Mench M.M. Fuel Cell Engines

Подождите немного. Документ загружается.

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

430 Hydrogen Storage, Generation, and Delivery

For a liquid pump, the density (ρ) is almost constant, and the minimum work per unit mass

can be evaluated as

int.rev.

− P

(8.2)

For gas-phase compression, the work to achieve the same increase in pressure compared

to the liquid pump is much higher, since the density is not constant. Since signiﬁcant

non–ideal gas effects are present when compression is to 71 MPa, a simple integration with

the ideal gas assumption is an underestimate but serves as a qualitative comparison. For an

isothermal compression of an ideal gas, we can show that

int.rev.





(8.3)

For a nonisothermal polytropic compression [where (P/ρ)

= constant] of an ideal gas, we

can show that

int.rev.

n − 1







(

n−1

)

− 1



(8.4)

The work required to compress the gas to the high pressures adds to the overall cost

and complexity of the compressed hydrogen energy solution and can exceed 20% of the

speciﬁc energy content of the compressed hydrogen itself. This greatly reduces the overall

well-to-wheel efﬁciency of the fuel cell system.

Another challenge with compressed hydrogen storage is the delivery of the hydrogen

to the fuel tank. Delivery of gas from a high-pressure to low-pressure source cannot be

accomplished rapidly or the temperature will quickly rise as it is compressed in the low-

pressure side, causing safety concerns and reducing storage capability.

Additionally, the safety of compressed hydrogen is always a question. The public

perception of hydrogen as a dangerous fuel is at least partially a result of the Hindenberg

disaster of 1937, in which the hydrogen-ﬁlled German zeppelin ignited while docking in

Lakehurst Naval Air Station in Manchester, New Jersey (Figure 8.2). After ignition from a

spark, lightning, or lighting (several theories abound), the ﬂame quickly spread, resulting

in the death of tens of passengers and crewmen. Although hydrogen was widely believed to

be the cause, it is now believed that the hydrogen fuel was not completely to blame, and the

ignitable surface coating on the craft also played a major role [6]. Hydrogen does have a

very wide ﬂammability limit of 4–77%, meaning that an air–hydrogen mixture is ignitable

from 4 to 77 mol %. However, since the diffusivity of hydrogen in air is so high and it is so

light, it quickly disperses upward into the atmosphere making continual combustion more

difﬁcult to sustain. Finally, the high-pressure hydrogen storage tank is a concern not only

due to ﬂammability but also perhaps even more from the stored energy of compression,

which can cause major damage if suddenly released in the event of an accident.

In summary, the main advantages of compressed hydrogen are as follows:

1. Reasonably advanced and developed technology that can be implemented in the

near term.

2. Favorable comparison to current cryogenic storage systems in terms of parasitic

energy requirements for preparation.

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

8.1 Modes of Storage 431

Figure 8.2 The Hindenberg disaster symbolizes the danger of the use of hydrogen to many, although

it is now believed the paint on the skin of the vessel played a major role in the ﬂame propagation.

The main disadvantages include the following:

1. Low volumetric and gravimetric storage density compared to gasoline storage and

lower than desired gravimetric and volumetric energy densities, even at 70 MPa

pressure.

2. Energy-intensive hydrogen compression process.

3. Safety concerns of high-pressure cylinders and hydrogen ﬂammability.

4. Slow reﬁll of hydrogen tank from a high-pressure source.

Ultimately, improvements in other hydrogen storage alternatives may eliminate high-

pressure storage as the main storage mode, considering the disadvantages noted. For the

near term or until other technologies develop, however, compressed storage remains a

viable option and the storage solution used in most automotive fuel cell systems.

