What Will Power the Hydrogen Economy?
Present and Future Sources of Hydrogen Energy
UCD- ITS- RR- 04- 10
July 12, 2004
Timothy E. Lipman
Final Report
Analysis and Report Prepared for
The Natural Resources Defense Council
Energy and Resources Group and
Institute of Transportation Studies
University of California - Berkeley
and
Institute of Transportation Studies
University of California - Davis
Report Publication:
Institute of Transportation Studies – Davis
One Shields Ave.
University of California
Davis, CA 95616
http:// www. its. ucdavis. edu/ publication. html
What Will Power the Hydrogen Economy?
i
Acknowledgments
This research has been funded by the Natural Resources Defense Council with additional
support from the UC Davis Hydrogen Pathways Program. I am appreciative of NRDC’s
timely support for this study, and of Antonia Herzog’s great assistance with this project.
I thank Lea Yancey for assisting in research on the hydrogen production from biomass
aspect of the report. I also thank Joan Ogden and Anthony Eggert for their comments
and assistance, and add special thanks to Jim Ohi and the National Renewable Energy
Laboratory ( NREL) for allowing us to include the renewable energy potential maps that
NREL has generated.
The author alone is responsible for the contents of this report.
What Will Power the Hydrogen Economy?
ii
Table Of Contents
Abbreviations and Acronyms......................................................................................... iii
Executive Summary .................................................................................................... v
I. Introduction ............................................................................................................. 1
II. Hydrogen Production Methods ............................................................................... 3
III. Hydrogen Distribution and Supply Systems......................................................... 28
IV. Hydrogen from Hydrocarbons and Carbon Sequestration .................................... 34
V. Environmental Impacts of Hydrogen Production Methods .................................... 35
VI. Renewable Hydrogen Potential and Regional Production Strategies .................... 38
VII. Conclusions ....................................................................................................... 41
References................................................................................................................. 47
Appendix A - Summary Tables of Hydrogen Production and Delivered Costs.......... A- 1
What Will Power the Hydrogen Economy?
iii
Abbreviations and Acronyms
AC = alternating current
ATR = autothermal reforming
C. = Celsius
CH4 = methane
CHP = combined heat and power
CNG = compressed natural gas
CO = carbon monoxide
CO2 = carbon dioxide
DC = direct current
DOE = U. S. Department of Energy
DTI = Directed Technologies Inc.
EIA = Energy Information Administration
EJ = exajoule or exajoules
EV = electric vehicle
FCV = fuel cell electric vehicle
FERCO = Future Energy Resources Corporation
ft2 = square foot or feet
GHG = greenhouse gas
GJ = gigajoule or gigajoules
GREET = Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation
HHV = higher heating value
HI = hydrogen iodide
H2 = hydrogen
H2SO4 = sulfuric acid
I2 = iodine
ICE = internal- combustion engine
IBGFC = integrated biomass gasification fuel cell
IRR = internal rate of return
kg = kilogram or kilograms
kW = kilowatt or kilowatts
LEM = lifecycle emissions model
LHV = lower heating value
m2 = square meter or meters
MCFC = molten carbonate fuel cell
MSW = municipal solid waste
MW = megawatt or megawatts
N2 = nitrogen
NOx = oxides of nitrogen
NM3 = normal cubic meter or meters
NRC = National Research Council
NREL = National Renewable Energy Laboratory
PEC = photo- electrochemical
What Will Power the Hydrogen Economy?
iv
Abbreviations and Acronyms ( cont’d)
PEM = proton- exchange membrane
PM = fine particulate matter
POx = partial- oxidation PSA = pressure swing adsorption
psi = pounds per square inch
PV = photovoltaic
Quad = quadrillion British thermal units
R& D = research and development
RFG = reformulated gasoline
scf = standard cubic foot or standard cubic feet
SMR = steam methane reforming
SO2 = sulfur dioxide
SOFC = solid oxide fuel cell
U. S. = Unites States
VOC = volatile organic compounds
WGS = water gas shift reaction
What Will Power the Hydrogen Economy?
v
Executive Summary
Hydrogen is an abundant element that can be produced in molecular form from many different
sources, and in many different ways. In the context of energy systems, hydrogen is best thought
of as an energy carrier, more akin to electricity than the fossil fuels that we extract from the
earth’s crust. Hydrogen can be produced from any “ hydrocarbon” fuel because by definition
these fuels contain both hydrogen and carbon. Hydrogen can also be produced from various
biological materials and from water. This “ water splitting” process is called electrolysis, and it is
the oldest electrochemical process known. Thought of in this way, hydrogen supplies are
virtually unlimited – unlike fossil fuels that someday will run out.
Potential Sources of Hydrogen
The main hydrogen production options currently known are as follows, including a short
technical and economic characterization of each production source.
Steam Methane Reforming
Steam reformation of natural gas ( or methane from other sources) produces a
hydrogen rich gas that is typically on the order of 70- 75% on a dry basis, along with
smaller amounts of methane ( 2- 6%), carbon monoxide ( 7- 10%), and carbon dioxide
( 6- 14%). Costs of hydrogen from steam methane reforming vary with feedstock cost,
scale of production, and other variables and range from about $ 2- 5 per kilogram at
present ( delivered and stored at high pressure). Delivered costs as low as about $ 1.60
per kilogram are believed to be possible in the future based on large centralized
production and pipeline delivery, and delivered costs for small- scale decentralized
production are projected to be on the order of $ 2.00- 2.50 per kilogram.
Gasification of Coal and Other Hydrocarbons
In the partial oxidation ( POx) process, also known more generally as “ gasification,”
hydrogen can be produced from a range of hydrocarbon fuels, including coal, heavy
residual oils, and other low- value refinery products. The hydrocarbon fuel is reacted
with oxygen in a less than stoichiometric ratio, yielding a mixture of carbon
monoxide and hydrogen at 1200° to 1350° C. Hydrogen can be produced from coal
gasification at delivered costs of about $ 2.00- 2.50 per kilogram at present at large
scale, with delivered costs as low as about $ 1.50 per kilogram believed to be possible
in the future.
Nuclear- Based Options
Various nuclear energy based hydrogen production schemes are possible, including
nuclear thermal conversion of water using various chemical processes such as the
sodium- iodine cycle, electrolysis of water using nuclear power, and high- temperature
electrolysis that additionally would use nuclear system waste heat to lower the
electricity required for electrolysis. Few cost studies of these schemes have yet been
conducted, but at large scale and in the future, nuclear thermal conversion of water is
believed to be capable of producing delivered hydrogen at costs of about $ 2.33 per
kilogram.
What Will Power the Hydrogen Economy?
vi
Electrolysis of Water
Hydrogen can be produced via electrolysis of water from any electrical source,
including utility grid power, solar photovoltaic ( PV), wind power, hydropower,
nuclear power, etc. Grid power electrolysis in the United States ( U. S.) would
produce hydrogen at delivered costs of $ 6- 7 per kilogram at present, with future
potential of about $ 4 per kilogram. Wind electrolysis- derived hydrogen would cost
about $ 7- 11 per kilogram at present, with future potential of delivered costs as low as
below $ 3 per kilogram. Solar hydrogen would be more expensive, on the order of
$ 10- 30 per kilogram at present, with future delivered costs of $ 3- 4 per kilogram
estimated to be possible.
Hydrogen from Biomass
Biomass conversion technologies can be divided into thermo- chemical and
biochemical processes. Thermo- chemical processes tend to be less expensive because
they can be operated at higher temperatures and therefore obtain higher reaction rates.
They also can utilize a broad range of biomass types. In contrast, biochemical
processes are limited to wet feedstock and sugar- based feedstocks. At medium
production scale and liquid distribution by tanker truck, current delivered costs of
hydrogen from biomass would be in the $ 5- 7 per kilogram range. However at larger
production scales and coupled with pipeline delivery, delivered costs as low as $ 1.50
to $ 3.50 per kilogram are believed possible. Pyrolysis of biomass, another production
option, also offers potentially low costs of delivered hydrogen, with costs as low as
about $ 1 per kilogram possible with large- scale production and pipeline delivery.
Other Hydrogen Production Options
Hydrogen can also be produced through various other methods, such as direct solar
thermal dissociation of water, from municipal solid waste “ landfill gas” and waste
gases from water treatment plants, and from hydrogen producing algae.
Figures ES- 1 and ES- 2 present ranges in hydrogen production and delivered hydrogen costs from
the technical literature. These results are directly taken from various studies, and have not been
adjusted for different assumptions in the studies ( with regard to interest rates, feedstock costs,
etc.) to make them more directly comparable.
What Will Power the Hydrogen Economy?
vii
Figure ES- 1: Ranges in Onsite Hydrogen Production Cost Estimates
0
1
2
3
4
5
6
7
8
9
10
11
12
NG SMR large
NG SMR small
Hydrogen energy station
Coal gasification large
Wind electrolysis current
Solar electrolysis current
Biomass gasification medium
NG SMR large w/ CO2 seq.
NG solar thermal
Coal gas. large w/ CO2 seq.
Nuclear electrolysis
Wind electrolysis future
Solar electrolysis future
Biomass gasification large
Biomass pyrolysis large
Hydrogen production cost ($/ kg)
High
Low
28.2
? ?
Nearer Term Technologies Longer Term Technologies
Note: Various sources – see Table A- 1 for details.
NG = natural gas; SMR = steam methane reforming; “?” = Costs of effective carbon sequestration from fossil fuels
are uncertain because sequestration technologies and methods are still in the R& D phase.
What Will Power the Hydrogen Economy?
viii
Figure ES- 2: Ranges in Delivered Hydrogen Cost Estimates
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
NG SMR med- large
NG SMR small ( onsite)
Hydrogen energy station
Coal gasification large
Wind electrolysis current
Solar electrolysis current
Biomass gasification med
NG SMR large w/ CO2 seq.
Coal gas. large w/ CO2 seq
Wind electrolysis future
Solar electrolysis future
Biomass gasification large
Biomass pyrolysis large
Delivered H2 Cost ($/ kg)
High
Low
28.19
Nearer Term Technologies Longer Term Technologies
Notes: Various sources - see Appendix for details. The ranges shown are taken from many different sources,
including those with assumptions that may be somewhat inconsistent with regard to production scale, interest rates,
etc. Wider and narrower ranges between high and low costs thus tend to reflect the relative numbers of studies for
each pathway, rather than inherent uncertainties in costs for each pathway.
Environmental Impacts of Hydrogen Production
In addition to the economics of production and distribution, additional important considerations
for hydrogen production methods include the environmental implications of various hydrogen
production methods. These include greenhouse gas ( GHG) emissions, local pollutant emissions,
soil and water emissions, and land, water, and other non- feedstock resource requirements.
In general, the GHG and air pollutant impacts of various hydrogen production pathways have
been reasonably well- studied, at least for the most prominent potential production pathways, but
other environmental considerations have been less well characterized. Additional studies are
therefore desirable, both to more fully characterize the potential environmental impacts of
hydrogen production in general, and to more carefully examine the environmental impacts of
hydrogen production for specific regions as these impacts will vary regionally to some extent.
Figure ES- 3 presents estimates of full fuel- cycle GHG emissions from various hydrogen
production and distribution pathways, for hydrogen used in fuel cell vehicles ( FCVs) relative to
the GHG emissions of conventional vehicles running on reformulated gasoline. As shown in the
figure, the GHG emissions associated with the production and use of hydrogen for vehicles can
vary greatly depending on the production method.
What Will Power the Hydrogen Economy?
ix
Figure ES- 3: Relative Fuel- Cycle Greenhouse Gas Emissions of Hydrogen Fuel Pathways
- 120
- 100
- 80
- 60
- 40
- 20
0
20
40
60
80
100
NG central - G H2
NG central - L H2
NG central w/ steam - G H2
NG central w/ steam - L H2
NG station - G H2
NG station - L H2
Electrol. US mix - G H2
Electrol. US mix - L H2
Electrol. CCNG - G H2
Electrol. CCNG - L H2
Electrol. Nuclear - G H2
Electrol. Nuclear - L H2
Electrol. Solar PV - G H2
Electrol. Solar PV - L H2
MeOH - Reformer
EtOH Corn - Reformer
EtOH Cellulosic - Reformer
GHGs % Change from Convetional RFG
GREET 1.6
LEM 2003
Notes: GREET 1.6 is the Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model. LEM
2003 is the Lifecycle Emission Model. CCNG = combined cycle natural gas power plant; EtOH = ethanol; G =
gaseous; L = liquid; NG = natural gas; MeOH = methanol; PV = photovoltaics; RFG = reformulated gasoline.
Renewable Hydrogen Potential in the U. S.
The overall technical potential for producing hydrogen from renewable sources in the U. S. is
great, with estimated potential of renewable hydrogen energy of more than the current petroleum
consumption in the U. S. for transportation. However, much of this renewable hydrogen
production potential is based on wind and solar resources that are unlikely to become
economically competitive with other sources, except perhaps in the long term. The most
economical sources of renewable hydrogen include dedicated energy crops, municipal solid
waste, agricultural and livestock residues, and certain wind resources that are located close
enough to demand centers to be relatively attractive.
Figure ES- 4 presents a hypothetical regional demand and supply picture for renewable hydrogen
for the year 2040. This scenario assumes that hydrogen demand in the U. S. reaches 10 Quads
( 10.5 exajoules) by this time, and that the demand in each U. S. state is proportional to current
gasoline consumption ( first set of bars). The figure shows three potential levels of renewables-based
hydrogen supply. The second set of bars – “ Renewable H2 ( economic)” -- shows the
amount of hydrogen that could be dispensed to consumers at a cost of $ 3.00 per kilogram or less
($ 21.30 per gigajoule and $ 1.50 per gallon of gasoline equivalent on a per- mile basis in 2x
What Will Power the Hydrogen Economy?
x
efficiency hydrogen vehicles). Some hydrogen in the Northeast and Southeast is supplied from
other nearby regions in this scenario. The third set of bars shows the total hydrogen demand in
each region, totaling 10 Quads, being met by renewable sources. The fourth set of bars –
“ Renewable H2 ( potential)” – shows the total technical potential of nearly 32 Quads of
production, with much of this potential in the Midwest and Central regions.
