CONCENTR  ATING

SOL  A R
ENERGY

20 A Bright Spot for Energy Production

21 Concentrating Solar Thermal Overview
30 Solar Thermochemical Energy Storage
36 Leveraging the Ammonia Industry
for Solar Energy Storage
42 Meet the Authors
SPECIAL SECTION: SOLAR ENERGY
A Bright Spot
for Energy Production
R
ecently, I had the opportunity to view a great swath of
the U.S. in less than six hours. Flying west from New
York to California, I spotted major cities and snow-
capped mountain ranges, but what stood out to me the most
was an enormous circular arrangement of mirrors all facing
one central point. From my aerial vantage, I recognized it as
a concentrating solar power (CSP) plant, and later learned it
was likely the Crescent Dunes CSP plant in Nevada.
The shiny field looked alien to my eyes. Hailing from
Pittsburgh, PA — a city that regularly ranks as one of the
cloudiest in the U.S. — I’m more accustomed to seeing
traditional methods of power production that predominate on
the East Coast. I’ve climbed to platforms on stacks at coal
and natural gas power plants, but none of those dusty gray
places looked like this bright spot in the desert.
Crescent Dunes produces enough electricity to power
75,000 homes. It is not the largest CSP plant in the world, or
in the U.S., but it utilizes a molten salt thermal storage sys-
tem that enables it to provide electricity even when the sun
is not shining. This is a distinct advantage over other forms
of renewable energy, such as wind and photo­voltaics (PV),
which are subject to intermittency and are less adaptable to
storage solutions. Although the cost of PV has plummeted
in recent years, the addition of bulky batteries for storage
increases its cost and practicality. Other CSP plants use mol-
ten salts for storage, but heat the fluid indirectly, requiring
expensive heat exchangers and additional heat-transfer flu-
ids. Crescent Dunes places molten salt directly at the central
receiver, where it is heated to over 1,000°F; the molten salt
can be stored and then used up to 10 hours later to produce
steam and generate electricity.
This is just one plant and one impressive application of
concentrating solar thermal (CST) technology. In their article
on pp. 21–28, Alexander Mitsos and Jeff Cumpston describe
CEP extends a special thanks to Nick AuYeung for help coordinating this special section.
20  www.aiche.org/cep  July 2017  CEP
Photo courtesy of SolarReserve.
how CST can be employed in various applications in addition
to electricity generation (i.e., CSP), including desalination,
enhanced oil recovery, and chemicals and fuels production.
Nick AuYeung and Peter Kreider note in their article on
pp. 30–35 that despite some commercial operations, “global
CSP capacity is still small compared to other renewable
energy technologies.” They explain, however, that CSP can
catch up to other forms of renewable energy by following
the example of Crescent Dunes and incorporating energy
storage. Crescent Dunes employs thermal methods, but
AuYeung and Kreider see thermochemical energy storage
(TCES) as a potentially better option, as it has “inherently
higher energy density ... because, in addition to sensible
heat, energy is stored as chemical potential.”
Keith M. Lovegrove, et al., zero in on ammonia TCES as
the solution to CSP storage needs (pp. 36–41). They describe
how the well-established Haber-Bosch process for ammonia
production could be employed in a CST receiver that would
also function as a reactor.
This special section emphasizes energy storage because
it will likely be the key to the success of CSP, whether the
solution is thermal, thermochemical, or something else.
With CSP plants in 23 countries around the world and 24
operational plants in the U.S. alone, design and economic
challenges are not holding CSP back.
The authors see a role for chemical engineers in push-
ing CST technologies even further. AuYeung and Kreider
urge chemical engineers to leverage their knowledge base
to advance CST and TCES technologies to help aid process
intensification and energy efficiency efforts in the chemical
process industries (CPI).
Elizabeth Pavone
Associate Editor, CEP
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
 Concentrating
Solar Thermal
Overview
Jeff Cumpston Concentrating solar thermal (CST) technology
Alexander Mitsos
RWTH Aachen Univ., AVT — Aachener can be used to generate electricity, desalinate
Verfahrenstechnik, Process water, and more. Find out how CST works
Systems Engineering (SVT)
and the role chemical engineers can play.

C
oncentrating solar thermal (CST) technologies The Solar Energy Generating System (SEGS), commis-
produce high-temperature, high-quality energy that sioned between 1984 and 1990 in the U.S., was the first use
can be used to drive a variety of engineering pro- of concentrating solar for power generation. The SEGS proj-
cesses. CST is attractive because of free and abundant solar ect consists of nine separate trough power plants that have
radiation, but significant engineering challenges need to be a total 354 MWe nameplate power generation capacity (2).
overcome. The project marked the beginning of a resurgence of interest
A CST system is composed of a field of distributed in CST technology.
reflectors that move over the course of a day to concentrate All of the components in a CST plant are designed to
sunlight onto a solar receiver (Figure 1). CST technologies
have different geometries and arrangements, but the same
principle of operation: A solar concentrator focuses solar
radiation onto a solar-thermal receiver that absorbs the
Receiver
solar radiation, converting it to thermal energy.
CST is not a new idea. Stories of its application go as
far back as antiquity. Folklore within the scientific commu-
nity suggests that Archimedes led attempts to focus solar
radiation onto Roman ships using many reflective surfaces.
The first modern commercial application of CST was the
implementation of a trough system in Egypt constructed in
1913 to produce hot steam to drive a pump for irrigation of
semi-arid farmland (1). Reflectors

u Figure 1. Reflectors move throughout the day to concentrate solar

radiation onto a central receiver at the PS10 concentrating solar power
(CSP) plant in Seville, Spain. Image courtesy Solúcar.

Copyright © 2017 American Institute of Chemical Engineers (AIChE) CEP  July 2017  www.aiche.org/cep  21
SPECIAL SECTION: SOLAR ENERGY

facilitate the absorption of well-focused sunlight by the solar receiver. In addition, heat losses at the solar receiver must be
receiver; thus, all the interactions that the solar radiation considered (as discussed in more detail later).
undergoes must be considered. This general discussion of CST technologies details the
The entire atmosphere — from the top of the Earth’s different types of solar collectors and the various applications
atmosphere to the concentrator, and between the concentra- to which CST can be applied, including electricity genera-
tor and receiver (Figure 2) — affects the quality of the solar tion (i.e, concentrating solar power [CSP]), desalination,
radiation. The night and day cycle, clouds, and aerosols in advanced oil recovery, and solar thermochemistry incorporat-
the atmosphere cause intermittencies in the solar radiation ing the production of solar fuels.
reaching the receiver. Intermittencies must be accounted for, CSP is an important application of CST technology. The
but energy storage technologies can mitigate their detrimen- term encompasses both CST plants that are used to gener-
tal effects on CST performance. ate electricity and concentrating photovoltaic (CPV) power
Designing and characterizing solar concentrators for plants. This article focuses on CST’s application for power
CST applications involves optimizing the concentration generation and does not discuss CPV.
ratio of the radiation considering the collector cost and
material limitations. A low concentration ratio implies low-­ Solar radiation and the atmosphere
temperature thermal energy and thus low exergy; a high con- Solar radiation incident at the outer edge of the Earth’s
centration ratio implies large thermal losses from the solar atmosphere has an intensity of about 1,370 W/m2 — the
so-called solar constant. Because of the
high temperature of the sun, the radia-
Solar Radiation
tion has a very high exergy. As viewed
Receiver from the Earth, the sun subtends a
Atmosphere half-angle of about 4.65 mrad. Propaga-
tion through the atmosphere affects the
Thermal intensity and the angular distribution of
Energy
Solar the incoming solar radiation. Aerosols
Concentrator
scatter the incoming radiation, increas-
ing the amount of circumsolar radiation,
which affects the concentration and
distribution of the image at the receiver
of a CST collector. Clouds attenuate and
scatter the radiation.
p Figure 2. Solar radiation from the sun passes through Earth’s atmosphere and is concentrated onto
a solar receiver. The thermal energy produced can be used, stored as thermal energy, or converted to Thus, it is important to under-
chemical energy and stored. stand the time dependence and spatial
distribution of clouds and aerosols.
Meteorological models and experimen-
tally derived data from ground stations
and satellites provide an overview and
inform predictions of time-dependent
changes in the atmosphere, which help
to define the expected functionality of
CST technologies.
Radiation incident on the surface of
the Earth is categorized as either diffuse
or direct. In direct solar radiation, most
of the angular distribution of the solar
source is preserved, and this direct radia-
tion can be concentrated and used in
CST applications. Diffuse solar radia-
tion cannot be concentrated because the
p Figure 3. This map of the global direct normal irradiation shows that the southwest U.S., northern
angular distribution is randomized and
Chile, Peru, central Australia, Saharan Africa, and South Africa are optimal regions for concentrating optics cannot focus it.
solar thermal (CST) technologies. Source: DNI Solar Map © 2016 Solargis (23). Therefore, CSP plants rely on direct

