Progress in Nuclear Energy 51 (2009) 500–525

Contents lists available at ScienceDirect

Progress in Nuclear Energy

journal homepage: www.elsevier.com/locate/pnucene

Nuclear heat for hydrogen production: Coupling a very high/high temperature

reactor to a hydrogen production plant
Rachael Elder*, Ray Allen
Department of Chemical and Process Engineering, University of Sheffield, Mappin Street, Sheffield S1 3JD, United Kingdom

a b s t r a c t

Keywords: Hydrogen has been dubbed the fuel of the future. As fossil fuel reserves become depleted and green-
Hydrogen house gas emissions are reduced inline with the Kyoto protocol, alternative energy sources and vectors,
Nuclear hydrogen production such as hydrogen, must be developed. Hydrogen produced from water splitting, as opposed to from
VHTR
hydrocarbons, has the potential to be a carbon neutral energy solution. There are several methods to
Coupling
extract hydrogen from water, three leading candidates being high temperature electrolysis, the SI
thermochemical cycle and the HyS hybrid thermochemical cycle. All three of these processes involve
a section requiring very high temperatures. The Very High Temperature Reactor (VHTR), a gas cooled
Generation IV reactor, is ideally suited for providing this high temperature heat. Nuclear hydrogen
production is being investigated around the world. The four leading consortiums are the Japan Atomic
Energy Agency (JAEA), PBMR/Westinghouse, GA, and AREVA NP/CEA/EDF. There are also many smaller
R&D efforts focussing on the development of particular materials and components and on process
flowsheeting.
A nuclear hydrogen plant involves four key pieces of equipment: the VHTR, the hydrogen production
plant (HPP), the intermediate heat exchanger (IHX) and the power conversion system (PCS). The choice
of all four items varies dramatically between programmes. Both pebble bed and prismatic fuel block
VHTRs are being developed, which can be directly or indirectly coupled to a HPP and PCS placed either in
series or parallel. Either a Rankine steam cycle or a Brayton gas turbine cycle can be employed in the PCS.
This report details the choices made and research being carried out around the world.
Predicted process efficiencies and plant costs are currently at a preliminary stage and are very similar,
regardless of the options chosen. The cost of hydrogen produced from water splitting using nuclear
technologies is around $2/kg H2. This is competitive with hydrogen produced by other methods,
particularly if carbon emissions are regulated and costed. The technological feasibility and testing of key
components will be one of the determining factors in plant viability.
Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction 1.1. Energy outlook

Nuclear power will play a significant part in the world’s At the end of last century, the world’s consumption of primary
energy supply throughout the next century. Decreasing supplies energy was 8380 million TOE (tonnes oil equivalent). Research by the
of fossil fuels, along side increasing instability in their availability, European Commission (IAEA, 1999) suggests that this value will rise
and the need to decrease greenhouse gas emissions in line with by 1.6% every year, and it is expected to have doubled by 2050 due to
the Kyoto protocol, mean the development of alternative energy population growth and the industrialisation of developing countries
sources and carriers is extremely important. Renewable forms of (Crabtree et al., 2004). In 2007, fossil fuels accounted for 80% of the
energy are not currently developed enough, or available on world energy supply and 67% of the world electricity generation
a large enough scale, to be a significant contributor to the world’s (IEA, 2007), despite the fact that they have a finite lifetime and their
energy supply. use produces three quarters of anthropogenic CO2 emissions (United
Kingdom Department of Environment Food and Rural Affairs, 2008).
Fossil fuels are distributed unevenly throughout the world, leading
to political and security issues. Where possible, fossil fuels should be
used as a valuable chemical feedstock and not burned for process
* Corresponding author. Tel.: þ44 114 2227500; fax: þ44 114 2227501. heating. The development of alternative energy sources and carriers
E-mail address: r.elder@sheffield.ac.uk (R. Elder). is essential to ensure a stable energy outlook for future generations.

0149-1970/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.pnucene.2008.11.001
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 501

1.1.1. Climate change Table 1

The Intergovernmental Panel on Climate Change (IPCC) stated in Generation IV reactors identified for development.

a review in early 2007 (IPCC, 2007) that: ‘‘Most of the observed Type Reactor Acronym Outlet temperature
increase in globally averaged temperatures since the mid-20th ( C)
century is very likely due to the observed increase in anthropogenic Thermal Very high temperature reactor VHTR >1000
greenhouse gas concentrations’’. Greenhouse gases (GHGs), the Supercritical water reactor SCWR 550
Molten salt reactor MSR 700
primary three of which are carbon dioxide, water and methane,
absorb solar radiation and so create a natural greenhouse blanket Fast Gas cooled fast reactor GFR 850
effect around the Earth, without which temperatures would be an Sodium cooled fast reactor SFR 550
Lead cooled fast reactor LFR 550 up to 800
estimated 30  C lower. Carbon dioxide is a long-term greenhouse
gas and, once added to the atmosphere, can remain in the atmo-
sphere for hundreds of years. The atmospheric concentrations of all
major GHGs have increased significantly since pre-industrial times and developing technologies coupling nuclear power to the trans-
due to human activities. For example, carbon dioxide concentra- port sector are vital for the future of the nuclear industry.
tions have risen by just over one third from 280 parts per million
(ppm) in around 1750, to 379 ppm in 2005. Without intervention, 1.2. Generation IV reactors
the Stern Review (Stern, 2007) states that CO2 levels could reach
550 ppm by 2050, leading to warming of at least 2  C. The outcomes Generation IV reactors are a new generation of nuclear reactors
of global warming include the rise of sea levels, glacial retreat, designed to broaden the opportunities for nuclear energy. They will
species extinction, an increase in the ranges of infectious diseases be inherently safer than previous generations, highly economical,
and an increased likelihood of severe weather patterns. A proliferation resistant and will produce minimal waste. Their
temperature rise of just 2.7  C could lead to the melting of the development is currently underway around the world and it is
Greenland ice cap. The broad agreement among climate scientists, predicted that by 2030 the first will be in operation. In the interim,
that global temperatures will continue to rise, has lead govern- Generation IIIþ reactors will be built, offering improved economics
ments and corporations to take actions to curtail global warming. over the current generation for near term deployment.
Many mitigation ideas have been suggested, including renewable In January 2000 the Generation IV International Forum (GIF)
energies, carbon taxes and offsets, development of new technolo- was established to investigate innovative nuclear energy system
gies and CO2 sequestration. After meetings of the World Climate concepts for meeting future energy challenges. The R&D consor-
Convention in 1988 and 1992, the Third World Summit resulted in tium has 11 members. The four main objectives are the following:
the creation of the Kyoto Protocol, an amendment to the United
Nations Framework Convention on Climate Change. Signatory 1. Advance nuclear safety.
countries agree to cut their GHG emissions by 5.2% compared with 2. Address nuclear non-proliferation and physical protection
1990 and 1995 levels by the years 2008 and 2012, respectively. issues.
Engaging in emissions trading will help some countries achieve the 3. Competitive economics.
goals. Britain is a leader in the field of climate change, setting 4. Minimise waste and optimise natural resource utilisation.
ambitious targets and policies, and is one of only a few nations on
target to meet their Kyoto commitment (United Kingdom Depart- The objectives are further divided into 8 goals, which are dis-
ment of Environment Food and Rural Affairs, 2008). In March 2007 cussed in The Technology Roadmap for Generation IV Nuclear
the UK produced a draft Climate Change Bill aimed at moving the Energy Systems (US DoE and GIF, 2002), published at the end of
UK to a low carbon economy. It aims to cut carbon emissions by 60% 2002. Six reactor types are identified for development, as shown in
by 2050, compared to the 1990 base level. The bill was introduced Table 1
in Parliament in November 2007 and completed its passage
through the House of Lords at the end of March 2008. It will shortly 1.3. The hydrogen economy
go to the House of Commons for consideration. If approved, the UK
will be the first country to have a long range, significant carbon Hydrogen has the potential to solve two major energy chal-
reduction target set into law. Even with a stabilisation or reduction lenges: reducing dependence on the decreasing reserves of fossil
in GHG emissions, anthropogenic warming will continue to rise for fuels and reducing pollution and greenhouse gas emissions.
centuries due to the timescales associated with climate processes. It Hydrogen is the third most abundant element on the Earth’s
is essential that new energy solutions are found which reduce GHG surface, where it is found primarily in water and organic
emissions as soon as possible. compounds. Hydrogen is high in energy, yet an engine that burns
pure hydrogen produces almost no pollution. Hydrogen is not
1.1.2. Nuclear power a primary fuel in the same sense as natural gas, oil and coal, rather,
Nuclear power currently accounts for only about 16% of the hydrogen is an energy carrier like electricity. The hydrogen
world’s electricity supply, and only just over 6% of the energy economy is the vision of building an energy infrastructure that uses
supply (IEA, 2007). In Britain, nuclear power accounted for 20% of hydrogen as an energy carrier. Hydrogen could be an effective
domestic electricity generation in 2005. These figures are relatively means of storing the energy generated by nuclear power or by
low considering the energy potential and the lack of greenhouse intermittent renewable energy sources such as solar power or
gas emissions. The main drawback of nuclear power is the radio- wind.
active waste produced, which, alongside heavily publicised inci- There are many conflicting views on the hydrogen economy. The
dents such as Chernobyl, has led to public concern and doubt. Also, proponents claim that our futures will depend on hydrogen,
nuclear power is currently used almost exclusively to produce produced from fossil fuels in the short term and nuclear and
electricity, which, without significant research and development, renewable sources in the longer term. The critics claim that
cannot be used on a large scale in the transport sector. Without hydrogen will never compete with electricity from fossil fuels or
public, and therefore governmental, support, nuclear power will direct from renewables (Kreith and West, 2004). Accepting that
struggle to find a foothold in future energy supplies. Developing fossil fuels will run out in the not too distant future, and consid-
reactor technology, increasing public awareness and acceptance, ering environmental concerns, everyone agrees that a new energy
502 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

solution is needed in order to provide a stable energy network for of crude oil, such as bituminous shells, residues and oil sands, the
future generations. It is doubtful that the electricity grid could cope use of hydrogen will increase dramatically. Hydrogen is also used in
with the dramatic increase in load that would result from turning to the Fisher Tropsch process to produce a synthetic petroleum
a solely electricity economy, to replace fossil fuels. It is likely, substitute from coal. Coal reserves are expected to last significantly
therefore, that in the future hydrogen and electricity will both be longer than oil reserves, indicating that increasing amounts of
used, tailored to suit individual needs. In the near term it may be hydrogen will be used in this process.
necessary to produce hydrogen from fossil fuels whilst other
technologies develop. Only through continued research and 1.3.1.2. Industrial. There are many industrial uses for hydrogen. The
development will a solution to the energy future be found. direct reduction of iron ore and the production of ammonia for
fertiliser are two examples where hydrogen is used on a large scale.
1.3.1. Markets for hydrogen Ammonia production consumes about half the hydrogen produced
The development of hydrogen technologies involves identifying today (Forsberg, 2007; IAEA, 1999).
potential markets and the constraints involved within them. The
markets can be split into two main categories (Forsberg, 2007): 1.3.1.3. Electrical. The demand for electricity varies from day to day
and season to season. As a result the price of electricity also varies
 large centralised demands for hydrogen and; dramatically. This creates the potential for a hydrogen market that
 decentralised demands. can produce electricity at times of high demand. A hydrogen
intermediate and peak electrical system (HIPES) using nuclear
There are currently two main technologies that easily couple to hydrogen has been proposed (Forsberg, 2007). It consists of
high volume, large scale, centralised hydrogen production: nuclear hydrogen production and storage and peak electrical production.
energy and fossil fuels with carbon dioxide sequestration. The Its feasibility is based on the low cost storage of large quantities of
electricity and heat from nuclear plants can be coupled with elec- hydrogen and the projected costs and efficiencies of fuel cells. The
trolysis or thermochemical cycles. Hydrogen from the steam fundamental problem with renewables is matching electricity
reforming of fossil fuels is a proven technology; however, the CO2 production to demand. If successful, HIPES could be an enabling
produced must be sequestered if there is to be an environmental technology for a nuclear hydrogen-renewables economy; the
benefit. This is not ideal in the long term and the technology still ability to store hydrogen for long periods provides a bridging
needs further development. There are four main markets for mechanism for many renewable energy sources.
hydrogen: transportation, industrial, electrical and commercial.
1.3.1.4. Commercial. Commercially, hydrogen will mainly be used
1.3.1.1. Transport. The transport sector currently runs almost to power fuel cells. Fuels cells convert the chemical energy of the
exclusively on petroleum. Greenhouse gases are not only released fuel fed to them into electricity (Larminie and Dicks, 2003). They
during burning in the internal combustion engine of vehicles, but differ from batteries in that they do not store the chemical
also in the fossil fuel extraction and conversion and the distribution energy, but instead rely on a fuel supply. A fuel cell consists of
of the petrol. There are two main alternatives to petroleum driven a negatively charged anode and a positively charged cathode
vehicles: battery-electric vehicles (EVs) and fuel cell vehicles separated by an electrolyte. Commonly, fuel cells are fuelled by
(FCVs). A comparison of four hypothetical fuel cycles possible for hydrogen, hydrocarbons or alcohols. The most common is the
light vehicles showed that electricity and hydrogen from renewable proton exchange membrane fuel cell (PEMFC) which uses
energy sources are more efficient (on a sun to tank basis) than hydrogen as a fuel, oxygen as the oxidant and a polymer
gasification and liquefaction of biomass (Pro et al., 2005). The membrane, such as Nafion, as the electrolyte. Hydrogen diffuses
comparison also showed that using photovoltaic energy to produce to the anode catalyst and dissociates into hydrogen ions (protons)
electricity and power an EV was slightly more efficient than using and electrons. The electrons travel round an external circuit,
the energy to electrolyse water to form hydrogen and power an providing electricity. The protons diffuse through the electrolyte,
FCV. Electrolysis may be the only way to supply sufficient hydrogen combining with the electrons and oxygen at the cathode to form
for FCVs in the near term. Modelling of the total annual electricity water. The efficiency of a fuel cell is dependent on the current
required to sustain hydrogen production for vehicles shows that through it. Currently, fuel cells have efficiencies of up to around
even with increased efficiency, it will still be necessary to signifi- 50%.
cantly increase electricity generation facilities (Kruger, 2001). EVs
are currently ahead of FCVs in delivering practical, on road 1.4. Hydrogen production
performance at a reasonable cost (Hammerschlag and Mazza,
2005). The main disadvantage of EVs is the battery life; currently Hydrogen does not occur in a free state on the Earth and so will
the maximum range is around 360 km per charge before the have to be manufactured. Even though hydrogen is generally
battery must be recharged for up to 4 h. Hybrid electric vehicles considered to be a clean fuel, it is important to recognise that the
(HEVs), such as the Toyota Prius, operate as an EV for short trips, but method of production plays a very significant role in the level of
also carry liquid fuel making long trips possible. The British environmental impact. There are two main sources of hydrogen,
Department of Transport concluded that further development of fossil fuels and water.
HEVs is the most effective method to reduce GHGs over the next A figure of merit assessment, giving the overall value of
couple of decades (Eyre and Fergusson, 2002). At the present both different routes to hydrogen (Ewan and Allen, 2005a), shows
EVs and FCVs need further development before they compete with a clear divide between renewable energy and more traditional
petrol vehicles in terms of cost and practicality. Only once this is fossil fuel based routes. The assessment is based on four factors:
achieved will they have enough consumer interest to become the carbon dioxide emissions, the primary energy availability, the
transport solutions for the future. land use implications and the hydrogen production cost. Whilst
In recent years there has been a shift from using hydrogen in renewable resources have the lowest CO2 residuals, they also
FCVs to using it to exploit heavy hydrocarbons, as well as biomass, have the highest land use and are the most expensive. Nuclear
to make synthetic motor fuels. Major products of hydrocracking are power, coupled with SMR and capture, or with thermochemical
jet fuel, diesel and LPG, which have a low sulphur content. As oil cycles, shows great potential for future large-scale hydrogen
reserves become depleted and we are left to rely on poorer grades production.
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 503