Figure ES- 4: Hypothetical Year 2040 Regional U. S. Hydrogen Demand of 10 Quads Per
Year and Renewable Production Potential
0.0
1.0
2.0
3.0
4.0
Northeast Southeast Midwest Central Pacific
H2 Demand/ Supply ( Quads/ year)
H2 Demand Renewable H2 ( economic) Renewable H2 ( 10 Quads) Renewable H2 ( potential)
14.83 14.19
Notes: “ Northeast” includes CT, ME, MA, NH, NJ, NY, PA, RI, and VT; “ Southeast” includes AL, DE, FL, GA,
KY, MD, MI, NC, SC, TN VA, and WV; “ Midwest” includes IA, IL, IN, KS, MI, MN, MO, NB, ND, OH, SD, and
WI; “ Central” includes AR, AZ, CO, ID, LA, MT, NV, NM, OK, TX, UT, and WY; “ Pacific” includes AK, CA, HI,
OR, and WA. However, the analysis did not consider production potential from AK and HI, and only CA, OR, and
WA production potentials are shown. “ Economic” renewable hydrogen potential includes resources that can be
used to produce delivered hydrogen to customers for $ 3.00 per kg or less.
Source: Meyers et al. ( 2003) and additional analysis conducted by the author.
Conclusions
In conclusion, hydrogen can be produced in a variety of different ways, from a large number of
potential feedstocks. Unlike the crude oil used to produce gasoline, the myriad potential sources
of hydrogen are generally well distributed around the U. S. and the world. Certain regions have
greater or lesser availability and suitability for specific hydrogen feedstocks and production
methods, but virtually all regions have at least a few and many have several possibilities for
hydrogen production.
The feedstock diversity for hydrogen is a critical characteristic in that it suggests the potential for
What Will Power the Hydrogen Economy?
xi
much greater reliance on local resources for energy production around the world, particularly
with regard to petroleum use in the transportation sector ( if practical hydrogen vehicles can be
realized). The diversity also raises another critical issue in that it suggests that in places such as
the U. S., where multiple hydrogen production options are possible, strategic decisions about
hydrogen production can be made based on multi- criteria analysis of the costs and benefits of
various options.
Based on this present and future hydrogen production and distribution assessment, the following
recommendations are made with regard to potential future directions for hydrogen energy
research and development.
Hydrogen Production:
Hydrogen from Coal – This pathway is potentially economically attractive,
especially without inclusion of externalities, but entails inherently large
greenhouse gas, air pollutant, and mining impacts. Carbon sequestration is a
potentially promising long- term option to mitigate GHG impacts, but at present
has large uncertainties with regard to cost and ultimate effectiveness.
Hydrogen from Natural Gas – This pathway is a front- runner as a transition fuel
in many areas, but faces distribution and supply concerns, as well as cost issues
with natural gas in the U. S. presently at ~ 2x historical cost levels. Natural gas is
probably best thought of as a transition fuel to cleaner and more renewable
hydrogen production options, and should be utilized near- term due to the
extensive natural gas infrastructure -- but in the context of a longer- term hydrogen
production strategy and cognizant of technological “ lock- in” concerns.
Distributed production of hydrogen from natural gas appears to be among the
most attractive near- term options, as it obviates the need for transportation of
hydrogen, but also precludes the possibility of carbon capture and sequestration.
Hydrogen from Nuclear Energy– This pathway faces critical and perhaps even
overwhelming ( in the U. S.) public acceptance, safety, and access to capital issues.
Costs could be competitive at production level but, like other centralized
production options, hydrogen production at large- scale nuclear plants implies high
distribution costs.
Hydrogen from Renewable Energy Sources – This is the only fully sustainable
“ end- game.” Renewable energy- based hydrogen offers multiple benefits
including fuel diversity and security, and clean and low- GHG hydrogen
production. Strategies that emphasize renewable feedstocks and distributed
production ( to avoid distribution costs and energy penalties), and that are
developed in the context of a regional resource base, may be critical to
maximizing the potential benefits of transitioning to hydrogen. Electrolyzer and
solar PV costs remain key issues for the electrolysis- based options, but hydrogen
from biomass using a wide range of potential renewable feedstocks offers another
attractive set of options.
What Will Power the Hydrogen Economy?
xii
Infrastructure Development:
Distributed vs. Centralized Production – In the early years of the hydrogen
transition, a mix of distributed production with small- scale reformers and
electrolyzers, coupled with some “ mobile vehicle refueler” and liquid hydrogen
delivery by truck, appears to be the best strategy. Once hydrogen demand has
reached the scale and level of concentration to begin to justify pipeline delivery,
then the development of larger- scale plants can be pursued, potentially with
carbon sequestration.
Integration of Hydrogen Infrastructure and Distributed Power Systems – The
potential for expanded use of hydrogen is relevant to both the transportation and
stationary power sectors, and as a result interesting synergies are possible. These
include co- production of electricity and hydrogen for vehicles with “ hydrogen
energy stations,” using distributed power generation to provide electricity for
hydrogen production through electrolysis, and even using hydrogen- powered
vehicles for backup power, peak power, and/ or grid ancillary services such as
spinning and non- spinning reserves.
Distribution Systems and Onboard Vehicle Storage – As highlighted by the recent
National Research Council ( 2004) study, further progress with regard to onboard
hydrogen storage is required in order for hydrogen- powered vehicles to become
more practical. While hydrogen storage as a compressed gas is the dominant
onboard storage system at present, this type of system suffers from bulkiness,
significant electricity requirements for hydrogen compression, and potential
safety concerns due to the high storage pressures. Other onboard storage systems
such as those based on metal and alkaline hydrides, cryogenic liquid hydrogen,
and carbon nanotubes are under development.
Policy / Institutional Roles and Priorities:
Incentives for “ Clean” Hydrogen Production – Given the many possible ways of
making hydrogen, and the widely varying environmental and social impacts that
these pathways imply, federal, state, and local governments should consider
incentivizing the production of hydrogen in ways that minimize social costs.
These incentives could be in the form of reduced or eliminated taxes on hydrogen
produced from renewables and other clean and “ climate- friendly” sources, and tax
breaks on investments in hydrogen production capital equipment.
Role of Government Agencies and the Private Sector – In addition to potentially
providing incentives for the development of clean hydrogen production systems
and distribution infrastructure, the public sector can play several other important
roles. These include public education and outreach on hydrogen safety issues,
development and implementation of codes and standards for hydrogen
infrastructure, support for hydrogen systems research and development,
What Will Power the Hydrogen Economy?
xiii
incorporating hydrogen vehicles into government fleets where practical, and
potentially using other government assets to assist in hydrogen systems
development ( e. g. allowing roadside “ rights- of- way” to be used for hydrogen
stations and pipelines).
In conclusion, hydrogen is a highly promising energy carrier and fuel for vehicles and stationary
power generation, but the potential expanded use of hydrogen involves a host of issues and
challenges. Primary among these is the fundamental issue of how the hydrogen is itself
produced and distributed. One of hydrogen’s chief advantages – the ability to be made in
various ways and with a diverse array of feedstocks – also complicates decision- making with
regard to planning for the development of hydrogen production and distribution infrastructure.
What Will Power the Hydrogen Economy?
1
I. Introduction
Hydrogen is the most abundant element in the known universe. The hydrogen molecule ( H2) is a
colorless gas at room temperature, and being light relative to other gases, it moves quickly unless
contained and disperses rapidly upward through and then out of the atmosphere. As well as
being produced “ artificially” for various commercial and industrial uses, hydrogen is formed
through various natural processes on earth. Since it readily rises through and escapes from the
atmosphere, it only tends to persist underground where it is formed by the action of bacteria on
ancient vegetable and animal remains. In fact, surprising recent research suggests that the
underground stores of hydrogen may be much more considerable than previously thought
( Freund et al., 2002). However, the economic and environmental costs of attempting to make
use of these underground stores of hydrogen are likely to be considerable, and probably
unattractive due to the deep underground and highly diffuse nature of the resource.
In the context of energy systems, hydrogen is best thought of as an energy carrier, more akin to
electricity than the fossil fuels that we extract from the earth’s crust. Hydrogen can be produced
from any “ hydrocarbon” fuel because by definition these fuels contain both hydrogen and
carbon. Hydrogen can also be produced from various biological materials and from water. This
“ water splitting” process is called electrolysis, and it is the oldest electrochemical process
known. Thought of in this way, hydrogen supplies are virtually unlimited – unlike fossil fuels
that someday will run out.
Hydrogen has been considered as a fuel for many years, but over the past ten to fifteen years
advances in fuel cell technology have spurred an enormous wave of interest in hydrogen. Fuel
cells have the remarkable ability to convert hydrogen to electricity efficiently by using special
electrolyte membrane materials and an electrochemical rather than combustion process. Over
the past decade or so, billions of dollars have been spent on research and development of
hydrogen technologies, mainly in the private sector but also among government research labs
and programs. These investments have spurred significant technological improvements, but
have also revealed the many difficulties of designing fuel cell and hydrogen storage and
dispensing systems that are practical, cost- effective, and safe to operate.
Fuel cells are being considered for a wide range of potential applications, from very small- scale
portable electronics at the level of watts or even milliwatts, all the way up to large powerplants in
sizes of one or more megawatts ( MW). The application that has garnered the most attention is
the prospect of incorporating fuel cell systems into electric vehicles ( EVs), meaning that they
could operate on hydrogen rather than by recharging their batteries. Particularly in recent
months, following the announcement of the Bush Administration of plans to spend $ 1.2 billion
over the next 5 years on the development of fuel cell vehicles ( FCVs) under the FreedomCAR
program, the merits of FCVs relative to other options for reducing pollution and oil consumption
have been hotly debated. Some analysts argue that the emphasis on FCVs and other hydrogen-powered
vehicles is well placed, given their tremendous potential to simultaneously address
several concerns related to energy use in the transportation sector. These include air pollutant
emissions, emissions of greenhouse gases ( GHGs), issues of oil dependency, and even the
prospects for reducing vehicle noise.
What Will Power the Hydrogen Economy?
2
Meanwhile, others argue that FCVs are a chancy proposition that may not “ pan out” and that
efforts instead should focus on or at least be complemented by emphasizing more conventional
technologies with a higher chance of success and with a more immediate impact. These include
technologies such as hybrid EVs that run on gasoline but that can be made more efficient by also
incorporating an electric motor and battery system that works in conjunction with the gasoline
engine ( Doniger et al., 2002; Weiss et al., 2003). Other hydrogen- based concepts include
hydrogen combustion engines for internal combustion engine ( ICE) vehicles and hybrid- electric
vehicles, hydrogen engines and/ or turbines for heavy- duty transportation applications ( fork- lifts,
maritime vessels, etc.) where they can substitute for diesel engines, and hydrogen gas turbines
for stationary applications.
Options for using hydrogen cleanly and efficiently thus exist apart from fuel cells, and use of
hydrogen as a fuel may well continue to expand even in scenarios where fuel cells do not
ultimately become widely adopted, or where they perhaps operate on natural gas and methanol to
a large extent ( for example, solid- oxide and direct- methanol fuel cells respectively), rather than
on hydrogen directly. Hydrogen used in this way could lead to development of a hydrogen
distribution infrastructure in advance of, or perhaps even in place of, the success of fuel cells in
some market segments, thereby helping to lead the development of the hydrogen economy in a
complementary fashion.
This report considers the important question of where the hydrogen will come from to operate
these various types of fuel cells and combustion systems that use hydrogen directly. At present,
natural gas is the leading “ feedstock” for hydrogen production, but hydrogen can be produced in
myriad ways from a large number of potential sources. Since natural gas is a non- renewable
resource, one that is not evenly distributed geographically, and that also is subject to market
price fluctuations and supply limitations, efforts to consider alternative sources of hydrogen
production are warranted. Also, unless and until efforts are successful to remove the carbon
dioxide ( CO2) and other GHGs from the hydrogen production process, hydrogen production or
“ reforming” from natural gas releases significant quantities of CO2 and methane ( CH4), thus
potentially contributing to climate change. These GHG emissions may be less than those
associated with conventional fuel- based baseline but they may still be significant cases ( such as
with hydrogen FCVs vs. gasoline vehicles -- see Section V below for GHG emissions estimates
for various hydrogen production pathways).
Technologies are evolving rapidly for hydrogen production from renewable and other sources,
including electrolyzers, small- scale reformers, and biomass and coal gasification systems. As
these other technologies begin to become competitive with natural gas reforming, as they may in
due course, they can complement the use of natural gas as a hydrogen source and add fuel
diversity and insulation against price fluctuations and supply limitations. Careful consideration
of the relative costs and benefits of various hydrogen pathways ( including environmental and
human health impacts of lifecycle pollutant emissions) are thus important so that hydrogen can
be produced in the most benign ways from an overall social perspective.
This report first considers various methods for producing hydrogen, including discussion of the
production potential for the United States ( U. S.). Next, hydrogen distribution and delivery
What Will Power the Hydrogen Economy?
3
options are briefly summarized and discussed, including discussion of the environmental
pollutant emissions implications and land and water use impacts of different production options.
Also discussed are potential regional hydrogen production strategies around the U. S., and
potential for regional hydrogen production from renewable sources. Finally, policy implications
are discussed in the conclusions section, in light of the hydrogen production option costs and
benefits that are reviewed and described in the report.
II. Hydrogen Production Methods
Unlike most conventional fuels that are found, extracted from the earth’s crust, and then refined,
hydrogen is not found but rather is created from other primary fuels and feedstocks. There are
believed to be significant quantities of hydrogen underground, formed by bacterial activity, but
these stores of hydrogen are predominantly deep underground, most likely at depths of about 20
kilometers ( Freund et al., 2002), and therefore probably not economical to extract. Hydrogen
production is thus in some ways more akin to electricity production than the production of fossil
fuels, but compared with electricity hydrogen has the advantage of being more readily storable,
or “ packable” in pipelines.