22  www.aiche.org/cep  July 2017  CEP Copyright © 2017 American Institute of Chemical Engineers (AIChE)
solar radiation, the intensity of which is described by the
direct normal irradiance (DNI), expressed in SI units as
Watts per square meter. Another important metric is the
global horizontal irradiance (GHI), which includes the dif-
fuse component of the solar radiation and is measured at a
plane parallel to the Earth’s surface.
Figure 3 is global map of the direct normal irradiation,
which is a time integral of DNI. High direct normal irradia-
tion makes the southwest U.S., northern Chile, Peru, central
Australia, Saharan Africa, and South Africa prime real estate
for development of CST technologies. Spain is a favorable
location for CST development in Europe.
Solar concentrators
Choosing a concentrator (i.e., reflector) type is one of the
chief optimization challenges of the fledgling solar thermal
industry. CST plants use four different types of concentra-
tors: linear Fresnel reflector (LFR), parabolic trough reflec-
tor, central receiver system with heliostats, and paraboloidal
dish reflector (Figure 4). Each
reflector type is defined by its
Multi-Reflector
ability to intercept and guide solar
radiation to a thermal receiver that (a)
(Figure 4b). Parabolic trough collectors are line-focus col-
lectors with a continuous surface geometry that is curved in
one dimension. The trough collector rotates about its long
axis and its receiver is fixed to the concentrator frame. Plants
in operation today generally align the collectors north-south.
East-west alignment offers an advantage of higher overall
annual solar collection, but comes with higher hourly varia-
tion in gathered solar energy due to cosine effects.
Parabolic trough collectors are capable of higher
concentration than a LFR, without requiring a secondary
concentrator. Similar to LFR systems, the receiver comprises
a large part of the energy transport network. Parabolic trough
concentrators are used in most CST plants currently in
operation.
Central receiver with heliostats (Figure 4c). Central
receiver systems are point-focus collectors composed of a
field of heliostats that surround a central receiver situated
at the top of a tower. Each heliostat tracks independently to
direct the solar radiation to the receiver. These systems, also
Single-Reflector
(b)

is engineered specifically for the

reflector type and application. LFR
and parabolic trough collectors are
classified as line concentrators and
Line Focus

track the sun along one axis. Cen-

tral receiver systems with heliostats
and paraboloidal dish collectors are
called point concentrators and track
the sun in two directions.
Linear Fresnel reflectors
(Figure 4a). A LFR is composed of
strips of mirrors. Each strip rotates
around the long axis to guide inci- (c) (d)
dent solar radiation to a stationary
receiver located above the reflec-
tors. The collectors are aligned
north-south.
LFRs are low-cost, low-­
Point Focus

concentration line-focus col-

lectors. Benefits include small
reflector size, low structural cost,
a stationary receiver position, and
noncylindrical receiver geometry
that reduces convection losses.
Receivers designed for use with
LFRs often incorporate a second-
ary concentrator to increase solar p Figure 4. Four types of concentrators are typically used in high-temperature CST applications: (a) linear-
concentration. Fresnel reflectors, (b) parabolic trough reflectors, (c) central receiver (or power tower) systems with heliostat
Parabolic trough collectors reflectors, and (d) paraboloidal dish reflectors. Source: Adapted from (24).

Copyright © 2017 American Institute of Chemical Engineers (AIChE) CEP  July 2017  www.aiche.org/cep  23
 SPECIAL SECTION: SOLAR ENERGY

known as power towers, are capable of higher solar concen-
tration than LFR and parabolic trough collectors.
The significant advantage of a central receiver system is
that much of the energy transport to the central power block
is performed optically, which removes the need for a large
pipe network to transport heat-transfer fluids (HTFs).
Paraboloidal dish collectors (Figure 4d). Paraboloidal
dish collectors are continuous-surface, point-focus col-
lectors. Dish collectors track the sun so that their aperture
planes are perpendicular to the solar vector at all times
during operation. Because the design geometry is that of
an ideal imaging reflective concentrator, these systems
can achieve the highest concentrations of the four types of
collectors.
Collector optical efficiency
Collector performance can be characterized by total
optical efficiency (ηopt), which is defined as the solar energy
that is incident on the receiver (as a fraction of the DNI)
multiplied by the reflector surface area. The total optical effi-
ciency is the product of efficiency coefficients that describe
the mirror reflectivity (ηrefl), cosine losses (ηcos), shading
(a) Absorber Tube
Glass Envelope
and blocking (ηshdblk), atmospheric attenuation (ηatm), and
receiver intercept (ηint).
Mirror reflectivity. The mirror reflectivity coefficient
is the fraction of the incident solar radiation reflected by a
given mirror or mirror field. Mirror reflectivity is affected
by the inherent reflectivity of the reflecting surface, the
cleanliness of the mirror, and any defects on the mirror
surface.
Cosine losses. The collectors discussed in this article
(except for the paraboloidal dish) intercept the sun at an
oblique angle, reducing the effective collection area by the
cosine of the angle.
Shading and blocking. The receiver, tower, adjacent
collectors, and other structures and formations may cast
a shadow on a portion of the collector field, reducing the
effective collector area. Blocking occurs when adjacent col-
lectors block solar radiation that has already been reflected.
Blocking does not usually occur in parabolic trough or
paraboloidal dish collector fields.
Atmospheric attenuation. The atmosphere between the
collector and the receiver attenuates the solar radiation, but
this is only significant for central receiver systems, where
the distances between col-
Bellows (b) Molten lector and receiver are often
Salt
hundreds of meters.
Receiver intercept. The
fraction of solar radiation
directed to the receiver that
Solar Radiation
Concentrated

is intercepted depends on
several factors: the geometry
Concentrated
Solar Radiation
of the receiver; the distribu-
tion of the solar radiation
(c) Pressure Vessel
that propagates toward the
Insulation
receiver, the so-called flux
distribution, which may be
Air
subject to broadening by
Solar Radiation
Concentrated

Inlet
atmospheric scattering; the
quality of the optical sur-
Air face of the collector; and the
Outlet
accuracy of collector tracking,
which depends on the location
Concentrated (d) of the sun. The complement
Absorber Solar Radiation
Secondary to this value is the receiver
Concentrator spillage, i.e., the fraction of
solar radiation that misses
p Figure 5. (a) Parabolic trough collectors employ the receiver.
cylindrical receivers through which a heat-transfer
medium propogates. (b) External cylindrical receiv- Divider The total optical effi-
ers are used with fields of heliostat collectors. (c) Plate Hot Salt ciency is not a fixed value
CST receivers can be used to heat air for use in because it depends strongly
high-temperature gas turbines. (d) The concentrating Cold Salt
on the position of the sun. A
solar power on demand (CSPonD) receiver integrates
the receiver and the storage medium for direct heat- useful metric for evaluating
ing of the solar salt. Source: Adapted from (3–6). the overall optical efficiency