1.4.1. Hydrogen from fossil fuels 2. Water splitting to produce hydrogen

96% of the hydrogen produced today comes from fossil fuels,
with steam reforming of natural gas accounting for a massive 48% In the long term, water is a much more viable source of
(Ewan and Allen, 2005a). Steam, at a temperature of 700–1100  C, is hydrogen than fossil fuels as no CO2 is released during processing.
mixed with methane gas in the presence of a nickel catalyst. There Two of the most promising production processes for extracting
are two competing reactions; in the first methane reacts stoichio- hydrogen from water are thermochemical cycles and high
metrically with water forming carbon monoxide and hydrogen, temperature electrolysis. Both these processes require a high
whereas in the second an excess of water leads to carbon dioxide temperature heat source, such as a nuclear or solar reactor. The
and hydrogen formation. Often a 300% excess of steam is used so state of the art of two processes is described in the following
that more CO2 is produced, moving the equilibrium in the water– sections.
gas shift reaction and so achieving a higher hydrogen yield and
avoiding carbon deposition due to the Boudouard reaction which is 2.1. Thermochemical cycles
also catalysed by nickel. If the reforming reaction is carried out in
the presence of a CO2 adsorbent, the reaction temperature is low- Thermochemical water splitting is the conversion of water into
ered and hydrogen purity increases with the same methane hydrogen and oxygen using a series of chemical reactions. All
conversion. chemical intermediates are recycled internally within the process
Hydrogen can also be formed by the thermal cracking of natural so that water is the only raw material and hydrogen and oxygen are
gas, the steam gasification of coal and the oxidation and cracking of the only products. The concept of thermochemical production of
heavy oils. The integrated gasification combined cycle (IGCC) hydrogen from water was first studied thermodynamically in the
integrates the steam–coal gasification process in a combined heat 1960s (Funk and Reinstrom, 1966). The decomposition by the use of
and power facility. It is currently considered the cleanest and one of thermal energy is Carnot limited. At standard temperature and
the most efficient coal fuelled techniques. Biomass gasification and pressure the free energy and enthalpy changes for the direct
pyrolysis can also be used to generate a fuel gas that can be splitting of water are DG ¼ 56.7 kCal/gmol and DH ¼ 68.3 kCal/
reformed into hydrogen. gmol, respectively. It is not until around 4400  C that the DG for the
reaction becomes negative (Funk, 2001). At such an extreme
1.4.2. Hydrogen from water temperature there are great problems with materials and separa-
The direct splitting of water to form hydrogen and oxygen tions rendering the direct decomposition infeasible. The work
requires a very large heat input and a temperature of around required for the one step process can be reduced by increasing the
4400  C. In the Plasma Arc Process water is heated to this high operating temperature, however, if an upper temperature of
temperature by an electric field, resulting in its cracking (Stein- 1100  C is imposed the reduction is modest. In a multi-step process
berg and Cheng, 1989). As the process consumes a lot of energy it it is, in theory, possible to reduce the work requirement to zero by
is very expensive and other water splitting technologies are operating reactions with positive entropy changes at high
much more viable. Indirect water splitting technologies, such as temperatures and reactions with negative entropy changes at low
thermochemical cycles and electrolysis, operate at lower temperatures. Considering a two-step cycle it is possible to calcu-
temperatures and have the potential to produce hydrogen from late ideal values of DH and DS in that cycle. Comparing those values
water on a large scale. These technologies are discussed in the with existing chemical products shows that no two-step cycle
next chapter. is possible within a 1100  C limit (De Beni, 1982). Consequently
all possible cycles will have three or more steps. In the 1970s 129
1.4.3. Roadmap for hydrogen production possible thermochemical cycles were published (Bamberger and
Assuming that the CO2 released in the gasification process is Richardson, 1976; Bamberger, 1978). The Joint Research Centre at
released to atmosphere and no cost penalty is paid, in the near ISPRA investigated 24 cycles in more detail, named Mark X. The
term hydrogen from water splitting will be more expensive than third phase of these cycles investigated the sulphur family. The
hydrogen produced from coal gasification (Engels et al., 1987). production of H2SO4 from SO2 is possible in several ways, one being
Water splitting is therefore a long-term energy conversion reaction with a halogen. The only suitable halogens are bromine
process which will come into its own when carbon emissions and iodine, as fluorine and chlorine introduce a high irreversibility
are regulated or correctly costed. This is highlighted in a coor- leading to high energy consumption elsewhere in the cycle (Beghi,
dinated action project recently carried out under the European 1986). The reduction of HBr is an electrochemical step, leading to
Union Framework 6 programme: ‘‘Innovative Routes for High a hybrid cycle. The sulphur family of cycles was also investigated
Temperature Hydrogen Production’’ (INNOHYP-CA). The state of elsewhere (Weirich et al., 1984).
the art in hydrogen production was evaluated (Ewan et al., High temperature processes for hydrogen production are
2007), with the most promising processes being divided into attractive from an efficiency perspective if you consider the second
three groups that make a consecutive market introduction law of thermodynamics, from which is derived h ¼ 1  ðTC =TH Þ,
possible. The first group, for short term pilot demonstration, where TC and TH are the cold and hot temperatures, respectively.
focuses on the enhancement of current state of the art However, some processes also require an input of electrical or
processes, such as solar steam methane reforming (SMR) and the mechanical energy. Electricity is produced with an efficiency hel,
SOLZINC process. The second group, CO2 free processes for so for a required electrical work Wel, an amount of heat equal to
which further development is needed, but that are likely to be Q 0 ¼ Wel =hel is required. The total energy required by the process
realised by 2020, includes the Hybrid Sulphur and Sulphur is the heat required to produce hydrogen, Q, plus that required for
Iodine thermochemical cycles and High Temperature Electrolysis. electricity, Q0 , giving QT ¼ Q þ ðWel =hel Þ. Normalising these values
The final groups, which may have benefits in the very long term, to the production of one unit of hydrogen allows a global process
are a second generation of CO2 free processes including the Zn efficiency to be calculated.
and CuCl thermochemical cycles. A road map of process devel- There has been much debate as to whether the higher heating
opment and cross cutting actions describing the period up to value (HHV) or lower heating value (LHV) of hydrogen should be
2022 was proposed. This report concentrates on the middle used in efficiency calculations. The HHV of 286 kJ/mol is the energy
group – namely water splitting, CO2 free processes, with a high gained from burning hydrogen in oxygen at ambient conditions,
temperature step involved. with the initial and final conditions the same at 1 bar and 25  C. The
504 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Table 2
Maximum thermodynamic efficiencies for the HyS and SI cycles.

Cycle h1 h2
HyS 0.67 0.67
SI 0.51 0.19

LHV of 236 kJ/mol assumes water is produced at 150  C, from an

initial state of 25  C, and the energy of vaporisation not recuper-
ated. In this thesis the HHV has been chosen as the basis. The HHV is
consistent with the first law of thermodynamics, whereas the LHV
is not. Also, using the LHV can lead to efficiencies of greater than
100%, for example in a condensation boiler in the gas industry. The
definition of efficiency used is given in equation below.

HHV HHV
h¼ ¼ (1)
QT Q þ Wel =hel
The efficiency achievable in order for thermochemical cycles to be
competitive with other routes to hydrogen is often discussed. The Fig. 2. The Hybrid Sulphur cycle.
US programme settled on a value of 47% (Brown, 2007), however,
no value has been specified in the European programme. All ther- where water is reacted with iodine and sulphur dioxide in an
mochemical cycles currently being investigated around the world exothermic reaction, at around 120  C, to form sulphuric and
still need improvement before this target is reached. The two cycles hydriodic acids. At certain reactant concentrations, involving an
considered to be the closest to commercialisation are the Sulphur excess of iodine, a phase separation occurs between the two acid
Iodine (SI) cycle and the Hybrid Sulphur (HyS) cycle. products leading to a H2SO4 phase principally devoid of HI and vice
A study of the limiting efficiencies for thermochemical cycles versa. In the second section, sulphuric acid is decomposed in an
(Ewan and Allen, 2005b) gives a useful basis for comparison of the endothermic 2-stage reaction first to SO3 and then to SO2. The first
cycles. The evaluation does not use process flowsheeting and stage occurs at a temperature of 400–500  C, whereas the second
instead is based entirely on thermodynamic considerations. The stage occurs at 800  C in the presence of a solid catalyst. The third
research uses two modified definitions of efficiency, which reflect section involves the decomposition of hydriodic acid to form
the fact that additional measures may be needed to realise the full hydrogen and iodine. This is a slightly endothermic reaction and
potential of the cycle. h1 assumes the additional sources of work are can be conducted in the liquid or gas phase.
included, whereas without these measures the lower efficiency, h2, The SI cycle was first investigated by General Atomic (GA) in the
will be the maximum available. Table 2 shows the values of h1 and 1970s, however, energy prices and economics put a stop to ongoing
h2 for the HyS and SI cycles. research (Norman et al., 1982). In the late 1990s GA restarted
As well as thermodynamic considerations, the eventual feasi- investigating thermochemical cycles, and, finding that the SI cycle
bility of a cycle will depend on the ease of separations, the had the highest predicted efficiency and the greatest potential for
components involved, and the cost of the process. The aim of further improvement, they selected it for further research (Mathias
research into all the cycles is to get as close as possible to the and Brown, 2003). It has since been developed in Japan, Korea and
maximum efficiencies shown. Europe as well as the USA. Current predicted efficiencies are in the
35–45% range. The European Union Framework 6 funded HYTHEC
2.1.1. The Sulphur Iodine cycle (Hydrogen from Thermochemical cycles) project investigated ways
The Sulphur Iodine (SI) cycle, or GA process, involves three to improve the process efficiency (Le Duigou et al., 2007).
separate sections, as shown in Fig. 1. The first is the Bunsen section, Through the US DoE and French-CEA I-NERI (International
Nuclear Energy Research Initiative) agreement, an integrated lab
scale (ILS) experiment of the SI cycle is being developed (Pickard
et al., 2007). The Bunsen section, developed by CEA, and the H2SO4
section, developed by Sandia National Laboratory, have both been
shipped to GA, where the HI section was developed. Closed loop
testing is due to begin this year.
Bench scale experiments of the SI cycle performed at the Japan
Atomic Energy Agency (JAEA) demonstrated a stable production of
hydrogen and further pilot-scale tests are underway (Kubo et al.,
2004a).

2.1.2. The Hybrid Sulphur cycle

The Hybrid Sulphur (HyS) cycle, or Westinghouse cycle, is
a combination of electrolysis and thermochemical processes
(Brecher et al., 1977), as depicted in Fig. 2. It consists of two main
steps, the electrolysis of water and sulphur dioxide at around 87  C
to give hydrogen and sulphuric acid, followed by the decomposi-
tion of sulphuric acid first to sulphur trioxide and steam and then
further to sulphur dioxide and oxygen. The sulphuric acid decom-
position stage is in common with the Sulphur Iodine cycle. It occurs
Fig. 1. The Sulphur Iodine cycle. at around 800  C so a high temperature heat source such as an
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 505

advanced nuclear reactor is needed. Having sulphur dioxide

present in the electrolyser reduces the required electrode potential
for water splitting from a theoretical value of 1.23–0.17 V, therefore
reducing the amount of energy required significantly compared to
traditional electrolysis. Optimisation of the flowsheet suggested
that an efficiency of 47% (LHV, 55% using HHV) is possible using
a decomposition temperature and pressure of around 930  C and
10 bar (Jeong et al., 2005). Another recent flowsheet optimisation
estimates the value to be 49.4% (HHV) (Gorensek et al., 2005). The
cost of hydrogen production for an Nth of a kind (NOAK) nuclear
hydrogen production plant is estimated to be $1.64/kg for a baseline
case. There is much scope for further improvements in efficiency by
reducing the electrode potential and investigating structural
materials stable at the high temperatures involved. The European
Union Framework 7 funded HycycleS project, which started in
January 2008, will investigate high temperature materials and
Fig. 3. The thermodynamics of high temperature electrolysis.
catalysts as well as the H2SO4 decomposition reactor and product
separator. At Savannah River National Laboratory, a single cell
electrolyser has been built and operated for 100 h (Summers et al.,
generation reactions, and significant development work is
2007). Testing of a multi-cell electrolyser is planned for this year.
required.
Westinghouse are also investigating the electrolyser and decom-
position reactor (McLaughlin et al., 2005).
2.1.4. Two-step thermochemical cycles
If an upper temperature limit is not applied to cycle selection,
2.1.3. Other multi-stage thermochemical cycles
the chemical constraints for chemical compounds to form via two-
2.1.3.1. The UT-3 cycle. The UT-3 cycle, developed at the University
step cycles are less stringent. The higher temperatures required
of Tokyo, consists of four reactions involving metallic bromides
could feasibly be supplied by solar power, however, are not
(Kameyama and Yoshida, 1978). The hydrolysis and bromination of
currently possible using nuclear technology. Metal oxide processes
Ca and Fe compounds are carried out in the gas–solid state. Much
involving a metal oxide redox pair show promise for two-step cycle
work was carried out on the cycle in the 80s and early 90s, leading
operation. Thermodynamic analysis of possible redox pairs indi-
to an adiabatic form of the cycle being developed which simula-
cated that Fe3O4/FeO and ZnO/Zn have the greatest potential for
tions showed had a higher efficiency than previous versions. The
two-step water splitting. Other pairs, such as Mn3O4/MnO and
combination of a HTGR (high temperature gas cooled reactor), the
Co3O4/CoO, are not thermodynamically able to produce hydrogen
UT-3 cycle, and an electric power generating system was suggested
in significant quantities over a wide temperature range (Kodama
to have an efficiency of around 49% (Sakurai et al., 1996). Further
and Gokon, 2007).
work showed that using a membrane gas separator was beneficial
to both the process efficiency and economics (Tadokoro et al., 1997)
and the Model Apparatus for the Study of Cyclic Operation in Tokyo 2.2. High temperature electrolysis
(MASCOT) pilot plant was operated in the late 80s. However, more
recent work has shown that these efficiencies were gross over Around 4% of the world’s hydrogen is currently produced by
estimates due to some of the assumptions made (Teo et al., 2005). conventional, low temperature, water electrolysis. An electric
The upper bound efficiency was shown to be 37.3% (LHV), equiva- current is passed through the water causing it to dissociate into
lent to 44.3% (HHV). Taking into account other process uncer- hydrogen and oxygen. Researchers have been investigating using
tainties, the efficiency was predicted to be only 13%, much too low solar energy and wind to power a conventional electrolyser to
to be economic. produce hydrogen (Glatzmaier and Blake, 1998). The efficiency of
The physiochemical properties of the solid and gaseous reac- converting electricity to hydrogen in an electrolyser is as high as
tants, for example the sintering of the solid, along with the toxicity 80%, however, the efficiency of converting heat to electricity is less
of the reactants, make the practicalities of the process very difficult than 40%, yielding an overall efficiency of less than 35%. High
(Lemort et al., 2006). A new flowsheet was developed which temperature electrolysis of steam, at 800–1000  C, has several
employed two asymmetric torus reactors with fluidised beds of advantages over the low temperature alternative; the thermody-
solid reactants in each leg. This has the advantage of avoiding the namic electric energy required is reduced, as shown in Fig. 3, and
energy intensive reactant preparation step and also improves the activation barrier at the electrolyte surfaces is easier to over-
the reaction kinetics. Using membrane separation techniques on come, resulting in an improved efficiency. However, significant
the gas streams, the efficiency was predicted to be 22.5%. Advances problems exist in constructing stacks that have long lifetimes for
in both membrane and process technology need to be made in the sophisticated catalysts required.
order for the process to be a viable route to hydrogen production. High temperature electrolysis (HTE) is based on the technology
of solid oxide fuel cells (SOFCs), which have been the subject of
2.1.3.2. The hybrid copper chloride cycle. The hybrid copper chlo- much R&D over the last 20 years. Whereas SOFCs consume
ride cycle is being developed at Argonne National Laboratory in the hydrogen and oxygen to produce heat and electricity, solid oxide
USA (Lewis and Masin, 2005). It has the advantage that the highest electrolytic cells (SOECs) consume electricity and steam and
temperature step operates at only 550  C, some 300  C lower than produce hydrogen and oxygen. The mechanism of HTE is shown in
the HyS and SI cycles. The cycle comprises three thermal reactions, Fig. 4. Before entering the electrolysis cell, water is heated to form
in which hydrogen, oxygen and HCl are generated, and an elec- steam. The steam is supplied to the cathode side, where the
trochemical step in which CuCl forms CuCl2 and metallic Cu. The application of a voltage breaks it down to give product hydrogen
idealized efficiency is estimated to be 42% (LHV, about 49% using and oxygen ions. The oxygen ions migrate through the electrolyte
HHV), however, there are significant gaps in the thermodynamic to the anode where they give up electrons to form product oxygen.
knowledge and on the optimum operation of the HCl and O2 The voltage required is about 0.3 V lower than in conventional
506 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Table 3
Comparison of thermal-to-hydrogen technologies (Yildiz and Kazimi, 2006).

Process Nuclear reactor Operating range

Minimum Maximum

Temp ( C) Efficiency Temp ( C) Efficiency
HTE AGR þ S CO2 650 0.52 750 0.56
HTE MHR þ GT 850 0.50 950 0.53
HyS MHR 800 0.53 900 0.57
SI MHR 800 0.38 900 0.52

greenhouse effect, otherwise the advantages gained by the clean

technology are negated by the power source. There are two main
contenders: nuclear power and solar energy.
As technology improves, lower temperature cycles may become
more attractive, therefore opening the door to other heat sources.

2.3.1. Solar
Concentrated solar radiation can be used as a high temperature
Fig. 4. High temperature electrolysis.
energy source for endothermic reactions in thermochemical cycles.
The large-scale concentration of solar energy is achievable using
three different optical configurations of parabolic reflectors:
electrolysers due to the high operating temperature. In addition the trough, tower and dish systems (Steinfeld, 2005). Using these
kinetics are faster than at room temperature and so polarisation systems the solar energy can be converted to thermal reservoirs at
losses are avoided. 1700  C. The high achievable temperatures mean that solar energy
HTE is being developed in both the USA and Japan. Idaho could be used to power the efficient 2-step thermochemical cycles
National Laboratory (INL) and Ceramatec have a partnership using metal oxides, such as the ZnO/Zn cycle (Kodama and Gokon,
developing the electrolyser (Herring et al., 2007a). Initial experi- 2007). The application of solar energy to a thermochemical cycle of
ments were carried out using a button cell, the results being then mixed iron oxides has been investigated (Roeb et al., 2006). The
used to create the first planar 10-cell stack. Hydrogen production iron oxides are coated on multi-channelled honeycomb ceramic
rates up to 90 NL/h were demonstrated (Herring et al., 2007b). supports which are capable of absorbing solar irradiation. This
A dual stack was then created which produced hydrogen at rates in allows the whole process to take place in a single receiver-reactor.
excess of 200 NL/h (O’Brien et al., 2007b). The stack was operated In the Sulphur Iodine cycle, both the extent and rate of H2SO4
for over 2000 h and experienced a w45% degradation in perfor- decomposition are improved at the very high temperatures
mance (Herring et al., 2007a). Post test evaluation of the cells is achievable with solar energy.
being carried out to investigate causes of degradation (Mawdsley The application of a combined energy supply of solar and fossil
et al., 2007). The dual stack represents a half module of the Inte- fuels would enable the Sulphur Iodine process to operate contin-
grated lab scale (ILS) experiment, which produced its first hydrogen uously (Giaconia et al., 2007b). A solar concentrator plant supplies
at the end of September 2007. Large-scale tests will be carried out a thermal load at medium temperatures (<550 C) and a methane
using this apparatus over the next year. CFD modelling using furnace provides the heat required by the high temperature sul-
FLUENT is being used to determine the best ways to reduce thermal phuric acid decomposition section. An additional section is added
stresses over the area of the cells and to assure good steam distri- to the flowsheet which combines the CO2 released with a fraction of
bution (Hawkes et al., 2007). During the half module test, the stack the hydrogen produced to form methanol.
was also operated in a co-electrolysis mode (called syntrolysis), High temperature electrolysis requires significant amounts of
where steam and carbon dioxide are supplied to the stack and electricity as well as high temperature heat. It is therefore less
a mixture of carbon monoxide and hydrogen called syn gas is suited to coupling with a solar reactor.
produced. This can be used as a feed in the Fisher Tropsch process The main disadvantage of solar energy is its varied concentra-
to produce synthetic fuels. The results were very promising and tion across the world. In Britain, where the sun index is low, there is
overall efficiencies of 43–48% were predicted (Stoots et al., 2007; not enough solar capacity for large-scale hydrogen production. The
O’Brien et al., 2007a). high temperature heat must therefore be provided by nuclear
Modelling work in Japan suggests that, when coupled to a HTGR, energy.
hydrogen production efficiencies of over 53% are accomplishable
(Fujiwara et al., 2008). A unit of 15 tubular cells was constructed 2.3.2. Nuclear
and achieved a hydrogen production rate of 130 NL/h. Good Coupling high temperature electrolysis or a thermochemical
agreement was seen between cell performance and simulations cycle to a high temperature nuclear reactor will provide the heat
prediction using the CFD code STAR-CD. The development of required without producing greenhouse gas emissions. Generation
several further cells is currently underway. IV reactors are ideal for coupling to a hydrogen production plant
(HPP) due to their inherent safety and very high temperatures.
Referring back to Table 1, the VHTR and GFR both reach tempera-
2.3. High temperature heat sources tures of over 800  C, high enough for use with the SI or HyS cycles.
An interesting comparison of thermal-to-hydrogen efficiency of
A key feature of high temperature electrolysis, and of many HTE, SI and HyS related technologies has been carried out and is
thermochemical cycles, is that they all have a stage requiring very summarised in Table 3 (Yildiz and Kazimi, 2006).
high temperatures. In fact, HTE, and the HyS and SI cycles all require The Modular Helium Reactor (MHR) is a type of VHTR, as
temperatures in excess of 800  C. It is important to obtain the heat explained in the next chapter. The first HTE option, which shows
from a source that does not contribute to emissions causing the high efficiencies, operates at a lower temperature than the others
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 507

3. Power conversion systems

There are two main types of power generation system used to

convert heat into work: the Rankine cycle and the Brayton cycle.
Both cycles are outlined below. The choice of which is coupled to
the nuclear reactor depends on several factors, and different
companies have made different selections. The choice of cycle can
also differ depending on whether the nuclear plant is for solely
electricity production, both electricity and hydrogen production, or
solely hydrogen production. This chapter presents the science
behind each cycle, and individual choices are covered in chapter 4
alongside their corresponding reactor configurations.