Hydrogen production is largely of interest due to the use of hydrogen being electrochemically
converted into electricity in fuel cells with high efficiency, or of being combusted relatively
cleanly and efficiently. Some fuel cell types would require relatively high- grade hydrogen as
input ( e. g., proton exchange membrane [ PEM] fuel cells with relatively low carbon monoxide
[ CO] tolerance), while other high- temperature fuel cells could actually input natural gas directly.
The temperatures internal to these high temperature fuel cells ( such as solid oxide and molten
carbonate fuel cells -- SOFCs and MCFCs) are sufficient to thermally dissociate or “ internally
reform” the CH4 into hydrogen and carbon compounds. Thus, the “ hydrogen” requirements of
various fuel cell types are rather diverse, and only some types of fuel cells require hydrogen to be
produced prior to being introduced into the fuel cell stack.
At present, hydrogen is produced mainly from natural gas and oil, but production possibilities
involving other sources such as from coal and through electrolysis have garnered attention in
recent years. At present, about half of the 500 billion normal cubic meters ( Nm3) of hydrogen
produced globally on an annual basis comes from natural gas ( U. S. DOE, 2003). Table 1, below,
presents recent data on global hydrogen production from various sources.
What Will Power the Hydrogen Economy?
4
Table 1: Annual Worldwide Hydrogen Production Share by Source
Source Nm3 ( billions)/ Year Share
Natural gas 240 48%
Oil 150 30%
Coal 90 18%
Electrolysis 20 4%
Total 500 100%
Source: U. S. DOE, 2003
Note: Nm3 are normal cubic meters of hydrogen.
Hydrogen Production from Fossil Fuels
Hydrogen production from fossil fuel sources is the most mature means of hydrogen production
today, and the least expensive. Large amounts of hydrogen are produced globally from fossil
sources at present, mostly for use in oil refineries and production of fertilizers and chemicals.
Approximately 500 billion Nm3, or about 50 million metric tons ( and about 9 million metric tons
in the U. S.), per year of industrial hydrogen is produced for these uses globally, almost entirely
by the steam reformation of natural gas ( Moore and Raman, 1998; U. S. DOE, 2003).
Hydrogen can be produced from natural gas and other fossil fuels through various reformation
processes ( as well as through electrolysis if the fossil fuels are first used to generate electricity,
as discussed below). These reformation processes include basic steam methane reforming
( SMR), catalytic or “ autothermal” reformer ( ATR) processes, and partial- oxidation ( POx)
processes. In general, SMR and ATR processes are suitable for reforming light hydrocarbons,
such as methane, butane, propane, and ( with a special catalyst) naphtha, while the POx process
can be applied to a broader range of hydrocarbon feedstocks ( see below).
The basic endothermic SMR process using natural gas as a feedstock and steam for process heat
and additional hydrogen, and produces a hydrogen rich gas that is typically on the order of 70-
75% on a dry basis, along with smaller amounts of methane ( 2- 6%), carbon monoxide ( 7- 10%),
and carbon dioxide ( 6- 14%) ( Hirschenhofer et al., 2000; Sircar et al., 1999). Additional
subsequent processes to improve the hydrogen yield and reduce impurities include exothermic
water- gas shift ( WGS) reactions and pressure- swing adsorption ( PSA) processes.
The WGS reactions, typically in two steps – a “ high temperature WGS” step and a “ low
temperature WGS” step – convert most of the carbon monoxide in the SMR product gas to
hydrogen and carbon dioxide. Subsequent PSA “ cleanup” steps produce high- purity hydrogen
( 99.999+%) for uses where high purity hydrogen is important. The PSA process typically
recovers on the order of 70- 85% of the incoming hydrogen from the WGS processes ( Sircar et
al., 1999), and the PSA waste gases include the removed impurities along with the un- recovered
hydrogen. Overall, hydrogen can be produced at up to 75- 80% efficiency with SMR systems
( higher heating value [ HHV] basis), based on the use of natural gas as a feedstock, but
efficiencies of around 70% are more likely for smaller- scale systems ( Simbeck, 2001; Williams,
2002).
What Will Power the Hydrogen Economy?
5
ATRs rely on catalytic reactions that occur adjacently along honeycomb- structured beds of
catalytic metals, such as platinum, in a reactor vessel. By injecting air into the reactor, 30 to
40% of the fuel is oxidized with the aid of the catalyst and this releases heat that then is used to
convert the remaining fuel along with steam into hydrogen, carbon dioxide, and carbon
monoxide. ATRs have the advantages of not needing external burners and being simpler and
less expensive than SMRs, but they suffer from somewhat lower efficiency levels of 65- 75%
( Rampe et al., 2000).
In the POx process, also known more generally as “ gasification,” hydrogen can be produced
from a range of hydrocarbon fuels, including coal, heavy residual oils, and other low- value
refinery products. The hydrocarbon fuel is reacted with oxygen in a less than stoichiometric
ratio, yielding a mixture of carbon monoxide and hydrogen at 1200° to 1350° Celsius ( C.)
( Moore and Raman, 1998). The gas mixture is then cooled, subjected to WGS reactions similar
to those described above, and finally purified by PSA ( if pure hydrogen is desired). Waste
products from this process include a low pressure waste hydrogen steam, and excess steam.
Unless these waste products can be used locally, the economics of hydrogen from POx processes
are not typically attractive, as hydrogen production yields are only about 55- 60% -- or about 75%
as high as those for SMR systems ( Hirschenhofer et al., 2000).
Considerable attention has been paid in recent years to developing very small- scale POx- type
reformers for use in producing hydrogen onboard vehicles from gasoline, but though prototype
vehicles have been built, these efforts have apparently yet to yield a practical system. This has
been due to the difficulties involved with of reforming complex fuels at very small scale,
meeting the demands the dynamic operational requirements of fuel cell vehicles, and providing
sufficient reformer gas clean- up onboard the vehicle to meet the low CO and sulfur tolerance
requirements of PEM fuel cells.
Hydrogen Production from Natural Gas
Hydrogen can be produced from fossil fuels at various scales through these reformation
processes. Most mature are SMR technologies for producing hydrogen at large scales of ~ 100-
120 million standard cubic feet ( scf) per day ( 241,000- 290,000 kilograms [ kg]/ day), but smaller-scale
reformers have also been developed for hydrogen production in a more decentralized
fashion ( such as at vehicle refueling stations), and for even smaller- scale applications such as for
providing hydrogen fuel to stationary fuel cells in the 5- 250 kilowatt ( kW) size range.
Large Scale Centralized Production
Assuming industrial natural gas costs of about $ 3.00 per gigajoule ( GJ), large SMR plants with
capacities on the order of 60 million scf per day ( 145,000 kg/ day) can produce hydrogen at
efficiency levels of 76% to 81%, and at costs of $ 0.78 to $ 1.04 per kg ($ 5.50 to 7.30/ GJ), by
varying estimates ( Ogden et al., 2003; Simbeck, 2001; Williams, 2002).
One principal advantage of large- scale hydrogen production from fossil fuels is that it is in
principle possible to capture or “ sequester” the carbon dioxide that is produced, whereas this is
unlikely to be economical for smaller- scale production. A research group at Princeton
University has been studying the prospects for hydrogen production with CO2 capture in recent
What Will Power the Hydrogen Economy?
6
years, and they estimate that CO2 capture would add about $ 0.24 per kg ($ 1.70/ GJ) to the cost of
hydrogen produced from large- scale SMR plants as above, raising the total hydrogen production
cost from ( by their estimate) $ 0.78 per kg ($ 5.50/ GJ) to $ 1.02 per kg ($ 7.20/ GJ) ( Williams,
2002).
The National Research Council ( NRC) has recently examined hydrogen production via SMR for
very large ( 1.1 million kg of hydrogen per day) and medium- sized plants ( 24,000 kg of hydrogen
per day), including cases where product CO2 is vented or sequestered. The NRC estimates are
shown in Table 2, below, including also estimates for small- scale distributed production at a
scale of 480 kilograms of hydrogen per day.
Table 2: National Research Council Estimates for Hydrogen Produced via SMR of N. Gas
Scale / Case
Hydrogen Production Cost ($/ kg and $/ GJ HHV)
Current Technology Future Technology
Large Scale – CO2 vented
( 1.1 million kg/ day)
$ 1.03/ kg ($ 7.26/ GJ) $ 0.92/ kg ($ 6.49/ GJ)
Large Scale – CO2 seq.
( 1.1 million kg/ day)
$ 1.31/ kg ($ 9.24/ GJ) $ 1.10/ kg ($ 7.76/ GJ)
Medium Scale – CO2 vented
( 24,000 kg/ day)
$ 1.38/ kg ($ 9.73/ GJ) $ 1.21/ kg ($ 8.53/ GJ)
Medium Scale – CO2 seq.
( 24,000 kg/ day)
$ 1.76/ kg ($ 12.41/ GJ) $ 1.55/ kg ($ 10.93/ GJ)
Small Scale - CO2 vented
( 480 kg/ day)
$ 3.51/ kg ($ 24.75/ GJ) $ 2.33/ kg ($ 16.43/ GJ)
Source: NRC, 2004
Small Scale – Decentralized Production
Moore and Raman ( 1998) of Air Products and Chemicals, Inc. have examined the costs of
medium- sized natural gas reformer systems constructed on- site. They estimate a facility cost of
$ 9.6 million for a station that can support refueling of 500 vehicles per day. These stations
would produce 2,700 kg of hydrogen per day, at an estimated net cost of $ 3.57 per kg
($ 25.14/ GJ) ( Moore and Raman, 1998).
Smaller- scale SMR production has also been examined, for example in the context of hydrogen
production from natural gas at gasoline service stations for refueling FCVs. At a production
scale of 470 kg of hydrogen per day ( or about enough for 100 FCVs), Simbeck and Chang
( 2002) estimate that hydrogen can be produced and sold for $ 4.40 per kg ($ 30.99/ GJ), assuming
a natural gas cost of $ 5.25 per GJ. Table 2, above, presents the NRC estimates for production at
a similar scale, including both “ current technology” and “ future technology” cases.
Also, the Praxair Corporation is working with DOE to develop practical small SMR units. The
program has the goal of developing systems that can produce hydrogen at rates of 58 to 290 kg
What Will Power the Hydrogen Economy?
7
per day ( 1,000 to 5,000 scf per hour), at an aggressive cost target of $ 1.19 per kg ($ 8.40/ GJ).
Achievement of this ambitious cost goal will require the use of “ design for manufacturing and
assembly” techniques, a high level of system integration, and volume production of system
components ( Bollinger and Aaron, 2002).
Small Scale - Hydrogen Energy Stations
Thomas et al. ( 2001) of Directed Technologies Inc. ( DTI) have analyzed small natural gas
production systems that also include the use of fuel cells to produce electricity for local uses
using the same natural gas reformer as the one supplying hydrogen for FCVs. They analyze
these systems based on the number of hydrogen system components that are produced, assuming
economies of scale in their production, and conclude that hydrogen could initially be sold at
$ 2.60 per kg ($ 18.31/ GJ), with hydrogen sales of $ 1.68 per kg ($ 11.83/ GJ) possible if system
components were manufactured in volumes of 10,000 units.
Lipman et al. ( 2003) examined similar systems using somewhat more conservative cost
estimates, and have found that the economics of producing hydrogen and electricity with small
hydrogen “ energy stations” are difficult at small service station locations where only 10 to 50
vehicles per day are serviced, but with hydrogen sales prices of $ 20 per GJ ($ 2.84/ kg) begin to
become favorable when at least 50 vehicles per day are refueled. However, at office building
type energy stations in “ high electricity cost” states such as California, with hydrogen production
being subsidized from the electricity cost savings from the use of the fuel cell, hydrogen can be
sold at lower costs and additional facility energy cost savings can be realized in some cases.
These improved economics at office building energy stations are possible due to higher electrical
loads ( 200- 500 kW), the use of larger fuel cell systems that have lower costs per kW, and
because cogeneration of hot water for local uses is possible.
Figure 1 presents a picture of a hydrogen energy station that is operating in Las Vegas. In
addition to producing electricity using a stationary fuel cell, the station is capable of dispensing
pure hydrogen and compressed natural gas ( CNG) for vehicle refueling, as well as
CNG/ hydrogen blends ( City of Las Vegas, 2002).
What Will Power the Hydrogen Economy?
8
Figure 1: The Las Vegas H2E- Station
Source: Air Products and Chemicals, Inc.
Thermal Decomposition of Hydrogen From Natural Gas
In addition to these reformation processes, hydrogen can be produced from natural gas by
thermally decomposing it into hydrogen and carbon. One method for doing this would use a
concentrating solar reactor, or heliostat, to provide the heat needed for the decomposition
reaction. Spath and Amos ( 2002) have analyzed one type of solar reactor system that would
allow concentrated sunlight from the heliostat to enter a quartz outer reactor tube, and then
irradiate a solid graphite reactor tube that would in turn heat an inner porous graphite reactor
tube. The natural gas would be fed through this heated inner graphite tube where the
decomposition reaction would then take place. Product hydrogen, carbon, and un- reacted natural
gas would then enter a baghouse filter in order to remove the carbon. Assuming that this co-product
“ carbon black” is sold for industrial purposes at $ 0.66 per kg, and assuming a target
natural gas conversion of 70%, Spath and Amos ( 2002) estimate that hydrogen produced in this
manner with a large 8,750 square meter ( 94,200 ft2) heliostat could be at costs on the order of
$ 2.56 per kg ($ 18/ GJ), also assuming a natural gas price of $ 3.72 per GJ. With smaller heliostats
of 2,188 and 4,375 square meters ( 23,550 and 47,100 ft2), the corresponding hydrogen
production costs are estimated to be $ 3.41 per kg ($ 24/ GJ) and $ 2.84 per kg ($ 20/ GJ),
respectively ( Spath and Amos, 2002).
What Will Power the Hydrogen Economy?