24  www.aiche.org/cep  July 2017  CEP Copyright © 2017 American Institute of Chemical Engineers (AIChE)
of a collector is the annual optical efficiency. To calculate
this, the range of sun positions that occur throughout the
year must be considered, and the energy incident on the
solar receiver for each position must be estimated. The
most straightforward way to do this is to model the opera-
tion of the collectors as a linear time series that spans a year
of operation:
where the subscript ann denotes an annual value, G is the
DNI value at time t, Qrcv is the thermal energy absorbed by
the receiver, and QDNI is the energy available in the form of
DNI over the corresponding time period. DNI values from
satellites, weather stations, or solar irradiance models for
each time step are required for the calculation. For simplic-
ity, engineers sometimes use a design-point solar position
and DNI to represent system performance, although care
must be taken when making this assumption.
Solar receivers
A solar-thermal receiver intercepts and absorbs solar
radiation. Solar radiation incident on the receiver is con-
verted to thermal energy. In most commercial power plants,
HTF travels through the receiver and transports and/or stores
the thermal energy.
Parabolic trough collectors typically have an arrange-
ment consisting of a central pipe coated with antireflective
material through which thermal oil is pumped, surrounded
by a glass-vacuum tube (Figure 5a) (3). The Gemasolar
Thermosolar power-tower plant uses an array of vertical
pipes arranged in a cylindrical shape, which transfer thermal
energy to a molten salt that serves as the energy storage
medium (4) (Figure 5b).
In a volumetric receiver, the thermal medium may
directly absorb the thermal energy. Solar-thermal volumetric
receivers have been used to heat air for power generation,
such as in DLR’s REFOS receiver (Figure 5c) (5). The
CSP-on-demand (CSPonD) concept employs a volumetric
receiver in which molten salts are directly illuminated with
solar radiation (Figure 5d) (6).
The receiver must intercept as much solar radiation as is
practical. To achieve this, the receiver must be large enough
to capture most of the concentrated solar radiation, while
not being so large as to incur prohibitive thermal losses
and manufacturing costs. Therefore, the dimensions of the
absorbing surface depend on the imaging properties and
the optical quality of the solar concentrator. The absorber
surface must have a high optical absorptivity over the solar
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
spectrum, which is achieved with a highly absorptive surface
layer that is resistant to high temperatures and accumulation
of pollutants like fine dust.
Heat transfer from the absorber surface to the surround-
ings needs to be minimized. Radiative heat loss from the
absorber surface depends on the operating temperature
of the receiver, and predominates for high-temperature
applications. The receiver geometry can significantly reduce
radiative losses by maximizing the amount of re-emitted
radiation that is intercepted by the receiver surface. Spec-
trally selective coatings that have a low radiative emissivity
at the design temperature of the receiver can also minimize
radiative losses.
Convective heat loss from the absorber surface due to
hot-air buoyancy effects or forced convection from wind
incident on the absorber surface can be significant. Con-
vective heat loss can be minimized by receiver geometry or
by enclosing the absorber surface within a space optically
accessible by solar radiation. A quartz window at the aper-
ture of a cavity receiver can be used to eliminate convec-
tion losses while still allowing solar radiation to enter the
cavity. Heat loss by natural convection is a complex phe-
nomenon that can be accurately captured by computational
fluid dynamics (CFD) models, but these are computation-
ally expensive.
High conductivity between the absorber surface and
the thermal medium can, to an extent, reduce the receiver
absorber surface temperature and corresponding heat loss.
The receiver wall should have a high conductivity and
forced convection from the inside wall to the heat-transfer
medium should be maximized. Maximizing the HTF veloc-
ity can enhance forced convection, although care must be
taken to ensure the fluid attains the desired temperature in
the process. For example, it is critical that thermal oil within
the receiver of parabolic trough collectors reach a velocity
that corresponds to turbulent flow.
Concentrated solar thermal applications
CST technology provides high-exergy thermal energy
that can be leveraged in a variety of applications, including
electricity generation, desalination, oil recovery, and fuel
and chemical production. CST is already employed com-
mercially, primarily for power generation, including stand-
alone CSP plants and conventional power plants with CST
collectors that contribute to the overall thermal energy input.
CST collectors are also currently used to aid the recovery of
heavy crude oil.
Electricity generation
The thermal energy from a CST plant can be used to
generate steam to produce electricity, which, as noted earlier,
is referred to as concentrating solar power (CSP). Steam can
CEP  July 2017  www.aiche.org/cep  25
SPECIAL SECTION: SOLAR ENERGY

Solar Collector Field Molten Salt System Power Block Steam Turbine boosted by government feed-in
Generator
tariffs, which reward CSP plant
Generator Turbines
owners well above market price for
Receiver the electricity they produce.
Tower
At the end of 2015, the U.S.
High-Pressure CSP plant installed capacity reached
Steam Reheat Condenser 1.7 GW. Emerging markets are
Steam
found in Morocco, South Africa, and
China.
Superheater Reheater
Thermal energy storage. CSP
Molten Salt
Loop Steam plants with energy storage can
Hot
Generator/Evaporator
Condensate complement intermittent renew-
Salt Tank able energy sources by providing
Feedwater Preheater
clean dispatchable electricity. For
example, CSP with energy storage
Heliostats
Cold Steam Generation System could aid grids with high solar-PV
Salt penetration, supplementing baseload
capacity. The combination of tech-
Thermal Storage System
nologies could reduce curtailment
during times of maximum solar irra-
p Figure 6. In this power-tower system, thermal energy is stored in molten salt, which is used to produce
steam for electricity generation. diance, helping to further penetration
of PV or other renewable electricity
be generated directly in the receiver, which Brightsource’s sources. In addition, the thermal energy storage system can
Ivanpah Solar Electric Generating System (ISEGS) does, or be combined with electric heaters to reduce curtailment by
via a heat exchanger where a more suitable HTF conveys the enabling thermal energy storage of electricity that might
thermal energy from the receiver. The HTF used in the latter otherwise be wasted (10).
configuration is a thermal oil stable up to 390°C. The CSP plants with thermal energy storage that are
A so-called solar salt — a molten eutectic salt mix- currently operational generally use solar salt as an energy
ture comprised of 60% sodium nitrate and 40% potassium storage medium. Vast Solar (an Australian CSP company),
nitrate — may also be used as a heat-transfer medium however, generates electricity from a multitower solar array
(Figure 6). Solar salt HTFs can operate at temperatures up that uses energy stored in liquid sodium. Thermo­chemical
to 565°C, enabling a higher energy conversion efficiency at energy storage (TCES) is also a storage option, and is dis-
the turbine. Solar salts also have the advantage of high heat cussed in detail on pp. 30–35.
capacity, so they can be stored hot in tanks for later use. The Integrated solar combined-cycle. An integrated solar
ability to store thermal energy helps to overcome intermit- combined-cycle (ISCC) scheme combines steam produced
tency issues — a major advantage of CSP over other renew- from solar-thermal collectors with steam from fossil fuel
able forms of power generation. power stations. The steam from the CST collectors can
Total global installed capacity of CSP plants is 4.8 GWe be used as a supplement to reduce fossil fuel costs and
(7), which lags behind solar-photovoltaic (PV) and wind emissions, or it can be used to increase overall electricity
electricity generation. The likely reason is that CSP plants generation.
are currently more costly than solar-PV, wind, and conven- The LCOE of electricity from CSP with an ISCC scheme
tional gas- or coal-fired power plants (using levelized cost is 35–40% less than the LCOE of a stand-alone CSP plant
of electricity [LCOE, $/kWh] as the metric). To counter (11). ISCC could be a cost-effective baseload electricity
this, the U.S. Dept. of Energy (DOE) funds the SunShot generation alternative that would help speed up the transition
initiative, which aims to reduce LCOE of a CSP plant to to sustainable energy production.
$60/MWh by 2020. The DOE reports that since its inception
in 2010, the program has helped lower the LCOE of CSP Desalination
plants from $210/MWh to $130/MWh (8). Water scarcity is a major global challenge that is more
Spain hosts 50 CSP plants with about 2.3 GW of pronounced in arid regions with high population density.
nameplate capacity; installed between 2007 and 2012, these These areas are often along the coast, which makes sea­
plants account for a majority of CSP installed globally water desalination a viable option for the production of
(9). The high penetration of CSP plants in Spain has been fresh water.