Cryogenic Liquid When cooled to about 20 K, hydrogen condenses into a liquid. Liquid

hydrogen has been used for various industrial processes since the 1940s, but application to

automotive use would require a completely new infrastructure. Figure 8.3 shows a com-

parison of the volumetric energy content of liquid versus gas-phase hydrogen storage, not

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

432 Hydrogen Storage, Generation, and Delivery

Figure 8.3 Comparison of energy content in liquid hydrogen (LH

) and gas-phase hydrogen stored

at different pressures. (Reproduced with permission from Ref. [7].)

including the storage cylinder volume. Obviously, cryogenic storage offers big advantages

compared to compressed hydrogen in terms of storage volume and reduced storage pres-

sure. Additionally, due to the lower pressures, it is possible to conform the shapes of the

storage vessel to the available space in automotive applications, increasing storage efﬁ-

ciency. Another potential advantage of cryogenic fuel compared to compressed hydrogen

is faster reﬁll capability, although the technical and safety concerns of public handling of a

cryogenic liquid at the fuel pump are nontrivial.

There are signiﬁcant limitations to the liquid hydrogen approach. Considering the

cryogenic nature of the ﬂuid and extremely low temperature, advanced thermally insulated

vessel technology must be used to reduce hydrogen boil-off rates. Boil-off is a result of

heat leakage into the storage vessel, which increases the temperature of the liquid pool

until the boiling point is reached. Additional input energy is absorbed as latent heat of

vaporization. The increased internal pressure of the sealed storage vessel as gasiﬁcation

occurs can delay the boil-off somewhat, but these vessels are generally not designed with

very high pressure capability. Therefore, when boil-off occurs, the excess hydrogen must

be vented or consumed to avoid overpressurization of the storage vessel. Venting hydrogen

to the environment poses safety and environmental concerns and results in loss of stored

fuel, which is a major inefﬁciency and practical annoyance. No one wants to leave their car

in the garage, go on vacation, and have an empty fuel tank on return. This also precludes

prefueling of vehicle rental or other ﬂeets in storage.

Another primary limitation of the cryogenic approach is the high energy loss required

to prepare liquid fuel (up to 30% of the total speciﬁc energy content [7]). Estimates of

future production costs of various fuel alternatives predict that liquid hydrogen production

will be signiﬁcantly more expensive than compressed hydrogen production, methanol, or

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

8.1 Modes of Storage 433

conventional oil resources [8]. This greatly reduces the overall well-to-wheel efﬁciency for

this option. Also, public reﬁlling stations for liquid (and likely gas-phase) hydrogen will

need to be automated, so that human interaction and possible contact with the cryogenic

fuel are eliminated.

In summary, the main advantages of cryogenic storage are as follows:

1. A relatively developed technology with demonstrated storage capability.

2. A high gravimetric and volumetric storage density.

3. Relatively quick reﬁll capability.

4. Efﬁcient transport from liquid hydrogen generation stations.

5. Some limited existing industrial infrastructure.

The main disadvantages include the following:

1. Use of a cryogenic fuel that will gain heat from the surroundings, causing boil-off

losses and necessitating highly advanced insulation.

2. The high energy loss over time and cost required to prepare liquid fuel.

3. Safety concerns to the public for delivery of cryogenic fuel.

Both pressurized storage and liquid hydrogen storage are commonly utilized in prototype

automotive units. Ultimately, however, neither cryogenic or high-pressure hydrogen storage

satisﬁes all the needs of the consumer, and alternate approaches or further advances in the

technology are needed.

Hydride Storage Metal and chemical hydrides also offer a potential solution to the hy-

drogen storage issue. In hydride storage, the hydrogen is stored as part of the solid material

matrix and released by chemical reaction or heat addition. Storage of hydrogen in this

manner offers the potential for safe, controllable and convenient storage. In general, hy-

dride storage comes in two varieties

: (1) chemical hydrides and (2) metal hydrides. Both

approaches have the potential to resolve hydrogen storage issues in mobile fuel cell ap-

plications but must achieve a higher gravimetric storage density or lower cost to compete

with other technologies. Another factor that must be considered with hydrides is the rate of

availability and release of hydrogen. Unlike compressed hydrogen storage, matching the

hydrogen release rate to the required fuel cell ﬂow rate is not a trivial matter, and proper

heat transfer and ﬂow systems must be employed.