9
Hydrogen Production From Coal
Hydrogen can also be produced via coal gasification POx processes, and while these systems are
less mature than SMR for hydrogen production, they are also relatively well established. For
example, Simbeck ( 2001) reports that over 15 coal gasifier systems are in operation around the
world that produce hydrogen for ammonia fertilizer manufacture, mostly in China but also in the
U. S., Germany, Japan, and India. There are various types of coal gasification systems available
at present, with primary differences being the gasifier type ( moving- bed, fluidized- bed, or
entrained- bed), the product gas temperature, and the composition of the product gas
( Hirschenhofer et al., 2000). All of these gasifier types use steam and air or oxygen to partially
oxidize coal into a gas product, and make use of the exothermic gasification reactions to produce
process heat. Table 3, below, shows a comparison of typical effluent streams from POx- based
reformation of coal, compared with SMR of natural gas.
Table 3: Typical Reformer Effluent for SMR of Natural Gas and POx Reformation of Coal
( Percent By Volume Basis)
SMR of Natural Gas
Reformer
Effluent
Reformer
Effluent
Shifted
Reformate
Fluidized- Bed POx
Reformation of Coal
Entrained- Bed POx
Reformation of Coala
H2 46.3 52.9 28.3 28.0
CO 7.1 0.5 33.1 47.9
CO2 6.4 13.1 15.5 6.2
CH4 2.4 2.4 4.6 0.2
N2 0.8 0.8 0.6 1.6
H2O 37.0 30.4 16.8 14.0
Total 100.0 100.0 98.9 97.9
Source: Hirschenhofer et al., 2000
Note: a Reformer effluent values shown are the average of four different entrained- bed coal reformation
technologies. These emissions vary somewhat by technology.
As with all hydrogen production pathways that use fossil fuels as feedstocks, but even more so
with the use coal, post- processing of the effluent gas is typically required to remove compounds
that are contaminants for most fuel cell types. These include hydrogen disulfide and other sulfur
compounds, ammonia, carbon monoxide, and tars, oils, and phenols. The various systems
needed to remove these compounds depend on the composition of the coal used but all add cost
and complexity to the production process, and in many cases have specific temperature
requirements that necessitate the extensive use of heat exchanger systems ( Hirschenhofer et al.,
2000).
Williams ( 2002) estimates that hydrogen can be produced via coal gasification at large scale for
costs as low as $ 0.88 per kg ($ 6.25/ GJ), assuming a coal price of $ 1.17 per GJ. However, the
CO2 emissions associated with producing hydrogen in this manner are considerable. Table 3
above provides a comparison of typical natural gas reforming and coal gasification effluent
compositions on a volume basis, with CO2 plus CO levels on the order of double those of
hydrogen in the coal systems, and with three to four times more hydrogen than CO2 plus CO
produced with the natural gas systems.
What Will Power the Hydrogen Economy?
10
This suggests that for coal gasification for hydrogen production to be practical in the long term,
these CO2 emissions ( once the CO is converted to CO2) would need to be captured and stored.
By Williams’ ( 2002) estimates, this would raise the cost of hydrogen production by $ 0.55 to
$ 1.25 per kg ($ 3.87 to $ 8.80/ GJ), or to a total of $ 1.43 to $ 2.13 per kg ($ 10.07 to $ 15.00 per GJ),
depending on whether or not hydrogen sulfide emissions are allowed to be captured and stored
along with the CO2 ( this reduces the costs of capture and storage and results in the lower- end
cost estimates).
Gray and Tomlinson ( 2002) and Ogden et al. ( 2004) calculate similar large- scale hydrogen
production costs from coal. By Gray and Tomlinson’s estimates, hydrogen can be produced at a
cost of $ 0.92 per kg ($ 6.48/ GJ) with CO2 emissions vented to the atmosphere, and $ 1.10 per kg
($ 7.75/ GJ) with CO2 sequestered at a cost of $ 10 per ton of carbon. With future “ advanced
gasification” techniques, they estimate that hydrogen could be produced from coal more
efficiently and with costs as low as $ 0.79 per kg ($ 5.56/ GJ), including the costs of carbon
sequestration at $ 10 per ton. At similar production scales ( e. g. about 600,000 kilograms of
hydrogen per day compared with 300,000 to 400,000 kilograms per day for Gray and
Tomlinson), Ogden et al. ( 2004) estimate hydrogen costs of $ 0.81 per kg ($ 5.69/ GJ) without CO2
sequestration, and $ 1.05 per kg ($ 7.36/ GJ) with CO2 sequestered.
The NRC estimates similar coal- to- hydrogen production costs as those of Gray and Tomlinson
( 2002) and Ogden et al. ( 2004), although at somewhat larger scale production of 1.2 million
kilograms of hydrogen per day. These estimates are $ 0.96 per kg ($ 6.77/ GJ) with current
technology and CO2 vented, $ 0.71 per kg ($ 5.01 per GJ) with future technology and CO2 vented,
$ 1.19 per kg ($ 8.39/ GJ) with current technology and CO2 sequestered, and $ 0.92 per kg ($ 6.49
per GJ) with future technology and CO2 sequestered ( NRC, 2004).
Finally, at somewhat smaller production scale of 150,000 kg of hydrogen per day ( compared
with approximately 300,000 to 500,000 kg per day in the above estimates), Simbeck and Chang
( 2002) estimate costs of $ 1.62 per kg ($ 11.41/ GJ), without CO2 sequestration.
Hydrogen Production from Nuclear Power
Nuclear power has been proposed for use in hydrogen production in a few different ways. First,
like any electricity- generation source, nuclear power can be used to produce hydrogen via the
electrolysis of water. This could be performed at ambient temperatures, or at higher
temperatures as “ hot electrolysis.” These nuclear power based electrolysis prospects are
discussed along with other electrolytic hydrogen production options in the following section.
Second, hydrogen could be produced through steam reformation of natural gas, using heat from
nuclear power plants to produce the steam. This reduces the natural gas requirements associated
with operating a boiler to raise the steam, and leads to hydrogen production with less natural gas
input. Third, hydrogen could be produced from water by using heat from nuclear plants to
thermally dissociate water molecules into hydrogen and oxygen.
Nuclear- assisted steam reforming is an option for hydrogen production, whereby the heat
What Will Power the Hydrogen Economy?
11
required for the endothermic reforming process is provided by the high- temperature heat from
the nuclear reactor rather than from the combustion of natural gas. In this case, the natural gas
input acts only as a source for hydrogen. This type of system is in the research and development
stage at present, with the main development activity occurring in Japan. In a research program
that was initiated in 2001, Japanese researchers have been experimenting with coupling a steam
reformer to a 30 MW high- temperature test reactor with a peak temperature of 950° C.
( Forsberg, 2003).
Hydrogen can also be produced from nuclear energy through thermo- chemical processes that use
high- temperature heat to dissociate water molecules. This can be accomplished in various ways,
and in fact over 100 potential cycles have been identified, with one leading concept being the use
of an iodine- sulfur process ( Brown et al, 2002; Forsberg, 2003). In this type of system, water is
combined with sulfur dioxide ( SO2) and iodine ( I2) to produce sulfuric acid ( H2SO4) and
hydrogen iodide ( HI). The sulfuric acid is decomposed into water, sulfur dioxide, and oxygen by
adding heat from the nuclear reactor at 800° to 1000° C., and the hydrogen iodide is converted to
hydrogen and iodine. The iodine and sulfur dioxide are then re- used in the first reaction, with
the oxygen being vented or potentially captured for industrial uses. The following reactions are
involved in this type of system:
Iodine- Sulfur Process Reactions
I2 + SO2 + 2 H2O  2 HI + H2SO4
H2SO4 + heat  H2O + SO2 + 1/ 2 O2
2 HI + heat  H2 + I2
Overall:
H2O + heat  H2 + 1/ 2 O2
By one estimate, the costs of producing hydrogen from this type of system could be as low as
$ 1.30 per kg ($ 9.15/ GJ), assuming a combined capital cost of $ 686 per kW ( thermal) for the
nuclear reactor ($ 371/ kW) and hydrogen plant ($ 315/ kW) and assuming 50% system efficiency
( Henderson, 2002). Another recent estimate by the NRC estimates somewhat higher costs of
$ 1.63 per kg ($ 11.50/ GJ), assuming a production plant capital cost of $ 2.5 million for a
production capacity of 1.2 million kilograms of hydrogen per day ( NRC, 2004). By another
estimate, the exact costs of producing hydrogen through this cycle are not specified but are
estimated to be about 60% of the costs of hydrogen production via electrolysis ( Forsberg, 2003).
The Process of Electrolysis to Produce Hydrogen
Electrolysis is the process of producing hydrogen by splitting water molecules. Electrolysis has
a long history, dating back to experiments by Carlisle and Nicholson around the year 1800, and
in fact is the oldest electrochemical process known ( Adreassen, 1998). However, it was not until
the early 1900s that commercial electrolyzers were developed for hydrogen production. The
uses for this early hydrogen production were primarily in the fertilizer industry, and there were
approximately 400 electrolyzers in operation in 1902 ( Adreassen, 1998).
The most mature electrolyzer technologies are based on alkaline electrolytes, typically composed
What Will Power the Hydrogen Economy?
12
of potassium hydroxide. However, additional electrolyzer technologies are now being developed
based on three more advanced technologies: 1) solid polymer electrolyte membrane- based
electrolysis ( the same membranes used in PEM fuel cells); 2) high- temperature steam
electrolysis; and 3) natural gas assisted steam electrolysis.
The solid polymer electrolyte- based PEM electrolyzers are currently characterized by
significantly higher capital costs than the alkaline designs, but these costs are expected to decline
as the technology matures and moves into higher volume production. However, PEM
electolyzers have the advantages of higher operational efficiencies than alkaline electrolyzers,
and the ability to operate reversibly as PEM fuel cells to produce electricity from the generated
hydrogen.
The high- temperature steam electrolyzers benefit from the fact that much lower electrolyzer cell
voltages are needed to split water at temperatures of 900° to 1000° C than are required at
ambient temperatures ( Ogden and Nitsch, 1993). These high- temperature electrolyzers would
use primarily ceramic electrode materials, but the high operational temperatures of these devices
present formidable material and fabrication problems, and their overall complexity has thus far
been a barrier to commercialization.
Natural gas assisted steam electrolysis has recently been investigated by the Lawrence Livermore
National Laboratory. In this scheme, natural gas is supplied to the anodes of high- temperature
steam electrolyzers instead of air. The natural gas on the anode side of the electrolyzer reduces
the amount of electricity that needs to be supplied, relative to conventional steam electrolysis.
This is because the oxygen ion charge- carrier in this type of system ( from the water molecule)
more readily crosses the membrane due to the affinity of the oxygen ion to react with and oxidize
the methane molecules, and this allows electrolysis to occur at lower cell voltages than with
conventional electrolyzer systems ( by up to a full volt or more per cell). This reduction in
electricity requirements with natural gas assisted electrolysis can result in hydrogen production at
60 to 75% efficiency, in terms of primary energy input, compared to about 40% efficiency for
conventional alkaline electrolysis using U. S. average grid electricity ( Pham et al., 2002).
Hydrogen can be produced from electrolysis at a range of scales. An important potential
advantage of electrolysis is that small- scale production of hydrogen by electrolysis is practical,
and this can reduce or even eliminate the need for hydrogen distribution. However, large- scale
electrolysis, such as in the nuclear power assisted case discussed below, would require the
hydrogen to be delivered by truck as compressed gas or liquid, or distributed with pipeline
networks. All of these options entail significant costs for hydrogen distribution, as discussed in
Section III of this report.
The following chemical equations show how hydrogen is produced from water using both the
alkaline electrolyte and solid polymer electrolyte technologies ( Adreassen, 1998).
What Will Power the Hydrogen Economy?
13
Alkaline Electrolyzer Electrode Reactions
Cathode: 2 H2O + 2 e- = H2 + 2 OH-Anode:
2 OH- = 1/ 2 O2 + H2O + 2 e-
Overall: H2O + energy = H2 + 1/ 2 O2
Solid Polymer Electrolyzer Electrode Reactions
Cathode: 2 H+ + 2 e- = H2
Anode: H2O = 1/ 2 O2 + 2 H+ + 2 e-
Overall: H2O + energy = H2 + 1/ 2 O2
Electrolyzers have the advantages of producing pure hydrogen, the ability to produce hydrogen
at relatively high pressures directly from the electrolyzer ( in the case of PEM electrolyzers), and
the ability to use any primary fuel that can be used to produce electricity. Disadvantages include
relatively high capital costs ( particularly for low- temperature PEM electrolyzers that like PEM
fuel cells require platinum- based catalysts), hydrogen production efficiencies that decline with
the rate of hydrogen production for a given system, and uncertain durability of the novel PEM
electrolyzers that are now starting to be commercialized.
The costs of producing hydrogen from electrolysis depend on the electrolyzer type ( and
corresponding capital cost and efficiency), as well as the costs of the required electricity and
other inputs. For large scale production using conventional alkaline electrolysis, Norsk Hydro
electrolyzers reports potential costs of $ 2.96 per kg of hydrogen ($ 20.9/ GJ), assuming a plant
that produces 4,000 Nm3 of hydrogen per hour, electrolyzer system capital costs of $ 13.0
million, and electricity costs of $ 0.029 per kWh ( Anreassen, 1998). 1
For PEM electrolyzers costs at present are high at about $ 3,000 to 4,000 per kW, or similar to
those of PEM fuel cells. This is not surprising as electrolyzers and fuel cells are similar devices
that essentially operate in reverse of one another. DOE has set development goals for PEM
electrolyzers of $ 600 per kW for systems that can produce about 25 kg ( 10,000 scf) of hydrogen
per day and that are manufactured in quantities of 10,000 units per year. Smaller units that can
produce about 2.5 kg ( 1,000 scf) of hydrogen per day have a cost goal of $ 1,200 per kW, again
in production volumes of 10,000 units per year. In larger volume production of 100,000 units
per year, costs for the 25 kg per day units have a cost target of $ 300 per kW ( Padro, 2002). Note
that these figures are simply cost targets, and not manufacturing cost estimates based on detailed
analysis.
Discussed below are various technologies to produce electricity that can then be used to directly
electrolyze water and produce hydrogen. Hydrogen production can be closely coupled to any
one of these methods, or from a “ grid mix” of different sources on a utility grid. Given the
current mix of powerplants in the U. S., direct electrolysis of grid power is not favored in terms of
energy use and greenhouse gas emissions, but the situation would improve with future reductions
in powerplant emissions and can be highly favorable when coupled with renewable and nuclear
power systems.