26  www.aiche.org/cep  July 2017  CEP Copyright © 2017 American Institute of Chemical Engineers (AIChE)
 These locations often also have high solar potential. CST can also be used to generate steam for steam
Thus, renewable energy sources such as CST could help reforming of biomass or fossil fuels into hydrogen. Molten
meet the energy demand of desalination by membrane (e.g., salt energy storage could be used to meet the demands of
reverse osmosis) and thermal (e.g., multiple-effect distilla- continuous operation (17).
tion [Figure 7]) techniques. Multiple-effect distillation could Extractive metallurgy involves the thermochemical
be used to cogenerate power and water. Coupling the pro- extraction of a variety of useful metals from other com-
duction of water and electricity allows for process
optimization; electricity could be produced to meet 1st Effect 2nd Effect 3rd Effect
high demand, while water could be produced when Pressure P1 Pressure P2 Pressure P3

demand for electricity lags (12). Seawater Vacuum Vacuum Vacuum

Enhanced oil recovery

About 70% of global oil reserves are heavy Steam

crude oil — an especially viscous oil. To aid extrac-

tion, enhanced oil recovery (EOR) injects steam
into the well to reduce the viscosity of the oil. EOR
can boost well productivity by up to 300%, increas-
ing crude oil recovery from 5–8% to 50–70%.
The process, however, uses about a quarter of the
energy that is retrieved as heavy crude oil.
Typically, natural gas is used to produce the
steam for injection in EOR. However, CST can also
supply the thermal energy required for the injection
process. CST EOR could help reduce gas costs to
increase margins, particularly when oil prices are
low. In areas where natural gas prices are high, CST
EOR is particularly effective. A 300-kWth CST
EOR plant was constructed in California in 2011,
and a 7-MWth CST EOR plant in Oman, Jordan,
has been operating since May 2013 (13).
Condensate
Fresh Brine
Water
p Figure 7. Multiple-effect distillation can be used for saltwater desalination. In each
effect (or stage) of the process, heat is transferred from input steam to a saltwater feed,
extracting fresh water through evaporation. Heat transfer from the steam occurs in the
first effect, and each stage after that transfers the excess sensible heat that is contained
in the evaporated steam to the saltwater feed, so that by the final stage, most of the
thermal energy from the steam has been used to evaporate water. The input steam can be
produced using CST.
Concentrated Solar Radiation

Fuel production and chemical processing

Oxygen Oxygen
CST plants with high concentration ratios
achieve high temperatures that can be lever-
aged in the production of synthetic fuels and
other chemicals. These capabilities are still in the
research phase. These research examples high-
Counter-Rotating
light the potential of CST in chemical engineering Ceramic Disks
applications.
Solar fuels are liquid hydrocarbons that are
generated by chemical adhesion of atoms from
a feedstock of carbon dioxide and water. CST is
employed as part of a metal-oxide thermo­chemical Carbon Carbon
Dioxide Dioxide
cycle that requires temperatures up to 1,500°C to
activate a ceramic catalyst that removes oxygen
from carbon dioxide (Figure 8) (14).
The application of CST in solar thermo­ CO to Processing into Fuel
chemical CO2 capture via calcium-oxide looping
is being investigated (15, 16). In this process, CST p Figure 8. Solar radiation is used to remove oxygen from carbon dioxide for processing
into solar fuels. The concentrated solar energy heats ceramic disks rotating in opposite
is used to achieve temperatures above 1,200°C,
directions. At 1,500°C, ceramic releases oxygen from its molecular lattice. The oxygen-­
which are required by the endothermic calcination deficient ceramic absorbs oxygen from CO2 at 1,100°C, creating CO for processing
step in the calcination/carbonation reaction. into fuels.