Fig. 5. Schematic of nuclear hydrogen future (US DoE NERI, 2006). 3.1. Rankine cycle

Almost all existing coal and nuclear power stations use the
and uses an Advanced Gas Reactor (AGR). There are 14 AGRs in Rankine cycle for electricity generations. Heat is applied externally
operation in the UK. The efficiencies described in the table below to a closed loop, usually of water, as shown in Fig. 6. The heated
assume coupling with a supercritical CO2 recompression Brayton water passes through a turbine, and is then condensed and pumped
cycle (see Section 3.2), however, no such systems have currently back to the heat exchanger. High efficiencies are achieved as the
been built. The second HTE option, which operates at higher water is liquid when pumped, which requires about 100 times less
temperature, utilises a helium gas turbine system. At the lower end energy than compressing a gas. The actual efficiency depends on
of the temperature range for a VHTR, the HyS cycle shows signifi- the temperatures employed. Typically the turbine inlet tempera-
cantly higher efficiencies than the SI cycle. However, as the ture is about 565  C, the creep limit of stainless steel, and the
temperature increases the efficiency of the SI cycle improves condenser temperature is around 30  C. As the turbine entry
greatly. The efficiencies given highlight the importance of the temperature is relatively low, the Rankine cycle is often used as
temperature achievable by the nuclear reactor and the coupling a bottoming cycle in combined cycle gas turbine power stations.
method used. Much work is being carried out around the world to The Rankine cycle is also sometimes referred to as a practical
investigate both of these. Carnot cycle, as the T–S diagram resembles that of a Carnot cycle if
A generalised schematic of a ‘‘nuclear hydrogen future’’ is given an efficient turbine is used. For a Carnot cycle, the lines on the T–S
in Fig. 5. Hydrogen is produced in a centralised production facility, diagram would be either horizontal or vertical, however, in order to
stored temporarily if necessary, and distributed to the end users. achieve acceptable efficiencies in the Rankine cycle, the fluid
A study by a US team, funded by the DOE Nuclear Energy Research pumped must be all liquid. There are four processes in a Rankine
Initiative (NERI), considered the economics and infrastructure of cycle, as seen in Fig. 6 and identified on the T–S diagram in Fig. 7a
a generic Nth of a kind (NOAK) commercial nuclear hydrogen plant (Sandler, 1999).
(U.S. Department of Energy Office of Nuclear Energy Science and
Technology Nuclear Energy Research Initiative (NERI), 2006). They  1–2, purple: the liquid is pumped from low to high pressure.
concluded that ‘‘.nuclear hydrogen production is a technically  2–3, red: the high pressure liquid enters the heat exchanger,
feasible and economically viable option for addressing future where it is heated to become a dry, saturated vapour.
national energy needs’’. No significant technical or economic  3–4, green: the vapour expands through a turbine, decreasing
barriers were identified. Further results from the study will be the temperature and pressure and generating power. Some
presented later in this report. condensation may occur.
There are four distinct areas of a nuclear hydrogen production  4–1, blue: the fluid is condensed at constant temperature and
plant: the nuclear reactor, the hydrogen production plant, the pressure.
power conversion system and the intermediate heat exchange loop.
These will be discussed in the following chapters. In an ideal process both the pump and turbine would be isen-
tropic and therefore maximise the work output. In reality, however,
these processes are non-reversible and entropy is increased. The
efficiency of the turbine is limited by water droplet formation; as
the water condenses, droplets hit the turbine blades at high speed
causing erosion which gradually decreases the turbine efficiency.
The simplest way to overcome this is by superheating the steam, as
shown by points 30 and 40 in Fig. 7b. This creates a dryer steam after
expansion, however, it also increases the condenser load.
There are two common modifications used to improve the
efficiency of the standard Rankine cycle. Fig. 8 shows the Rankine
cycle with reheat, in which two turbines are used in series. High
pressure steam from the boiler passes through the first turbine and
then goes back to the boiler to be reheated. It then goes through the
second turbine which operates at a lower pressure. This both
improves the efficiency and prevents vapour condensation on the
turbine blades.
Fig. 9 shows the regenerative Rankine cycle. A regenerator, or
open feedwater heater, is used as a mixing chamber. The feedwater
is the water leaving the pump (state 2), which is mixed with
Fig. 6. Rankine cycle. a portion of the high temperature steam leaving the turbine (state
508 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Fig. 7. T–S diagram of (a) a typical Rankine cycle and (b) a Rankine cycle with superheating.

4), to form a saturated liquid (state 7). The regenerator is shown by cycle with reheating, regeneration and intercooling. The maximum
the orange lines in Fig. 9. A second pump is then used to increase efficiency is attained when equal pressure ratios are maintained
the temperature and pressure before the liquid goes to the boiler. across each compressor and each turbine stage. The main advan-
This regenerative Rankine cycle is commonly used in modern tage of multi-staging is that the high efficiencies associated with
power stations. low pressure ratio regenerative cycles can be achieved without the
very large regenerator required for a single-stage cycle with the
3.2. Brayton cycle same power output. Thermodynamic analysis shows that the high
efficiencies obtained by using two compressors are not significantly
The coupling of a Brayton cycle to a nuclear power plant has yet improved if more compression stages are used (Herranz et al.,
to be demonstrated, however, the technology is used in other areas. 2006). The cycle shown in Fig. 11 is ideal – the turbines and
The Brayton cycle, also known as the gas turbine, differs from the compressors are isentropic and the heaters and coolers are isobaric.
Rankine cycle in that the fluid is vapour in the entire cycle (Sandler, In practice, there are pressure losses in the ducts between
1999). For coupling to a high temperature nuclear reactor, helium is components and across the heat exchangers. This reduces the
the most common working fluid. In a simple Brayton cycle, efficiency as the pressure drop across the turbine is decreased. On
a turbine and compressor are mounted on the same shaft, as shown the T–S diagram the lines for the turbines and compressors would
in Fig. 10. The gas is first heated, then the turbine extracts work not be vertical and the lines for the heaters and coolers would jump
from the high temperature and pressure gas before it is cooled and between nearby isobars.
passed through the compressor. The turbine and compressor stages can be on a single shaft or
In the same way as for the Rankine cycle, the power and effi- several shafts, which can be mounted either horizontally or verti-
ciency of the Brayton cycle can be increased by using different, cally. Horizontal mounting minimises the load on the bearings,
more complicated configurations. Reheating can be used to however, it has a larger footprint.
increase the power of the cycle. After passing through the turbine, The Brayton cycle can be operated either directly or indirectly
the gas is reheated before going through a second turbine. Inter- coupled to the nuclear reactor. Direct coupling has the advantage of
cooling utilises two stages of compressors, with an additional simplicity, the use of only one hot helium loop and no need for an
cooler in between them. This decreases the overall energy required intermediate heat exchanger. Indirect coupling, however, presents
by the compression stages, however, does utilise more energy from less safety issues and allows more flexibility.
the heat source. Reheating and intercooling are both examples of Research is also being carried out investigating operating
multi-staging, where more than one turbine and/or compressor are a Brayton cycle with supercritical CO2 as the working fluid. The
used. Regeneration, or recuperation, improves the efficiency by main advantage is comparable efficiency to the helium Brayton
utilising waste heat. The hot gas leaving the turbine passes through cycle but at significantly lower temperatures (Dostal et al., 2004). In
a heat exchanger to preheat the fluid re-entering the heat source. fact, the temperature required for comparable efficiency is only
A thermodynamic evaluation showed that regenerative reheating 550  C, at a pressure of 20 MPa, making it compatible with many
significantly enhances the thermal performance of a reference different nuclear reactor types. The high efficiency is achieved by
Brayton cycle (Herranz et al., 2007). Fig. 11 shows an ideal Brayton taking advantage of the abrupt property changes near the critical

Fig. 8. Rankine cycle with reheat.

 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 509

Fig. 9. Regenerative Rankine cycle.

point of CO2 and so reducing the compression work. Recent according to the degree of complexity and proof of technology.
advances in turbo machinery have made it possible for the first Group A shows the lowest efficiencies, however, are only moder-
time to affordably evaluate and experimentally demonstrate the ately complex and involve predominately proven technology. In
technology required and Sandia National Labs have set up a test contrast, group C shows the highest efficiencies, but are highly
loop to do a series of experiments on an appropriate scale. Two complex and not nuclear-proven technology.
papers will be presented on this work later in the year (Wright The majority of research centres have chosen to use either
et al., 2008a,b). proven Rankine cycle technology at the expense of high efficiency,
or to use the as yet unproven Brayton cycle which gives higher
efficiencies. Clearly, as technology develops, higher efficiencies are
3.3. Combined cycles
achievable, particularly if a combined cycle is employed. The cycles
chosen by individual centres are discussed in the following chapter.
As the Brayton cycle operates at high temperatures and pres-
sures it is ideal for combining with a Rankine cycle. In such
a combined cycle, a topping Brayton cycle and bottoming Rankine 4. Nuclear technologies for hydrogen production
cycle are used, as shown in Fig. 12. This has the advantage of high
process efficiencies, however, has not yet been demonstrated in Nuclear heat has great potential for coupling to a hydrogen
combination with nuclear technology. production plant (HPP) utilising HTE, the HyS cycle or the SI cycle.
Referring back to Table 1, there are two types of nuclear technology
which can achieve the high temperatures required: the VHTR and
3.4. Comparison of power conversion systems
the GFR. A great deal of research is being carried out around the
world on VHTRs, whereas the GFR has received less interest to date.
The influence of the type of cycle, coupling method, steam
The VHTR is a high temperature gas cooled reactor (HTGR) which
parameters, reheating and intercooling (IC) has been investigated
operates at outlet temperatures of 750  C and higher.
thermodynamically (Frohling et al., 2002). The results of the study,
based on a 200 MWt reactor, are shown in Fig. 13. The efficiencies
shown are first-order approximations, and whilst they demonstrate 4.1. The high temperature gas cooled reactor (HTGR)
the trends a more detailed study is required as well as a cost
analysis. There is significant potential for increasing the efficiency HTGR technology has been developing over the last 50 years and
by increasing the reactor core and steam temperatures. Improving so far seven plants have been designed, constructed and operated.
the efficiencies of the turbo-machinery and pumps would also have The HTGR concept evolved from early air and CO2 cooled reactors
a significant effect. The cycles are arranged into three groups, (LaBar et al., 2004). The use of helium as a coolant, along with

Fig. 10. A simple Brayton cycle.

510 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Fig. 11. Ideal Brayton cycle with regeneration and intercooling.

a graphite moderator, offers enhanced neutronic and thermal effi- was a first-of-its-kind pebble bed reactor which had an average
ciencies. Table 4 summarises the HTGRs built to date and those outlet temperature greater than 950  C for more than 10 years
currently being developed. Korea have also started a HTGR pro- (IAEA, 1999). A large number of fuel elements were tested during
gramme, however. have not yet chosen the final reactor design. The this time. The AVR was followed by the Thorium high temperature
early reactors demonstrated key safety features and the use of reactor (THTR), which was designed to demonstrate the viability of
TRISO coated fuel particles, where ceramic coatings are used to sub systems and hardware designs (Breitenfelder et al., 1980). In its
contain the fission products at high temperatures. This fuel design short active lifetime, many of the safety features were demon-
is a key feature of HTGRs. The value of the inherent safety features strated, however, it was shut down prematurely, primarily due to
of this type of small HTGR were only fully recognised after the 1979 political reasons. The Germans also began design of the HTR-100
Three Mile Island accident in the USA. Following the early devel- and HTR-500 reactors, with electrical outputs of 100 MWe and
opment of HTGRs, General Atomic received orders for several large 500 MWE, respectively. The thermal rating of the HTR-500 was
2000–3000 MWt plants. The orders were, however, cancelled a massive 1390 MWt (Wachholz, 1988). Neither were constructed
following the oil embargo in the 1970s and the resulting reduced as the German nuclear programme was shut down.
energy demand and reduced need for new electricity generation. The Fort St Vrain reactor was successful in demonstrating the
Recently there has been much renewed interest in HTGRs, partic- overall systems performance as well as design features of the HTGR
ularly as they are included in the Generation IV programme. At the (Bramblett et al., 1980). Many problems were encountered during
end of the last century, the HTTR and HTR-10 were built in Japan start-up and power rise experiments, particularly connected with
and China, respectively. Current reactor designs are based on the the circulators and with temperature fluctuations. In 1989 the plant
experience gained from the early reactors. was shut down due to continuing problems and it became the first
nuclear plant in the USA to be decommissioned.
4.1.1. The early reactors
Much of the technology employed today is based on that from 4.1.2. Reactor developments
the AVR, THTR and Fort St Vrain reactors. In the 1970s and 1980s In order to maintain the coated fuel temperatures below
Germany was a world leader in HTGR technology. The 46 MWt AVR damage limits, original HTGR designs were limited to about

Fig. 12. Single shaft, direct, combined cycle.

 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 511

Switzerland (Kuijper et al., 2006). Benchmark calculations and cold

critical experiments for fresh low enriched uranium (LEU) particles
were carried out. The results were fed into the HTR-N project, part
of the European Union 5th Framework Programme, investigated
high temperature reactor nuclear physics, and waste and fuel cycles
(Kuijper et al., 2006). The Proteus facility is now being used to
investigate Light Water Reactors (LWRs).

4.1.4. Computations
HTGR designs present computational challenges due to their
core physics and thermal–hydraulic characteristics, for example,
the gas flow in a high temperature, high burn up, porous ceramic
core is unique to the HTGR design. Validation of the HTGR codes
used is essential for the assessment of design uncertainties and
therefore for the development and licensing of the reactors.
Computational benchmarks have been proposed by the IAEA
coordinated research project (CRP-5), calculated by comparisons of
different models and data available (Colak et al., 2006; Wu and Yu,
2007). Data from the HTTR, HTR-10, ASTRA and PBMM (Pebble Bed
Micro Model) were included. The benchmarks provide a useful
framework for assessing computational uncertainties. As well as
experimental studies, the HTR-N project also involved the valida-
Fig. 13. Efficiency comparison for different PCSs (Frohling et al., 2002). tion of computational tools and the qualification of models. The
activities were centred on the two existing HTGRs, i.e. the HTR-10
and HTTR, as well as results obtained from HTR-PROTEUS.
200 MWt due to passive heat decay removal. However, plants of
Computational results agreed well with the data produced exper-
this size were not predicted to be economically competitive. This
imentally. Analyses were also performed on a number of concep-
led to the development of an annular core concept, which gives
tual HTGR designs, derived from reference pebble bed and
a maximum reference core power of 600 MWt. The resulting HTGR
hexagonal block type HTGRs. Several HTGR concepts were shown
design represents a fundamental change in reactor design and
to be promising as systems for the incineration of plutonium and
safety philosophy. The PBMR, GTHTR300, HTR-PM, GT-MHR and
possibly minor actinides. The 5th Framework programme funded
ANTARES designs, which are currently being developed, all use the
several other projects on HTGR technology which led on to the
annular core design. These reactors, as well as the HTTR and HTR-
RAPHAEL IP, funded by the 6th framework programme and
10, are discussed in Sections 4.1.1–4.1.5.
described in Section 4.1.10.
4.1.3. Critical facilities
4.1.5. HTR-10 and HTR-PM
Many features of HTGR designs have been tested in critical
The HTR-10 is built on the site of Tsinghua University’s Institute
facilities. There are two centres outside of the main reactor pro-
of Nuclear Energy Technology (INET) near Beijing in China and is
grammes, the ASTRA and HTR-PROTEUS facilities. The ASTRA
part of the Chinese national High Technology R&D Programme
facility at the Kurchatov Institute in Russia is used for experimental
(Zhang and Yu, 2002; Xu et al., 2005). Design began in 1992,
investigations of the neutronics and criticality safety of HTGR
construction commenced in 1995, and the reactor was completed
reactors (Ponomarev-Stepnoi et al., 2003). The assembly is essen-
in 2000. First criticality was reached in December 2000 and full
tially a reactor-grade graphite block shaped as a cylinder, the
power operation began in January 2003. The objectives of the HTR-
central cavity of which is filled with spherical elements. Fuel
10 are to verify and demonstrate the technical and safety features of
elements, along with graphite and absorbing elements, are
the modular HTGR and to establish an experimental base for
employed in different core configurations. The facility was used to
developing nuclear process heat applications (Hu et al., 2006).
simulate early HTGR designs in Russia and has since been used to
The HTR-10 is a pebble bed reactor and uses fuel pebbles of
simulate the PBMR reactor and the GT-MHR project. The HTR-
graphite spheres of around 6 cm diameter dispersed with TRISO
PROTEUS experimental programme was carried out from 1992 to
fuel particles. The reactor core is a 1.8 m diameter void in the
1996 at the Proteus critical facility in the Paul Scherrer Institut in
graphite reflector and contains around 27,000 fuel elements (IAEA,
2003). In the side reflector are holes which guide the control rod
Table 4
system. The fuel elements move through the core in a ‘‘multi-pass’’
HTGR plants operated and those in development.
pattern, enabling a relatively uniform burn up distribution in the
Reactor Location Power (MWt) He temp. Core Operation years core. Fuel pebbles are continuously discharged through a pneu-
in/out ( C)
matic pulse single exit gate at the bottom of the core and are either
Dragon UK 20 350/750 Cylindrical 1965–1975
discarded or returned to the top depending on their burn up. The
Peach Bottom USA 115 377/750 Cylindrical 1967–1974
AVR Germany 46 270/950 Pebble bed 1968–1988
helium coolant flows downwards through the bed at a rate of
Fort St Vrain USA 842 400/775 Hexagonal 1976–1989 4.3 kg/s (Kuijper et al., 2006). Despite its small power size, the HTR-
THTR Germany 750 270/750 Pebble bed 1985–1989 10 is nearly a 1:1 scale test of a modular HTGR and so is a repre-
HTTR Japan 30 395/950 Hexagonal 1998 sentative test for the passive decay heat removal. Five safety
HTR-10 China 10 250–300/ Pebble bed 2000
verification experiments were carried out in October 2003 which
700–950
PBMR SA 500 350/950 Pebble bed In development demonstrated and verified many inherent safety features of the
ANTARES France 600 400/850 Hexagonal In development modular HTGR.
GT-MHR USA/Russia 550–600 490/950 Hexagonal In development There are two operational phases for the HTR-10. In the first,
HTR-PM China 450 250/750 Pebble bed In development
which was employed at start-up and operated for several years, the
GTHTR300 Japan 600 589/850 Hexagonal In development
plant operated with a helium outlet temperature of 700  C and inlet
512 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