1 Costs converted from Norwegian kroner to dollars using a conversion factor of 1 dollar = 6.9 kroner.
What Will Power the Hydrogen Economy?
14
Hydrogen From Electrolysis Using Grid Power
Hydrogen can be produced via electrolysis using utility grid power, with the regional “ generation
mix” being important to the resulting economics and environmental impacts. Particularly where
regional utility grids include large proportions of coal- based electricity generation, the
environmental impacts of producing hydrogen in this manner may be unattractive. Where utility
grids are “ greener” and include significant amounts of renewable or other clean electricity
production, the environmental impacts of producing hydrogen from grid power can be more
attractive. The important point is thus that the environmental impacts of grid power electrolysis
can vary widely, and also vary depending on the overall efficiency of the electrolyzer system.
The NRC has analyzed the costs of producing hydrogen from grid power at two different scales
and based on current and future electrolyzer technology. At medium scale of 24,000 kilograms
of hydrogen produced per day and with wholesale electricity costs of $ 0.045 per kWh, the NRC
estimates current hydrogen production costs of $ 4.70 per kg ($ 33.15/ GJ) and future potential
costs of $ 2.30 per kg ($ 16.22/ GJ). At smaller production scale of 480 kilograms of hydrogen per
day, and with commercial electricity costs of $ 0.07 per kWh, the NRC estimates current costs of
$ 6.58 per kg ($ 46.41/ GJ) and future potential costs of $ 3.93 per kg ($ 27.72/ GJ) ( NRC, 2004).
Hydrogen From Electrolysis Using Nuclear Power
Nuclear energy production is relatively capital- intensive, and only practical in large plant sizes,
but has relatively low operating costs ( plant decommissioning and long- term waste
disposal/ storage costs being another matter, and one that can significantly increase the overall
costs of nuclear energy). With these relatively low operating costs once installed, nuclear energy
is best suited for “ baseload” electricity generation. At times of low or “ off- peak” demand,
nuclear power plants and other baseload power plants could supply electricity for electrolysis,
particularly if demand dips below the level of power that the baseload plants are capable of
supplying. Thus, particularly compared to the alternative of turning down baseload power
plants, hydrogen production from nuclear power via electrolysis may be attractive.
In addition to simply using the electricity generated from nuclear power for electrolysis, hot
electrolysis is another option and one that is favorable for use in conjunction with nuclear power
because of the availability of the necessary heat from nuclear reactors. Electrolysis at these
higher temperatures ( 700° to 1000° C.) requires less electrical input because some of the
electrical input required for ambient temperature electrolysis is replaced with thermal energy.
This thermal energy is less costly than the electrical energy that it replaces, and these systems
can in theory produce hydrogen at reduced costs. However, these hot electrolysis systems are in
an early stage of development, and are characterized by high capital costs that are on the order of
twice the costs of lower- temperature electrolyzer systems ( Forsberg and Peddicord, 2002), as
well as likely high associated hydrogen delivery costs due to the remote locations of most
nuclear plants.
Hydrogen From Electrolysis Using Renewable Sources
Hydrogen can be produced from renewable sources via three primary pathways. First, electricity
generated from renewable sources, such as wind and solar photovoltaics ( PV), can be used to
split water molecules via electrolysis as discussed here. Second, hydrogen can be thermally
What Will Power the Hydrogen Economy?
15
dissociated from water using renewable sources that generate large amounts of heat, such as
solar thermal technologies, without the addition of electricity. Third, biomass and other
feedstocks can be used to produce hydrogen via gasification processes, similar to the coal
gasification processes described above. The second and third options will be discussed in the
following sections.
Hydrogen Production by Electrolysis from Wind Power
Hydrogen production from wind power is a particularly interesting proposition for two primary
reasons. First, among renewable sources of electricity, wind power is particularly competitive
economically with electricity costs of 4 to 5 cents per kWh in good sites. This means that
hydrogen produced with electrolyzers that are connected to wind farms can generate hydrogen at
lower costs than with some other options such as solar power. Second, the availability of wind
power tends to be highly variable diurnally, and also fluctuates seasonally. This makes wind
power a somewhat difficult resource to depend on, but at the same time hydrogen production
from wind power that is available at off- peak times can help to in effect even- out this wind
power potential. Coupled with the use of fuel cells, producing hydrogen from off- peak wind
power can make wind a more dispatchable resource by acting as a storage or “ buffer” system.
Similar to pumped hydro storage, hydrogen production in this way can make effective use of
wind power at times when it is otherwise not valuable, such as late at night or at times of the year
when electricity demands are relatively low.
Hydrogen storage costs are highly variable depending on the type of storage system, the
hydrogen flow rate, and the duration of storage. The primary potential storage methods include
compressed gas storage, liquid storage, metal hydride storage, and underground storage. Of
these, underground storage is generally the most economical, with costs on the order of $ 0.10-
0.20 per kg of hydrogen stored, but this type of storage is only possible with the presence of
suitable underground caverns. Compressed gas and liquid storage are the next most economical,
with compressed gas storage being cheaper for low flow rates and storage times ( on the order of
$ 0.30- 2.00 per kg for 14 days of storage depending on flow rate) and liquid storage having costs
of about $ 0.50- 2.00 per kg depending on flow rate but irrespective of storage duration. Metal
hydride storage costs scale linearly with storage duration and flow rate, and are only
economically competitive with the other methods for low flow rates and storage times ( Mann et
al., 1998).
Based on future projections that were made about 10 years ago, Ogden and Nitsch ( 1993)
estimate that hydrogen could be produced at “ excellent” wind sites ( with wind speeds of about
20 miles per hour or 10 meters per second) for approximately $ 1.56 per kg ($ 11.00/ GJ) assuming
a 6% real discount rate and $ 2.27 per kg ($ 16.00/ GJ) using a 12% discount rate. At “ good” wind
sites ( with wind speeds of about 17 miles per hour or 8.5 meters per second), the corresponding
hydrogen production costs are estimated to be $ 2.41 per kg ($ 17.00/ GJ) assuming a 6% real
discount rate and $ 3.55 per kg ($ 25.00/ GJ) using a 12% real discount rate. These estimates are
for “ stand- alone” hydrogen production from wind power, unassisted by additional power from
the utility grid.
Mann et al. ( 1998) estimate that hydrogen can be produced from wind power in a stand- alone
configuration for $ 7.10 per kg ($ 50.00/ GJ) near- term, and $ 4.00 per kg ($ 28.20/ GJ) by the 2010
What Will Power the Hydrogen Economy?
16
timeframe. Coupling wind/ hydrogen systems with the electrical grid appears to improve the
economics of hydrogen production, especially for the near term, where some of the electricity for
electrolysis comes from the grid. Such systems are estimated to potentially produce hydrogen
for $ 3.90 per kg ($ 27.50/ GJ) near term and $ 3.00 per kg ($ 21.10/ GJ) circa 2010. These
estimates assume wind power costs of $ 900 per kW in 2000 and $ 700 per kW in 2010, and wind
capacity factors of 35% ( 2000) and 40% ( 2010).
Padro ( 2002) has also published estimates of hydrogen costs from wind power, with estimates as
low as $ 1.14 per kg ($ 8.00/ GJ) in small stand- alone settings, and from $ 1.86 to 2.63 per kg
($ 13.00 to $ 18.50/ GJ) in small to medium scale grid- tied applications and with various system
design and economic assumptions. Padro also includes estimates of hydrogen production costs
from wind power that include the effects of a 15% “ internal rate of return” target and a 37%
taxation rate ( see Table A- 1 for details).
Finally, the NRC reports estimates of hydrogen production costs from wind power that range
from $ 10.69 per kg ($ 75.39/ GJ) for current technology and stand- alone production, to $ 6.81 per
kilogram ($ 48.03/ GJ) for current technology and grid- assisted production. With future
improvements, the NRC is projecting particularly large production cost reductions for stand-alone
systems ($ 2.86 per kg of hydrogen or $ 20.17 per GJ), with significant reductions also
projected for the grid- tied systems ($ 3.50 per kg of hydrogen or $ 24.68 per GJ). These estimates
are for production at scales of 480 kg per day for grid- tied production and 1,200- 16,00 kg per
day for stand- alone production ( NRC, 2004).
Hydrogen From Wind Production Potential in the U. S
The potential for hydrogen production from wind power in the U. S. is well distributed, with
significant resources in the Midwestern, Northeastern, and Western parts of the country ( see
Figure 2). Ogden and Nitsch ( 1993) estimate that the total annual production of hydrogen in the
U. S. from wind power could be approximately 220 billion kg per year ( 31.3 EJ/ year), with
leading states in terms of their production potential including North Dakota ( 24.7 billion
kg/ year), Texas ( 24.3 billion kg/ year), Kansas ( 21.8 billion kg/ year), South Dakota ( 21.1 billion
kg/ year), Montana ( 20.9 billion kg/ year), Nebraska ( 17.7 billion kg/ year), Wyoming ( 15.3 billion
kg/ year), and Oklahoma ( 14.8 billion kg/ year). With the total hydrogen requirements for a U. S.
fleet of light- duty FCVs in 2010 estimated by Ogden and Nitsch ( 1993) at approximately 33.8
billion kg/ year ( 4.8 EJ/ year), this wind hydrogen potential alone is approximately 6.5 times the
requirement for light- duty vehicles in the U. S. if the fleet were to be converted to operate on
hydrogen fuel cells.
What Will Power the Hydrogen Economy?
17
Figure 2: Hydrogen From Wind Power Availability Map of the U. S.
What Will Power the Hydrogen Economy?
18
Hydrogen Production by Electrolysis from Solar Power
Hydrogen can be produced from solar power by generating electricity with solar photovoltaic
( PV) systems, and then using the electricity to electrolyze water, or through high temperature
processes where water is dissociated thermally without the addition of electricity. The solar
electrolytic processes can either be at ambient temperature, or at higher temperatures where the
water is first converted to steam before being electrolyzed. The electrolysis- based methods of
producing hydrogen from solar energy are discussed below, and the thermal solar systems are
discussed in a subsequent section.
With solar PV electrolysis, hydrogen is produced when the solar resource is available by using
the generated electricity to split water with the use of an electrolyzer, as described above. The
costs of producing hydrogen through solar PV electrolysis are higher than with most other
methods, but are expected to decline with continued reductions in PV module costs and
electrolyzer system costs. Glatzmaier et al. ( 1998) estimate that solar PV systems in the 10
MWe size range can produce hydrogen for $ 6.39 per kg, $ 12.21 per kg, and $ 25.84 per kg
($ 45/ GJ, $ 86/ GJ, and $ 182/ GJ) assuming PV system costs of $ 750 per kW, $ 2,000 per kW, and
$ 5,000 per kW, respectively, and electrolyzer costs of $ 450 per kW of direct current ( DC)
power. These estimates assume an electrolyzer efficiency of 82%, and a capacity factor of 0.28.
Ogden and Nitsch ( 1993) estimated similar costs for solar hydrogen production in the early
1990s. Their estimates range from $ 6.39 to $ 23.71 per kg ($ 45- 167/ GJ), depending on interest
rate and other assumptions, but they also forecast much lower potential future costs based on PV
module and electrolyzer component cost reductions. They project that hydrogen produced in
sunny areas with a solar insolation of about 270 watts per square meter could have future costs of
$ 1.42 to $ 2.27 per kg ($ 10- 16/ GJ) using a real discount rate of 6%, and $ 2.13 to $ 3.55 per kg
($ 15- 25/ GJ) with a more conservative 12% real discount rate. These estimates assume future PV
module costs of $ 30 to $ 55 per square meter, and PV module efficiencies of 12 to 18%. Lower
costs could be achieved with higher PV module efficiencies of closer to 20%, lower PV module
costs, or a combination of the two.
In addition to using solar PV systems to meet the electricity needs of electrolysis, additional
solar- based technologies such as solar dish- Stirling systems and solar “ power towers” can also
produce electricity for renewable hydrogen production. With solar dish- Stirling systems,
concentrated solar power is used to produce heat to power a Stirling engine, which then operates
a generator to produce alternating current ( AC) power. The solar “ power tower” systems also
produce AC power using concentrated solar energy and a generator. The AC power from either
of these systems is then rectified to DC before being used in the electrolyzer cells, with hydrogen
production then proceeding as above ( Glatzmaier et al., 1998).
Solar power can also produce hydrogen through high- temperature steam electrolysis, where heat
from concentrating solar systems first produces steam to reduce the electricity requirements for
electrolysis, and then supplies the electricity required for electrolysis of the steam. One concept
for this type of system would use a high- temperature electrolyzer similar in design and
construction to a solid- oxide fuel cell, operating “ in reverse” as an electrolyzer ( Glatzmaier et al.,
1998). A solar “ power tower” used in conjunction with the electrolyzer would produce both heat
What Will Power the Hydrogen Economy?
19
to raise steam as well as AC power, which would then be converted to DC to operate the
electrolyzer. Glatzmaier et al. ( 1998) estimate that this type of system could produce hydrogen
for $ 5.68 to $ 6.25 per kg ($ 40- 44/ GJ) assuming electrolyzer costs of $ 500 per kW, and $ 7.67 to
$ 11.42 per kg ($ 54- 79/ GJ) assuming electrolyzer costs of $ 2,000 per kW, depending on the
current density at which the electrolyzer is operated.
Mann et al. ( 1998) estimate that hydrogen can be produced from solar PV systems in a stand-alone
configuration for $ 17.60 per kg ($ 124.00/ GJ) near- term, and $ 7.50 per kg ($ 52.80/ GJ) by
the 2010 timeframe. With solar PV/ hydrogen production systems coupled to the electrical grid,
estimated costs are considerable lower because some of the electricity comes from the utility grid
at lower cost. Such systems are estimated to potentially produce hydrogen for $ 7.40 per kg
($ 52.10/ GJ) near term and $ 4.50 per kg ($ 37.10/ GJ) circa 2010. These estimates assume solar
power costs of $ 3,133 per kW in 2000 and $ 1,662 per kW in 2010, and PV plant capacity factors
of 28% ( near term) and 30% ( in 2010).
