

200-kWe PAFC unit, there have been three generations of PC25 units, with continual

reduction in size and weight and increased durability. The ﬁrst-generation PC25A weighed

in at 30,800 kg and almost 80 m

in total system volume, for a gravimetric and volumetric

power density of 6.5 W/kg and 2.5 W/L, respectively. The third-generation PC25C, built

only three years later in 1995, weighed 17,700 kg with a system volume of only only 51 m

for a gravimetric and volumetric power density of 11.3 W/kg and 3.92 W/L, respectively.

Although the fuel stack materials, design, and operation itself were improved, much of the

reduction in size and weight was achieved through total system improvement. Signiﬁcant