Chemical Hydrides Chemical hydrides are manufactured hydrogen-containing materials

that are chemically reacted with water, releasing heat and hydrogen gas. The hydrogen

release process occurs upon a hydrolysis reaction with water. For example, the sodium

borohydride reaction is [9]

NaBH

+ 2H

O → NaBO

+ 4H

+ heat (8.5)

Usually the chemical hydride reacts in a water solution, and some of the hydrogen released

is from the water itself. After hydrogen release, the reaction product can typically be

Liquid-hydrogen-carrying fuels are also sometimes referred to as liquid hydrides.

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

434 Hydrogen Storage, Generation, and Delivery

rehydrided

fuel outspent

fuel in

NaBO

Gas

Pump

spent fuel

recovery

NaBO

NaBH

+ H

Chemical Plant

Figure 8.4 Schematic of chemical hydride recycle process.

recycled and reused (see Figure 8.4). Other potential chemical hydrides include calcium

hydride (CaH

), lithium aluminum hydride (LiAlH

), and magnesium hydride (MgH

The hydrogen release rate must match that required by the fuel cell, so there must

be an ability to control the hydrogen production reaction. The chemical hydride reaction

rate can be controlled by temperature via heat transfer from recycled waste heat, solution

concentration, or ﬂow rates. Rapid production of hydrogen is not typically limiting, however

[10], and the technical challenge lies in achieving the proper control of the heat transfer

and ﬂow system response.

Advantages of the chemical hydride reaction approach are the ability to rapidly and

controllably generate hydrogen, storage and transportation of hydrogen in solid form with

high hydrogen storage density, and ability to recycle and recharge the spent fuel. However,

the disadvantages of this approach are the expense of the hydride regeneration reaction and

added complexity of the hydrogen production control needed, compared to compressed gas

cylinders.

Metal Hydrides Metal hydrides have been in use for many years. Perhaps the most

well-known application is the rechargeable nickel metal hydride (NiMH) batteries that

are used in hybrid electric vehicle and laptop computer applications. These applications

use metal hydrides for electricity generation and not hydrogen storage, however. The key

feature of the metal hydride is that the hydrogen bond between the metal and hydrogen

can be reversibly controlled by thermodynamic (temperature and pressure) conditions, and

therefore the hydrogen can be liberated from the metal matrix at desired rates needed for

stack operation through controlled heat addition. Similarly, the discharged metal hydride

can be reversibly recharged with hydrogen by reﬁlling the storage vessel in a controlled

temperature and pressure environment. The generic thermodynamic equilibrium reaction

for the metal hydride process is [9]

M + 0.5 × H

↔ MH

+ heat (8.6)

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

8.1 Modes of Storage 435

Hydrogen in

Radiator Coolant pump

Hydrogen out

PEFC

Hydride

storage tank

Flow

control

H recycle

Coolant flow loop

PEFC

Coolant bypass

Waste

heat

M H + heat

H + M

Figure 8.5 Schematic of metal hydride fueling/refueling ﬂow diagram in a fuel cell car.

Note that this is an expression of thermodynamic equilibrium, and the proportion of hydride

metal and hydrogen is thus a reversible reaction. In chemical hydrides, the hydrogen

generation is accomplished through an irreversible chemical reaction of the fuel, and the

spent fuel cannot be simply recharged in situ. The solid metal hydride itself is usually in the

form of small solid particles or powder. This is a result of a process known as decrepitation,

where the volume change of the metals during charging and discharging results in crushing

of the initial metal particles.

Before use for storage, the hydride elements must undego an activation procedure to

ensure maximum storage capacity is reached. This process involves removal of impurities

and particle decrepitation to maximize the surface area for the reaction, which is a key

factor in storage potential.