Finally, the NRC estimates and projects hydrogen production costs from solar PV including
estimates of $ 28.19 per kg ($ 198.81/ GJ) for current technology and stand- alone production,
$ 6.18 per kg ($ 43.58/ GJ) for future technology and stand- alone production, $ 9.71 per kg
($ 68.48/ GJ) for current technology and grid- tied production, and $ 4.37 per kg ($ 30.82/ GJ) for
future technology and grid- tied production. These estimates are for production scales of 2,400
kilograms of hydrogen per day for the stand- alone systems and 480 kilograms of hydrogen per
day for the grid- tied systems ( NRC, 2004).
Hydrogen From Solar Production Potential in the U. S
Solar hydrogen potential in the U. S. is more geographically constrained than either wind power
or biomass, with the greatest resource in the southwest and additional resources in the west and
Midwest ( see Figure 3). Leading states for solar PV based hydrogen production potential
include Texas ( 4.41 exajoules [ EJ]/ year), California ( 2.77 EJ/ year), New Mexico ( 2.57 EJ/ year),
Arizona ( 2.41 EJ/ year), Nevada ( 2.24 EJ/ year), Montana ( 2.19 EJ/ year), and Colorado ( 2.01
EJ/ year), with the total estimated resource even larger than the wind power potential at 47.56
EJ/ year ( Ogden and Nitsch, 1993). Using Ogden and Nitsch’s ( 1993) estimate of 4.8 EJ/ year for
a fleet of light- duty FCVs in the U. S., this solar hydrogen potential is approximately 10 times
larger than that requirement.
What Will Power the Hydrogen Economy?
20
Figure 3: Hydrogen From Concentrating Solar Power Availability Map of the U. S.
What Will Power the Hydrogen Economy?
21
Hydrogen Production by Electrolysis from Hydropower
Hydrogen can also be produced from hydropower by generating electricity with turbines that
operate along waterways. With low “ off- peak” hydropower rates that can be as low as $ 0.02-
0.03 per kWh, hydrogen production from this source can be economical. Hydrogen production
in this manner is an interesting alternative to “ pumped storage,” where each day water is pumped
to higher elevation reservoirs “ off peak” and then released “ on peak” for delivery, particularly if
there is a use for the hydrogen in the nearby area. However, it is important to note that for
electricity markets in the U. S. there tends to be high demand for this inexpensive but constrained
resource. Production of hydrogen would have to compete with other electricity uses to secure
the hydropower supply.
Direct Hydrogen Production from Renewable Resources
In addition to being used to produce electricity for electrolysis, renewable resources can be used
to produce hydrogen directly. This can be done through biomass conversion/ gasification, and
through other biological process. These methods are briefly discussed below, along with the
potential biomass- to- hydrogen production potential in the U. S.
Hydrogen From Biomass Gasification
Hydrogen production from biomass is potentially attractive for several reasons. First, biomass is
a near- term option for use of a renewable resource that can be produced in large quantities and
indefinitely. The International Institute for Applied Systems Analysis ( IIASA) estimates that
biomass potentials are about 250 EJ on a global basis at present ( about 3.5 EJ in the U. S.), with
potential of over 350 EJ by 2050 ( Mann and Overend, 2003). These values compare with the
approximate global use of fossil fuels at present of about 300 EJ, and projected global non-electric
fuel demand of about 286 EJ by 2025 and 289 EJ by 2050 ( Ogden and Nitsch, 1993,
based on Intergovernmental Panel on Climate Change scenarios). In other words, in terms of
energy content, the near- term global biomass potential is roughly on par with global fossil fuel
use at present, and could be expanded to be in excess of the total non- electric energy needs of the
world in the 2025 to 2050 timeframe.
Estimates for the U. S. suggest that 7- 8 EJ of biomass could be supplied by 2020, depending on
the biomass price and with contributions from all of the primary sources -- energy crops and crop
residues, and forest, animal, and urban residues ( EIA, 2002; Mann and Overend, 2003). Second,
since biomass growth removes CO2 from the atmosphere, producing hydrogen from biomass
fundamentally releases that CO2 back into the atmosphere with no net additions -- depending on
the use of fertilizer additions and other factors -- unlike fossil fuels burning or reforming that
steadily adds CO2 to the atmosphere. Third, unlike some renewable sources of hydrogen that are
geographically constrained due to resource availability, biomass resources are relatively widely
distributed both within the U. S. and globally. Fourth, biomass production can also produce
valuable co- products, in addition to waste heat, such as adhesives, carbon black, activated
carbon, polymers, fertilizers, ethanol, various acids, Fischer- Tropsch diesel fuel, waxes, and
methanol ( Mann and Overend, 2003). Exploiting these co- product values, as they are more fully
explored and developed, is likely to improve the economics of hydrogen production from
biomass.
Biomass conversion technologies can be divided into thermo- chemical and biochemical
What Will Power the Hydrogen Economy?
22
processes. Thermo- chemical processes tend to be less expensive because they can be operated at
higher temperatures and therefore obtain higher reaction rates. They also can utilize a broad
range of biomass types. In contrast, biochemical processes are limited to wet feedstock and
sugar- based feedstocks. Milne et al. ( 2002) provide a state- of- the art assessment of these
processes and of hydrogen from biomass technologies in general, and Spath et al. ( 2000) present
a detailed analysis of the economics of hydrogen production from three leading types of thermo-chemical
systems.
With regard to thermo- chemical hydrogen production systems, “ producer gas” is the name given
to the combustible gases that are produced from biomass. This producer gas consists of CO,
CO2, nitrogen ( N2), and hydrogen, and is about 10- 15% of the heating value of natural gas
( Larson, 1993). These producer gases are similar to the “ synthesis gas” that is produced from
coal for application in synthesizing chemicals and fuels. Hydrogen yields can be increased by
subjecting producer gas to WGS reactions, and high purity hydrogen can be produced by adding
a PSA step to the end of the production process.
The first common thermo- chemical option for hydrogen production from biomass is indirectly
heated gasification coupled with steam reforming. These are then followed by WGS reaction,
and a PSA step to produce pure hydrogen if desired. A second option is oxygen- blown
gasification coupled with steam reforming, whereby pure oxygen is reacted rather than air.
Finally, another option is pyrolysis coupled with steam reforming, along with co- product
production ( Mann, 2003). As with gasification/ SMR, these processes can be coupled with PSA
or other hydrogen purification steps to produce pure hydrogen, such as for vehicle refueling, or
they could be used in conjunction with electricity production systems ( such as some types of
high- temperature fuel cells) that can operate directly on producer gas.
The production of hydrogen from biomass through the gasification processes begins with pre-processing,
depending on the type of biomass, to dry and condense the biomass material prior to
gasification. Then in the gasification step, in indirectly heated gasification hot sand circulates
between the char combustor and gasification vessel, and provides heat for the endothermic
gasification process. The gasification producer gas is then cooled ( and pressurized if a PSA
purification unit is to be used) and the processed in SMR and WGS units to provide high
hydrogen yields ( Spath et al., 2000). Overall thermal efficiencies of hydrogen production from
these indirectly heated systems can be as high as 69% on an HHV basis ( Larson, 1993).
One type of indirectly heated gasifier was developed by Battelle Columbus Laboratories, with
the rights now owned by Future Energy Resources Corporation ( FERCO). FERCO is now
demonstrating the technology at a plant in Burlington, Vermont. In an economic analysis of this
system type, Spath et al. ( 2000) estimate plant- gate hydrogen selling prices that range from
$ 1.12/ kg ($ 7.90/ GJ) to $ 1.19/ kg ($ 8.41/ GJ) with no internal rate of return ( i. e., production costs
only) and from $ 2.31/ kg ($ 16.28/ GJ) to $ 2.87/ kg ($ 20.18/ GJ) with a 20% internal rate of return,
depending on production scale ( see Table A- 1 for details).
The second type of gasifier is the oxygen- blown type, similar to the above but incorporating the
use of an air- separation unit to produce relatively pure oxygen for use in the gasifier. These
systems also differ in that the gasifier is fired directly rather than indirectly, and in that the
What Will Power the Hydrogen Economy?
23
gasifier is operated at much higher pressure. The thermal efficiency of hydrogen production is
lower with this type of gasifier system than with the indirectly heated type, on the order of 57%
on an HHV basis ( Larson, 1993). With regard to production costs, again depending on
production scale and internal rate of return, estimated hydrogen selling prices for these oxygen-blown
gasifier systems can range from $ 1.19/ kg ($ 8.40/ GJ) to $ 1.20/ kg ($ 8.48/ GJ) with 0%
internal rate of return and from $ 2.67/ kg ($ 18.77/ GJ) to $ 3.52/ kg ($ 24.81/ GJ) with a 20%
internal rate of return ( Spath et al, 2000).
The NRC has also estimated the current and potential future costs of hydrogen from oxygen-blown
biomass gasification, at a medium production scale of 24,000 kilograms of hydrogen per
day. These cost estimates are for $ 4.63 per kg ($ 32.65/ GJ) for current technology and $ 2.21 per
kg ($ 15.59/ GJ) with future technology, with the produce CO2 vented. If the CO2 is captured and
sequestered, the NRC estimates costs of $ 5.08 per kg ($ 35.83/ GJ) for current technology and
$ 2.53 per kg ($ 17.84/ GJ) with future technology ( NRC, 2004).
Hydrogen From Pyrolysis of Biomass
The third type of thermo- chemical hydrogen production from biomass is based on pyrolysis,
which uses heat at temperatures of 400° to 800° C. in to volatilize organic solids to gases and
liquids. Since biomass is typically 70- 90% volatile matter, pyrolysis can be used effectively for
biomass gasification ( Larson, 1993). In one pyrolysis- based scheme, the biomass is first dried
and then converted to an oil through “ fast pyrolysis” where a fluidized bed reactor is used to
convert the heated biomass particles into bio- oils, gases, and char. The char and gases are
combusted to supply heat, and the bio- oils are cooled, condensed, and then converted to
hydrogen through SMR and WGS processes. This process also produces a phenol co- product
that can be used in the manufacture of phenol formaldehyde resins that are used as wood
adhesives. Overall thermal efficiency of biomass- to- hydrogen production is estimated at 47.9%
for this type of pyrolysis system, on an HHV basis ( Iwasaki, 2003).
Including credit for the phenol co- product, estimated hydrogen costs from this type of pyrolysis
system are relatively low, and range from $ 0.75 to $ 0.93 per kg ($ 5.30/ GJ to $ 6.57/ GJ) with 0%
internal rate of return, to from $ 1.39 to $ 1.62 per kg ($ 9.79/ GJ to $ 11.41/ GJ) with a 15% internal
rate of return ( Spath et al, 2000). By another estimate, hydrogen can be produced from biomass
via pyrolysis for $ 1.09 per kg ($ 7.70/ GJ), assuming that the phenol co- product is sold for $ 0.44
per kg ( French et al., 2000).
Hydrogen From Biological Gasification
Other than these thermo- chemical processes, the other primary type of biomass conversion
system is biological gasification, or anaerobic digestion, of wet feedstocks using fermentative
bacteria. This process produces “ biogas”, consisting of methane and CO2. The biogas stream
can then be converted to a hydrogen- rich gas stream through SMR and WGS processes, as in the
above cases. Besides the production of fuel gases, biomass digesters also yield co- product
effluents that can be used in fertilizers ( Larson, 1993).
Hydrogen From Agricultural and Livestock Residues
As well as dedicated energy crops, hydrogen can be produced from both agricultural residues
and manure from livestock. These are both relatively economic sources, with delivered
What Will Power the Hydrogen Economy?
24
hydrogen costs as low as $ 2.30 per kg ($ 16.32 per GJ) for production from manure and $ 2.65 per
kg ($ 18.80 per GJ) for production from agricultural residues, by one set of estimates ( Meyers et
al., 2003).
Biomass and High Temperature Fuel Cells
Finally, one interesting type of biomass- based system for electricity production from hydrogen
would combine a biomass gasifier, a molten- carbonate fuel cell ( a high- temperature fuel cell type
with similar thermal characteristics to the gasifier), and a turbine- based bottoming cycle. These
integrated biomass gasification fuel cell ( IBGFC) systems would produce electricity from
biomass residues with overall projected efficiency levels of 55%, or higher in the overall thermal
efficiency sense if used in conjunction with waste heat recovery. These systems would not
produce pure hydrogen at any stage, but would use hydrogen- rich gas from the biomass gasifier
to operate the high- temperature fuel cell. Applications of this technology include converting
sugar cane bagasse to electricity with efficiencies of 44 to 47% ( lower heating value [ LHV]
basis), with the potential for readily increasing this to 53% efficiency assuming a biofuel
containing 15% moisture ( Knight et al., 1998).
Hydrogen From Biomass Production Potential in the U. S.
Hydrogen can be produced from biomass residues in most areas of the U. S. ( see Figure 4), and
additional hydrogen production possible from dedicated energy crops. Forecasts for potential
hydrogen production from biomass in the U. S. have been made out to 2020, and other forecasts
have been made for global production as far out as 2050. The 2020 U. S. potential has been
estimated at 28.8 billion kg per year ( 4.1 EJ/ year) of hydrogen from 7 EJ of biomass energy, by
one analysis that assumes a 50% energy conversion ratio on an LHV basis ( Mann and Overend,
2003). Another, nearer- term analysis suggests the potential for production of 19.4 billion kg per
year ( 2.76 EJ/ year) of hydrogen from biomass energy crops in the U. S., with the greatest
potential resources in Texas, North Dakota, Kansas, Montana, South Dakota, Iowa, and Colorado
( Ogden and Nitsch, 1993).
The amount of biomass that can be supplied depends on the price of the biomass fuel, as higher
prices allow more of the resource to be economically harvested or recovered. The U. S. DOE’s
Energy Information Agency ( EIA) estimates that the 7 EJ of biomass by 2020 suggested above
would correspond to a field- edge biomass price of about $ 3 per GJ, with up to 7.5 EJ possible at
prices of $ 4 per GJ and only small increases possible at higher prices ( EIA, 2002). This includes
urban and mill residues, energy crops, forest residues, and agricultural residues. In comparison,
a field- edge biomass price of $ 2 per GJ would yield about 3 EJ of biomass supply, or less than
half of the supply at $ 3 per GJ ( EIA, 2002).