Copyright © 2017 American Institute of Chemical Engineers (AIChE) CEP  July 2017  www.aiche.org/cep  27
SPECIAL SECTION: SOLAR ENERGY
pounds. The process requires the addition of a reductant
and heat. Conventional operations use coal as the reducing
agent and, via combustion, the source of heat. Process heat
produced from CST could, for example, reduce the carbon
consumed to extract iron from hematite by over 80% (18).
The role of chemical engineers
CST achieves high temperatures, and high-temperature
operation requires materials that can handle long-term expo-
sure to harsh conditions. Chemical engineers could help to
develop such materials at an economical cost.
Chemical engineers can contribute their understanding of
chemical processes and chemical pathways to design reac-
tions that temper difficult engineering requirements, making
them more compatible with CST (19). For example, the use
Literature Cited
1. Ragheb, M., “Solar Thermal Power and Energy Storage Historical
Perspective,” Univ. of Illinois at Urbana-Champaign, www.ragheb.
co/NPRE 498ES Energy Storage Systems/Solar Thermal Power and
Energy Storage Historical Perspective.pdf (Sept. 10, 2014).
2. National Renewable Energy Laboratories, “Concentrating Solar
Power Projects,” U.S. Dept. of Energy, Golden, CO, www.nrel.gov/
csp/solarpaces (accessed Mar. 24, 2017).
3. Lüpfert, E., et al., “Experimental Analysis of Overall Thermal
Properties of Parabolic Trough Receivers,” Journal of Solar Energy
Engineering, 130 (2) (Mar. 10, 2008).
4. Romero, M., and A. Steinfeld, “Concentrating Solar Thermal
Power and Thermochemical Fuels,” Energy and Environmental Sci-
ence, 5 (11), pp. 9234–9245 (2012).
5. Bräuning, T., et al., “Solar-Hybrid Gas Turbine-Based Power
Tower Systems (REFOS),” Journal of Solar Energy Engineering,
124 (3), p. 21 (2002).
6. Slocum, A. H., et al., “Concentrated Solar Power on Demand,”
Solar Energy, 85 (7), pp. 1519–1529 (2011).
7. Sawin, J. L., et al., “Renewables 2016: Global Status Report,”
Renewable Energy Policy Network for the 21st Century, www.
ren21.net/wp-content/uploads/2016/06/GSR_2016_Full_Report.pdf
(accessed Mar. 14, 2017).
8. U.S. Dept. of Energy, “Concentrating Solar Power,” DOE, Wash-
ington, DC, https://energy.gov/eere/sunshot/concentrating-solar-
power (accessed Mar. 14, 2017).
9. SolarPACES, “Spain,” The International Energy Agency, Paris,
France, www.solarpaces.org/country-information/spain (accessed
Mar. 14, 2017).
10. Lizarraga-Garcia, E., et al., “Optimal Operation of a Solar-
Thermal Power Plant with Energy Storage and Electricity Buy-Back
from Grid,” Energy, 51, pp. 61–70 (Mar. 2013).
11. Alqantani, B., and D. Patiño-Echeverri, “Integrated Solar
Combined Cycle Power Plants: Paving the Way for Thermal Solar,”
Applied Energy, 169, pp. 927–936 (May 2016).
12. Ghobeity, A., et al., “Optimal Time-Invariant Operation of a Power
and Water Cogeneration Solar-Thermal Plant,” Solar Energy, 85 (9),
pp. 2295–2320 (Sept. 2011).
13. GlassPoint Solar, Inc., “Solar-Powered Oil Production Overview,”
28  www.aiche.org/cep  July 2017  CEP
of a manganese-ferrite mixture in a water-splitting metal-
oxide redox cycle for hydrogen production can lower the
operating temperature of the metal oxide redox cycles from
up to 1,600°C to between 750°C and 800°C (20).
Optimized process scheduling using detailed process
models can be applied to CSP plants to maximize revenue
(21), and, similarly, to optimize energy use scheduling in
CST desalination plants (12). Desalination membrane tech-
nology is an ongoing field of research that requires careful
choice of materials (22).
A general understanding of chemical process engineering
and optimization facilitates engineering of chemical process
parameters, equipment, and operations, which helps to
broaden the range of applications economically and techni-
cally feasible with CST. CEP
www.glasspoint.com/wp-content/uploads/Solar-EOR-Overview.pdf
(2016).
14. Sandia National Laboratories, “Sunshine to Petrol: Solar Recy-
cling of Carbon Dioxide into Hydrocarbon Fuels,” U.S. Dept. of
Energy, Albuquerque, NM, http://energy.sandia.gov/wp-content/gal-
lery/uploads/S2P_SAND2009-5796P.pdf (accessed Mar. 14, 2017).
15. Matthews, L., and W. Lipinski, “Thermodynamic Analysis of
Solar Thermochemical CO2 Capture via Carbonation/Calcina-
tion Cycle with Heat Recovery,” Energy, 45 (1), pp. 900–907
(Sept. 2012).
16. Nikulshina, V., et al., “CO2 Capture from Atmospheric Air via
Consecutive CaO-Carbonation and CaCO3-Calcination Cycles in a
Fluidized-Bed Solar Reactor,” Chemical Engineering Journal,
146 (2), pp. 244–248 (Feb. 2009).
17. Pye, J., et al., “An Exergy Analysis,” in “Proceedings of Solar-
PACES 2014,” Beijing, China (Sept. 2014).
18. Steinfeld, A., “High-Temperature Solar Thermochemistry for CO2
Mitigation in the Extractive Metallurgical Industry,” Energy, 22 (2),
pp. 311–316 (Feb.–Mar. 1997).
19. Lanchi, M., et al., “Thermal Characterization of a Cavity Receiver
for Hydrogen Production by Thermochemical Cycles Operating at
Moderate Temperatures,” Solar Energy, 92, pp. 256–268
(June 2013).
20. Varsano, F., et al., “The Carbonatation Reaction of Layered
Na(Mn1/3Fe2/3)O2: A High Temperature Study,” Solid State Ionics,
187 (1), pp. 19–26 (Apr. 2011).
21. Vasallo, M. J., and J. M. Bravo, “A MPC Approach for Optimal
Generation Scheduling in CSP Plants,” Applied Energy, 165,
pp. 357–370 (Mar. 2016).
22. Jhaveri, J. H., and Z. V. P. Murthy, “A Comprehensive Review
on Anti-Fouling Nanocomposite Membranes for Pressure Driven
Membrane Separation Processes,” Desalination, 379, pp. 137–154
(Feb. 2016).
23. Solargis, “World Map of Direct Normal Irradiation,” http://solargis.
com/assets/graphic/free-map/GHI/Solargis-World-GHI-solar-
resource-map-en.png (accessed Mar. 14, 2017).
24. Cumpston, J. S., “Optimisation of Collector Locations and Pipe
Networks for Dish-Rankine Power Plants,” The Australian National
Univ., Canberra, Australia (Oct. 2016).
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
SPECIAL SECTION: SOLAR ENERGY
Solar Thermochemical
Energy Storage
Nick AuYeung Thermochemical energy storage
Oregon State Univ.
Peter Kreider could be the key to widespread
Australian National Univ. concentrating solar power (CSP) deployment.
T
hermal energy from the sun can be stored as chemi-
cal energy in a process called solar thermochemical
energy storage (TCES). The thermal energy is used
to drive a reversible endothermic chemical reaction, storing
the energy as chemical potential. During periods of high
solar insolation, an energy-consuming reaction stores the
thermal energy in chemical bonds; when energy is needed,
the reverse reaction recombines the chemical reactants and
releases energy.
Incorporating storage into concentrating solar power
(CSP) systems enables dispatchable generation, whereby
utilities produce power to match demand. This efficient
method of power production overcomes intermittency chal-
lenges faced by other forms of renewable energy produc-
tion. It also reduces the cost of solar energy through higher
utilization.
Thermochemical storage has inherently higher energy
density than latent- or sensible-heat storage schemes
because, in addition to sensible heat, energy is stored as
chemical potential. The endothermic reactions that could
be employed for solar TCES can operate at significantly
higher temperatures than current state-of-the-art CSP stor-
age systems (e.g., molten salt storage). Higher-temperature
operation enables the use of high-efficiency power cycles.
A storage scheme with higher energy density and higher
30  www.aiche.org/cep  July 2017  CEP
power cycle efficiency could greatly reduce the cost of solar-
derived power. The challenge in developing solar TCES,
however, is in achieving high enough energy and power den-
sity to offset increasingly complex process control, service
life, and material compatibility issues inherent in the use of a
high-temperature, chemically reactive system.
Global CSP capacity is still small compared to other
renewable energy technologies, but it has been expand-
ing with an annual growth rate of about 40% since 2006
(Figure 1). Breakthroughs in storage technologies could help
to increase CSP’s share of the energy portfolio.
Solar TCES subsystems
Incorporating solar TCES into CSP installations requires
integration of an additional subsystem — essentially a
chemical plant for the reversible energy storage reaction(s).
The subsystem receives heat from the solar collection field,
which it can use to drive the endothermic energy storage
reaction when power is not in demand.
Because the purpose of the chemical process is energy
storage, a critical component of the subsystem is the storage
tanks. Thermochemical storage mechanisms have a higher
energy density than thermal methods, which could help lower
capital costs by reducing storage tank volumes (Figure 2).
When energy is required from storage, the TCES sub-
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
system delivers heat to the power block. The reactants must 103
Wind
contact each other to react, while heat is simultaneously
extracted and sent to the power block. In an ideal design,
102
the heat-transfer fluid would also be a reactant (e.g., a gas)
that would directly contact the other reactant(s), such as in a
packed-bed reactor. Photovoltaics (PV)

Capacity, GWe
101
Solar TCES subsystems are classified as direct or
indirect (Figure 3). In a direct solar TCES subsystem, the
endothermic reactor receives heat directly from the solar 100
receiver, while in an indirect subsystem, the endothermic
reactor receives heat from the solar collection field via a Concentrating Solar Power (CSP)
heat-transfer fluid. The receiver and reactor are coupled in a 10–1
direct subsystem; heat-transfer fluids for extreme tempera-
tures may not be available or practical for some processes.
10–2
Although heliostat-tower (i.e., power tower or central 1980 1985 1990 1995 2000 2005 2010 2015 2020
receiver with heliostats) collectors are often associated with
solar TCES, TCES could conceivably be implemented with p Figure 1. From 1990 to 2005, wind and photovoltaics (PV) were
deployed at much higher rates than concentrating solar power (CSP). Since
both dish and trough designs, depending on the process tem- then, the growth curve for CSP has begun to more closely mirror those
perature requirements. Subsystems might include intermedi- other forms of renewable energy. Data courtesy John Pye.
ate separators, heat exchangers, and storage vessels related 35
to the type of chemical reaction used. For example, a metal Thermochemical
oxide gas/particle TCES subsystem would include a reduc- 30 Latent Heat

tion reactor (for storage when the sun is shining), a cyclone Sensible Heat
Intrinsic Energy Density, GJ/m3

