Wind Energy and The Hydrogen Economy
Wind Energy and The Hydrogen Economy
Wind Energy and The Hydrogen Economy
www.elsevier.com/locate/solener
a
University of Florida, Department of Mechanical and Aerospace Engineering, 232 MAE-B, P.O. Box 116300,
Gainesville, FL 32611-6300, United States
b
Connecticut Global Fuel Cell Center, 44 Weaver Road, Unit-5233, Storrs, CT 06269-5233, United States
c
Clean Energy Research Institute, University of Miami, P.O. Box 248294, Coral Gables, FL 33124, United States
Received 12 November 2002; received in revised form 29 November 2004; accepted 4 January 2005
Available online 12 March 2005
Communicated by: Associate Editor Ali Raissi
Abstract
The hydrogen economy is an inevitable energy system of the future where the available energy sources (preferably
the renewable ones) will be used to generate hydrogen and electricity as energy carriers, which are capable of satisfying
all the energy needs of human civilization. The transition to a hydrogen economy may have already begun. This paper
presents a review of hydrogen energy technologies, namely technologies for hydrogen production, storage, distribution,
and utilization. Possibilities for utilization of wind energy to generate hydrogen are discussed in parallel with possibil-
ities to use hydrogen to enhance wind power competitiveness.
2005 Elsevier Ltd. All rights reserved.
Keywords: Hydrogen energy; Wind energy; Fuel cells; Solar energy; Hydrogen production; Hydrogen storage
0038-092X/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.solener.2005.01.002
648 S.A. Sherif et al. / Solar Energy 78 (2005) 647–660
fossil fuels is not. Hydrogen is another clean, efficient on location, electricity may be used directly or trans-
and versatile energy carrier, which supplements electric- formed into hydrogen. For large-scale storage, hydrogen
ity very well. Together these two carriers may satisfy all can be stored underground in ex-mines, caverns or aqui-
the energy needs and form an energy system that is per- fers. Fuel cells may be available in MW power plant size
manent and independent of energy sources (Bockris, or several kW suitable for distributed power generation.
1975; Bockris and Veziroglu, 1985; Bockris et al., 1991). Together with renewable energy sources, such as solar
Hydrogen has some unique characteristics that make and wind, electricity and hydrogen form a clean energy
it an ideal energy carrier (Veziroglu and Barbir, 1992), system capable of permanently satisfying all the energy
namely: needs of human civilization.
• It can be produced from and converted into electric- 1.1. Hydrogen energy technologies
ity at a relatively high efficiency.
• Raw material for hydrogen production is water, Most of the technologies required for hydrogen pro-
which is available in abundance. Hydrogen is a com- duction, storage, and utilization have already been
pletely renewable fuel, since the product of hydrogen developed. Few of them are at a level where they can
utilization (either through combustion or through compete with the existing energy technologies. The fol-
electrochemical conversion) is pure water or water lowing is a review of the technologies for hydrogen pro-
vapor. duction, storage, distribution, and utilization.
• It can be stored as liquid, gas, or solid (metal
hydrides). 1.1.1. Technologies for hydrogen production
• It can be transported over large distances using pipe- Production of hydrogen requires feedstock (logical
lines, tankers, or rail trucks. sources being hydrocarbon fuels, CXHY, and water,
• It can be converted into other forms of energy in H2O) and energy. The amount of energy required to
more ways and more efficiently than any other fuel, produce hydrogen is always greater than the energy that
i.e., in addition to flame combustion (like any other can be released by hydrogen utilization.
fuel) hydrogen may be converted through catalytic Presently, hydrogen is mostly being produced from
combustion, electro-chemical conversion, and fossil fuels (natural gas, oil, and coal). Hydrogen is used
hydriding. in refineries to upgrade crude oil (hydrotreating and
• Hydrogen as an energy carrier is environmentally hydrocracking), in the chemical industry to synthesize
compatible. It produces small amounts of NOx if it various chemical compounds (ammonia, methanol,
is burned with air at high temperatures. etc.), and in metallurgical processes as a reduction or
protection gas.
Fig. 1 shows a global energy system in which electric- Technologies for hydrogen production from fossil
ity and hydrogen are produced from available energy fuels have been developed and are used to produce
sources and used in many applications. Both hydrogen industrial hydrogen. These include steam reforming of
and electricity complement renewable energy sources natural gas, partial oxidation of hydrocarbons and coal
particularly well, by presenting them to the end user in gasification. Depending on the cost of fuel, hydrogen
a convenient form and at a convenient time. Depending can be produced for 6–14 $/GJ (Steinberg and Cheng,
1988). However, these technologies depend on fossil
fuels and emit CO2. The only method that can generate
hydrogen from fossil fuels without generation of CO2 is
direct thermal and catalytic cracking of hydrocarbons.
This method has been used to produce carbon, but for
cost effective hydrogen generation it is still in the early
development phase (Muradov, 1998).
Water electrolysis is relatively efficient (>70%), but
because it needs electricity, hydrogen produced by
water-electrolysis is expensive (>$20/GJ assuming a cost
of about $0.05/kW h). However, there is a potential to
generate relatively inexpensive hydrogen from hydro-
power and nuclear plants.