What Will Power the Hydrogen Economy?
25
Figure 4: Hydrogen From Biomass Residues Availability Map of the U. S.
What Will Power the Hydrogen Economy?
26
Solar Thermal Processes
In addition to production via either ambient temperature or high temperature electrolysis,
hydrogen can be produced from solar power by using concentrated solar energy to super- heat
steam to temperatures of over 2000 Kelvin ( 1727° C.) where it thermally dissociates into
hydrogen and oxygen. The production of hydrogen in this manner is favored at higher
temperatures, with greater amounts of hydrogen produced as temperatures increase toward 3000
Kelvin ( 2727° C.) and at lower pressures ( Glatzmaier et al., 1998). Thus, in the absence of
materials and operating constraints and overall energy balance considerations, solar thermal
hydrogen production should be conducted at high temperature and low pressure levels. In order
to make the process somewhat more efficient, the enthalpy of the separated hydrogen and
oxygen high temperature can be recovered with a heat exchanger to raise the temperature of
incoming water, thereby reducing the requirements for providing heat from the solar thermal
system ( such as a solar power tower).
Glatzmaier et al. ( 1998) describe general designs of solar thermal systems for hydrogen
production, but they do not quantitatively analyze such systems because of their early stage of
development and lack of knowledge of the potential materials and methods for producing
hydrogen in this manner. Ogden and Nitsch ( 1993) do provide estimates of hydrogen production
from solar thermal systems, assuming production in the Southwestern U. S. They estimate that
hydrogen could be produced in the future (“ post 2000”) with solar thermal systems at costs of
$ 2.56 to $ 3.55 per kg ($ 18- 25/ GJ) assuming a 6% discount rate, and $ 3.83 to $ 5.11 per kg ($ 27-
36/ GJ) assuming a 12% discount rate.
Hydrogen Production from Other Renewable Sources
Hydrogen can be produced from municipal solid wastes ( MSW) and landfill gases. Capital costs
for these systems are typically higher than for biomass plants due to the additional processing
requirements for this type of feedstock, but due to no or even negative feedstock costs ( i. e.,
avoided disposal costs) levelized lifecycle costs for MSW plants could be comparable to those of
other biomass- based hydrogen production facilities ( Williams, 2002). Meyers et al. ( 2003)
estimate that delivered hydrogen can be produced from MSW and landfill gases for about $ 2.50
per kilogram across much of the U. S.
Landfill gas in particular has been successfully utilized to produce electricity with stationary
hydrogen fuel cells in several of the landfill gas recovery sites in the U. S. Phosphoric acid fuel
cells have been the most used fuel cell technology to- date in these settings, but solid oxide and
molten carbonate fuel cells show greater promise for use with landfill gas because they are more
tolerant of impurities and would integrate better with land fill gas reformer technologies
( Xenergy, 2002).
Finally, hydrogen can also be produced through natural biological processes by eukaryotic green
algae. The algae species Chlamydomonas reinhardtii produces molecular hydrogen, particularly
when the algae is deprived of inorganic sulfur and saturated with light, and when oxygen
evolution from photosynthesis is separated from hydrogen production ( Melis et al., 2000). In
recent research, novel methods have been developed to regulate this hydrogen production
process from green algae ( Melis et al., 2000). This research is noteworthy because it shows that
there are practical methods for producing hydrogen from algae continuously for several days,
What Will Power the Hydrogen Economy?
27
rather than in the trace amounts for short periods that had been previously detected.
Melis et al. ( 2000) have demonstrated hydrogen production levels of 140 milliliters of hydrogen
per liter of algae culture, over an 80- hour period, and they believe that significant increases in
this production rate are possible with continued research. Commercial systems for producing
hydrogen in this way have not yet been developed, but DOE has established development goals
for algal hydrogen production at $ 16 per kg ($ 113/ GJ) in the mid- term, with a longer- term
“ continuous” system target of $ 1.30 to $ 2.30 per kg ($ 9.15- 16.20/ GJ) ( Padro, 2002).
Summary of Hydrogen Production Costs from Various Methods
Hydrogen can be produced from a wide range of sources, with onsite production costs as low as
$ 0.75- 1.00 per kg ($ 5.28- 7.04/ GJ) from sources such as large- scale natural gas reformation,
biomass gasification, and coal gasification. Other methods such as wind and nuclear electrolysis
can produce hydrogen for costs that may be competitive in some settings, and solar electrolysis
costs continue to decline but are still relatively high, on the order of $ 6.50 to $ 25.80 per kg ($ 46-
182/ GJ) depending on a variety of factors such as the solar resource and the interest rate applied
to the project. See the following section for a further discussion of hydrogen costs, including
delivered hydrogen cost estimates that include the costs of distribution to points of end use.
Figure 5, below, presents a summary of onsite hydrogen production costs. Table A- 1 in the
Appendix presents the cost estimates from which this chart is derived, including further details
with regard to the assumptions behind each estimate. Note that this figure presents onsite
production costs only. See Figures 6 and 7 and Table A- 2 for delivered hydrogen cost estimates.
What Will Power the Hydrogen Economy?
28
Figure 5: Ranges in Onsite Hydrogen Production Cost Estimates
0
1
2
3
4
5
6
7
8
9
10
11
12
NG SMR large
NG SMR small
Hydrogen energy station
Coal gasification large
Wind electrolysis current
Solar electrolysis current
Biomass gasification medium
NG SMR large w/ CO2 seq.
NG solar thermal
Coal gas. large w/ CO2 seq.
Nuclear electrolysis
Wind electrolysis future
Solar electrolysis future
Biomass gasification large
Biomass pyrolysis large
Hydrogen production cost ($/ kg)
High
Low
28.2
? ?
Nearer Term Technologies Longer Term Technologies
Note: Various sources – see Table A- 1 for details.
NG = natural gas; SMR = steam methane reforming; “?” = Costs of effective carbon sequestration from fossil fuels
are uncertain because sequestration technologies and methods are still in the R& D phase.
III. Hydrogen Distribution and Supply Systems
The technologies for distributing and supplying hydrogen for various uses are generally well
known and established, with the exception of very high- pressure ( 10,000 pounds per square inch
[ psi]) hydrogen storage and dispensing systems and other innovative hydrogen storage systems.
These very high- pressure systems are currently in the research and development stage, and offer
high storage densities but at slightly higher compression energy losses. On the other hand, lower
pressure compressed- gas storage systems for hydrogen of 3,600- 5,000 psi are well proven and
are currently being used in many prototype FCVs around the globe.
Hydrogen distribution and supply systems can be classified as being associated with
“ centralized” production, with significant needs for distribution, and “ decentralized” production
with much lower or no needs for distribution but potentially significant requirements for storage
( depending on the relationship between the operation of the production system and end- use
requirements). The various hydrogen production methods discussed in the previous section thus
imply some differences with regard to hydrogen delivery, particularly with regard to the scale
and distribution of production, along with the relative purity of the hydrogen produced. Further,
these systems can be distinguished by their potential for near- term deployment, versus being
longer- term options.
What Will Power the Hydrogen Economy?
29
With regard to near- term options for hydrogen distribution, centralized production systems can
be coupled with either truck delivery by liquid or high pressure gas “ tube trailers,” pipeline
delivery, or for hydrogen- powered vehicles in particular, refillable hydrogen “ trailers” that can
be filled with compressed gas onsite and then placed at off- site locations until refilling is
required. For on- site production from SMR and electrolysis, or in the case of chemical plant by-product
hydrogen, hydrogen can simply be used as it is produced, or compressed, stored, and
dispensed for refueling hydrogen- powered vehicles and/ or producing electricity with stationary
or automotive fuel cell systems. 2
Longer- term options for hydrogen distribution and supply include the above possibilities along
with more futuristic concepts. These include supplying hydrogen as a component of
superconductive electricity delivery, using hydrogen as the cooling fluid for the
superconductivity as well as a distributed fuel ( Gehl and Rastler, 2001), as well as systems
associated with hydrogen production with carbon sequestration and large- scale renewable power
systems. In the case of large- scale hydrogen production through electrolysis ( with either
renewable or nuclear power), either the hydrogen or the electricity could be transported to the
point of end use. In the latter case, electricity would be transmitted to the point of end- use,
particularly during “ off- peak” periods when electrical transmission systems are unconstrained,
where electrolyzers would then produce hydrogen for local uses.
Delivered Hydrogen Cost Estimates
Several studies of delivered hydrogen costs have been conducted. The discussion below focuses
on the most recent and comprehensive of these, and the results of some additional analyses are
reported in Table A- 2. In general, these studies are difficult to compare in “ apple to apple” terms
because of the many assumptions that go in to each study, and therefore the ranges of estimates
discussed here and shown in Figure 7 and Table A- 2 are simply reported “ as is.” In other words,
as with the production cost estimates discussed above, no attempt has been made to correct the
studies for differing assumptions to make them more directly comparable.
In one recent and comprehensive study, Simbeck and Chang ( 2002) of SFA Pacific, Inc. have
performed a detailed economic analysis of different hydrogen production and distribution
systems, with a particular focus on near- term options. Included are various options for both
central plant and distributed production of hydrogen, including natural gas, petroleum coke, coal,
biomass, grid power electrolysis, and other renewable feedstocks, and liquid hydrogen, pipeline,
tube trailer, and decentralized distribution.
They conclude that for large- scale options ( on the order of 150,000 kg per day), natural gas
based production at can produce delivered hydrogen for $ 3.66 to $ 5.00 per kg ($ 25.77 to
$ 35.21/ GJ), petroleum coke based production coupled with pipeline delivery can produce
delivered hydrogen costing $ 5.35 per kg ($ 37.68/ GJ), coal- based production can produce
delivered hydrogen for $ 4.51 to $ 5.62 per kg ($ 31.76/ GJ to $ 39.58/ GJ), biomass- based
2 This latter possibility has come to be known as “ vehicle- to- grid” or V2G power. See Kempton and Letendre,
( 1997); Kempton et al. ( 2002); and Lipman et al. ( 2004) for details with regard to various possibilities for different
types of electric- drive vehicles.
What Will Power the Hydrogen Economy?
30
production can produce delivered hydrogen for $ 4.98 to $ 6.29 per kg ($ 35.07/ GJ to $ 44.30/ GJ),
and large- scale electrolysis with commercial grid power can produce hydrogen for $ 7.62 to
$ 9.13 per kg ($ 53.66- 64.30/ GJ). For smaller- scale, distributed options ( on the order of 500 kg of
hydrogen produced per day), estimated delivered costs of hydrogen are $ 4.40 per kg ($ 30.99/ GJ)
for natural gas SMR, and $ 12.12 per kg ($ 85.35/ GJ) for water electrolysis ( using grid power at
$ 0.07/ kWh).
In general, Simbeck and Chang ( 2002) find that for the centralized options at the scale examined,
the costs of hydrogen delivery are lowest for liquid hydrogen distribution by tanker truck,
followed by tube trailer delivery, followed by delivery by pipeline. With regard to centralized
versus decentralized production of hydrogen from natural gas, decentralized production is found
to be competitive to the centralized options, with costs identical to tube trailer delivery, and
somewhat more than liquid hydrogen delivery by tanker truck and somewhat less than pipeline
delivery.
With regard to additional recent studies of delivered hydrogen costs from natural gas, Ogden
( 1999) estimates that the delivered cost of pure hydrogen as a transportation fuel are about the
same for onsite “ advanced” SMR and centralized SMR with “ high density” pipeline delivery,
assuming relatively large- scale dispensing of 2,400 kg per day ( 1 million scf/ day). In these
cases, delivered hydrogen costs of about $ 1.70 per kg ($ 12/ GJ) are estimated, versus about $ 2.13
per kg ($ 15/ GJ) for SMR production with “ low density” pipeline delivery, about $ 2.70 per kg
($ 19/ GJ) for centralized SMR and liquid hydrogen delivery by truck, and about $ 3.55 per kg
($ 25/ GJ) with onsite hydrogen production from advanced electrolysis ( Ogden, 1999).
Based on the hydrogen production model developed by SFA Pacific, the NRC has recently
estimated the total delivered costs of hydrogen from various production methods. These are
based on the production costs discussed above and in Table A- 1, and assume that large central
station production options are coupled with pipeline delivery and that medium- scale production
options are coupled with liquid hydrogen delivery by tanker truck. See Figure 6 below for a
graphical depiction of these NRC cost estimates, including breakdown of costs by production,
delivery, and dispensing to FCVs, as well as approximate carbon- related costs ( e. g., a $ 50 per
ton carbon tax and/ or carbon separation and sequestration costs of $ 10 per ton of carbon).
What Will Power the Hydrogen Economy?
31
Figure 6: One “ Internally Consistent” Set of Delivered Hydrogen Cost Estimates
0
2
4
6
8
10
12
14
16
Delivered Hydrogen Cost ($/ kg)
CS NG- C
CS NG- F
CS NG- C seq
CS NG- F seq
CS coal- C
CS coal- F
CS coal- C seq
CS coal- F seq
CS nuc- F
MS NG- C
MS NG- F
MS NG- C seq
MS NG- F seq
MS Bio- C
MS Bio- F
MS Bio- C seq
MS Bio- F seq
MS Elect- C
MS Elect- F
Dist NG- C
Dist NG- F
Dist Elect- C
Dist Elect- F
Dist NGASE
Dist WT- C
Dist WT- F
Dist WTgr- C
Dist WTgr- F
Dist PV- C
Dist PV- F
Dist PVgr- C
Dist PVgr- F
Production Distribution Dispensing CO2 cost
28.19
Source: NRC, 2004. Notes: Bio = biomass gasification; C = current technology; CS = large central station; Dist = distributed production at small scale; Elect =
grid- power electrolysis; F = future technology; gr = grid assisted electrolysis; MS = medium- scale; NG = natural gas SMR; PV = solar photovoltaic powered
electrolysis; seq = with carbon sequestration; WT = wind turbine powered electrolysis. CO2 cost = carbon disposal cost and/ or carbon tax of $ 50 per ton.