In the metal hydrides investigated for hydrogen storage, the hydrogen storage step

[forward reaction in Eq. (8.6)] is exothermic. Therefore, storage at a refueling station

requires cooling of the tank during charging. During discharge, the hydrogen liberation is

endothermic, so that external heat must be used to maintain the temperature and pressure

level in the storage vessel. This can be accomplished by recycled waste heat from the fuel

cell but requires a more complex heat and ﬂow control system compared to compressed

gas storage (see Figure 8.5).

Equilibrium charging and discharging of the metal hydride is governed by thermody-

namics, and a common metric of comparison of the various metal hydrides is an isothermal

pressure–composition loop. A generic isothermal metal hydride charging and discharging

cycle diagram is shown in Figure 8.6. This hydrogen storage plot represents the central

ﬁgure of merit for metal hydrides, deﬁning the storage fraction and pressure at a given

temperature. The vertical axis is gas-phase hydrogen pressure, which can greatly vary for

different hydrides. For automotive storage applications, a range of 0.3–3 MPa is possible

[9]. Along the horizontal axis is the hydrogen–metal (H–M) mole ratio. The maximum

value of the H–M ratio achievable to date is less than 2.0, with most observed values closer

to 1.0. This low value greatly reduces the gravimetric storage density of these materials.

Typical weight storage percentages for these materials are around 2–3% for normally

achievable operating pressure and temperature. Also note from this plot that there is

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

436 Hydrogen Storage, Generation, and Delivery

Log pressure

H/M mole ratio (H/M max)

Charge/

discharge

Pressure plateau

Hysteresis

Figure 8.6 Schematic of generic isothermal metal hydride charging and discharging cycle.

(Reproduced with permission from Refs. [9] and [11].)

generally a hysteresis between the absorption and desorption pressures (P

and P

respectively). In order to achieve suitable operation in fuel cell vehicles, the absorption

and desorption pressures must be sufﬁciently low at reasonable temperatures for the charg-

ing and discharging process. Materials that desorb with suitably fast kinetics at pressures

and temperatures below 1 MPa and 100

◦

C are sought for automotive fuel cell applications.

The desorption temperature is particularly important since the waste heat from a PEFC

must be used in the process and is not normally available at temperatures above 90

◦

C. The

pressure must be sufﬁciently low to avoid bulky storage containers and overlap with the

problems of compressed hydrogen storage.

A bevy of elemental metals, solid solutions, intermetallic compounds, and other metal

mixtures have potential for hydrogen storage and the possibilities continue to evolve.

Therefore, a comprehensive listing is not given here. Instead, a brief summary of some of

the more fundamental trends is given here.

Most elemental metals do not possess suitable desorption temperatures or pressures for

fuel cell storage needs, requiring temperatures greater than 200

◦

C. Therefore, intermetallic

compounds are chosen where the hydriding properties of each metal can be mixed to

achieve the desired temperature and pressure desorption and absorption characteristics.

The approach has led to the successful development of several classes of intermetallic

compounds that have desirable thermodynamics of absorption and desorption. Compounds

AB, AB

, and AB

are common, where A is a strongly hydriding element and B is a

weakly hydriding element. For example, TiFe, ZrMn

, and LaNi

represent AB, AB

, and

compounds, respectively. More complex intermetallic compounds of the same general

formulations have been developed as well according to

= AB

(x −a)

(8.7)

Where a + x equals 2 or 5 for the AB

or AB

compounds, respectively. An example of this

for an AB

compound is LaNi

4.8

0.2

. Although the potential to mix and match hydrides

in this manner opens up a nearly inﬁnite degree of combinations, the required automotive

storage density levels of 6 wt. % have not yet been achieved.

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

8.1 Modes of Storage 437

Besides gravimetric storage density limitations, the control of the heat transfer during

storage and hydrogen evolution, several other technical challenges must be addressed [9]:

1. Cycle Durability The charge/discharge capacity of the hydride must maintain high

values for many cycles. This is mainly a function of the chemical stability of the

metallic compounds and the presence of poisons in the absorption hydrogen source

stream.