Water electrolysis is particularly suitable to be used
in conjunction with photovoltaics (PVs) and wind en-
ergy. In general, there is a good match between the
polarization curves of PV cells and electrolyzers. Experi-
Fig. 1. Hydrogen/electricity energy system. ence from PV/electrolysis pilot plants shows that they
S.A. Sherif et al. / Solar Energy 78 (2005) 647–660 649
can be matched directly with no power tracking elec- structure near Beynes, France. Imperial Chemical Indus-
tronics, and can yield a relatively high efficiency (>93% tries of Great Britain stores their hydrogen in the salt
coupling efficiency) (Steeb et al., 1990). Electricity pro- mine caverns near Teeside, United Kingdom (Pottier
duced from photovoltaics is expensive and hydrogen and Blondin, 1995).
produced from such electricity is even more expensive. Above ground pressurized gas storage systems are used
Nevertheless, this technology is well developed that today in natural gas business in various sizes and pres-
eventually could be used on a large scale for electricity sure ranges from standard pressure cylinders (50 l,
and hydrogen generation. The cost of PV-cells in the 200 bars) to stationary high-pressure containers (over
1990s did not drop in spite of a multi-fold increase in 200 bars) or low-pressure spherical containers (>
production volume and a reduction in manufacturing 30,000 m3, 12–16 bars). This application range may be
cost. Projections from the late 1980s showed that the similar for hydrogen storage.
cost could be as low as $0.2/Wp (Ogden and Williams, Vehicular pressurized hydrogen tanks made of ultra-
1989), but the price of PV modules is still over $4/Wp, light but strong new composite materials that allow
and installed can be as high as $15/Wp (Ventre, 2000). pressures in excess of 200 bars have been developed
This would be prohibitively expensive to generate elec- and used in prototype automobiles and buses. Storage
tricity and hydrogen on a large scale. density of higher than 0.05 kg of hydrogen per 1 kg of
As the ‘‘Wind Energy and Hydrogen’’ section indi- total weight is easily achievable (Mitlitsky, 1996).
cates, wind power may be produced at a very low cost Liquid hydrogen storage: Hydrogen liquefaction is an
in the regions with enough wind resources. It can be energy intensive process. It requires amounts of energy
used to generate hydrogen on a small or a large scale. equal to about one third of the energy in liquefied hydro-
Other pilot scale methods for hydrogen production gen. Hydrogen liquefaction and use of liquid hydrogen
include direct thermal decomposition or thermolysis is usually practiced only where achieving high storage
(Baykara and Bilgen, 1989), thermochemical cycles density is absolutely essential, such as in aerospace
(Yalcin, 1989), and photolysis (photo-chemical, photo- applications. Some prototype hydrogen-powered auto-
electrochemical, and photo-biological processes) (Willner mobiles as well as commercially-available automobiles
and Steinberger-Willner, 1988; Das and Veziroglu, 2001). also use specially developed liquid hydrogen tanks (Bra-
Unfortunately, all of these processes have severe techni- ess and Strobl, 1996).
cal difficulties and are far from industrial applications. Metal hydride storage: Hydrogen can form metal hy-
For a comprehensive review of hydrogen production drides with some metals and alloys. During the forma-
technologies the reader is referred to Goswami et al. tion of the metal hydride, hydrogen molecules are split
(2003). and hydrogen atoms are inserted in spaces inside the lat-
tice of suitable metals and/or alloys. In such a way an
1.1.2. Technologies for hydrogen storage effective storage comparable to the density of liquid
Hydrogen as an energy carrier must be stored to hydrogen is created. However, when the mass of the
overcome daily and seasonal discrepancies between en- metal or alloy is taken into account then the metal
ergy source availability and demand. Hydrogen can be hydride gravimetric storage density is comparable to
stored either as gas or as liquid. Also, it can also be the storage of pressurized hydrogen. The best achievable
stored in metal hydrides, chemical hydrides, glass micro- gravimetric storage density is about 0.07 kg of H2/kg of
spheres, or cryo-adsorbers. metal for a high temperature hydride such as MgH2.
Large underground hydrogen storage in caverns, aqui- During the storage process, heat is released which must
fers, depleted petroleum, and natural gas fields is likely be removed in order to achieve the continuity of the
to be technologically and economically feasible (Taylor reaction. During the hydrogen release process, heat must
et al., 1986). Hydrogen storage systems of the same type be supplied to the storage tank. An advantage of storing
and the same energy content will be more expensive by hydrogen in hydriding substances is the safety aspect. A
approximately a factor of three than natural gas storage serious damage to a hydride tank (e.g. collision) would
systems, due to hydrogenÕs lower volumetric heating not pose fire hazard, since hydrogen would remain in
value. Technical problems, specifically for the under- the metal structure. Table 1 provides some hydriding
ground storage of hydrogen other than the expected substances as hydrogen storage media. Table 2 provides
losses of the working gas in the amount of 1–3% per year a list of hydrogen storage types and densities.
are not anticipated. The city of KielÕs public utility has Novel hydrogen storage methods: Hydrogen can be
been storing town gas with a hydrogen content of physically adsorbed on activated carbon and be
60–65% in a gas cavern with a geometric volume of ‘‘packed’’ on the surface and inside the carbon structure
about 32,000 m3 and a pressure of 80–160 bars at a more densely than if it has been just compressed.
depth of 1330 m since 1971 (Carpetis, 1988). Gaz de Amounts of up to 48 g H2 per kg of carbon have been
France, the French National Gas Company, has stored reported at 6.0 MPa and 87 K (Schwartz and Aman-
hydrogen-rich refinery by-product gases in an aquifer kwah, 1993). The adsorption capacity is a function of
650 S.A. Sherif et al. / Solar Energy 78 (2005) 647–660
Table 1
Hydriding substances as hydrogen storage media
Medium Hydrogen content, Hydrogen storage capacity, Energy density, Energy density,
kg/kg kg/liter of vol. kJ/kg kJ/liter of vol.