of 250  C. The secondary circuit included a steam turbine cycle for (Takizuka et al., 2004). The deployment roadmap for the GTHTR300
electricity generation with the capability for district heating. The contains three phases: plant design and technologies R&D which
reactor core and steam turbine were housed side by side in separate was carried out at JAEA up to the end of 2007 and was funded by
steel pressure vessels. In the second phase, named HTR-10GT, the Japanese government; prototype plant demonstration which
a direct helium gas turbine Brayton cycle replaces the steam cycle. will be carried out from 2008 to 2018 and funded mainly by the
The reactor will be operated with a core outlet temperature of private sector; and finally the full deployment phase. The primary
900  C and an inlet of 300  C. This phase is currently at the design system is housed in three steel pressure vessels: the reactor, the
optimisation stage (Guojun et al., 2007). The gas turbine, which is power conversion unit (PCU) and the heat exchanger vessel.
vertically mounted and housed in a separate pressure vessel, A containment vessel for the reactor is not necessary as it is
employs 2-stage compression (Wu and Yu, 2007). designed to be severe-accident-free, that is no large amount of
In 2004, INET signed an agreement with the China Huaneng fission product release from fuels in any postulated accident.
Group (CHNG) and China Nuclear Engineering and Construction A passive heat removal system is in place in case of the loss of
Corporation (CNECC) to build a HTGR demonstration plant using the forced cooling. The core design is essentially the same as that used
pebble bed modular technology, named HTR-PM. The HTR-PM will in the HTTR, although the fuel columns are arranged in an annular
be similar in design to the HTR-10, however, will consist of two ring of internal diameter 3.6 m and external diameter 5.5 m. Each
250 MWt reactors, each having an annular core with one central fuel column contains eight layers of fuel blocks, which in turn
column of moveable graphite spheres (Zhang et al., 2006; Wu and contain 57 fuel rods each. A new refuelling method termed
Yu, 2007). Original plans were to use a Brayton cycle for electricity ‘‘sandwich shuffling’’ has been proposed. The design of the PCU is
generation, however, the technology was deemed to be too uncer- based on a regenerative, closed Brayton cycle using helium gas (Yan
tain at that time and so a Rankine steam cycle will be used instead. et al., 2003). No intercooling is used, as this simplifies the cycle,
Once tested and verified in the HTR-10GT project, the gas turbine however, does give a 2% efficiency reduction (Kunitomi et al., 2004).
may be used in the HTR-PM in the future. There are many technical The system includes one turbine and one compressor on
and engineering problems which need to be solved over the next a common, horizontally mounted shaft. There are several major
few years in order to achieve the target of completion in 2012. R&D issues which need to be overcome in order to demonstrate the
feasibility of the design.
4.1.6. HTTR and GTHTR300
The High Temperature Engineering Test Reactor (HTTR), con- 4.1.7. PBMR
structed at the Japan Atomic Energy Agency (JAEA, formerly the The Pebble Bed Modular Reactor (PBMR), under development by
Japan Atomic Energy Research Institute, JAERI) site at Oarai in PBMR Pty (Ltd), Eskom and partners around the world, has been in
Japan, is a graphite moderated helium cooled reactor with development since 1994 (Koster et al., 2003). The South African
a maximum outlet temperature of 950  C. Development work government provided significant funding for the project and stated
started as early as 1969 under the Japanese Long-term Programme an objective of producing 4000 MWe from PBMRs (Matzner and
for Research, Development and Utilisation of Nuclear Energy. The Wallace, 2005). The initial design was for a 268 MWt reactor with
reactor first attained criticality in November 1998 (Ogawa and a core consisting of fuel spheres and a dynamic central reflector
Nishihara, 2004). This was followed by ‘‘rise to power’’ tests which column of graphite spheres. However, as more insight on graphite
resulted in full power operation with an outlet temperature of behaviour was gained, the core structure was redesigned to allow
850  C in December 2001, and high temperature operation at a possible mid-life replacement of the reflector layer adjacent to the
950  C in April 2004 (Nakagawa et al., 2004; Fujikawa et al., 2004). core. This allowed the design to implement a fixed central column.
The objectives of the HTTR are to establish HTGR and nuclear heat The core outer diameter is 3.7 m and inner diameter 2 m. Inside the
utilisation technologies, to carry out basic research on high core are around 452,000 fuel pebbles consisting of TRISO fuel
temperature irradiation and to develop the technology for high particles encased in a graphite sphere (Wallace et al., 2006). The
temperature heat applications such as hydrogen production helium coolant flows at a rate of 185 kg/s. Online refuelling is a key
(Shiozawa et al., 2004). feature of the PBMR and on average each fuel pebble will make six
The active core of the reactor is 2.3 m in diameter and contains passes through the reactor. The aim is to operate uninterrupted for
30 fuel columns and 7 control rod guides. It is surrounded by 6 years before the reactor is shut down for a nominal 30 days of
a reflector which also contains control rods. Each fuel column is scheduled maintenance. The reactor is housed in a building, part of
made up of hexagonal shaped pin-in-block type fuel elements, which is a thick reinforced concrete containment structure called
consisting of a graphite block with 33 fuel rods inserted in holes the citadel. Much of the PBMR can be built in the factory instead of
leaving small gaps for the helium coolant to pass by the fuel. The onsite leading to lower costs and time frames.
fuel rods are TRISO particles imbedded in graphite. The helium A direct helium Brayton cycle is to be coupled to the PBMR for
flows downwards through the reactor at a rate of 10.2 kg/s for electricity production (Matzner et al., 2006). The helium turbine is
950  C operation. To control over-reactivity, burnable poison is a high-technology item and has some similarity with a jet engine.
applied to some of the fuel blocks. The reactor operates in batch Mitsubishi Heavy Industries, one of the world’s major producers of
mode so after each cycle part of the fuel is replaced. The HTTR has power turbines, is being employed to use their experience and
two modes of loop operation, single and parallel loaded mode. In expertise to develop the gas turbine. As the PBMR is the first reactor
the single loaded operation, only the main cooling system operates that will use a direct Brayton cycle, simulations of its unique
and the primary pressurised water cooler takes the full 30 MWt features are very important (Venter and Mitchell, 2007). Flownex,
from the hot helium. In parallel loaded operation, the secondary a CFD software package, was adapted to include all the major
cooling loop removes 10 MWt and the remaining 20 MWt goes to components of the PCS, including the pebble bed reactor itself. One
the primary loop. of the distinguishing features of the PCS chosen is that three
In April 2001 JAEA began the design and development of a HTGR separate shafts are used. In order to gain a better understanding of
with a gas turbine electricity generation system, called the the dynamic behaviour of the system, a functional model, termed
GTHTR300 (Kunitomi et al., 2004). It will produce 600 MWt energy the Pebble Bed Micro Model (PBMM), was created (Greyvenstein
and the objective is to establish a feasible plant design and helium and Rousseau, 2003). The objective of the model was not to address
gas turbine technology with the ultimate goal of demonstration of specific helium related issues, or test individual components, but to
a prototype in the 2010s for commercialisations in the 2020s produce the same overall performance characteristics and control
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 513

During operation only the control rods in the reflector are in use.
Studies on an indirect, multi-shaft Brayton cycle were conducted
and work is being conducted at INL to improve the efficiency of the
cycle (Oh and Moore, 2005). Simulations suggest that using
supercritical CO2, instead of helium, in the Brayton cycle increases
the cycle efficiency. More recently, a direct Brayton cycle was
deemed to be more efficient and economical (INL, 2007).
For electricity production, the MHR will operate with a power
level of 600 MWt and an outlet helium temperature of 850  C. The
inlet helium temperature is 490  C, which allows high efficiency
operation of the Brayton cycle and acceptable operating conditions
for a reactor vessel made from a Cr–Mo steel. Using 850  C as an
upper temperature eliminates the need for turbine blade cooling
and ensures acceptable performance of the fuel particles during
normal operation. For hydrogen production, it is desirable to
increase the helium outlet temperature in order to improve effi-
ciency and economics. Design points of 950  C and 590  C were
selected for the outlet and inlet, respectively. This allows a suffi-
ciently high coolant flow and convective heat transfer rate.

Fig. 14. Schematic of the PBMM. 4.1.9. ANTARES

The French ANTARES programme is a research partnership
between AREVA NP, CEA and EDF aimed at the development of an
topology as that of the prototype plant. Nitrogen is therefore used industrial HTGR prototype. All three partners are also involved in
as the working fluid, allowing the use of conventional turbines and the RAPHAEL project. It distinguishes itself from other concepts due
compressors, and an electric heater is used as the heat source. A to its flexible design, employing an indirect cycle which allows
schematic of the PBMM is shown in Fig. 14. As well as operation of decoupling of the nuclear reactor from different possible applica-
the Brayton cycle, control strategies for start-up, load following and tions (Billot et al., 2006). The reactor consists of an annular core
load rejection were demonstrated. Flownex was successfully used with hexagonal block type fuel assemblies. A significant R&D pro-
to aid the design of the PBMM and to simulate the results (Matzner, gramme is currently underway, looking at materials development,
2004). fuel fabrication and many other aspects. The programme is in the
The Helium Test Facility (HTF), built at NECSA in Pretoria, is used pre-conceptual design stage and the preliminary design is expected
to test system components up to 1100  C at high pressure (Matzner, to be ready in 2010. It is based on the GT-MHR concept.
2004). A heat transfer test facility (HTTF) is being built in associa-
tion with North-West University and should be in operation this 4.1.10. RAPHAEL
year (Rousseau and van Staden, 2006). Its purpose is to validate In April 2005, the RAPHAEL IP (ReActor for Process heat,
correlations that are used to model heat transfer and fluid flow Hydrogen And ELectricity generation Integrated Project) was
phenomena. It does not represent a scale model of the reactor, but launched as part of the European Union 6th Framework pro-
is some representative sections of the core suitable for performing gramme. It addresses the viability and performance issues of
tests. innovative HTGRs designed to provide both electricity and process
The construction of the 400 MWt PBMR demonstration power heat. The project was initiated based on the success of several
plant (DPP) on the Koeberg site is expected to be completed during smaller projects, such as HTR-N mentioned above and HTR-E. The
2010 (Greyvenstein et al., 2006). The DPP includes a horizontal HTR-E project, funded by the 5th Framework programme, had
single shaft recuperative Brayton cycle with pre- and intercooled the objective to develop innovative technologies needed for the
two stage compression. The DPP will be followed by the components and systems of a HTGR with a direct cycle, with
construction of commercial units, the first of which are to be reference to existing industrial projects at that time (GT-MHR and
produced by 2014 (Tennenbaum, 2006). The PBMR Advanced PBMR). Design recommendations were made for the helium
Electricity Plant (AEP) consists of a 500 MWt PBMR reactor coupled turbine, recuperator, bearings and seals (Besson et al., 2006). The
to a combined cycle, involving a topping Brayton cycle and a bot- RAPHAEL consortium includes 33 organisations from 10 European
toming Rankine cycle. The envisaged Brayton cycle builds on countries. The project explores the performance of fuel, materials
knowledge gained from the DPP and so should avoid first of a kind and components, the reactor physics models, the nuclear safety and
engineering issues (Matzner et al., 2006). waste disposal issues, the overall system integration and potential
interfaces with hydrogen production. In collaboration with the
4.1.8. GT-MHR HYTHEC project, interface requirements have been described. The
The GT-MHR (Gas Turbine Modular Helium Reactor), being main priorities are the increase of performance and the adaptation
developed by General Atomic (GA), couples a HTGR contained in of technologies to the needs of heat supply for industrial purposes
one pressure vessel with a high efficiency Brayton cycle gas turbine (Hittner et al., 2006). The ambition of RAPHAEL is to bring much
system in an adjacent pressure vessel (LaBar et al., 2004). The plant needed R&D support and so be a leading force in the development
is designed for a 60-year life and the entire reactor confinement of HTGRs.
structure is underground. The reactor employs block type fuel
elements, each of which contains 210 blind fuel holes, 108 coolant 4.1.11. NGNP
holes and 3126 fuel compacts (Richards et al., 2006b). A fuel The US DoE Next Generation Nuclear Plant (NGNP) is planned to
compact is made up of TRISO coated particles in a carbonaceous be a ‘‘demonstration of the technical, licensing, operational and
matrix. The reactor core consists of 102 fuel columns in three commercial viability of HTGR technology for the production of
annular rings, with 10 fuel blocks in each column. The outer process heat, electricity and hydrogen’’ (INL, 2006, 2007). Three
reflector contains 36 control rods, whilst the core contains 12. teams, led by Westinghouse, AREVA NP and GA, are developing the
514 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Table 5
Nuclear hydrogen programmes.

Company JAEA GA PBMR/Westinghouse AREVA NP

Country Japan USA SA/USA France
Reactor GTHTR300 MHR-GT PBMR ANTARES
Power 600 550–600 500 600

PCS Brayton Brayton Rankine Rankine

HPP SI HTE/SI HyS/HTE HTE/SI
Cycle configuration Direct PCS, series Direct PCS, parallel Indirect, series Indirect, parallel
indirect HPP indirect HPP HPP and PCS HPP and PCS
IHX Helical coil S&T Single-stage PCHE Two stage PCHE PCS: helical coil S&T
Process: PCHE or fin plate

basis for selection of the specific design and operational charac- however, much more is needed before design and construction of
teristics of the NGNP. The work included completion of specific a plant can be realised.
studies to address key aspects of the NGNP, followed by cost esti- The remainder of this report, therefore, concentrates on
mates and schedules for design, construction and operation. Key coupling an HTGR with a hydrogen production plant.
decisions on the choice of a conceptual design are scheduled for
mid 2008. In the project, emphasis is put on initiating design and 5. Nuclear hydrogen production plant
licensing work as early as practical to help focus R&D activities. It is
expected that operation will commence in 2018 and continue for All of the HTGR programmes detailed in Section 4.1, with the
two to three years after which equipment performance will be exception of that of the Chinese, are developing a hydrogen
confirmed by disassembly and inspection. production plant to couple to the nuclear reactor. The Chinese HTR-
10 and HTR-PM programmes are concentrating solely on electricity
4.1.12. Korea production. The choice of hydrogen production plant (HPP), the
The Korea Atomic Energy Research Institute (KAERI) have power conversion system (PCS) used and the plant layout differ
a strong programme developing nuclear technologies, with the significantly between programmes. Indirect or direct coupling can
aims to be on par with advanced countries by the early 2000s in be used, as can series and parallel placement of the HPP and PCS.
terms of safety and economic value, and to establish a base in the Table 5 gives a summary of the different programmes, including the
national energy supply. KAERI aim to demonstrate massive scale selection of the intermediate heat exchanger (IHX), a key compo-
hydrogen production using a VHTR by the early 2020s. A key target nent in the system. The following sections describe each of these
is to obtain an operating licence for a Nuclear Hydrogen Develop- programmes in detail.
ment and Demonstration (NHDD) reactor (Chang et al., 2007). There are also significant research programmes investigating
There are four candidate core designs being explored for the HPPs using HTE and the SI and HyS cycles in countries and regions
200 MWt demonstration plant, two prismatic modular reactors and which do not have programmes researching coupling, HTGRs and
two pebble bed reactors. The two design types are scaled down Generation IV reactors. These are briefly described in Section 5.8.
versions of the GT-MHR and PBMR. The pre-conceptual design is
expected to be ready this year, the engineering design finished by 5.1. JAEA
2014 and construction to begin in 2016 ready for operation in 2020.
KAERI have links with both JAEA in Japan and INET in China and in JAEA are currently the only institution to have a working HTGR
2005 set up a programme with the US involving GA as part of the I- and a nuclear hydrogen programme, putting them ahead of the rest
NERI contract. of the world in terms of nuclear hydrogen development. Originally,
they focussed their nuclear hydrogen programme on three distinct
4.2. Gas Cooled Fast Reactor (GFR) R&D areas:

All the HTGRs discussed above operate in a once-through cycle 1. Design work for steam methane reforming.
and the particle fuel employed allows a significant increase in 2. Large-scale SI cycle experiments.
discharge burn up and so reduced waste generation per unit of 3. Lab scale high temperature electrolysis experiments.
produced energy in comparison to current LWRs. However, in the
long term, adaptation of a closed cycle will be necessary to mini- Two sets of lab scale experiments of high temperature elec-
mise waste and efficiently utilise uranium resources. A gas cooled trolysis were carried out, the first with 12 tubular solid oxide cells
fast reactor (GFR), which can achieve high uranium utilisation and in series, producing hydrogen at 44 Ncm3/cm2 h and 950  C, and
recycling of actinides whilst benefiting from the high efficiency of the second with a planar cell (dimensions 80 mm  80 mm),
gas cooled reactors, is therefore of considerable interest (Hejzlar producing hydrogen at 38 Ncm3/cm2 h and 850  C (Hino et al.,
et al., 2005). Research was carried out on GFRs in the 1970s, 2004). However, neither experimental setup kept its integrity
however, like other nuclear technologies, it was stopped until through one thermal cycle. Long term, costly developments were
recently. Currently the most extensive investigations are being deemed necessary, and, considering the return on investment, the
carried out by CEA and Argonne National Laboratory under the study on HTE was halted.
I-NERI programme. The reference design employs a steel reactor The HTTR has thus far only been used for electricity production.
vessel with a high power density core coupled to a direct helium However, the design includes an intermediate heat exchanger
Brayton cycle. The high power density and no need for moderator (IHX), installed in a reactor containment vessel, which can supply
lead to a very compact design, but pose challenges for decay heat 10 MW of thermal energy to a heat utilisation system (Ohashi et al.,
removal. A guard containment and low pressure drop core address 2004). This corresponds to a third of the HTTR power. Steam
this problem, allowing decay heat removal by circulation of helium methane reforming will be the first heat application coupled to the
under elevated pressure. To date much R&D has been carried out, HTTR. This will be the world first demonstration of hydrogen
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 515

production coupled to a HTGR. First, a mock-up test facility using an

electric heater as the heat source was created in order to demon-
strate the control technology. It is approximately a 1/30 scale model
of the HTTR hydrogen production system. Performance tests were
carried out at the same reaction conditions as those of the HTTR
hydrogen production system and hydrogen was stably produced at
a rate of 120 m3/h. Fluctuations of the helium outlet temperature
from the steam generator and the pressure difference in the steam
reformer can both be controlled within allowable ranges (Inaba
et al., 2005). Construction of the HTTR hydrogen production system
is now underway and demonstration tests are planned for 2009.
In the long run, JAEA are aiming to produce hydrogen from
water, and so eliminate the carbon emissions associated with SMR.
Efforts are being concentrated on the Sulphur Iodine cycle for mass
scale production. The coupling technologies between the HTTR and
SMR system have many similarities with those that will be used for Fig. 15. Process scheme for the GTHTR300C (Sakaba et al., 2007).
the coupling of the SI cycle and so are an important step in
development of the GTHTR300C, the name given to the GTHTR300
reactor with combined electricity and hydrogen generation. The secondary helium loop supplies heat to the H2SO4, SO3 and HI
GTHTR300H system will produce much more hydrogen than elec- decomposers. The plant aims to produce 1000 Nm3/h of hydrogen
tricity. R&D on the SI process was split into several stages: Initial lab at a thermal efficiency of around 40%. If achieved, this will confirm
scale demonstrations before 1998, bench scale demonstrations the economic prospects of commercial water splitting nuclear
from 1999 to 2005, then pilot-scale tests and finally HTTR-IS hydrogen production systems. Two methods are proposed to
process tests (The Japanese refer to the Sulphur Iodine cycle as the increase the yield of the HI decomposition reactor. The first
IS cycle.) (Kubo et al., 2004a). The knowledge gained through HTTR employs absorbing the iodine in the HI vapour with cobalt, which
and SI cycle improvements will then be combined in the design and entails the addition of a cobalt recovery column to the process
construction of the GTHTR300C (Shiozawa, 2007). flowsheet. The second option, which gives a higher hydrogen yield
The bench scale studies were split into three main sections: and higher efficiency, involves the application of a hydrogen
permeable membrane reactor (HPMR). This reduces the tempera-
1. Study on advanced HI processing. ture required in the HI decomposition reactor.
2. Bench scale demonstration of the whole process. An important design objective for the GTHTR300C is the use of
3. Screening of corrosion resistant materials. technologies already accumulated in JAEA so that any new tech-
nological development is limited and the investment risk is mini-
HI processing is one of the areas requiring the most develop- mised (Sakaba et al., 2007). The aim is that the system will be
ment in the SI cycle as it has a high energy demand and compli- technically feasible and economically attractive by 2030. The design
cated separations. The Japanese are investigating using of the GTHTR300C is almost the same as that of the HTTR-IS system,
electroelectrodialysis to concentrate the HIx solution. Process and uses Co to absorb I2 in the HI decomposition reactor. The
simulations suggest that this can significantly increase process technology developments underway for the GTHTR300 reactor will
efficiency (Hwang et al., 2003). The bench scale experimental also be employed. The layout of the GTHTR300C is shown in Fig. 15.
facility is made of glass, quartz and fluorine resin equipment and The helically coiled intermediate heat exchanger (IHX) is the same
contains over ten chemical reactors and separators (Kubo et al., as that used in the HTTR-IS and is installed between the reactor
2004b). Continuous, closed cycle operation was demonstrated for pressure vessel (RPV) and gas turbine (GT) system. Despite the
20 h producing hydrogen at a stable rate of 31.5 L/h. The Bunsen smaller size, no major changes are made to the GT from the
section was operated at 70  C, the H2SO4 section at 850  C and the GTHTR300 design. The primary coolant pressure employed for
HI section at 500  C, with all three being at atmospheric pressure. the GTHTR300C is around 5 MPa, 2 MPa lower than that in the
The pilot plant is designed using industrial materials (identified GTHTR300 (Kunitomi et al., 2007). The higher pressure is not
during Section 3 of the bench scale studies above) and aims to required as there is no gas circulator which requires down-sizing.
produce 30 Nm3/h hydrogen using electrically heated helium as the Using a lower pressure helps extend the lifetime of the heat
process heat supplier (Kubo et al., 2004a), as well as obtaining exchangers and despite increasing the turbine costs, the saving in
engineering data for constructing and operating larger plants. A costs for the heat exchangers, and pressure vessels means no
major part of the pilot-scale demonstration is designing and economic penalty is paid.
building the reactors from corrosion resistant materials and testing Table 6 summarises the R&D stages at JAEA and highlights the
their structural integrity. Hydrogen production tests will verify the depth of the research programme.
developed control protocol and be used to verify an analytical code
system (Terada et al., 2007). R&D so far has concentrated particu- 5.2. GA
larly on the H2SO4 and SO3 decomposers. The construction and the
operation of the pilot-scale plant are planned to take 6 years, the At GA two technologies have been selected as leading candi-
first half being construction and the second half operation. dates for coupling to the MHR: the SI cycle and HTE (Richards et al.,
During the pilot-scale tests, the design study for the HTTR-IS is 2006b). The combination of the hydrogen production plant and the
also being carried out. The aims of the HTTR-IS are to: establish reactor is entitled the H2-MHR. For the H2-MHR the outlet
procedures on safety design and evaluation, develop technology on temperature of the helium from the MHR is increased to 950  C in
key components, establish the control technology, and to gain order to improve efficiency and economics. This also necessitates
construction, operation and maintenance experience (Sakaba et al., a 100  C increase in outlet helium temperature from the HPP, to
2007). Basic design of the HTTR-IS system is planned to be 590  C, to ensure acceptable fuel performance. Current schedules
completed by 2010. Process heat from the HTTR will be transferred suggest that construction of a H2-MHR will begin in 2015, with
to the SI plant through an intermediate heat exchanger. The operation commencing in 2020.
516 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Table 6
R&D work at JAEA.

R&D stage Lab Bench Pilot HTTR-IS GTHTR300C GTHTR300H

H2 production rate (Nm3/h) w0.001 w0.03 w30 w1000 w25,000 w50,000
Heat supply Electric Electric Helium heated Helium heated Helium heated Helium heated
electrically by HTTR by GTHTR300 by GTHTR300
Power for SI plant (MWt) 10 170 371
Materials Glass Glass Industrial Industrial Industrial Industrial
Process pressure Atm Atm High High High High
Electricity production (MWe) 202 87
Net efficiency 40–44% 46–49% 42–49%

5.2.1. SI-based H2-MHR 5.2.2. HTE-based H2-MHR

In the SI-based H2-MHR the power conversion unit of the MHR The HTE-based H2-MHR design includes a direct PCS and an
is replaced with an intermediate heat exchanger (IHX) and a helium indirect HTE plant positioned in parallel, as depicted in Fig. 17
circulator on the cold leg of the primary helium circuit, as shown in (Richards et al., 2006b). The MHR supplies both the heat to generate
Fig. 16 (US DoE NERI, 2006). Heat transfers through the IHX to the steam and the electricity required by the electrolysers. Approxi-
secondary helium loop, which in turn transfers heat to the mately 90% of the thermal energy from the MHR is used to generate
hydrogen production plant (HPP). The secondary helium loop electricity, with the remainder being transferred through the
includes a 115 m horizontal separation between the IHX and the intermediate heat exchanger (IHX) to the steam generator. Steam is
HPP. This is deemed to be an acceptable distance to protect the supplied to both the anode and cathode sides of the electrolyser.
MHR from any potential hydrogen explosions in the HPP. The SI That on the cathode side is split into hydrogen and oxygen, whereas
flowsheet being developed through the I-NERI programme, as that on the anode side is used as a sweep gas to collect the oxygen.
discussed in Section 2.1.1, is used for the HPP. In 2006, the flowsheet A small quantity of electricity is generated from the oxygen/steam
included a reactive distillation step in the HI processing section. stream which is used to provide power in the plant house. The plant
This requires a significant amount of electricity for shaft work, produces 268,000 tonnes H2/year at an efficiency of 55.8% (HHV),
including heat pumps with vapour recompression. For a plant with a product hydrogen pressure of 4.95 MPa.
consisting of four MHR units (2400 MWt), the total electricity The solid oxide electrolyser (SOE) module design is based on the
requirement of the HPP is about 800 MWe, the bulk of which is planar cell technology being jointly developed by INL and Ceram-
needed for the process pumps. If the electricity is generated at 48% atec. A module will contain 40, 500-cell stacks, with a total elec-
efficiency using additional GT-MHR modules, the overall efficiency tricity consumption of 500 kWe. It is estimated that the hydrogen
is estimated to be 45% (HHV) using AspenPlus. This produces production cost will be $1.92/kg.
approximately 368,000 tonnes H2/year at a pressure of 4 MPa. The
flowsheet is being continually refined, and the design is likely to
change over the coming years. Of particular note is the recent idea 5.3. PBMR
of switching back to using extractive distillation as opposed to
reactive distillation as it is a proven technology. PBMR have a partnership with Westinghouse and the Shaw
The primary helium loop and sulphuric acid decomposition will Group developing the HyS cycle for hydrogen production. The cycle
be operated at a pressure of around 7 MPa. The secondary helium is configured to optimally use the thermal and electrical energy
circuit will be operated at a slightly higher pressure to ensure that from a PBMR, and was chosen due to its simplicity compared to
any radioactive leak from the MHR and/or chemical leak from the other cycles (Correia et al., 2006). Several different heat transfer
HPP do not cross the heat transfer surfaces and so contaminate the configurations have been evaluated:
secondary helium circuit. Both helium circuits have a slipstream
helium purification train to prevent contamination of the hydrogen 1. Indirect Rankine cycle in series with IHX (indirectly coupled to
produced with tritium or other radioactivity. The intermediate heat HPP).
exchanger (IHX) design consists of 40 printed circuit heat 2. Direct Brayton in series with IHX (indirectly coupled to HPP).
exchangers (PCHE), a concept developed by Heatric. Each module 3. Direct Brayton cycle in parallel with IHX (indirectly coupled to
has a duty of 15 MWt and weighs around 5 tonnes. HPP).
Hydrogen production costs, taking into account capital and 4. Separate nuclear plants for thermal coupling and electricity
operating and maintenance costs, are estimated to be $1.97/kg. generation.

Each configuration has advantages and disadvantages. The first

option has fewer components on the primary nuclear loop and
utilises standard Rankine technology, however, presents more
complicated safety considerations. Option 2 is ideal for process
applications with low energy demands relative to electricity
generation, however, the Brayton cycle is less efficient at lower
temperatures and is sensitive to energy demands on the process
side. Parallel operation, as in option 3, allows individual control
over the two sections, however, presents problems in control and
mixing of the two hot helium streams. The final option allows
optimisation of the two plants individually, however, the HyS
process requires high temperature and, due to constraints on the
reactor inlet temperature, the hot helium would need to be cooled
resulting in waste heat. Analysis showed that the highest efficiency
was achieved when no additional electrical power over that needed
Fig. 16. Process schematic of the SI-based H2-MHR (US DoE NERI, 2006). for the HPP was produced and that there is no significant efficiency
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 517

Fig. 17. Process schematic of the HTE-based H2-MHR (Richards et al., 2006b).

difference between using a high pressure Rankine cycle or a Bray- 2006). Within the programme, the CEA is responsible for the
ton cycle (McLaughlin et al., 2005). majority of the work developing the SI cycle and HTE, as well as the
A version of option 1 was selected for detailed analysis (Lahoda coupling of the HPP to the nuclear reactor.
et al., 2006), as shown in Fig. 18. The design is termed the ‘‘Four The SI cycle is the reference process for the coupling studies
Pack’’ as it consists of four 500 MWt PBMRs, each coupled to (Barbier et al., 2006). Analysis on the cogeneration of heat and
a modular HyS plant through an IHX and a helium heat transfer electricity has been conducted. The main issue of cogeneration is
loop. Each HyS plant consumes around 200 MWt energy, with the producing a flexible cogeneration ratio between electricity and
remaining 1200 MWt from the PBMRs powering a single Rankine hydrogen production. For the first option considered, a direct
cycle. About 400 MWt of low grade rejection heat from the HyS Brayton cycle operated in parallel with an indirect SI process,
plants is used as water preheat in the PCS which generates about flexibility could be obtained by control of the helium inventory.
600 MWe of electrical energy at an efficiency of greater than 38%. However, the resulting circuit pressure decrease would increase
Process simulations were carried out using ChemCad in order to circulator power consumption and so decrease the efficiency. The
size the vessels and processes in the system. The catalytic decom- second option considered, parallel indirect operation of the PCS and
position reactor is the most challenging piece of equipment to HPP, involves separate circulators for the two systems and therefore
design, mainly due to materials selection problems encountered flexibility can be controlled by the rotation speed of each circulator.
with the high operating temperatures. Eight bayonet style heat Further studies are being carried out to investigate this option.
exchangers are employed, as well as a H2SO4 concentration column. For the main safety and economic analysis, it is assumed that the
Initial estimates suggest that the cost of hydrogen will be in the 600 MW HTGR thermal power is totally dedicated to the HPP. An
range $2–$3/kg, however, a number of technology developments additional 100 MWe is required from the grid. Table 7 shows the
need to be successfully addressed. energy requirements of each section of the HPP. The design of the
HPP consists of ten ‘‘shops’’ operated in parallel. This allows design
using conventional codes as individual equipment items are
5.4. AREVA NP/CEA smaller, and ensures plant availability and acceptable chemical
reactor dimensions. Heat transfer between the nuclear island and
The AREVA NP programme has identified HTE and the SI cycle as the HPP is performed via helium through a 200 m long insulated
the two most promising candidates for nuclear hydrogen produc- pipe. Safety constraints were analysed using a defence in depth
tion and aims to build a commercial plant by 2040 (Rodriguez et al., (DID) concept. Economic analysis puts the cost of hydrogen
production in the range 4–5V/kg, about 30% of which is energy
consumption. Taking into account expected improvements due to
R&D, a cost in the range 2–3V/kg is predicted. Of this 50–60% is due
to energy consumption.
In 2002 the CEA launched an integrated research programme on
the SI cycle, the work being split between the AREVA NP pro-
gramme (Billot et al., 2006), the HYTHEC project (Le Duigou et al.,
2007) and the CEA-GA-SNL I-NERI agreement. The goal is to choose
the most promising way to produce nuclear hydrogen by 2008

Table 7
Energy needs of each section of a 1 kmol/s HPP (Vitart et al., 2008).

Section Total energy Heat Electricity

requirement (MW) required (MWt) required (MWe)
I: Bunsen 34 0 17
II: H2SO4 389 385 2
III: HI 376 214 81

Total 799 599 100

Fig. 18. Schematic of the Four Pack PBMR/HyS process.
518 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Fig. 19. The HYTHEC/RAPHAEL coupling scheme (de Lorenzo et al., 2006).