25
separator (or similar equipment) to separate the charged
particles from the hot gas, heat exchangers for heat recovery 20
and heat delivery to the power block, an oxidation reactor (to
produce energy when the sun is not shining), and integrated 15
storage vessels that minimize heat losses and the distance
extremely hot gas/particles must travel. Simple chemical 10
reactions that avoid or minimize the need for complicated
5
separation steps would be expected to have the highest
energy and exergy efficiencies.
0
Molten H2 Cu/Mg/Si Wood CaO/CaCO3 Gasoline
Salts (100 bar) Phase Change
Power block Material
Most CSP installations use steam Rankine cycles to con-
p Figure 2. The energy density of thermochemical storage is inherently
vert thermal energy to electrical energy, but there are some higher than sensible or latent energy storage because energy is stored as
efforts to explore the use of organic Rankine cycles. Existing chemical potential.
Direct Subsystem Indirect Subsystem

Heat
Exchanger Endothermic
Receiver
Endothermic

Reactor
Receiver

Reactor

Product Storage
Separation Vessel

Product Separation
Pump Heat
Exchanger
Pump
Exothermic Power Storage Exothermic Power
Reactor Block Vessel Reactor Block

p Figure 3. Direct subsystems (left) combine the receiver and endothermic reactor in the same vessel. In an indirect subsystem (right), a heat-transfer
fluid transports heat from the receiver to the endothermic reactor. In both designs, the exothermic reactor exchanges heat with a power block when power is
in demand. Source: Adapted from (2).