MgH2 0.070 0.101 9933 14,330
Mg2NiH4 0.0316 0.081 4484 11,494
VH2 0.0207 3831
FeTiH1.95 0.0175 0.096 2483 13,620
TiFe0.7Mn0.2H1.9 0.0172 0.090 2440 12,770
LaNi5H7.0 0.0137 0.089 1944 12,630
R.E.Ni5H6.5 0.0135 0.090 1915 12,770
Liquid H2 1.00 0.071 141,900 10,075
Gaseous H2 (100 bar) 1.00 0.0083 141,900 1170
Gaseous H2 (200 bar) 1.00 0.0166 141,900 2340
Gasoline – – 47,300 35,500
hydrogen, the recompression stations could be spaced line engines. Basically, the only products of hydrogen
twice as far apart. In economic terms, most of the stud- combustion in air are water vapor and small amounts
ies found that the cost of large-scale transmission of of nitrogen oxides. However, the emissions of NOx in
hydrogen is about 1.5–1.8 times larger than that of nat- hydrogen engines are typically one order of magnitude
ural gas transmission. However, transportation of smaller than emissions from comparable gasoline en-
hydrogen over distances greater than 1000 km is more gines. Small amounts of unburned hydrocarbons, CO2,
economical than transmission of electricity (Öney and CO have been detected in hydrogen engines due to
et al., 1994). lubrication oil (Nornbeck et al., 1996).
In order to match the consumption demand, hydro- Hydrogen use in turbines and jet engines is similar to
gen can be regionally transported and distributed, both use of conventional jet fuel. The use of hydrogen avoids
as gas or liquid, by pipelines or in special cases in con- the problems of sediments and corrosion on turbine
tainers by road and rail transportation. Gaseous and blades, which prolongs life and reduces maintenance.
liquid hydrogen carriage is subject to strict regulations Gas inlet temperatures can be pushed beyond normal
ensuring public safety, which in some countries is very gas turbine temperatures of 800 C, thus increasing the
constraining. The transportation of hydrogen in a dis- overall efficiency. The only pollutants resulting from
continuous mode, whether in gaseous or liquid state, is the use of hydrogen in turbines and jet engines are nitro-
currently used by occasional or low volume users. The gen oxides.
cost of discontinuous transport is very high (up to 2–5
times the production cost). 1.1.4.2. Direct steam generation by hydrogen/oxygen
Hydrogen in the gas phase is generally transported in combustion. Hydrogen combusted with pure oxygen re-
pressurized cylindrical vessels (typically at 200 bars) ar- sults in pure steam, i.e., 2H2 + O2 ! 2H2O. This reac-
ranged in frames adapted to road transport. The unit tion would develop temperatures in the flame zone
capacity of these frames or skids can be as great as above 3000 C. Therefore, additional water has to be in-
3000 m3. Hydrogen gas distribution companies also in- jected so that the steam temperature can be regulated at
stall such frames at the user site to serve as a stationary a desired level. Both saturated and superheated vapor
storage. can be produced. The German Aerospace Research
Establishment (DLR) has developed a compact hydro-
1.1.4. Technologies for hydrogen utilization gen/oxygen steam generator (Sternfeld and Heinrich,
Technologies for hydrogen conversion into other use- 1989). Such a device is close to 100% efficient, since there
ful energy forms have already been developed and dem- are no emissions other than steam and little or no ther-
onstrated. In almost all the cases hydrogen is converted mal losses. It can be used to generate steam for spinning
more efficiently than any other fuel. An important as- reserve in power plants, for peak load electricity genera-
pect is the fact that hydrogen conversion creates little tion, in industrial steam supply networks and as a
or no emissions (mainly water or water vapor). These micro steam generator in medical technology and
technologies are the driver for development of technolo- biotechnology.
gies for hydrogen production and storage.
1.1.4.3. Catalytic combustion of hydrogen. Hydrogen
1.1.4.1. Combustion of hydrogen in internal combustion and oxygen in the presence of a suitable catalyst may
engines and turbines. Hydrogen is a very good fuel for be combined at temperatures significantly lower than
internal combustion engines. Hydrogen-powered inter- flame combustion (ambient-500 C). Catalytic burners
nal combustion engines are on average about 20% more require considerably more surface area than conven-
efficient than comparable gasoline engines. The thermal tional flame burners. Therefore, the catalyst is typically
efficiency of an engine can be improved by increasing dispersed in a porous structure. The reaction rate and
either the compression ratio or the specific heat ratio. resulting temperature are easily controlled by control-
In hydrogen engines both ratios are higher than in a ling the hydrogen flow rate. The only product of cata-
comparable gasoline engine due to hydrogenÕs lower lytic combustion of hydrogen is water vapor. Due to
self-ignition temperature and ability to burn in lean mix- low temperatures there are no nitrogen oxides formed.