What Will Power the Hydrogen Economy?
32
Figure 7 presents ranges of production plus delivery and dispensing costs from the technical
literature. Note that the ranges in the delivered cost estimates are somewhat narrower than the
ranges in the production- only estimates shown in Figure 5, above, primarily because fewer
delivered hydrogen cost studies have been presented in the public domain. Finally, see Table A- 2
for more details of these and other delivered hydrogen cost estimates.
Figure 7: Ranges in Delivered Hydrogen Cost Estimates
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
NG SMR med- large
NG SMR small ( onsite)
Hydrogen energy station
Coal gasification large
Wind electrolysis current
Solar electrolysis current
Biomass gasification med
NG SMR large w/ CO2 seq.
Coal gas. large w/ CO2 seq
Wind electrolysis future
Solar electrolysis future
Biomass gasification large
Biomass pyrolysis large
Delivered H2 Cost ($/ kg)
High
Low
28.19
Nearer Term Technologies Longer Term Technologies
Notes: Various sources - see Appendix for details. The ranges shown are taken from many different sources,
including those with assumptions that may be somewhat inconsistent with regard to production scale, interest rates,
etc. Wider and narrower ranges between high and low costs thus tend to reflect the relative numbers of studies for
each pathway, rather than inherent uncertainties in costs for each pathway.
Figure 8, below, presents a map of the existing hydrogen production facilities in the U. S., along
with the major natural gas pipelines in the country. Most of the current hydrogen production
facilities in the U. S. are in the Midwest and Gulf Coast regions.
What Will Power the Hydrogen Economy?
33
Figure 8: Current Hydrogen Production Facilities in the U. S.
What Will Power the Hydrogen Economy?
34
IV. Hydrogen from Hydrocarbons and Carbon Sequestration
Hydrocarbon sources of hydrogen are attractive due to their relatively low costs. However, since
one potential “ selling point” of hydrogen is the potential for low environmental impacts, the
GHG emissions from these sources are an important concern. One potential solution is to
produce hydrogen from hydrocarbons at large, centralized sites and to “ sequester” the CO2
emissions in some way to prevent these emissions from reaching the atmosphere.
CO2 emissions can be sequestered in a variety of different ways. These include terrestrial,
microbial, underground, and oceanic storage solutions. Terrestrial and microbial solutions
include enhancing photosynthesis and other processes in biological systems as a means of
increasing CO2 uptake from the atmosphere. Underground storage in geologic formations
includes the prospects of storing carbon in mined coal beds and salt domes, deep aquifers, and
depleted oil and gas reservoirs, and as mineral carbonates or bicarbonates. Oceanic storage
techniques include pipeline “ droplet plumes,” pipelines towed by ships, dry ice- based transfer
schemes, and deep- water platform pumping stations with underwater CO2 “ lakes.”
These systems all have varying associated issues and potential costs, with costs in particular
being uncertain at this time. Cost predictions of $ 30 per ton of CO2 in the long term have been
estimated ( equivalent for reference purposed to an additional $ 13 per barrel of oil or $ 0.25 per
gallon of gasoline), but sequestration costs would be higher in the near term ( Lackner, 2003).
With regard to underground storage options, coal beds offer potentially low cost carbon storage
but the technologies for doing this are immature. Salt domes are also potentially attractive
storage sites, but use of these structures would entail relatively high costs. Deep saline aquifers
offer large potential storage capacity, but uncertain storage integrity, while depleted oil and gas
reservoirs are expected to provide greater storage integrity but are of limited capacity. The
neutralization of carbonic acid to form carbonates and bicarbonates is another option, and one
with an expected high degree of safety but also relatively high costs ( Lackner, 2003).
The oceanic storage options are mainly classified by the method of carbon dioxide insertion,
along with the potential for carbon dioxide mixing in the ocean or the prospect of creating deep
ocean carbon dioxide “ lakes” where the CO2 would pool at the bottom of the ocean. The droplet
plume option would have CO2 delivered into the ocean at the end of a land- based pipeline, at
depths of perhaps 1,000- 2,000 meters. This would have uncertain effects on oceanic ecosystems,
and would also result in some leakage back to the atmosphere. Pipelines towed by ships could
inject CO2 at similar depths and with similar potential drawbacks. Meanwhile, dry ice deposition
of CO2 would rely on simple technology, but entails relatively high costs and uncertain
environmental impacts and leakage concerns. Finally, using deep- water pumping stations to
inject CO2 to depths of 3,000 meters or greater would create deep- ocean CO2 “ lakes.” These
lakes would likely separate the CO2 from the atmosphere for thousands of years, but the
technology for sequestering carbon with this method is at an immature stage.
Finally, carbon/ GHG emission sequestration strategies have thus far focused on CO2, but other
GHG emissions associated with hydrogen production are of concern as well. Emissions of CH4
and other GHGs associated with hydrogen production from fossil fuels could potentially also be
What Will Power the Hydrogen Economy?
35
mitigated but -- since these options have been less studied -- with even less pathway and cost
certainty than for carbon dioxide.
V. Environmental Impacts of Hydrogen Production Methods
Additional important considerations for potential hydrogen production methods other than costs
and feedstock requirements include the environmental implications of various hydrogen
production methods. These include GHG emissions, local pollutant emissions, soil and water
emissions, and land, water, and other non- feedstock resource requirements.
Greenhouse Gas Emissions Estimates from Hydrogen Fuel Pathways
Various hydrogen production strategies have dramatically differing implications with regard to
the greenhouse gas emissions implications of end- use, as shown in Figure 9 below. The results
reported in the figure compare the results of two well- known GHG analysis models, for various
hydrogen production and distribution pathways. These models include the Greenhouse Gases,
Regulated Emissions, and Energy Use in Transportation ( GREET) Model developed by Argonne
National Lab ( ANL, 2001), and the Lifecycle Emission Model ( LEM) developed at UC Davis
( Delucchi, 2003). The results shown are CO2- equivalent emissions of several combined GHGs,
for hydrogen used to fuel FCVs in comparison with using reformulated gasoline ( RFG) in
conventional vehicles.
These results show that the GHG emissions implications of hydrogen production and use in
vehicles can vary enormously relative to the emissions from the use of gasoline in conventional
vehicles. Depending on the hydrogen production and distribution pathway, emissions can be as
high as 80% of the level of emissions from conventional vehicles ( using the U. S. average grid
power mix to produce liquid hydrogen via electrolysis), and as low as 0% with biomass and
some other renewable pathways. If one goal of the use of hydrogen is to reduce GHG emissions,
then the method of hydrogen production and distribution chosen is clearly critical to achieving
that goal.
What Will Power the Hydrogen Economy?
36
Figure 9: Greenhouse Gas Emissions from Hydrogen Fuel Cell Vehicle Refueling Pathways
- 120
- 100
- 80
- 60
- 40
- 20
0
20
40
60
80
100
NG central - G H2
NG central - L H2
NG central w/ steam - G H2
NG central w/ steam - L H2
NG station - G H2
NG station - L H2
Electrol. US mix - G H2
Electrol. US mix - L H2
Electrol. CCNG - G H2
Electrol. CCNG - L H2
Electrol. Nuclear - G H2
Electrol. Nuclear - L H2
Electrol. Solar PV - G H2
Electrol. Solar PV - L H2
MeOH - Reformer
EtOH Corn - Reformer
EtOH Cellulosic - Reformer
GHGs % Change from Convetional RFG
GREET 1.6
LEM 2003
Notes: GREET 1.6 is the Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model. LEM
2003 is the Lifecycle Emission Model. CCNG = combined cycle natural gas power plant; EtOH = ethanol; G =
gaseous; L = liquid; NG = natural gas; MeOH = methanol; PV = photovoltaics; RFG = reformulated gasoline.
Air Pollutant Emissions and Soil and Water Impacts from Hydrogen Fuel Pathways
Air pollutant emissions and soil and water impacts of hydrogen production and distribution
pathways also can vary widely for various methods, and unlike GHG emissions have important
local and regional environmental quality implications. These impacts have generally been
studied in less detail than potential GHG emissions, particularly with regard to soil and water
impacts. The social and environmental welfare impacts of air pollutant emissions also vary in
part depending on population exposure levels, adding even more uncertainty with regard to these
potential impacts.
Figure 10, below, presents an analysis of major air pollutant emissions from the same hydrogen
production and distribution pathways shown in Figure 9. Shown are total pollutant emissions of
oxides of nitrogen ( NOx), fine particulate matter less than 10 microns in diameter ( PM), and
volatile organic compounds ( VOC) from the full fuel cycle for each pathway. As shown in the
figure, the relative air pollutant emissions from transportation uses of hydrogen also vary widely
depending on the hydrogen production pathway, with most pathways producing significant
reductions but some – particularly electrolyzer pathways coupled with either average grid mix or
combined cycle natural gas plants – producing significant increases in NOx and PM emissions.
See ANL ( 2001) for more details of these estimates, disaggregation of results by “ in basin” and
What Will Power the Hydrogen Economy?
37
“ out of basin” emissions, and additional results for emissions of CO and oxides of sulfur.
Additional air pollutant emissions analyses from hydrogen pathways include those conducted by
Delucchi ( 2003) and Contadini ( 2002).
Figure 10: Air Pollutant Emissions from Hydrogen Fuel Cell Vehicle Refueling Pathways
- 150
- 100
- 50
0
50
100
150
200
250
NG central - G H2
NG central - L H2
NG central w/ steam - G H2
NG central w/ steam - L H2
NG station - G H2
NG station - L H2
Electrol. US mix - G H2
Electrol. US mix - L H2
Electrol. CCNG - G H2
Electrol. CCNG - L H2
Electrol. Nuclear - G H2
Electrol. Nuclear - L H2
Electrol. Solar PV - G H2
Electrol. Solar PV - L H2
MeOH - Reformer
EtOH Corn - Reformer
EtOH Cellulosic - Reformer
Major Air Pollutants % Change from Convetional RFG
NOx
PM
VOC
320% 458% 450%
Source: GREET Model v1.6 ( ANL, 2001). CCNG = combined cycle natural gas power plant; EtOH = ethanol; G =
gaseous; L = liquid; NG = natural gas; MeOH = methanol; PV = photovoltaics; RFG = reformulated gasoline.
With regard to land and water resource requirements, Ogden and Nitsch ( 1993) have estimated
land and water resource requirements for renewable hydrogen production options and found that
resource requirements for wind, solar PV, solar thermal electric, biomass, and hydroelectric
based production are relatively well- defined, except for significant uncertainties in land
requirements for wind and hydroelectric power. In general, hydroelectric and wind power
require the largest amounts of land, and hydroelectric Table 4, below, presents the results of the
Ogden and Nitsch analysis.
What Will Power the Hydrogen Economy?
38
Table 4: Land and Water Requirements for Renewable Hydrogen Production Methods
Land Requirements
( m2/ GJ/ yr)
Water Requirements
( liters/ GJ HHV)
Electrolytic Production
Solar PV 1.89 63
Solar thermal electric 5.71 63
Wind 6.3- 33 63
Hydroelectric 11- 500 > 63
Biomass Production 50 37,000- 74,000
Notes: HHV = higher heating value; GJ = gigajoules; m2 = square meters; PV = photovoltaic; yr = year.
Source: Ogden and Nitsch, 1993
VI. Renewable Hydrogen Potential and Regional Production Strategies
The overall technical potential for producing hydrogen from renewable sources in the U. S. is
great, with approximately 32 quadrillion British thermal units ( Quads) per year of renewable
hydrogen production possible in the 2030- 2050 timeframe according to one recent estimate
( Meyers et al., 2003). This is an enormous quantity of potential energy production, equal to 33.6
EJ per year, and this compares with current total transportation sector petroleum consumption in
the U. S. of about 27 Quads per year ( Davis and Diegel, 2003). However, much of this renewable
hydrogen production potential is based on wind and solar resources that are unlikely to become
economically competitive with other sources, even during this extended timeframe. As
discussed above, more economical sources of renewable hydrogen include dedicated energy
crops, MSW, agricultural and livestock residues, and certain wind resources that are located
close enough to demand centers to be relatively attractive.
Figure 11 presents a hypothetical regional demand and supply picture for renewable hydrogen
for the year 2040. This scenario assumes that hydrogen demand in the U. S. reaches 10 Quads
( 10.5 EJ) by this time, and that the demand in each U. S. state is proportional to current gasoline
consumption ( first set of bars). The figure shows three potential levels of renewables- based
hydrogen supply. The second set of bars – “ Renewable H2 ( economic)” -- shows the amount of
hydrogen that could be dispensed to consumers at a cost of $ 3.00 per kg or less ($ 21.30 per GJ
and $ 1.50 per gallon of gasoline equivalent on a per- mile basis in 2x efficiency hydrogen
vehicles). Some hydrogen in the Northeast and Southeast is supplied from other nearby regions
in this scenario. The third set of bars shows the total hydrogen demand in each region, totaling
10 Quads, being met by renewable sources. If all of this hydrogen demand were supplied by
hydrogen from renewables, the average nationwide cost of that hydrogen is estimated to be about
$ 4.00 per kg ($ 28.40 per GJ) in this future scenario. This scenario implies considerable transport
of hydrogen from states in the Midwest and Central that are closest to the Northeast, Southeast,
and Pacific regions. The fourth set of bars – “ Renewable H2 ( potential)” – shows the total
technical potential of nearly 32 Quads of production, with much of this potential in the Midwest
and Central regions ( Meyers et al., 2003).
What Will Power the Hydrogen Economy?
39
Figure 11: Hypothetical Year 2040 Regional U. S. Hydrogen Demand of 10 Quads Per Year
and Renewable Production Potential
0.0
1.0
2.0
3.0
4.0
Northeast Southeast Midwest Central Pacific
H2 Demand/ Supply ( Quads/ year)
H2 Demand Renewable H2 ( economic) Renewable H2 ( 10 Quads) Renewable H2 ( potential)
14.83 14.19
Notes: “ Northeast” includes CT, ME, MA, NH, NJ, NY, P