(Vitart et al., 2008). As well as flowsheet analysis, the R&D pro- H2, 1665 kJ/mol H2 of which is provided through internal heat
gramme involves the design and building of reactors and the recovery with the remaining 504 kJ/mol H2 coming from the VHTR.
development of specific analytic methods to determine the The 319 MW of energy from the VHTR therefore produces
composition of liquid and gas phases (Doizi et al., 2007). As part of a hydrogen flow of 633 mol/s.
the I-NERI contract, CEA have built a lab scale Bunsen reactor,
which has been shipped to the US for combination with the other 5.6. US DoE Hydrogen Program and NHI
sections. Reactive distillation is a strong contender for decompos-
ing the HI in the SI cycle, however, currently shows low efficiencies The US DoE Hydrogen Program was launched in 2004 to
due to poor HI to H2 conversion in the vapour phase (Goldstein implement the President’s Hydrogen Fuel Initiative (US DoE, 2007).
et al., 2005). Advanced, inorganic hydrogen separation membranes Its mission is to ‘‘research, develop, and validate hydrogen
are being developed to improve the process scheme. Corrosion production, storage, and fuel cell technologies and to overcome the
tests are also being carried out to develop suitable materials for non-technical barriers to the commercialisation of these technol-
construction. The HYPRO project is intended to follow on from the ogies – with the ultimate goals of reducing oil use and carbon
lab scale I-NERI experiments. The objective is to demonstrate emissions in the transportation sector and enabling clean, reliable
hydrogen production at a pilot scale on the helium loop facility energy for stationary and portable power generation’’. The program
(HELITE) that will be built at CEA Cadarache and able to deliver is wide ranging, with research into hydrogen production from fossil
1 MWt. The time schedule for HYPRO is around 2012. Initially the fuels, biomass and water. One such research area is the Nuclear
feasibility study will focus on the sulphuric acid decomposition, Hydrogen Initiative (NHI), the goal of which is to demonstrate the
and in particular a high temperature heat exchange reactor for SO3 economic, commercial-scale production of hydrogen using nuclear
decomposition (CERSO3). energy. Five major National Laboratories, Sandia (SNL), Savannah
The CEA have also started a programme investigating HTE. Both River (SRNL), Idaho (INL), Argonne (ANL) and Oak Ridge (ORNL), are
developments of the cell, such as thermodynamic models and involved, as well as General Atomic (GA), Ceramatec and many
catalyst and electrode development, as well as the whole system, universities. The French-CEA are working alongside GA and SNL
for example stack architecture and scale effects, are being investi- under the I-NERI agreement (International Nuclear Energy
gated. The programme aims to produce a HTE demonstration Research Initiative). The broad division of research is described in
module able to produce 200 l H2/h in conjunction with the Table 8.
Norwegian University of Science and Technology.
5.7. Korea

5.5. European programmes The Hydrogen Energy division of the Korea Institute of Energy
Research (KIER) and the Korean Atomic Energy Research Institute
The HYTHEC and RAPHAEL European projects specialise in the SI (KAERI) are developing both the SI cycle and HTE as possible
cycle and VHTR development, respectively, and have worked in technologies for the Nuclear Hydrogen Development and
conjunction on the coupling between the two (de Lorenzo et al.,
2006). The scheme developed is based on a self sustaining concept Table 8
US DoE NHI.
where all the electricity needed is supplied by the VHTR, making
the net electricity balance zero. The layout involves coupling the SI Company/laboratory Main area of research Comments
cycle in parallel to both a Brayton cycle and a bypass in an indirect GA SI cycle See Section 2.1.1
secondary circuit, as demonstrated in Fig. 19. Helium is used as the Sandia
circulating fluid in the secondary circuit as well as the primary CEA

circuit due to its high heat capacity. Parallel indirect operation INL HTE See Section 2.2
allows great flexibility – the electricity and hydrogen production SRNL HyS cycle See Section 2.1.2
can be altered as a function of the amount of helium bypassed. The
ORNL Membrane applications
three way valve is a key component of this design which does not
INL
feature in other layouts. It regulates the helium flow by iteratively
solving the system equations to ensure the electricity demand is ANL Other thermochemical cycles See Section 2.1.3
Many universities
exactly met. The energy requirement of the SI cycle is 2169 kJ/mol
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 519

Demonstration reactor (NHDD, see Section 4.1.12). The conceptual Table 9

flow diagram for HTE includes a single shaft, direct Brayton cycle IHX parameters.

operated in parallel with an indirect loop for steam production JAEA HYTHEC/ AREVA NP GA
(Shin et al., 2007). Efficiencies of around 45% are expected from RAPHAEL
modelling, however, much important experimental work is yet to mP (kg/s) 324.2 209.95 240 321
be done. A lab scale study of the SI cycle is in process with the aim of TIn,P ( C) 950 950 850 950
PIn,P (MPa) 5.00 5.5 5.5 7.03
producing 5 l H2/h. Experimental studies on many aspects of the
TOut,P ( C) 850 387 350 590
cycle are being carried out (Hong et al., 2007a,b), however, no POut,P (MPa) 5.4 7.00
coupling strategy has yet been published. mS (kg/s) 80.3 218.89 614 321
TIn,S ( C) 500 350 300 565
5.8. Other national programmes PInS (MPa) 5.15 5 5.5 7.10
TOut,S ( C) 900 890 800 925
POut,S (MPa) 4.9 7.07
5.8.1. Italy Q (MWt) 170 613.3 608 600
In 2005 the Italian ENEA (Ente per le Nuove tecnologie, l’Energia LMTD ( C) 154 48 50 24
e l’Ambiente, the National Agency for New Technologies, Energy HEX design Helical shell – Plate fin/tubular/ Printed
and tube printed circuit circuit
and the Environment) launched an ambitious project under their
Hydrogen and Fuel Cells division named TEPSI (Tarquini, 2007). The
three year project has three main strands:
continue. The Japanese are the only nation with a stand alone
1. Hydrogen production with thermochemical cycles fed by programme and are currently world leaders in nuclear hydrogen
concentrated solar energy. development.
2. Development of new materials for hydrogen storage. The power conversion system (PCS) and the hydrogen produc-
3. Zero-emission integrated hydrogen and electric power tion plant (HPP) employed vary dramatically between programmes.
production by coal hydro-gasification. For the PCS, PBMR and AREVA NP have chosen the more conven-
tional Rankine technology, whereas GA and JAEA have opted for
The first strand includes research into the SI cycle, the mixed Brayton cycles. JAEA have concentrated their R&D efforts on the SI
ferrites cycle and solar reactors. Although aimed at the solar cycle, whereas the others are also looking at high temperature
production of hydrogen, much of the SI cycle research is also electrolysis and Westinghouse have chosen the HyS cycle. Predicted
relevant to nuclear hydrogen production. Bench scale Bunsen efficiencies and costs are currently at a preliminary stage and are
reaction experiments have been carried out (Giaconia et al., 2007a), very similar, regardless of the options chosen. The technological
as well as investigations of the HI decomposition (Spadoni et al., feasibility of key components is likely to be the determining factor
2007) and membrane distillation experiments (Caputo et al., 2007). for which design comes out on top.

5.8.2. Australia
The Australian Commonwealth Scientific and Industrial 6. Intermediate heat exchanger
Research Organisation (CSIRO) Hydrogen Technology programme is
concentrated mainly on fuel cells, however, research is also being In contrast to the Power Conversion System (PCS), which can be
carried out on electrolysis and solar thermochemical hydrogen both directly and indirectly coupled to the nuclear reactor, all
production. The recently published report on Australia’s hydrogen flowsheets use indirect coupling between the Hydrogen Production
economy roadmap (Australian Academy of Science, 2008) considers Plant (HPP) and the reactor. This is primarily due to control and
nuclear and hydrogen energy technologies to be only of low safety constraints and means that the Intermediate Heat Exchanger
priority and that Australia should ‘‘monitor international develop- (IHX) is a very important piece of equipment. The IHX has to
ment and follow as needed’’. In contrast, high priority is placed on transfer several hundred MW from the primary to the secondary
advanced coal technologies and photovoltaics. loop with high efficiency, operate at high temperature and be
compact enough to be enclosed in a single vessel. It must also
5.8.3. Germany prevent water and process gas ingress into the nuclear core and
The German Aerospace Centre (DLR) are investing the decom- minimise hydrogen and tritium permeation between the primary
position of sulphuric acid, an important stage in the SI and HyS and secondary loops. There is currently no industrial experience
cycles, using a solar reactor (Noglik et al., 2007). They are also with the size and operating conditions required. The temperature
investigating other thermochemical cycles that can be operated in and environmental conditions (e.g. the presence of impurities in
conjunction with solar power, however, do not consider nuclear the helium which may cause oxidation reactions and carbonisation)
power and coupling. means that materials other than those commonly used in heat
exchangers are required.
5.9. Discussion Helium is the heat transfer fluid on both sides of the IHX, with
the inlet temperature on the primary loop being about 950  C. The
There are four major programmes across the world to build inlet temperature on the secondary side depends on the configu-
nuclear hydrogen plants, as well as many smaller R&D support ration employed. Table 9 shows the parameters chosen by several
efforts. Three of the four major programmes, those of GA, PBMR/ leading nuclear hydrogen programmes, following the notation
Westinghouse and AREVA NP, are vying for funding through the US given in Fig. 20. The data for the PBMR/Westinghouse design was
DoE Next Generation Nuclear Plant (NGNP) programme. A decision not available.
on the chosen design is expected later this year. Sections from each As shown in Table 9, there are several different types of heat
of the three plans may be incorporated into the final design. The exchanger being considered for the IHX. Before a selection is made,
NGNP programme has a strong emphasis on early design and all aspects must be considered, including materials, design, fabri-
licensing and plans to have an operational plant in 2018. If unsuc- cation, testing, safety and control. For most cases, the small LMTD
cessful in being chosen for major components of the NGNP, it is across the IHX means that a large heat transfer area is required, and
unclear where funding will come from for the programmes to therefore a HEX of compact surface geometry, such as a plate type, is
520 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

designs. After the preliminary testing is complete, the HELITE loop

at CEA Caderache will be used to validate and optimise the selected
design (Billot et al., 2006).
The IHX will be housed inside a containment vessel, and so is
itself not the primary pressure boundary. The containment vessel is
at a much lower temperature, and so can be built from GR91 steel.
A safety valve will be required between the secondary side of the
IHX and the HPP.
Fig. 20. Schematic of the intermediate heat exchanger.

6.3. GA
the only economically viable option. Due to the competitive nature
of the stage the research is currently at, most vendors have not The GA IHX design is based on the Printed Circuit Heat
disclosed specific design information. The following sections Exchanger (PCHE) concept developed by Heatric corporation
describe the IHX designs of selected nuclear hydrogen programmes. (Richards et al., 2006a). It consists of metal plates that are diffusion
bonded to restore the properties of the base metal. A technique that
is similar to that used for etching printed electric circuits is used to
6.1. JAEA
chemically mill fluid flow channels into the plates. The design
consists of 40 Heatric-type modules, manufactured from a high
The Japanese design is unique in comparison to the others as it
temperature Ni-based superalloy. Candidate materials are Inconel
has a significantly higher LMTD due to the high, narrow temper-
617 and Hastelloy XR. The IHX vessel is manufactured from SA533
ature range employed on the primary side (Kunitomi et al., 2007).
steel and insulated with Kaowool to maintain operating tempera-
This is possible due to the particular location of the IHX. JAEA have
tures below 350  C and prevent creep damage.
selected a helical shell and tube construction for the GTHTR300C.
Hastelloy XR is used for the heat transfer tubes and tube bundles.
Hastelloy XR is a nickel-based superalloy developed by JAEA and 6.4. PBMR/Westinghouse
Mitsubishi Materials Co. and is approved for high temperature
applications (Sakaba et al., 2006). The design of the GTHTR300C- The PBMR design for electricity production utilises a direct
IHX is similar to that used in the HTTR, the main difference being Brayton cycle and therefore eliminates the use for an IHX. However,
the size and the pressures used. A small pressure difference for hydrogen production, an indirect, series PCS and HPP are used,
between the primary and secondary sides of only 0.15 MPa is used so an IHX is required. Westinghouse are responsible for the
to avoid creep of the heat transfer tube. The small pressure development of the IHX and have conducted materials research as
difference is required as the creep strength is low at the high well as design evaluations.
temperatures involved, and is controlled using a differential Inconel 617 and Hayness 230, both Nickel-based superalloys
pressure control system. After experiments on, and comparison that can be used up to around 1000  C, are candidate materials
with, the HTTR-IHX, a lifetime of 20 years is assured for the (Fazluddin et al., 2004). However, the design qualification with
GTHTR300C-IHX. respect to environmental effects and deformation mechanisms is
not mature, and substantial experimental work and modelling is
6.2. European programmes still required. Mo-TZM is being considered as an alternative mate-
rial as it has significant high temperature strength, and is well
The IHX is a common element in the HYTHEC and RAPHAEL capable of withstanding operational conditions up to at least
projects. Information from RAPHAEL about the sizing of the IHX 1200  C. It does, however, form a brittle carburised layer in
was fed into the HYTHEC project which designed the coupling carburising atmospheres and costs three times more than the Ni-
strategy. As a main partner in both RAPHAEL and the I-NERI based alloys.
contract, AREVA NP are carrying out significant research on the IHX. The Westinghouse design employs two stages of heat
The IHX is the only component of the ANTARES design which does exchangers (INL, 2007). The first operates at temperatures between
not use conventional and well-proven technology. One of the key 710 and 900  C and is expected to be replaceable. The second
areas of the research is materials development. Two nickel-based operates at temperatures below 710  C and is expected to have
alloys, alloy 617 and alloy 230, were selected for investigation of a 60-year lifetime. The high temperature IHX could be either metal
mechanical properties, ageing, environmental effects and fabric- or ceramic, whereas the low temperature one would be metal. As
ability (Billot et al., 2006). A decision is expected this year as to the heat transfer area required is large both heat exchangers are of
which material is most suitable. the compact design and a printed circuit (PCHE) design has been
The helical tubular concept, inherited from past German selected.
developments where a 10 MW HEX was built (Maus et al., 1984), is
being studied and developed so a more modern, compact version 6.5. Discussion
can be assessed. The design is thought to be feasible based on
previous experience (INL, 2006). In parallel to the work on the The IHX presents an engineering challenge due to the high
tubular design, and taking advantage of the significant improve- temperatures, large heat transfer areas and small available space
ments to plate exchangers over the last 20 years, several plate type involved. Metallic heat exchangers are being used by all the pro-
designs which have the potential to be more compact are also being grammes and Ni-based super alloys are being developed and
investigated. Both printed circuit and plate fin designs are being investigated as they represent the most cost effective materials able
explored. The different designs were scheduled to be tested in 2007 to operate at the high temperatures required. All the programmes,
in the HEFUS 3 helium loop at ENEA that provides representative with the exception of the Japanese, have selected to use compact
fluid dynamics and conditions, although no results have yet been heat exchangers, mainly of the printed circuit design although fin
published. The homogeneity of flow distribution will be tested in plate designs are still being considered. Due to the lower heat
the large flowrate loop PAT at EDF this year, and the CLAIRE loop at transfer area required, the Japanese are using the larger but more
CEA will be used to test the thermo-mechanical performance of the conventional and well-proven tubular design. Further high
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 521

temperature testing is required before any IHX design is finalised 7.2. Hydrogen production plant
and ready for coupling to a VHTR.
The hydrogen production plant (HPP) is designed and operated
based on existing non-nuclear industrial standards which are
7. Safety
applied to conventional chemical plants. The HPP is not expected as
the reactor core cooling system. In the case of abrupt thermal load
Safety is very important in all industrial plants, however, is
change of the HPP due to malfunction, precautions must be in place
particularly important when nuclear technology is involved. Past
to ensure this does not affect the reactor. In the JAEA design, three
incidents, such as Chernobyl and Three Mile Island, left the public
control valves in the primary helium circuit are used to mitigate the
sceptical about nuclear power. New Generation IV nuclear plants
turbulence of the inlet helium temperature to the gas turbine
therefore have a strong emphasis on inherent safety features.
(Kunitomi et al., 2007). In the HTTR-IS design, a steam generator is
Hydrogen is a highly explosive gas, so the coupling of a hydrogen
placed downstream of the HPP in the secondary circuit to act as
production plant to a nuclear reactor presents even more safety
a thermal absorber and mitigate any temperature fluctuations to
considerations. The main areas of concern and development are
within 10  C (Inagaki et al., 2007). A temperature rise of 15  C or
detailed in the following sections.
greater would cause the reactor to scram. In the HYTHEC flowsheet
a bypass emergency cooler is included as a heat sink (de Lorenzo
7.1. VHTR et al., 2006).
Chemical plant hazards, such as chemical toxicity and corro-
All the VHTRs described in this report use TRISO (tristructural- sivity, are also important in the HPP and conventional chemical
isotropic) fuel particles. The particles consist of a fuel kernel plant design standards are applied. Leakage and spillage are the
composed of uranium oxides, coated with four layers of three most typical causes for the occurrence of fire after component
isotropic materials (Matzner and Wallace, 2005). The innermost failure. The basic principles to avoid them are sound operating
layer is porous carbon, which allows the fission products to procedures and regular maintenance schedules. Explosive atmo-
collect without creating internal pressure. The second layer is spheres should be prevented and ignition sources excluded.
dense pyrolytic carbon (PyC), which is followed by a layer of SiC A Failure Mode Effects and Criticality Analysis (FMECA) has been
to retain fission products at elevated temperatures. The final layer carried out on the H2SO4 and HIx processing section of the SI cycle
is also dense PyC. The outer layers create a compound barrier (de Lorenzo et al., 2006).
against fission product release, effectively providing each kernel
with its own miniature pressure vessel, and so are fundamental 7.3. Tritium and hydrogen migration
to exceptional safety. The TRISO particles are designed not to
crack due to stresses for temperatures up to 1600  C. It is Hydrogen and its isotopes, deuterium and tritium, can permeate
therefore important that reactors are designed with this in mind. through solid metal. Tritium is radioactive and is produced in the
The safety features of the TRISO fuel were demonstrated over reactor core in several ways: as a ternary fission product, by acti-
extended time periods in the early German AVR and THTR reac- vation reaction of lithium and boron in graphite components, and
tors. In prismatic design reactors, such as the GT-MHR, ANTARES He-3 fractions of the He coolant. A small amount of the tritium
and JAEA designs, the TRISO fuel particles are imbedded in produced tends to permeate through the IHX to the secondary
graphite blocks, whereas in pebble bed design, such as the PBMR helium loop and the HPP. Furthermore, it is probable that it will mix
and Chinese designs, the TRISO particles are embedded in with the product hydrogen. In the same way that tritium can
graphite spheres. permeate from the reactor to the HPP, hydrogen can permeate in
Both the GT-MHR and GTHTR300 achieve passive safety by the opposite direction.
designing for a core cool-down that limits the peak fuel As part of the HYTHEC project, preliminary calculations were
temperature to 1600  C (INL, 2006; Kunitomi et al., 2007). Under carried out to estimate the amount of tritium permeation (de
normal operating conditions, the control rods are employed to Lorenzo et al., 2006). In contrast to normal operating conditions,
stop the reactor. In case of failure of the control rods, a backup the leakage requires a pressure drop between the primary and
system is also in place. Decay heat is conducted radially through secondary loops of the IHX. The secondary loop is purposefully
the core and pressure vessel and then radiated to passive air- operated at a pressure above that of the primary loop to avoid this
cooled panels in the reactor cavity building. The air-cooled panels happening. A total tritium release into the primary helium stream
also eliminate the need for high temperature concrete. Direct core was calculated to be 6.43  105 Ci/s, leading to a tritium
cooling is not necessary in any accident due to the inherent safety contamination of the hydrogen of 4.066  102 Ci/tonne H2. This
characteristics of the high thermal capacity of the core and the value is significantly lower than allowed maximum limits,
strong negative coefficient of reactivity. Under any postulated however, is based on many assumptions and not experimental
depressurisation accident the TRISO fuel particles will retain the values.
fission products. The Japanese have carried out testing on various metals
In Pebble bed reactors, the fuel pebbles are located in an annular including Hastelloy XR, the material used for the tubes of the IHX,
cavity in the reactor vessel. They are dropped from several points to investigate the permeability of hydrogen isotopes (Masui et al.,
above the core annulus and proceed vertically downwards until 1979; Takeda et al., 2004; Sakaba et al., 2006). Permeability values
they are removed at the bottom. This online refuelling makes were obtained for hydrogen, deuterium and tritium. The perme-
reactor shut-downs unnecessary and allows the reactor to operate ation rate depends on the hydrogen partial pressure and the
with no excess reactivity. The graphite core of the reactor provides surface condition of the metal. The amount of tritium permeation
a high heat capacity and slow thermal response. The reactor has through the reaction tube is expected to be reduced by counter-
a negative coefficient of reactivity, which inherently shuts down permeation of tritium and hydrogen. A helium purification system
the core when the temperature exceeds normal operating condi- is included on the primary helium loop to reduce the quantity of
tions. Passive heat decay is provided in the same way as the pris- impurities such as hydrogen. R&D is planned with the aim to
matic reactors, through conduction and radiation to the cavity reduce permeation through the IHX tubes by controlling the
cooling system, which has a capacity to absorb the heat for more formation of an oxidised film layer by adjusting the coolant
than 72 h (Koster et al., 2003). chemistry.
522 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