tures. However, the use of hydrogen in internal combus- The device is inherently safe since the reaction cannot
tion engines results in 15% loss of power due to the migrate into the hydrogen supply, because there is no
lower energy content in a stoichiometric mixture in the flame, in addition, to the fact that hydrogen concentra-
engineÕs cylinder. The power output of a hydrogen en- tion is above the higher flammable limit (75%). Possible
gine can be improved by using advanced fuel injection applications of catalytic burners are in household appli-
techniques or liquid hydrogen (Nornbeck et al., 1996). ances such as cooking ranges and space heaters. Fraun-
One of the most important advantages of hydrogen hofer Institute for Solar Energy Systems, Germany, has
as a fuel for internal combustion engines is that hydro- developed and demonstrated several household appli-
gen engines emit less pollutants than comparable gaso- ances using catalytic burners (Ledjeff, 1990).
652 S.A. Sherif et al. / Solar Energy 78 (2005) 647–660
1.1.4.4. Electrochemical electricity generation (fuel trolyte. These cells operate at 900–1000 C where
cells). Fuel cells are one of the most attractive and ionic conduction by oxygen ions takes place.
most promising hydrogen technologies. In a fuel cell Recently, low temperature (600 C) solid oxide fuel
hydrogen combines with oxygen without combustion cells are being developed (Kinoshita et al., 1988; Blu-
in an electrochemical reaction (reverse of electrolysis) men and Mugerwa, 1993).
and produces direct current (DC) electricity. Depending
on the type of the electrolyte used, there are several types A typical fuel cell consists of the electrolyte, in con-
of fuel cells: tact with porous electrodes, on both sides. A schematic
representation of a fuel cell with reactant and product
• Alkaline fuel cells (AFC) use 85 wt% KOH as the gases, and ions flow directions for the major types of fuel
electrolyte for high temperature operation (250 C) cells are shown in Fig. 2. The electrochemical reactions
and 35–50 wt% for lower temperature operation occur at the three-phase interface—porous electrode/
(<120 C). The electrolyte is retained in a matrix electrolyte/reactants. The actual electrochemical reac-
(usually asbestos), and a wide range of electrocata- tions that occur in the above listed types of fuel cells
lysts can be used (such as Ni, Ag, metal oxides, and are different (as shown in Table 3) although the overall
noble metals). This fuel cell is intolerant to CO2 pres- reaction is the same, i.e., H2 + 1/2O2 ! H2O. Low tem-
ence in either the fuel or the oxidant (Kinoshita et al., perature fuel cells (AFC, PEMFC, PAFC) require noble
1988). electrocatalysts to achieve practical reaction rates at the
• Polymer electrolyte membrane or proton exchange anode and cathode. High temperature fuel cells (MCFC
membrane fuel cells (PEMFC) use a thin (30 lm) and SOFC) can also utilize CO and CH4 as fuels. The
proton conductive polymer membrane (such as per- operating temperature is high enough so that CO and
fluorosulfonated acid polymer) as the electrolyte. CH4 can be converted into hydrogen through the
The catalyst is typically platinum with loadings of water-gas shift and steam reforming reactions,
about 0.3 mg/cm2. If the hydrogen feed contains respectively.
small amounts of CO, Pt–Ru alloys are used. The The electrolyte not only transports dissolved reac-
operating temperature is typically between 60 and tants to the electrode, but it also conducts ionic charge
80 C. between the electrodes and thereby completes the cell
• Phosphoric acid fuel cells (PAFC), use 100% con- electric circuit, as shown in Fig. 2.
centrated phosphoric acid as the electrolyte. The
matrix used to retain the acid is usually SiC, and
the electrocatalyst in both the anode and cathode is
platinum. The operating temperature is typically water oxygen
vapor from
between 150 and 220 C (Kinoshita et al., 1988; Blu- air
men and Mugerwa, 1993).
• Molten carbonate fuel cells (MCFC) have the electro-
lyte composed of a combination of alkali (Li, Na, K)
porous
carbonates, which is retained in a ceramic matrix of cylinder
LiAlO2. Operating temperatures are between 600 reaction or plate
and 700 C where the carbonates form a highly con- zone
ductive molten salt, with carbonate ions providing
ionic conduction. At such high operating tempera-
tures, noble metal catalysts are typically not required
(Kinoshita et al., 1988; Blumen and Mugerwa, 1993). hydrogen
• Solid oxide fuel cells (SOFC) use a solid, nonporous
metal oxide, usually Y2O3-stabilized ZrO2 as the elec- Fig. 2. Schematic representation of catalytic burner.
Table 3
Fuel cell reactions
Fuel cell type Anode reaction Cathode reaction
Alkaline H2 + 2OH ! 2H2O + 2e 1/2O2 + H2O + 2e ! 2OH
Proton exchange H2 ! 2H+ + 2e 1/2O2 + 2H+ + 2e ! H2O
Phosphoric acid H2 ! 2H+ + 2e 1/2O2 + 2H+ + 2e ! H2O
Molten carbonate H2 + CO@3 ! H2O + CO2 + 2e
1/2O2 + CO2 + 2e ! CO@
3
Solid oxide =
H2 + O ! H2O + 2e 1/2O2 + 2e ! O=
S.A. Sherif et al. / Solar Energy 78 (2005) 647–660 653
Cell potential
where DG0 = Gibbs free energy at 25 C and atmo-
spheric pressure, n = number of electrons involved in
the reaction, F = FaradayÕs constant, E0 = reversible po-
tential at 25 C and atmospheric pressure (V).