2. Purity of Absorption Hydrogen Source Stream The storage capacity of the hy-

drides are very sensitive to even minute levels of gas-phase impurities. Gas-phase

impurities can retard kinetics, reduce storage capacity, or cause irreversible corro-

sion of the storage alloy matrix.

Although the raw materials and processing costs for most of the hydriding metals are

relatively high, they can potentially be recycled after use to recover some costs, similar

to the planned recycling of platinum and other materials from PEFCs. Despite the high

potential for safe, reversible, and low-pressure hydrogen storage, metal hydrides have not

met the needs of automotive hydrogen storage due to low gravimetric energy densities and

other economic and technical challenges, but new materials hold promise for eventually

achieving this solution.

Example 8.2 How Much Would the Metal Hydride Gas Tank Weigh? From Example

8.1, we found that an automotive fuel cell system needs about 7.5 kg of hydrogen to replace

a gasoline fuel tank in automotive systems. Considering an AB

metal compound (ZrMn

with a hydrogen storage capacity of 1.77 wt. % as shown in ref. [9], determine the minimum

weight of a ZrMn

hydride storage tank needed.

SOLUTION

7.5kg

0.0177

= 960 kg

That is over a ton of metal powder, greater than 2100 lb! Some smaller cars actually weigh

less than this. In comparison, 15 gal of gasoline weighs 52 kg, or almost 20 times less.

COMMENTS: This represents the minimum weight of the metal hydride, since some sig-

niﬁcant bulk will also be required to contain the hydride powder. If the DOE goal of 6 wt. %

storage were met, the hydride powder would still weigh 150 kg, or roughly 330 lb. Even if

these goals are achieved, there is still room for improvement.

Carbon Storage Much has been made of the potential for hydrogen storage on specialized

carbon nanoconﬁgurations (activated carbon, exfoliated graphite, fullerenes, nanotubes,

nanoﬁbers, and nanohorns) or other special conﬁgurations. Very high gravimetric storage

densities have been reported, even approaching 50% for these conﬁgurations [12]. However,

initial reports were erroneous due to experimental difﬁculties in quantiﬁcation of the actual

stored hydrogen content. Storage of a few weight percent of hydrogen on carbon structures

has been observed in low-temperature (77 K) or high-pressure (>10 MPa) environments.

In October 2006, the DOE ceased research funding of single-walled nanotubes (SWNTs)

for hydrogen storage due to the inability for these materials to reach 6% storage by material

weight at room temperature [13]. However, research on metal-doped carbon nanotubes or

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

438 Hydrogen Storage, Generation, and Delivery

other alternatives is still ongoing [14, 15]. Ultimately, the goal is to achieve >6% weight

storage at room temperature and pressure. The keys to achieving high storage seem to be

increasing the surface area of carbon and tailoring the bond energy between the carbon

and hydrogen. As a rough rule of thumb, the hydrogen storage percentage increases 1%

with each 500 m

/g of carbon surface area [16]. Several different carbon-based storage

approaches have shown at least qualitative agreement with this trend. Using this as an

approximation, a surface area of at least 3000 m

/g will be needed to satisfy the storage

requirements of 6%. Being able to achieve this at low pressure and ambient temperature

will require signiﬁcantly more development and understanding of the nature of bonding

between the hydrogen and substrate.

Several other approaches to hydrogen storage on carbon have potential but are not

yet developed, including boron-doped carbon nanostructures, which show some increased

storage capacity at room temperature [17]. Additionally, carbide-derived carbons (CDCs),

produced by high-temperature chlorination of carbides, have high potential to achieve high

storage capacity since the average pore size, size distribution, and total pore volume can

be controlled with great sensitivity to achieve the high surface area needed for hydrogen

storage [18]. Metal organic frameworks (MOFs), which are highly ordered and linked

carbon networks, also have a potential for low-cost hydrogen storage [19]. The MOFs have

demonstrated 7.5 wt. % at 77 K under high pressures (∼7 MPa) and a desirable surface

area greater than 5000 m

/g [20].