7.4. Explosion the US, the HTR-PM in China and the GTHTR300 in Japan. All five are
currently in the design and component testing stage, and aim to be
The explosion of a combustible gas such as hydrogen is a very operational around 2020. The VHTR is a Generation IV reactor
severe problem for reactor safety. The overpressure caused by which incorporates many inherent safety features. The US DoE Next
a blast can damage safety related components. The simplest Generation Nuclear Plant (NGNP) is a government funded pro-
counter measure is to have a long distance between the reactor and gramme aiming to produce an operational VHTR. It is currently in
the HPP, however, this leads to an increase in construction costs of the process of selecting the best design from those put forward by
the piping and a drop in helium temperature and so lower effi- GA (the GT-MHR), PBMR/Westinghouse and AREVA NP (ANTARES).
ciency. Other countermeasures, such as reducing the leak of There are also many small research programmes investigating
combustible gases and mitigating overpressures caused with particular components and features of the VHTR.
barriers should be considered. The placement of the reactor With the exception of the Chinese, all the countries with active
underground aids this. The rupture of gas pipes was found to be the R&D on VHTRs also have a nuclear hydrogen programme. As the
main cause of combustible gas leaks (Inagaki et al., 2007). Double only institution with both a working HTGR and nuclear hydrogen
walled tubes with an inert gas, such as nitrogen, in the cavity are research, the Japan Atomic Energy Agency (JAEA) are world leaders
being designed and tested with a view to protection from leakage, in the field and have the most all-encompassing research pro-
therefore enabling the HPP to be housed close to the reactor. gramme. Significant research is being carried out into high
temperature electrolysis, and both the SI and HyS thermochemical
7.5. Isolation valves cycles. As well as the major nuclear programmes, there are also
many other research centres investigating these processes.
Isolation valves are a key safety component when coupling There are four key components of a nuclear hydrogen plant: the
a nuclear reactor to a hydrogen production plant. In the case of reactor, the hydrogen production plant (HPP), the power conver-
a pipe rupture, it is essential to isolate the primary loop to prevent sion system (PCS) and the intermediate heat exchanger (IHX). The
the release of helium contaminated by fission products. As the IHX coupling between these components is also very important. The
is designed to have a pressure gradient across it, if depressurisation PCS and HPP employed vary dramatically between programmes.
occurs in the secondary loop, tube breakage is likely to occur. The For the PCS both direct and indirect coupling are possible and either
amount of primary coolant released will depend on the number of a Rankine steam cycle or a Brayton gas turbine cycle can be used.
failed tubes/components and the shut off time of the isolation PBMR and AREVA NP have chosen the more conventional Rankine
valve. Calculations suggest that the shut off time must be less than technology, whereas GA and JAEA have opted for Brayton cycles.
10 s (Yasuno et al., 1980). PBMR are also developing Brayton cycle technology and plan to use
Two isolation valves are placed in the primary helium loop of it in the future. JAEA have concentrated their R&D efforts on the SI
the Japanese design, one to prevent radioactive material release cycle, whereas the others are also looking at high temperature
from the reactor to the HPP and the other to prevent ingress of electrolysis and Westinghouse have chosen the HyS cycle. All
combustible gases from the HPP or IHX to the reactor (Inagaki et al., current designs involve indirect coupling of the HPP to the reactor
2007). These valves must operate at very high temperatures. JAEA via an IHX. Materials of construction for the IHX, as well as other
have been conducting design and testing work on an angle valve, components, present an engineering challenge due to the high
focussing on minimising helium leakage and the prevention of temperatures involved and much R&D is centred on developing
deformation of the valve seat (Nishihara et al., 2004). The structural new alloys.
integrity of a new coating material was confirmed. Further work is Predicted process efficiencies and plant costs are currently at
being carried out to investigate the durability. a preliminary stage and are very similar, regardless of the options
Some testing of valves has also been carried out at PBMR, chosen. The cost of hydrogen produced from water splitting using
however, only up to 350  C (Matzner, 2004). Further testing at nuclear technologies is around $2/kg H2. The technological feasi-
higher temperatures is planned. One of the work packages in the bility and testing of key components will be one of the determining
RAPHAEL project, WP-CT1, is dedicated to heat exchangers, valves factors in plant viability.
and vessels (Besson et al., 2006). A design study will be carried out
based on past experience in Germany as well as on current tech-
Acknowledgements
nologies, for example the AGR.

This report has been produced under the EPSRC’s ‘‘Keeping the
7.6. Licensing
Nuclear Option Open’’ programme whose support is gratefully
acknowledged.
Licensing of the nuclear plants is very important and will only
be carried out when stringent safety criteria are set and followed.
The above sections have laid out some of the primary concerns References
which need to be addressed before a nuclear hydrogen plant
becomes operational. Australian Academy of Science, 2008. Towards development of an Australian
scientific roadmap for the hydrogen economy.
Bamberger, C.E., 1978. Hydrogen production from water by thermochemical cycles:
8. Conclusions a 1977 update. Cryogenics 18, 170–183.
Bamberger, C.E., Richardson, D.M., 1976. Hydrogen production from water by
thermochemical cycles. Cryogenics 16, 197–208.
Nuclear hydrogen production has the potential to provide Barbier, D., Bertrand, F., Haubensack, D., Aujollet, P., Alpy, N., 2006. Massive H2
a carbon neutral energy solution for the future. Very High production (I/S cycle) with VHTR heating: preliminary designs & safety
Temperature Reactors (VHTRs) can be coupled to a hydrogen approach for coupling. In: HTR2006: 3rd International Topical Meeting on High
Temperature Reactor Technology, Johannesburg, South Africa.
production plant (HPP) utilising either high temperature electrol-
Beghi, G.E., 1986. A decade of research on thermochemical hydrogen at the Joint
ysis (HTE), the Sulphur Iodine (SI) thermochemical cycle or the Research Centre, ISPRA. International Journal of Hydrogen Energy 11, 761–771.
hybrid sulphur (HyS) cycle. Besson, D., Breuil, E., Verrier, D., Briottet, L., Tochon, P., Buckthorpe, D., Casalta, S.,
There are currently two small-scale, operational VHTRs in the Chauvet, V., Cucini, D., von Lensa, W., Falcand, C., Gronek, M., Worlitz, F., 2006.
RAPHAEL IP project high and very high temperature component developments.
world, with five more, of a much larger scale, in development. These In: HTR2006: 3rd International Topical Meeting on High Temperature Reactor
are the PBMR in South Africa, ANTARES in France, the GT-MHR in Technology, Johannesburg, South Africa.
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 523

Billot, P., Hittner, D., Vasseur, P., 2006. Outlines of the French R&D Programme for Gorensek, M.B., Summers, W.A., Buckner, M.R., Qureshi, Z.H., 2005. Conceptual
the development of high and very high temperature reactors. In: HTR2006: 3rd Design and Projected Performance for a Hybrid Sulfur Process. Cincinnati, OH,
International Topical Meeting on High Temperature Reactor Technology, United States. American Institute of Chemical Engineers, New York, United
Johannesburg, South Africa. States.
Bramblett, G.C., Fisher, C.R., Swart, F.E., 1980. Operational experience at Fort St Greyvenstein, G.P., Rousseau, P.G., 2003. Design of a physical model of the PBMR
Vrain. Specialists Meeting on Gas-cooled Reactor Safety and Licensing Aspects. with the aid of Flownet. Nuclear Engineering and Design 222, 203–213.
Lausanne, Switzerland. Greyvenstein, R., Correia, M., Kriel, W., 2006. South Africa’s opportunity to maxi-
Brecher, L., Spewock, S., Warde, C., 1977. The Westinghouse sulphur cycle for the mise the role of nuclear power in a global hydrogen economy. 3rd International
thermochemical decomposition of water. International Journal of Hydrogen Topical Meeting on High Temperature Reactor Technology: HTR2006. Johan-
Energy 2, 7–15. nesburg, South Africa.
Breitenfelder, R., Wachholz, W., Weicht, U., 1980. Accident analysis and accident Guojun, Y., Yang, X., Zhengang, S., Huidong, G., 2007. Characteristic analysis of rotor
control. Specialists Meeting on Gas-cooled Reactor Safety and Licensing dynamics and experiments of active magnetic bearing for HTR-10GT. Nuclear
Aspects, Lausanne, Switzerland. Engineering and Design 237, 1363–1371.
Brown, L.C., 2007. Evolution of the sulfur–iodine flowsheet, 1977–2007. AIChE Hammerschlag, R., Mazza, P., 2005. Questioning hydrogen. Energy Policy 33,
Annual Meeting. Salt Lake City, USA. 2039–2043.
Caputo, G., Felici, C., Tarquini, P., Giaconia, A., Sau, S., 2007. Membrane distillation of Hawkes, G.L., O’Brien, J.E., Stoots, C.M., Hawkes, B., 2007. 3D CFD model of a multi-
HI/H2O and H2SO4/H2O mixtures for the sulfur–iodine thermochemical process. cell high temperature electrolysis stack. AIChE Annual Meeting 2007, Salt Lake
International Journal of Hydrogen Energy 32, 4736–4743. City, USA.
Chang, J., Kim, Y.W., Lee, K.Y., Lee, Y.W., Lee, W.J., Noh, J.M., Kim, M.H., Lim, H.S., Hejzlar, P., Pope, M.J., Williams, W.C., Driscoll, M.J., 2005. Gas cooled fast reactor for
Shin, Y.J., Bae, K.K., Jung, K.D., 2007. A study of a nuclear hydrogen production Generation IV service. Progress in Nuclear Energy 47, 271–282.
demonstration plant. Nuclear Engineering and Technology 39, 111–122. Herranz, L., Linares, J., Moratilla, B., 2007. Assessment of regenerative reheating in
Colak, U., Brey, L., Methnani, M., 2006. Status of the IAEA coordinated research direct brayton power cycles for high-temperature gas-cooled reactors. Nuclear
project on HTGR core physics and thermal hydraulics benchmarks (CRP-5). In: Technology 159, 15–24.
HTR2006: 3rd International Topical Meeting on High Temperature Reactor Herranz, L., Linares, J., Moratilla, B., Lopez, R., 2006. Thermal assessment and second
Technology, Johannesburg, South Africa. law analysis of indirect brayton power cycle for high-temperature gas-cooled
Correia, M., Greyvenstein, R., Silady, F., Penfield, S., 2006. PBMR as an Ideal Heat nuclear reactors. In: HTR2006: 3rd International Topical Meeting on High
Source for High-temperature Process Heat Applications. Miami, FL, United Temperature Reactor Technology, Johannesburg, South Africa.
States. American Society of Mechanical Engineers, New York, NY 10016-5990, Herring, J., Stoots, C., O’Brien, J., Hartvigsen, J., Housley, G., 2007a. Recent prog-
United States. ress in high temperature electrolysis. AIChE Annual Meeting 2007, Salt Lake
Crabtree, G.W., Dresselhaus, M.S., Buchanan, M.V., 2004. The hydrogen economy. City, USA.
Physics Today 57, 39–44. Herring, J.S., O’Brien, J.E., Stoots, C.M., Hawkes, G.L., Hartvigsen, J.J., Shahnam, M.,
De Beni, G., 1982. Thermochemical processes for hydrogen production by water 2007b. Progress in high-temperature electrolysis for hydrogen production
splitting: from theory to practice. Journal of the Electrochemical Society 129, using planar SOFC technology. International Journal of Hydrogen Energy 32,
67–72. 440–450.
de Lorenzo, D., Cedillo, J., Borgard, J., Corgnale, C., 2006. HYTHEC D6: I–S Cycle – Hino, R., Haga, K., Aita, H., Sekita, K., 2004. 38. R&D on hydrogen production by
VHTR reactor coupling with energy recovery. First Safety Assessment. high-temperature electrolysis of steam. Nuclear Engineering and Design 233,
FP6-HYTHEC-D6. 363–375.
Doizi, D., Dauvois, V., Roujou, J.L., Delanne, V., Fauvet, P., Larousse, B., Hercher, O., Hittner, D., Bogusch, E., Besson, D., Buckthorpe, D., Chauvet, V., Futterer, M.A., van
Carles, P., Moulin, C., Hartmann, J.M., 2007. Total and partial pressure Heek, A., von Lensa, W., Phelip, M., Pirson, J., Scheuermann, W., Verrier, D., 2006.
measurements for the sulphur–iodine thermochemical cycle. International RAPHAEL, a European project for the development of HTR/VHTR technology for
Journal of Hydrogen Energy 32, 1183–1191. industrial process heat supply and cogeneration. In: HTR2006: 3rd Interna-
Dostal, V., Driscoll, M.J., Hejzlar, P., 2004. A supercritical carbon dioxide cycle for tional Topical Meeting on High Temperature Reactor Technology, Johannesburg,
next generation nuclear reactors. MIT-ANP-TR-100. South Africa.
Engels, H., Funk, J.E., Hesselmann, K., Knoche, K., 1987. Thermochemical hydrogen Hong, S.-D., Kim, B.-K., Bae, K.-K., Choi, H.S., Hwang, G.J., 2007a. Evaluation of the
production. International Journal of Hydrogen Energy 5, 291–295. membrane properties with changing iodine molar ratio in HIx (HI–I2–H2O
Ewan, B., Graf, D., Monnerie, N., Sattler, C., Le Naour, F., Stein, W., Hinkley, J., 2007. mixture) solution to concentrate HI by electro-electrodialysis. Journal of
INNOHYP CA final report. INNOHYP CA-FR-CEA/07-05. Membrane Science 291, 106–110.
Ewan, B.C., Allen, R.W.K., 2005a. A figure of merit assessment of the routes to Hong, S.-D., Kim, J.-K., Kim, B.-K., Choi, S.-I., Bae, K.-K., Hwang, G.-J., 2007b. Evalu-
hydrogen. International Journal of Hydrogen Energy 30, 809–819. ation on the electro-electrodialysis to concentrate HI from HIx solution by using
Ewan, B.C.R., Allen, R.W.K., 2005b. A representation of limiting efficiencies for two types of the electrode. International Journal of Hydrogen Energy 32,
thermochemical cycles. AIChE Annual Meeting, Cincinnati, USA. 2005–2009.
Eyre, N., Fergusson, M., 2002. Fuelling road transport – implications for energy Hu, S., Liang, X., Wei, L., 2006. Commissioning and operation experience and safety
policy (Online). Available from: http://www.dft.gov.uk (cited April 2007). experiments on HTR-10. In: HTR2006: 3rd International Topical Meeting on
Fazluddin, S., Smit, K., Slabber, J., 2004. The use of advanced materials in VHTRs. High Temperature Reactor Technology, Johannesburg, South Africa.
HTR:2004 2nd International Meeting on High Temperature Reactor Technology, Hwang, G., Onuki, K., Nomura, M., Kasahara, S., Kim, J., 2003. Improvement of the
Beijing, China. Thermochemical water-splitting IS (iodine–sulfur) process by electro-electro-
Forsberg, C.W., 2007. Future hydrogen markets for large-scale hydrogen production dialysis. Journal of Membrane Science 220, 129–136.
systems. International Journal of Hydrogen Energy 32, 431–439. IAEA, 1999. Hydrogen as an energy carrier and its production by nuclear power.
Frohling, W., Unger, H., Dong, Y., 2002. Thermodynamic assessment of plant effi- IAEA-TECDOC-1085.
ciencies for HTR power conversion systems. In: HTR2002: 1st International IAEA, 2003. Evaluation of high temperature gas cooled reactor performance:
Topical Meeting on HTR Technology, Petten, The Netherlands. benchmark analysis related to initial testing of the HTTR and HTR-10. IAEA-
Fujikawa, S., Hayashi, H., Nakazawa, T., Kawasaki, K., Iyoku, T., Nakagawa, S., TECDOC-1382.
Sakaba, N., 2004. Achievement of reactor-outlet coolant temperature of 950  C IEA, 2007. Key World Energy Statistics 2007.
in HTTR. Journal of Nuclear Science and Technology 41, 1245–1254. Inaba, Y., Ohashi, H., Nishihara, T., Sato, H., Inagaki, Y., Takeda, T., Hayashi, K.,
Fujiwara, S., Kasai, S., Yamauchi, H., Yamada, K., Makino, S., Matsunaga, K., Takada, S., 2005. Study on control characteristics for HTTR hydrogen production
Yoshino, M., Kameda, T., Ogawa, T., Momma, S., Hoashi, E., 2008. Hydrogen system with mock-up test facility: system controllability test for fluctuation of
production by high temperature electrolysis with nuclear reactor. Progress in chemical reaction. Nuclear Engineering and Design 235, 111–121.
Nuclear Energy 50, 422–426. Inagaki, Y., Ohashi, H., Inaba, Y., Sato, H., Nishihara, T., Takeda, T., Hayashi, K.,
Funk, J., Reinstrom, R., 1966. Energy requirements in the production of hydrogen Ogawa, M., 2007. Research and development on system integration technology
from water. Industrial and Engineering Chemistry Process Design and Devel- for connection of hydrogen production systems to an HTGR. Nuclear Tech-
opment 5, 336–342. nology 157, 111–119.
Funk, J.E., 2001. Thermochemical hydrogen production: past and present. Interna- INL, 2006. Next generation nuclear plant materials research and development
tional Journal of Hydrogen Energy 26, 185–190. program plan. INL/EXT-06-11701.
Giaconia, A., Caputo, G., Ceroli, A., Diamanti, M., Barbarossa, V., Tarquini, P., Sau, S., INL, 2007. Next generation nuclear plant pre-conceptual design report. INL/EXT-07-
2007a. Experimental study of two phase separation in the Bunsen section of the 12967.
sulfur–iodine thermochemical cycle. International Journal of Hydrogen Energy IPCC, 2007. Climate Change 2007 – The Physical Science Basis. Cambridge University
32, 531–536. Press.
Giaconia, A., Grena, R., Lanchi, M., Liberatore, R., Tarquini, P., 2007b. Hydrogen/meth- Jeong, Y.H., Hohnholt, K.J., Yildiz, B., Kazimi, M.S., 2005. Optimization of the hybrid
anol production by sulfur–iodine thermochemical cycle powered by combined sulfur cycle for hydrogen generation. ICAPP ’05. Seoul, Korea.
solar/fossil energy. International Journal of Hydrogen Energy 32, 469–481. Kameyama, H., Yoshida, K., 1978. Test of one-loop flow scheme for the UT-3 ther-
Glatzmaier, G.C., Blake, D.M., 1998. Analysis of Hydrogen Production Methods Using mochemical hydrogen production process. 2nd World Hydrogen Energy
Electrolysis and Concentrated Solar Energy. Albuquerque, NM, USA. ASME, Conference, Zurich, Switzerland.
Fairfield, NJ, USA. Kodama, T., Gokon, N., 2007. Thermochemical cycles for high-temperature solar
Goldstein, S., Borgard, J.M., Vitart, X., 2005. Upper bound and best estimate of the hydrogen production. Chemical Reviews 107, 4048–4077.
efficiency of the iodine sulphur cycle. International Journal of Hydrogen Energy Koster, A., Matzner, H.D., Nicholsi, D.R., 2003. PBMR design for the future. Nuclear
30, 619–626. Engineering and Design 222, 231–245.
524 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525