The reversible potential changes with temperature
and pressure; in general it is lower at higher tempera-
tures (reaching 1.0 V at 1000 K), and it is higher at
Current density
higher pressures or higher concentrations of reactants.
The actual voltage of an operational fuel cell is always Fig. 4. Typical fuel cell polarization curve.
lower than the reversible potential due to various irre-
versible losses, such as activation polarization, concen- The fuel cell efficiency is a function of cell voltage.
tration polarization, and ohmic resistance. While The theoretical fuel cell efficiency is:
ohmic resistance is directly proportional to the current,
activation polarization is a logarithmic function of the gFC ¼ DG=DH ð2Þ
current, and is thus more pronounced at very low cur- where DH is hydrogenÕs enthalpy or heating value
rent densities. Concentration polarization is an exponen- (higher or lower). The theoretical fuel cell efficiency,
tial function of the current and thus becomes a limiting defined as a ratio between produced electricity and
factor at high current densities. Fig. 3 shows a typical higher heating value of hydrogen consumed is therefore
fuel cell polarization curve with pronounced regions of 83%. The lower heating value of hydrogen results in an
predominant irreversible losses. Fig. 4 shows actual efficiency of 98%. Since the actual voltage of an opera-
polarization curves of some representative fuel cells. tional fuel cell is lower than the reversible potential,
The fuel cells are typically operated in a range between the fuel cell efficiency is always lower than the theoretical
0.6 and 0.8 V. The Space Shuttle fuel cell (alkaline) is de- one. Generally, the fuel cell efficiency is a product of sev-
signed to operate at 0.86 V and 410 mA/cm2. PEM fuel eral efficiencies:
cells have the highest achievable current densities, be-
gFC ¼ gTh gV gF gU ð3Þ
tween 1 and 2 mA/cm2 at 0.6 V with pressurized hydro-
gen and air. where gTh = thermal efficiency, i.e., ratio between Gibbs
free energy of the reaction and heating value of the fuel,
DGr/DHfuel (similar to internal combustion engines, the
load
fuel cell efficiency is often expressed in terms of lower
heating value), gV = voltage efficiency, defined as a ratio
e- between actual voltage (V) and thermodynamic voltage
(E), i.e., V/E, gF = Faradaic efficiency, or ratio between
depleted fuel and depleted oxidant and
product gases out product gases out the actual current and current corresponding to the rate
at which the reactant species are consumed, I/nFm,
H2 OH- O2
AFC where m is the rate (in moles/s) at which the reactants
H2O H2O are consumed, gU = fuel utilization, or ratio between
H2 O2 PEMFC the amount of fuel actually consumed in the electro-
H+
H2O PAFC chemical reaction and fuel supplied to the fuel cell.
H2 For a hydrogen/oxygen or hydrogen/air fuel cell
CO3= O2 MCFC
CO2 operating with 100% fuel utilization, the efficiency is a
CO2
H2O function of cell voltage only. For such a fuel cell the effi-
H2 O= O2
SOFC
ciency in an operating range between 0.6 V and 0.8 V is
H2O between 0.48 and 0.64 (Fig. 5).
In order to get useable voltages (i.e., tens or hundred
fuel in oxidant in Volts), the cells are combined in a stack. The cells are
physically separated from each other and electrically
anode electrolyte cathode
connected in series by a bipolar separator plate. Fig. 6
shows a schematic representation of a typical fuel cell
Fig. 3. Operating principle of various types of fuel cells. stack.
654 S.A. Sherif et al. / Solar Energy 78 (2005) 647–660
Alkaline fuel cells have been used in the space pro- 1.2. Hydrogen safety
gram (Apollo and Space Shuttle) since 1960s. Phospho-
ric acid fuel cells are already commercially available in Hydrogen poses risks if not properly handled or con-
container packages for stationary electricity generation. trolled. The risk of hydrogen must be considered relative
High temperature fuel cells, such as molten carbonate to the common fuels such as gasoline, propane or natu-
and solid oxide fuel cells, have been developed to a pre- ral gas. The specific physical characteristics of hydro-
commercial/demonstration stage for stationary power gen are quite different from those common fuels. Some
S.A. Sherif et al. / Solar Energy 78 (2005) 647–660 655
of those properties make hydrogen potentially less Hydrogen has a flame velocity seven times faster than
hazardous, while other hydrogen characteristics could that of natural gas or gasoline. A hydrogen flame would
theoretically make it more dangerous in certain therefore be more likely to progress to a deflagration or
situations. even a detonation than other fuels. However, the likeli-
Since hydrogen has the smallest molecule it has a hood of a detonation depends in a complex manner on
greater tendency to escape through small openings than the exact fuel/air ratio, the temperature and particularly
other liquid or gaseous fuels. Based on properties of the geometry of the confined space. Hydrogen detona-
hydrogen such as density, viscosity and diffusion coeffi- tion in the open atmosphere is highly unlikely.