Copyright © 2017 American Institute of Chemical Engineers (AIChE) CEP  July 2017  www.aiche.org/cep  31
SPECIAL SECTION: SOLAR ENERGY
Table 1. Thermochemical storage reactions are evaluated based on their
decomposition temperature and theoretical gravimetric energy densities.
Volumetric energy density is a more informative metric, but depends on process
design, storage strategies, and the structure of the chemical constituents.
Source: Data from (5, 12–14).
Reaction Decomposition Theoretical Gravimetric
Temperature at
1 atm, °C
Mn3O4 + ½ O2  Mn2O3 900
3 CoO + ½ O2  Co3O4 900
CaO + CO2  CaCO3 885
SrO + CO2  SrCO3 1,175
CaO + H2O  Ca(OH)2 507
Metal Hydrides 600–1,100
(1) 2 H2SO4(l)  2 H2O(g) + O2(g) + 2 SO2(g) 850 (charge)
2 H2O(l) + 3 SO2(g)  2 H2SO4(aq) + S(s,l)
(2) S(s,l) + O2(g)  SO2(g) 1,200 (discharge)
N2 + H2  NH3 200
Rankine systems typically employ a molten salt storage
system. CSP installations with molten salt storage operate
at about 600°C, which limits the choice of power block and
limits the achievable thermal efficiencies to about 40%.
The development of noncorrosive salts or TCES
systems capable of operating at higher temperatures is an
active area of research that could unlock the use of higher-
efficiency power block arrangements, such as combined
cycles or supercritical CO2 (sCO2) cycles. Combined cycle
power systems, composed of a Brayton (air) cycle followed
by a Rankine (steam) cycle, are currently employed com-
mercially for electricity generation. They achieve thermal
efficiencies of 50–60% and operate at temperatures of
800–1,500°C.
Supercritical CO2 cycles have received interest from the
CSP community as a potential next-generation power block
option. Since sCO2 has a higher density than steam, sCO2
turbines can be much smaller than conventional air or steam
turbines of the same power level. Operating temperatures for
sCO2 turbines are 800–1,000°C and thermal efficiencies are
reported to be about 50% (1). Given the smaller turbine size,
the lower thermal efficiency (compared to combined cycles)
may be a worthwhile concession.
TCES is typically compared to other forms of thermal
energy storage using volumetric energy density and volu-
metric power density. These metrics must exceed those
for commercialized molten salt or phase-change materials
to be considered a viable commercial pursuit. Exergetic
efficiency is another key metric, which ultimately impacts
the cost competitiveness of a TCES system. The U.S. Dept.
of Energy (DOE) has identified a thermochemical storage
subsystem capital cost target of $15/kWhth as a means of
achieving electricity prices of $0.06/kWhe (2).
32  www.aiche.org/cep  July 2017  CEP
Storage chemistries
Table 1 summarizes some chemical
reactions that have been studied for stor-
ing solar thermal energy. Most reactions
involve decomposition of a solid X to
Energy Density, form another solid Y and a gas; the gas
kJ/kg charged species
is recombined with solid Y when energy
202 is in demand. Solid-gas chemistries,
875 however, are inherently unsteady state
3,175 due to consumption of the solid reactant,
which complicates process control. Other
2,272
reactions, such as ammonia synthesis, are
1,930
a combination of gas and liquid phases.
~ 2,800 Metal oxides. Metal oxide oxid­-
9,600 ation/reduction reactions have received
attention because air provides an abun-
dant source of the oxygen reactant. Using
2,706 air also eliminates the need to store the
decomposed reactants between charge
and discharge steps. The challenges of concentrating solar
energy to achieve temperatures above 1,000°C and finding
materials capable of long-term exposure to these extreme
conditions make lower-temperature redox reactions an
attractive option. Manganese-based (3–5) and cobalt-based
(6) oxides, for example, decompose in air at temperatures
around 900°C.
Metal oxides are receiving substantial attention in the
TCES community. While the theoretical and experimental
performance of many pure and mixed metal oxides have
been explored, barium, cobalt, and manganese oxides are
some of the most promising performers based on their oper-
ating temperatures, material costs, and preliminary measure-
ments of chemical performance in cyclical oxidation and
reduction reactions (4, 5).
Current research efforts focus on mixed-metal oxides.
Perovskites are metal oxides of the form ABO3, where A
and B are two or more different metals. Perovskites with
more than one metal occupying the A and/or B sites can
have unique properties. Researchers at Sandia National
Laboratories have developed perovskite materials based
on LaxSr1-xCoyMn1-yO3 (LSCM) and LaxSr1-xCoyFe1-yO3
(LSCF) that have an experimentally demonstrated energy
density of 245 kJ/kg (7). More recently, they demonstrated
higher energy densities, in the vicinity of 390 kJ/kg, with
CaTi0.2Mn0.8O3 and CaAl0.2Mn0.8O3 (8).
Carbonates. Carbonates also garner attention as stor-
age media, in part because of their use as carbon capture
materials. Unlike metal oxides, however, carbonates require
storage of the gas-phase reactant (CO2), which must be
considered as part of an overall process design and techno­
economic assessment. Calcium carbonate (CaCO3) has a
reasonable decomposition temperature (~ 850°C) and low
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
cost. Recently, the suitability of SrCO3-based materials with
decomposition temperatures around 1,150°C for a com-
bined cycle power block have been explored (9).
The challenge of metal oxide carbonation chemistry is the
loss of surface area during cyclical operations. Over time, the
active materials tend to sinter, become denser, and eventually
become deactivated. Additionally, product layers form on the
surface during carbonation, which impedes diffusion of CO2,
slows the reaction kinetics, and ultimately reduces the con-
version that can be achieved over a reasonable time frame.
To make matters worse, the less-dense carbonate species will
clog the pores of the calcium oxide material due to volu-
metric expansion (lower density) of the calcium carbonate.
Materials called polymorphic spacers, which exhibit negative
thermal expansion, have been found to help prevent sintering
(10). Another strategy is the use of a steam hydration treat-
ment step to regenerate active surface area (11).
Metal hydroxides and hydrates. Group 1 and Group 2
metals have also been used to form hydroxides and hydrates.
Hydration of a Group 1 or Group 2 metal oxide is strongly
exothermic, while decomposition is conversely endothermic.
The advantage of this approach is that water is readily avail-
able and can be stored in its condensed phase. Unfortunately,
the use of water in a high-temperature process also requires
vaporization, which consumes energy. Salt-based hydroxides
and hydrates will likely see more use in seasonal thermo-
chemical storage schemes for low-­temperature, off-grid
residential applications, which have been a major develop-
ment thrust in the European Union.
Metal hydrides. Metal hydrides are intriguing because of
750-kWth 100-kWe
Concentrated Engine
Solar
Thermocline
Low-Temperature
Hydride High-Temperature
Storage Hydride
Thermochemical
Reactor
p Figure 4. EMC Solar designed a conceptual TCES system in which a
high-temperature hydride decomposes and a low-temperature hydride
absorbs the hydrogen. To charge the system, sunlight enters the calcium
hydride thermochemical reactor to release hydrogen. As hydrogen is
released, it flows toward the low-temperature metal hydride. Along the
way, the hydrogen passes through a thermocline, where it heats a sensible
energy storage medium (e.g., sand). In times of power demand, hydrogen
is absorbed by calcium in the high-temperature reactor to release heat. The
thermocline serves as a preheater for the hydrogen, which is released from
the low-temperature metal hydride. Source: Adapted from (12).
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
their fast kinetics and high energy densities. In a conceptual
system (Figure 4), a high-temperature hydride decomposes
and a low-temperature hydride absorbs the hydrogen. At
the core of making this concept a reality is finding a cheap
low-temperature hydride — a challenge that has also been
encountered in the hydrogen storage community.
It is worth working to overcome these challenges
because of the inherent benefits of a self-regulating system
that transfers hydrogen between a low-temperature and a
high-temperature metal hydride (12). During discharge,
the temperature decreases in the high-temperature reactor,
causing hydrogen to be absorbed by the high-temperature
metal hydride to release heat. As discharge progresses, the
system pressure decreases and hydrogen evolves from the
low-temperature metal hydride and flows toward the high-
temperature metal hydride. Conversely, when the system
is charging, the high-temperature metal hydride is heated;
hydrogen evolves and flows toward the low-temperature
metal hydride where it is absorbed.
Other TCES reactions. General Atomics has investigated
the use of a cycle based on the disproportionation of sulfur
dioxide to form sulfuric acid and elemental sulfur aided by
an iodine catalyst (Figure 5). Sulfuric acid can be thermally
decomposed to sulfur dioxide, while elemental sulfur can be
burned in oxygen to release 9,281 kJ/kg of thermal energy.
Unique to this chemistry is that sulfur can be combusted at a
higher temperature (~ 1,200°C) than the solar charging step
(~ 850°C). A disadvantage of this approach is the need for
three reaction steps and the toxicity of the sulfur constitu-
ents. Cost projections, however, indicate that this concept
could meet the SunShot metrics of less than $15/kWhth for
the subsystem (13).
Ammonia synthesis as a TCES medium is also under
development (see pp. 36–41). This method leverages fluid
components that exhibit automatic phase separation based
on the difference in density between the liquid ammonia and
the nitrogen and hydrogen gases. Storage of the gas phase
remains a challenge.
Solar Power
+ O2 Block
Field SO2
SO2 + H2O
H2SO4 SO2 Sulfur
Decomposition Disproportionation Combustion
H2SO4 Sulfur
p Figure 5. In General Atomic’s sulfur energy storage concept, sulfuric
acid decomposes into SO2, H2O, and O2. The SO2 disproportionates into
either sulfuric acid or elemental sulfur. If energy is needed, the elemental
sulfur can be combusted, which is strongly exothermic, to release a
substantial amount of energy.
Article continues on next page
CEP  July 2017  www.aiche.org/cep  33
SPECIAL SECTION: SOLAR ENERGY
Reactor designs
Reactor design is an area of solar TCES that will require
more attention. Ströhle, et al., at ETH-Zurich have analyzed
thermochemical storage from a design perspective to under-
stand how best to integrate the subsystem with the power
block. Using a manganese oxide-based redox chemistry
as a representative test case, they determined two general
concepts for integration of a TCES subsystem with a power
block (3):
• a parallel configuration, in which during charging the
heat-transfer fluid flow is split between the storage sub­
system and the power block (Figure 6)
• a serial configuration, in which the heat-transfer fluid
goes through the storage subsystem first and then flows
through the power block (Figure 7).
The choice of a parallel or serial configuration for the
integration of the storage and power block subsystems dic-
tates the type of reactor best suited for TCES. For example,
a fluidized bed has a relatively uniform temperature profile;
the outlet gases exit hot and then could flow into a power
block that would harness the heat (i.e., a serial configura-
tion). In a parallel scheme, fluids would flow into either a
reactor (perhaps a packed-bed reactor) or the power block,
with the goal of exiting at the lowest temperature possible
to extract the maximum amount of energy available. The
(a) Charging
Heat Thermochemical
Exchanger Storage Reactor
Solar Power
Field Block
(b) Discharging
Heat Thermochemical
Exchanger Storage Reactor
Solar Power
Field Block
p Figure 6. In the parallel configuration, energy from the solar field is
absorbed by a heat-transfer fluid during the charging step (a), which drives
a reaction in the thermochemical storage reactor and provides thermal
input to the power block. During discharging (b), the reverse reaction
heats a heat-transfer fluid, which supplies energy to the power block.
Source: Adapted from (3).
34  www.aiche.org/cep  July 2017  CEP
exit temperatures of the reactor and the power block would
ideally be approximately equal and as close to ambient
as possible.
Simulations of these manganese oxide storage systems
indicate that a significant portion of the usable stored energy
is actually the sensible heat of the materials — up to 91% in
packed-bed parallel configurations (3). This finding could
inform future designs of hybrid TCES systems that use
low-cost inert materials (e.g., ceramics) for sensible energy
storage and optimized, high-performing chemically reactive
materials to enhance performance.
The reactor must also be integrated with a solar energy
source, using either a direct or indirect subsystem. The DLR
(the German acronym for the Institute of Solar Research)
laboratory in Jülich, Germany, employs an indirect sub­
system: silcon carbide receivers heat air, which then
exchanges heat downstream with the reaction.
The DLR laboratory has also investigated direct heating
methods using rotary kilns, which receive concentrated
sunlight and carry out the decomposition reaction. Neises,
et al., demonstrated 30 cycles using Co3O4/CoO in a rotary
kiln, with roughly 400 kJ/kg stored per cycle (6). More
recently, Singh, et al., demonstrated the use of a honey-
comb reactor concept in a solar furnace for the Co3O4/CoO
cycle (15).
(a) Charging Heat
Exchanger
Solar Power
Field Block
Thermochemical
Storage Reactor
(b) Discharging Heat
Exchanger
Solar Power
Field Block
Thermochemical
Storage Reactor
p Figure 7. In the serial configuration, energy from the solar field is
absorbed by a heat-transfer fluid during the charging step (a), which first
drives a reaction in the thermochemical storage reactor and then provides
thermal input to the power block. During discharging (b), the reverse reac-
tion heats a heat-transfer fluid, which supplies energy to the power block.
Source: Adapted from (3).
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
Looking to the future of solar TCES
Metal oxide redox cycles that use air as the source of
oxygen are currently among the most attractive chemis-
tries for solar TCES, primarily because they do not require
storage of the gaseous reactant. Work is needed to develop
new materials for energy storage, as well as to demonstrate
a complete TCES subsystem. Pilot-scale trials will need to
demonstrate that heat can be stored indefinitely and effi-
ciently released on-demand.
Questions linger related to the integration of the stor-
age subsystem with the power block, process control, and
the interplay between sensible heat and chemical energy
release in functional designs. High-temperature power
blocks are more efficient, but higher temperatures expo-
nentially increase radiative losses and limit the choices
for materials of construction. Challenges associated with
high-­temperature materials and the desire to achieve higher
conversion efficiencies are driving researchers toward an
800–1,000°C operating temperature range. These tempera-
tures could supply heat to sCO2-based power blocks — a
substantial improvement over the current state-of-the-art
molten salt storage systems, which operate at lower tempera-
tures (around 600°C).
TCES could find application beyond CSP and power
generation, such as for storing and supplying heat. The
Literature Cited
1. Ahn, Y., et al., “Review of Supercritical CO2 Power Cycle Technol-
ogy and Current Status of Research and Development,” Nuclear
Engineering and Technology, 47 (6), pp. 647–661 (2015).
2. U.S. Dept. of Energy, “SunShot Initiative — Concentrating
Solar Power: Efficiently Leveraging Equilibrium Mechanisms for
Engineering New Thermochemical Storage (CSP: ELEMENTS),”
Funding Opportunity Announcement (FOA) #DE-FOA-0000805,
DOE, Washington, DC (Apr. 23, 2013).
3. Strohle, S., et al., “The Effect of the Gas-Solid Contacting Pattern
in a High-Temperature Thermochemical Energy Storage on the
Performance of a Concentrated Solar Power Plant,” Energy & Envi-
ronmental Science, 9 (4), pp. 1375–1389 (2016).
4. Carillo A. J., et al., “Thermochemical Energy Storage at High
Temperature via Redox Cycles of Mn and Co Oxides: Pure Oxides
versus Mixed Ones,” Solar Energy Materials and Solar Cells, 123,
pp. 47–57 (Apr. 2014).
5. Carillo A. J., et al., “Thermochemical Heat Storage Based on the
Mn2O3/Mn3O4 Redox Couple: Influence of the Initial Particle Size
on the Morphological Evolution and Cyclability,” Journal of Mate-
rials Chemistry A, 2 (45), pp. 19435–19443 (2014).
6. Neises, M., et al., “Solar-Heated Rotary Kiln for Thermochemical
Energy Storage,” Solar Energy, 86 (10), pp. 3040–3048 (2012).
7. Babiniec, S. M., et al., “Investigation of LaxSr1-xCoyM1-yO3-δ
(M = Mn, Fe) Perovskite Materials as Thermochemical Energy
Storage Media,” Solar Energy, 118, pp. 451–459 (Aug. 2015).
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
thermal energy for TCES could be obtained from waste
heat or geothermal sources and used for residential heat-
ing applications and/or industrial drying, separations, or
endothermic reactions. Industrial processes that produce
substantial amounts of reactive gases such as carbon diox-
ide might find that carbonation of a solar-charged Group 1
or Group 2 metal oxide in a cycle similar to those described
previously could provide heat and eliminate the need to
burn fossil fuels.
Although the solar thermochemical research com-
munity has largely focused on temperatures greater than
600°C for solar TCES, these high temperatures present
storage chemistries, service life, and materials of construc-
tion challenges. However, many thermal separations (e.g.,
thermal distillation) account for a significant portion of the
overall energy consumption of a chemical process. Existing
storage chemistries that work at low temperatures could be
integrated into such processes with appropriate design and
demonstration effort.
Process intensification and energy efficiency are major
thrusts in the continuing evolution of the chemical process
industries (CPI). As industry looks to cut operating costs and
minimize fossil energy consumption, chemical engineers
should be on the lookout for opportunities to incorporate
TCES into existing processes. CEP
8. Babiniec, S. M., et al., “Doped Calcium Manganites for Advanced
High-Temperature Thermochemical Energy Storage,” International
Journal of Energy Research, 40 (2), pp. 280–284 (Dec. 2015).
9. Rhodes, N. R., et al., “Solar Thermochemical Energy Storage
Through Carbonation Cycles of SrCO3/SrO Supported on SrZrO3,”
Chemsuschem, 8 (22), pp. 3793–3798 (Oct. 2015).
10. Zhao, M., et al., “Novel Calcium Looping Absorbent Incorporated
with Polymorphic Spacers for Hydrogen Production and CO2
Capture,” Energy & Environmental Science, 7 (10) (Aug. 2014).
11. Meier, A., et al., “Multitube Rotary Kiln for the Industrial Solar
Production of Lime,” Journal of Solar Energy Engineering-­
Transactions of the ASME, 127 (3), pp. 386–395 (Apr. 2005).
12. Harries, D. N., et al., “Concentrating Solar Thermal Heat Storage
Using Metal Hydrides,” Proceedings of the IEEE, 100 (2),
pp. 539–549 (2012).
13. Wong, B., et al., “Sulfur Dioxide Disproportionation for Sulfur
Based Thermochemical Energy Storage,” Solar Energy, 118,
pp. 134–144 (Aug. 2015).
14. National Institute of Standards and Technology, “NIST Chemis-
try WebBook,” U.S. Dept. of Commerce, Washington, DC,
http://webbook.nist.gov/chemistry (2016).
15. Singh, A., et al., “Solar Thermochemical Heat Storage via the
Co3O4/CoO Looping Cycle: Storage Reactor Modelling and Experi-
mental Validation,” Solar Energy, 144, pp. 453–465 (2017).
CEP  July 2017  www.aiche.org/cep  35
SPECIAL SECTION: SOLAR ENERGY