Kreith, F., West, R.E., 2004. Fallacies of a hydrogen economy: a critical analysis of Ogawa, M., Nishihara, T., 2004. Present status of energy in Japan and HTTR project.
hydrogen production and utilization. Journal of Energy Resources Technology Nuclear Engineering and Design 233, 5–10.
Transactions of the ASME 126, 249–256. Oh, C.H., Moore, R.L., 2005. Brayton cycle for high-temperature gas-cooled reactors.
Kruger, P., 2001. Electric power requirement for large-scale production of hydrogen Nuclear Technology 149, 324–336.
fuel for the world vehicle fleet. International Journal of Hydrogen Energy 26, Ohashi, H., Inaba, Y., Nishihara, T., Inagaki, Y., Takeda, T., Hayashi, K., Katanishi, S.,
1137–1147. Takada, S., Ogawa, M., Shiozawa, S., 2004. Performance test results of mock-up
Kubo, S., Kasahara, S., Okuda, H., Terada, A., Tanaka, N., Inaba, Y., Ohashi, H., test facility of HTTR hydrogen production system. Journal of Nuclear Science
Inagaki, Y., Onuki, K., Hino, R., 2004a. A pilot test plan of the thermochemical and Technology 41, 385–392.
water-splitting iodine–sulfur process. Nuclear Engineering and Design 233, Pickard, P., Gelbard, F., Moore, B., Vernon, M., Parma, E., Besenbruch, G., Russ, B.,
355–362. Buckingham, R., Brown, L., 2007. II.J.1 Sulfur iodine thermochemical cycle. DOE
Kubo, S., Nakajima, H., Kasahara, S., Higashi, S., Masaki, T., Abe, H., Onuki, K., 2004b. Hydrogen Program FY2007 Annual Progress Report.
A demonstration study on a closed-cycle hydrogen production by the ther- Ponomarev-Stepnoi, N.N., Kukharkin, N.E., Bobrov, A.A., Chuniaev, E.I., Garin, V.P.,
mochemical water-splitting iodine–sulfur process. Nuclear Engineering and Glushkov, E.S., Kompaniets, G.V., Krutov, A.M., Lobyntsev, V.A., Polyakov, D.N.,
Design 233, 347–354. Smirnov, O.N., Zimin, A.A., 2003. Experiments on HTR criticality parameters at
Kuijper, J.C., Raepsaet, X., De Haas, J.B.M., Von Lensa, W., Ohlig, U., Ruetten, H.J., the ASTRA facility of the Kurchatov Institute. Nuclear Engineering and Design
Brockmann, H., Damian, F., Dolci, F., Bernnat, W., Oppe, J., Kloosterman, J.L., 222, 215–229.
Cerullo, N., Lomonaco, G., Negrini, A., Magill, J., Seiler, R., 2006. HTGR reactor Pro, B.H., Hammerschlag, R., Mazza, P., 2005. Energy and land use impacts of
physics and fuel cycle studies. Nuclear Engineering and Design 236, 615–634. sustainable transportation scenarios. Journal of Cleaner Production 13,
Kunitomi, K., Katanishi, S., Takada, S., Takizuka, T., Yan, X., 2004. Japan’s future HTR 1309–1319.
– the GTHTR300. Nuclear Engineering and Design 233, 309–327. Richards, M., Shenoy, A., Schultz, K., Brown, L., Fukuie, M., Harvego, E., 2006a. The
Kunitomi, K., Yan, X., Nishihara, T., Sakaba, N., Mouri, T., 2007. JAEA’s VHTR for modular helium reactor for hydrogen production. 15th Pacific Basin Nuclear
hydrogen and electricity cogeneration: GTHTR300C. Nuclear Engineering and Conference, Sydney, Australia.
Technology 39, 9–20. Richards, M., Shenoy, A.S., Venneri, F., Brown, L., Fukuie, M., Harvego, E., 2006b.
LaBar, M.P., Shenoy, A.S., Simon, W.A., Campbell, E.M., 2004. The gas-turbine Modular helium reactor deployment for flexible energy demands. In: HTR2006:
modular helium reactor. Nuclear Energy 43, 165–175. 3rd International Topical Meeting on High Temperature Reactor Technology,
Lahoda, E.J., McLaughlin, D.F., Mulik, R., Kriel, W., Kuhr, R., Bolthrunis, C., Corbett, M., Johannesburg, South Africa.
2006. Estimated costs for the improved HyS flowsheet. 3rd International Rodriguez, G., Robin, J., Billot, P., Berjon, A., Cachon, L., Carles, P., Leybros, J., Le
Topical Meeting on High Temperature Reactor Technology: HTR2006, Johan- Naour, F., Pra, F., Terlain, A., Tochon, P., 2006. Development of a key component
nesburg, South Africa. of the iodine sulfur thermochemical cycle: the SO3 decomposer. WHEC 16.
Larminie, J., Dicks, A., 2003. Fuel Cell Systems Explained. Wiley. Lyon, France.
Le Duigou, A., Borgard, J.M., Larousse, B., Doizi, D., Allen, R.W.K., Ewan, B.C., Roeb, M., Sattler, C., Kluser, R., Monnerie, N., De Oliveira, L., Konstandopoulos, A.G.,
Priestman, G.H., Elder, R.H., Devonshire, R., Ramos Sanchez, V., Cerri, G., Agrafiotis, C., Zaspalis, V.T., Nalbandian, L., Steele, A., Stobbe, P., 2006. Solar
Salvini, C., Giovannelli, A., De Maria, G., Corgnale, C., Brutti, S., Roeb, M., hydrogen production by a two-step cycle based on mixed iron oxides. Journal of
Noglik, A., Rietbrock, P., Mohr, S., de Oliveira, L., Monnerie, N., Schmitz, M., Solar Energy Engineering Transactions of the ASME 128, 125–133.
Sattler, C., Orden Martinez, A., de Lorenzo Manzano, D., Cedillo Rojas, J., Rousseau, P., van Staden, M., 2006. Introduction to the PBMR heat transfer test
Dechelotte, S., Baudouin, O., 2007. HYTHEC: An EC funded search for a long facility. 3rd International Topical Meeting on High Temperature Reactor Tech-
term massive hydrogen production route using solar and nuclear technologies. nology: HTR2006, Johannesburg, South Africa.
International Journal of Hydrogen Energy 32, 1516–1529. Sakaba, N., Kasahara, S., Onuki, K., Kunitomi, K., 2007. Conceptual design of
Lemort, F., Lafon, C., Dedryvere, R., Gonbeau, D., 2006. Physicochemical and ther- hydrogen production system with thermochemical water-splitting iodine–
modynamic investigation of the UT-3 hydrogen production cycle: a new tech- sulphur process utilizing heat from the high-temperature gas-cooled reactor
nological assessment. International Journal of Hydrogen Energy 31, 906–918. HTTR. International Journal of Hydrogen Energy 32, 4160–4169.
Lewis, M., Masin, J., 2005. An Assessment of the Efficiency of the Hybrid Copper- Sakaba, N., Ohashi, H., Takeda, T., 2006. Hydrogen permeation through heat transfer
Chloride Thermochemical Cycle. Cincinnati, OH, United States. American Insti- pipes made of Hastelloy XR during the initial 950  C operation of the HTTR.
tute of Chemical Engineers, New York, United States. Journal of Nuclear Materials 353, 42–51.
Masui, K., Yoshida, H., Watanabe, R., 1979. Hydrogen permeation through iron, Sakurai, M., Bilgen, E., Tsutsumi, A., Yoshida, K., 1996. Adiabatic UT-3 thermo-
nickel, and heat resisting alloys at elevated temperatures. Transactions of the chemical process for hydrogen production. International Journal of Hydrogen
Iron and Steel Institute of Japan 19, 547–552. Energy 21, 865–870.
Mathias, P., Brown, L.C., 2003. Thermodynamics of the sulfur–iodine cycle for Sandler, S., 1999. Chemical and Engineering Thermodynamics, third ed. Wiley.
thermochemical hydrogen production. 68th Annual Meeting of the Society of Shin, Y., Park, W., Chang, J., Park, J., 2007. Evaluation of the high temperature
Chemical Engineers, Japan. electrolysis of steam to produce hydrogen. International Journal of Hydrogen
Matzner, D., 2004. PBMR existing and future R&D test facilities. In: HTR2004: 2nd Energy 32, 1486–1491.
International Topical Meeting on High Temperature Reactor Technology, Bei- Shiozawa, S., 2007. Status of Japan’s HTR Program. International Congress on
jing, China. Advances in Nuclear Power Plants.
Matzner, D., Kriel, W., Correia, M., Greyvenstein, R., 2006. Cycle Configurations for Shiozawa, S., Fujikawa, S., Iyoku, T., Kunitomi, K., Tachibana, Y., 2004. Overview of
a PBMR Steam and Electricity Production Plant. Reno, NV, United States. HTTR design features. Nuclear Engineering and Design 233, 11–21.
American Nuclear Society, La Grange Park, IL 60526, United States. Spadoni, A., Lanchi, M., Felici, C., Favuzza, P., Liberatore, R., Tarquini, P., Mazzocchia,
Matzner, D., Wallace, E., 2005. The PBMR’s progress. Nuclear Engineering Interna- C., Tito, C.A., 2007. Kinetic analysis of the catalytic decomposition of hydrogen
tional 50, 17–21. iodide in the S–I thermochemical cycle. AIChE Annual Meeting, Salt Lake City,
Maus, W., Mendte, K., Teubner, H., 1984. The He/He heat exchanger – design and UT, US.
semitechnical testing. Nuclear Engineering and Design 78, 195–214. Steinberg, M., Cheng, H.C., 1989. Modern and prospective technologies for hydrogen
Mawdsley, J., Carter, J., Yildiz, B., Call, A., Kropf, A., Ferrandon, M., Myers, D., Maroni, production from fossil fuels. International Journal of Hydrogen Energy 14,
V., 2007. Post-test evaluation of the oxygen electrode from a solid oxide elec- 797–820.
trolysis stack and electrode materials development. AIChE Annual Meeting Steinfeld, A., 2005. Solar thermochemical production of hydrogen – a review. Solar
2007, Salt Lake City. Energy 78, 603–615.
McLaughlin, D.F., Paletta, S.A., Lahoda, E.J., Kriel, W., Nigra, M.M., McLaughlin, G.T., Stern, N., 2007. The Economics of Climate Change: The Stern Review. Cambridge
2005. Hydrogen costs for the PBMR thermal reactor and the westinghouse University Press, London.
process. AIChE Annual Meeting 2005. Cincinnati, OH, United States, American Stoots, C.M., O’Brien, J.E., Hartvigsen, J., 2007. Results of recent high temperature
Institute of Chemical Engineers, New York, NY 10016-5991, United States. coelectrolysis studies at the Idaho National Laboratory. AIChE Annual Meeting
Nakagawa, S., Tachibana, Y., Takamatsu, K., Ueta, S., Hanawa, S., 2004. Performance 2007, Salt Lake City.
test of HTTR. Nuclear Engineering and Design 233, 291–300. Summers, W., Steimke, J., Steeper, T., Hobbs, D., Colon-Mercado, H., Herman, D.,
Nishihara, T., Sakaki, A., Inagaki, Y., Takami, K., 2004. Development of high 2007. II.J.2 Hybrid sulfur thermochemical process development. DOE Hydrogen
temperature isolation valve for the HTTR hydrogen production system. Trans- Program FY2007 Annual Progress Report.
actions of the Atomic Energy Society of Japan 3, 381–387. Tadokoro, Y., Kajiyama, T., Yamaguchi, T., Sakai, N., Kameyama, H., Yoshida, K., 1997.
Noglik, A., Roeb, M., Rzepczyk, T., Sattler, C., Pitz-Paal, R., 2007. Solar Thermo- Technical evaluation of UT-3 thermochemical hydrogen production process for
chemical Generation of Hydrogen: Development of a Receiver Reactor for the an industrial scale plant. International Journal of Hydrogen Energy 22, 49–56.
Decomposition of Sulfuric Acid. Long Beach, CA, United States. American Society Takeda, T., Iwatsuki, J., Inagaki, Y., 2004. Permeability of hydrogen and deuterium of
of Mechanical Engineers, New York, NY 10016-5990, United States. Hastelloy XR. Journal of Nuclear Materials 326, 47–58.
Norman, J.H., Besenbruch, G.E., Brown, L.C., O’Keefe, D.R., Allen, C.L., 1982. Ther- Takizuka, T., Takada, S., Yan, X., Kosugiyama, S., Katanishi, S., Kunitomi, K., 2004.
mochemical water splitting cycle, bench-scale investigations and process R&D on the power conversion system for gas turbine high temperature reactors.
engineering. Nuclear Engineering and Design 233, 329–346.
O’Brien, J.E., McKellar, M., Stoots, C.M., Herring, J.S., Hawkes, G.L., 2007a. Parametric Tarquini, P., 2007. Review and current status of ENEA’s hydrogen program. ENEA-
study of large-scale production of syngas via high temperature co-electrolysis. HYTHEC workshop Rome, Italy.
AIChE Annual Meeting 2007, Salt Lake City, USA. Tennenbaum, J., February 10 2006. South Africa’s PBMR: world’s most versatile
O’Brien, J.E., Stoots, C.M., Herring, J.S., Hartvigsen, J.J., 2007b. Performance of planar nuclear system. EIR Science and Technology, 34–47.
high-temperature electrolysis stacks for hydrogen production from nuclear Teo, E.D., Brandon, N.P., Vos, E., Kramer, G.J., 2005. A critical pathway energy efficiency
energy. Nuclear Technology 158, 118–131. analysis of the UT-3 Cycle. International Journal of Hydrogen Energy 30, 559–564.
 R. Elder, R. Allen / Progress in Nuclear Energy 51 (2009) 500–525 525

Terada, A., Iwatsuki, J., Ishikura, S., Noguchi, H., Kubo, S., Okuda, H., Kasahara, S., Weirich, W., Knoche, K., Behr, F., Barnert, H., 1984. Thermochemical processes for
Tanaka, N., Ota, H., Onuki, K., Hino, R., 2007. Development of hydrogen water splitting – status and outlook. Nuclear Engineering and Design 78,
production technology by thermochemical water splitting IS process pilot test 285–291.
plan. Journal of Nuclear Science and Technology 44, 477–482. Wright, S.A., Fuller, R., Noall, J., Vernon, M., Radel, R., Prickard, P.S., 2008a. Super-
U.S. Department of Energy Office of Nuclear Energy Science and Technology Nuclear critical CO2 Brayton cycle compression and control near the critical point.
Energy Research Initiative (NERI), 2006. Centralized hydrogen production from ICAPP08, Anaheim, California, USA.
nuclear power: infrastructure analysis and test-case design study. WSRC-MS- Wright, S.A., Pickard, P.S., Vernon, M., 2008b. Initial status and test results
2005-00693. from a supercritical CO2 Brayton cycle test loop. ICAPP08. Anaheim,
United Kingdom Department of Environment Food and Rural Affairs, 2008. Climate California, USA.
change bill. Bill 97. Wu, Z., Yu, S., 2007. HTGR projects in China. Nuclear Engineering and Technology
US DoE, 2007. DoE Hydrogen Program FY2007 annual progress report. 39, 103–110.
US DoE, GIF, 2002. A technology roadmap for Generation IV nuclear energy systems. Xu, Y., Hu, S., Li, F., Yu, S., 2005. High temperature reactor development in China.
GIF-002-00. Progress in Nuclear Energy 47, 260–270.
US DoE NERI, 2006. Centralized hydrogen production from nuclear power: infra- Yan, X., Takizuka, T., Takada, S., Kunitomi, K., Minatsuki, I., Mizokami, Y., 2003. Cost
structure analysis and test-case design study. WSRC-MS-2005-00693. and performance design approach for GTHTR300 power conversion system.
Venter, P.J., Mitchell, M.N., 2007. Integrated design approach of the pebble BeD Nuclear Engineering and Design 226, 351–373.
modular reactor using models. Nuclear Engineering and Design 237, 1341–1353. Yasuno, T., Mitake, S., Ezaki, M., Suzuki, K., 1980. The reactor safety study of
Vitart, X., Carles, P., Anzieu, P., 2008. A general survey of the potential and the main experimental multi-purpose VHTR design. Specialists Meeting on Gas-cooled
issues associated with the sulfur–iodine thermochemical cycle for hydrogen Reactor Safety and Licensing Aspects, Lausanne, Switzerland.
production using nuclear heat. Progress in Nuclear Energy 50, 401–409. Yildiz, B., Kazimi, M.S., 2006. Efficiency of hydrogen production systems using
Wachholz, W., 1988. The present state of the HTR concept based on experience alternative nuclear energy technologies. International Journal of Hydrogen
gained from AVR and THTR. Technical committee meeting on design require- Energy 31, 77–92.
ments, operation and maintenance of gas-cooled reactors. San Diego, USA, Zhang, Z., Yu, S., 2002. Future HTGR developments in China after the criticality of
International Atomic Energy Agency. the HTR-10. Nuclear Engineering and Design 218, 249–257.
Wallace, E., Matzie, R., Heiderd, R., Maddalena, J., 2006. From field to factory – Zhang, Z., Wu, Z., Sun, Y., Li, F., 2006. Design aspects of the Chinese modular high-
taking advantage of shop manufacturing for the pebble bed modular reactor. temperature gas-cooled reactor HTR-PM. Nuclear Engineering and Design 236,
Nuclear Engineering and Design 236, 445–453. 485–490.