cient in air, the propensity of hydrogen to leak through The lower detonability fuel/air ratio for hydrogen is
holes or joints of low pressure fuel lines may be only 13–18%, which is two times higher than that of natural
1.26–2.8 times faster than a natural gas leak through gas and 12 times higher than that of gasoline. Since
the same hole (and not 3.8 times faster as frequently as- the lower flammability limit is 4% an explosion is possi-
sumed based solely on diffusion coefficients). Experi- ble only under the most unusual scenarios, e.g., hydro-
ments have indicated that most leaks from residential gen would first have to accumulate and reach 13%
natural gas lines are laminar (Thomas, 1996). Since nat- concentration in a closed space without ignition, and
ural gas has over three times the energy density per unit only then an ignition source would have to be triggered.
volume the natural gas leak would result in more energy Should an explosion occur, hydrogen has the lowest
release than a hydrogen leak. explosive energy per unit stored energy in the fuel, and
For very large leaks from high-pressure storage a given volume of hydrogen would have 22 times less
tanks, the leak rate is limited by the sonic speed. Due explosive energy than the same volume filled with gaso-
to higher sonic velocity (1308 m/s) hydrogen would ini- line vapor.
tially escape much faster than natural gas (sonic velocity Hydrogen flame is nearly invisible, which may be
of natural gas is 449 m/s). Again, since natural gas has dangerous, because people in the vicinity of a hydrogen
more than three times the energy density than hydrogen, flame may not even know there is a fire. This may be
a natural gas leak will always contain more energy. remedied by adding some chemicals that will provide
Some high strength steels are prone to hydrogen the necessary luminosity. The low emissivity of hydro-
embrittlement. Prolonged exposure to hydrogen, partic- gen flames means that near-by materials and people will
ularly at high temperatures and pressures, can cause the be much less likely to ignite and/or hurt by radiant heat
steel to lose strength, eventually leading to failure. How- transfer. The fumes and soot from a gasoline fire pose a
ever, most other construction, tank and pipe materials risk to anyone inhaling the smoke, while hydrogen fires
are not prone to hydrogen embrittlement. Therefore, produce only water vapor (unless secondary materials
with proper choice of materials, hydrogen embrittlement begin to burn).
should not contribute to hydrogen safety risks. Liquid hydrogen presents another set of safety issues,
If a leak should occur for whatever reason, hydrogen such as risk of cold burns, and the increased duration of
will disperse much faster than any other fuel, thus reduc- leaked cryogenic fuel. A large spill of liquid hydrogen
ing the hazard levels. Hydrogen is both more buoyant has some characteristics of a gasoline spill, however it
and more diffusive than either gasoline, propane or nat- will dissipate much faster. Another potential danger is
ural gas. a violent explosion of a boiling liquid expanding vapor
Hydrogen/air mixture can burn in relatively wide vol- in case of a pressure relief valve failure.
ume ratios, between 4% and 75% of hydrogen in air. The In conclusion, hydrogen appears to pose risks of the
other fuels have much lower flammability ranges, viz., same order of magnitude as other fuels. In spite of pub-
natural gas 5.3–15%, propane 2.1–10%, and gasoline lic perception, in many aspects hydrogen is actually a
1–7.8%. However, the range has a little practical value. safer fuel than gasoline and natural gas. As a matter
In many actual leak situations the key parameter that of fact, hydrogen has a very good safety record, as a
determines if a leak would ignite is the lower flammabil- constituent of the ‘‘town gas’’ widely used in Europe
ity limit, and hydrogenÕs lower flammability limit is four and USA in the 19th and early 20th century, as a com-
times higher than that of gasoline, 1.9 times higher than mercially used industrial gas, and as a fuel in space pro-
that of propane and slightly lower than that of natural grams. There have been accidents, but nothing that
gas. would characterize hydrogen as more dangerous than
Hydrogen has a very low ignition energy (0.02 mJ), other fuels.
about one order of magnitude lower than other fuels.
The ignition energy is a function of fuel/air ratio, and
for hydrogen it reaches minimum at about 25%-30%. 2. Wind energy and hydrogen
At the lower flammability limit hydrogen ignition energy
is comparable with that of natural gas (Swain and As discussed above hydrogen complements the
Swain, 1992). renewable energy sources. An energy system where
656 S.A. Sherif et al. / Solar Energy 78 (2005) 647–660
hydrogen is derived from renewable sources is self-suffi- uled energy is not delivered. These penalties, known as
cient, clean, and represents a permanent energy solution generation imbalance charges, are costing wind plant
for sustainable development. Wind power is one of the operators as much as $0.10/kW h of undelivered energy
renewable sources. In some locations today, wind is cost (OÕBryant, 2000).
competitive with conventional, fossil fuel, or nuclear-
generated electricity. It is the fastest growing renewable 2.1. Issues
energy sector with annual growth of 27%, which means
doubling the installed capacity every three years. There Wind electricity may be converted to hydrogen
is more than 25,000 MW of installed wind turbines through water electrolysis. Alkaline (KOH) electrolyzers
worldwide (till the end of 2001) (Swisher, 2001). Over are commonly used, but recently the electrolyzers with
the past 20 years the generating capacities of individual proton exchange membrane (PEM) are making progress
units have grown from 25 kW to about 2500 kW. With (Anderson et al., 2002). Although water electrolysis is a
better designs, resulting from better understanding of mature technology its use in conjunction with wind
the structural stresses that the wind turbines are sub- power has some specifics, which will be discussed below.