Liquid Fuel Storage The use of a nonhydrogen liquid fuel, such as methanol, as a hydro-

gen carrier is a common approach to reduce fuel storage volume in portable applications

using PEFCs, as discussed in Chapter 6. For these applications, the reduced performance

of the fuel cell when using the alternative fuel is acceptable because of the reduced overall

system complexity and size.

For larger power applications, the use of a nonhydrogen liquid fuel is also an option.

Due to the low power density associated with DAFC application, however, DAFCs are not

viable for large applications. The generation and transportation of a variety of liquid fuels

such as methanol, ammonia, or synthetic liquid fuels as hydrogen carriers remain important

options, discussed in Section 8.3.

For high-temperature solid oxide or molten carbonate systems, the liquid fuel can be

internally reformed, as discussed in Chapter 7. This unique capability reduces the burden

of fuel storage in these systems, although SOFCs and MCFCs are generally not considered

for mobile applications. In stationary systems, however, this capability allows for operation

directly from the natural gas grid.

8.2 MODES OF GENERATION

Hydrogen generation can be achieved by a variety of chemical, electrochemical, biological,

and other methods. In 2005, about 90% of hydrogen worldwide was generated from chem-

ical reformation of carbon-based fuels. Much of the remaining hydrogen was generated via

electrolysis. Biological production of hydrogen holds great promise for an environmentally

friendly solution but is not yet practiced in large volumes comparable to the other available

approaches. Each approach has merits and limitations, and the choice of what is “best”

depends on the economics of the location involved.

c08 JWPR067-Mench December 15, 2007 21:19 Char Count=

8.2 Modes of Generation 439

8.2.1 Chemical Hydrogen Production

Chemical generation of hydrogen from liquid- or gas-phase carbon-based fuels is known as

fuel reformation. While there are many manifestations of reformation, a generic schematic

of the process is shown in Figure 8.7. Some carbon-containing fuel, such as natural gas

or methanol, enters the reformer and the carbon is oxidized to produce hydrogen, carbon

dioxide, and other species. In principle, any carbon-containing species can be reformed to

produce hydrogen.

Reformation is well developed and convenient but produces carbon dioxide, a green-

house gas, and carbon monoxide, a major poison for low-temperature PEFCs. Although

this approach produces carbon dioxide and carbon monoxide, since hydrogen fuel cells

can potentially be more efﬁcient than direct use of the reformed fuel in a conventional

process, less overall fuel is used so there can still be a net beneﬁt to the environment.

Carbon sequestration from large reformation plants can also be used to further minimize

the impact of the carbon emissions to the environment.

Because of the undesired CO byproduct, the hydrogen produced for PEFCs generally

needs to be further processed to reduce CO levels. One way to achieve this is through a

water–gas shift (WGS) reaction:

CO + H

O → CO

+ H

+ heat (8.8)

This is an exothermic process that liberates 41 kJ/mol of CO and can usually reduce the CO

levels in the stream to around 1%, which is still too high for PEFC operation but is suitable

for PAFC operation. A WGS reactor is often used downstream of or in parallel with the

fuel reformer to reduce CO levels. Another methods to remove the CO is the preferential

oxidation reaction:

CO +

→ CO

+ heat (8.9)

In this case, air or oxygen is introduced to the fuel stream to promote oxidation of CO to

. This reaction can be used to reduce the CO content to the 10–100-ppm level, which

is still harmful for PEFC operation. Further cleanup can be accomplished through selective

membrane ﬁltering and catalyst absorption beds. Obviously, the CO cleanup required for

PEFC operation is extensive and represents a major drawback of the reformation approach

for supplying hydrogen for PEFCs. Besides water management and heat rejection, increased

Oxidizer

(air/H

Carbon containing fuel

(e.g., CH , CH OH)

+ CO

+ CO + N

+ H

Reformer

Figure 8.7 Generic fuel reformation process.