jected to, the reliability of wind turbines has dramati-
cally improved over the last 20 years. Capital cost of 2.1.1. Intermittent operation
todayÕs wind turbines is less than $1000/kW, and elec- Direct coupling of an electrolyzer with a wind turbine
tricity costs as low as 4 cents per kW h in areas with implies intermittent operation with a highly variable
good wind resources (Baldwin, 2002). Prototype tur- power output (as shown in Fig. 7). The problem, partic-
bines are exceeding capacities of 5 MW. ularly with alkaline electrolyzers, is that at very low
One of the inherent drawbacks of wind power is that loads the rate at which hydrogen and oxygen are pro-
the wind velocity is highly intermittent, on a second by duced (which is proportional to current density) may
second basis, as well as hourly, daily, and even season- be lower than the rate at which these gases permeate
ally. Fig. 7 shows typical fluctuations in wind velocity through the electrolyte, and mix with each other. This
and power over a four-day period (Dutton et al., may create hazardous conditions inside the electrolyzer.
2000). It is generally accepted that, although power from Hydrogen flammability limits in oxygen are between
the wind turbine or wind-farm fluctuates significantly 4.6% and 93.9%, but the alarms and automatic shut-
with time, the grid can readily absorb most of the wind down of the electrolyzer are set at much safer concentra-
power produced, so long as wind power is less than 20% tions. This is more pronounced in the alkaline than in
of the maximum load (Paynter et al., 1991). Because of a PEM electrolyzers. Hydrogen permeation rate at 80 C
highly intermittent nature of its source, wind turbines through Nafion 117, typically used in PEM electrolyzers,
operate with relatively low capacity factor, which in a should be less than 1.25 · 104 cm3/s/cm2 at atmospheric
good location (with a high mean wind velocity) with pressure, corresponding to a current density of 0.002 A/
an efficient wind turbine may reach 33–38%, but rarely cm2 (Kocha et al., 2002), which is rather negligible com-
exceeds 40%. In some regions, the transmission provid- pared to 1 A/cm2, a typical current density in PEM elec-
ers see wind as a generating source that cannot be called trolyzers at full power. Oxygen permeation rate through
up on demand, and impose hefty penalties when sched- hydrated Nafion membrane is considerably lower (Sakai
et al., 1986).
Another problem related to operation with a highly
variable power source is thermal management. The elec-
trolyzer takes time to reach its normal operating temper-
ature, but due to intermittent operation it may operate
most of the time at a temperature below nominal, which
results in a lower efficiency.
2.1.2. Efficiency
The efficiency of an electrolyzer is inversely propor-
tional to the cell potential, which in turn is determined
by the current density, and that in turn directly corre-
sponds to the rate of hydrogen production per unit of
electrode active area. A higher voltage would result in
more hydrogen production, but at a lower efficiency.
Typically, cell voltage is selected at about 2 V, but a
lower nominal voltage (as low as 1.6 V) may be selected,
Fig. 7. An example of wind speed and power time series (rated if the efficiency is more important than size (and capital
power 10 kW) (Dutton et al., 2000). cost) of the electrolyzer.
S.A. Sherif et al. / Solar Energy 78 (2005) 647–660 657
Fig. 12. Schematic diagram of grid independent integrated wind-hydrogen energy system.
Dillon, A.C., Jones, K.M., Heben, M.J., 1996. Carbon nano- Series E-295. Kluwer Academic Publishers, Dordrecht, The
tube materials for hydrogen storage. In: Proc. 1996 US Netherlands, pp. 181–194.
DOE Hydrogen Program Review, vol. II, National Renew- Pottier, J.D., Blondin, E., 1995. Mass Storage of Hydrogen. In:
able Energy Laboratory (NREL), Golden, CO, pp. 747– Yurum, Y. (Ed.), Hydrogen Energy System, Utilization of
763. Hydrogen and Future Aspects, NATO ASI Series E-295.
Dutton, A.G., Bleijs, J.A.M., Dienhart, H., Falchetta, M., Hug, Kluwer Academic Publishers, Dordrecht, The Netherlands,
D., Prischich, D., Ruddell, A.J., 2000. Experience in the pp. 167–180.
design, sizing, economics, and implementation of autono- Rambach, G., Hendriks, C., 1996. Hydrogen Transport and
mous wind-powered hydrogen production systems. Int. J. Storage in Engineered Glass Microspheres. In: Proc. 1996
Hydrogen Energy 25, 705–722. US DOE Hydrogen Program Review, vol. II, National
Goswami, D.Y., Ingley, H.A., Goel, N., Mirabal, S.T., 2003. Renewable Energy Laboratory (NREL), Golden, CO, pp.
Hydrogen production. In: Goswami, D.Y. (Ed.), Advances 765–772.
in Solar Energy, vol. 15, pp. 405–458. Sakai, T., Takenaka, H., Torikai, E., 1986. Gas diffusion in
Jensen, C., 1996. Hydrogen storage via polyhydride complexes. dried and hydrated nafions. J. Electrochem. Soc. 133 (1),
In: Proc. 1996 US DOE Hydrogen Program Review, vol. II, 88–92.
National Renewable Energy Laboratory (NREL), Golden, Schwartz, J.A., Amankwah, K.A.G., 1993. Hydrogen storage
CO, pp. 787–794. systems. In: Howell, D.G. (Ed.), The future of energy gases.
Keith, G., Leighty, W., 2002. Transmitting 4000 MW of New US Geological Survey Professional Paper 1570, US Gov-
Windpower from N. Dakota to Chicago: New HVDC ernment Printing Office, Washington, pp. 725–736.
Electric Lines or Hydrogen Pipeline. In: Proc. 14th World Spinadel, E., Gil, J., Gamallo, F., 2002. Wind-powered
Hydrogen Energy Conference, Montreal, Canada, June. electrolysis for stand alone operation: An approach to a
Kinoshita, K., McLarnon, F.R., Cairns, E.J., 1988. Fuel Cells: specific wind converter design. In: Proc. 14th World
A Handbook, US Department of Energy, DOE/METC88/ Hydrogen Energy Conference, Montreal, Canada, June.
6069, Morgantown, WV. Steeb, H., Brinner, A., Bubmann, H., Seeger, W., 1990.
Kocha, S., Plasse, P., Onishi, L, Wheeler, D., Bett, J., 2002. In- Operation experience of a 10 kW PV-electrolysis system in
situ electrochemical measurement of hydrogen crossover different power matching modes. In: Veziroglu, T.N.,
currents in PEMFC. In: Proc. Fuel Cell Technology: Takahashi, P.K. (Eds.), Hydrogen Energy Progress VIII,
Opportunities and Challenges, 2002 AIChE Spring National vol. 2. Pergamon Press, New York, pp. 691–700.
Meeting, New Orleans, March, pp. 307–312. Steinberg, M., Cheng, H.C., 1988. Modern and prospective
Ledjeff, K., 1990. New hydrogen appliances. In: Veziroglu, technologies for hydrogen production from fossil fuels. In:
P.K., Takahashi, P.K. (Eds.), Hydrogen Energy Progress Veziroglu, T.N., Protsenko, A.N. (Eds.), Hydrogen Energy
VIII, vol. 3. Pergamon Press, New York, pp. 429–444. Progress VII, vol. 2. Pergamon Press, Oxford, pp. 699–
Mitlitsky, F., 1996. Development of an advanced, lightweight, 740.
high pressure storage tank for on-board storage of com- Sternfeld, H.J., Heinrich, P., 1989. A demonstration plant for
pressed hydrogen. In: Proc. Fuel Cells for Transportation the hydrogen/oxygen spinning reserve. Int. J. Hydrogen
TOPTEC, Alexandria, VA, SAE, Warrendale, PA. Energy 14, 703–716.
Muradov, N.Z., 1998. CO2 free production of hydrogen by Swain, M.R., Swain, M.N., 1992. A comparison of H2, CH4,
catalytic pyrolysis of hydrocarbon fuels. In: Bolcich, J.C., and C3H8 fuel leakage in residential settings. Int. J.
Veziroglu, T.N. (Eds.), Hydrogen Energy Progress XII, vol. Hydrogen Energy 17 (10), 807–815.
1, Int. Association for Hydrogen Energy, Coral Gables, FL, Swisher, R., 2001. Fundamentals of Wind Energy: Market
pp. 523–532. Overview. American Wind Energy Association, presented at
Nornbeck, J.M., Heffel, J.W., Durbin, T.D., Tabbara, B., Windpower 2001.
Bowden, J.M., Montano, M.C., 1996. Hydrogen fuel for Taylor, J.B., Anderson, J.E.A., Kalyanam, K.M., Lyle, A.B.,
surface transportation, SAE. Philips, L.A., 1986. Technical and economic assessment of
OÕBryant, M., 2000. Making variable supply legitimate. Wind- methods for the storage of large quantities of Hydrogen.
power Monthly (February), 40. Int. J. Hydrogen Energy 11 (1), 5–22.
Ogden, J.M., Williams, R.H., 1989. Solar Hydrogen: Moving Thomas, C.E., 1996. Preliminary hydrogen vehicle safety
beyond Fossil Fuels. World Resources Institute, Washing- report. The Ford Motor Company, Contract No. DE-
ton, DC. AC02-94CE50389, US Department of Energy.
Öney, F., Veziroglu, T.N., Dülger, Z., 1994. Evaluation of Ventre, G., 2000. Florida Solar Energy Center, private
pipelines transportation of hydrogen and natural gas communication, May 2000.
mixtures. Int. J. Hydrogen Energy 19, 813–822. Veziroglu, T.N., Barbir, F., 1992. Hydrogen: The wonder fuel.
Paynter, R.J.H., Lipman, N.H., Foster, J.E., 1991. The Int. J. Hydrogen Energy 17 (6), 391–404.
potential of hydrogen and electricity production from wind Willner, I., Steinberger-Willner, B., 1988. Solar hydrogen
energy. Energy Research Unit, Rutherford Appleton Lab- production through photobiological, photo-chemical and
oratory, September. photoelectrochemical assemblies. Int. J. Hydrogen Energy
Pottier, J.D., 1995. Hydrogen transmission for future energy 13 (10), 593–604.
systems. In: Yurum, Y. (Ed.), Hydrogen Energy System, Yalcin, S., 1989. A review of nuclear hydrogen production. Int.
Utilization of Hydrogen and Future Aspects, NATO ASI J. Hydrogen Energy 14 (8), 551–561.