Leveraging
the Ammonia Industry
for Solar Energy Storage
Keith M. Lovegrove Concentrating solar power (CSP) with energy storage
ITP Thermal Pty Ltd.
Adrienne S. Lavine could help meet our zero-emission energy needs.
Univ. of California, Los Angeles This article explains how the Haber-Bosch process
Hamarz Aryafar
California State Polytechnic
for ammonia synthesis could be used for
Univ., Pomona low-cost energy storage in CSP plants.
Chen Chen
Univ. of California, Los Angeles

T
he Paris Climate Agreement aims to limit the human-
induced global temperature rise to less than 2°C.
Meeting that goal will require a global energy system
with net zero emissions by 2050 or sooner. Renewable
energy sources, such as wind and photovoltaics (PV), will
help make a zero-emission energy system a reality. How-
ever, the variable nature of these energy sources limits their
practical use for electricity generation.
Integrating energy storage with energy production is the
key to a zero-emission energy system future. Energy storage
can be built into a concentrating solar power (CSP) system,
without increasing the cost of the delivered energy. The stan-
dard industry approach of producing electricity with energy
storage is to heat a molten mixture of potassium and sodium
nitrate salts to high temperatures using the solar concentra-
tors. The thermal energy can be stored or used on-demand in
heat exchangers to produce steam to power steam turbines.
The CSP industry is expanding, but is still a decade behind
PV in deployed capacity.
This article introduces the concept of leveraging the
Haber-Bosch ammonia synthesis process as a potential
lower-cost energy storage alternative to molten salts in
CSP plants.
Ammonia as an energy storage medium
Solar thermochemical energy storage (TCES) exploits a
chemically reversible reaction by using solar energy to heat
an endothermic reactor. The reaction products are stored,
and when energy is needed, it can be recovered as heat given
off by an exothermic reactor that reconstitutes the origi-
nal reactants. In addition to ammonia, reactions involving
hydroxides, carbonates, hydrides, and sulfates, as well as
methane reforming, have been investigated as TCES media
(see pp. 30–35) (1–3).
Annual global production of ammonia is more than
160 million m.t., making ammonia the second-most-­
common chemical commodity produced (after ethylene).
The Haber-Bosch process for ammonia synthesis has been
used industrially for over 100 years, and at scale in most
industrialized countries. The U.S., for example, has more
than 50 ammonia plants (4).
An industrial ammonia plant, as a whole, is a major
energy-consuming system. Typically, plants use natural gas
as a feedstock and as fuel to drive the operation. Much of the
plant is dedicated to the energy-intensive process of steam
reforming of natural gas to produce hydrogen and nitrogen
gas mixtures. (For more details on the ammonia production

36  www.aiche.org/cep  July 2017  CEP Copyright © 2017 American Institute of Chemical Engineers (AIChE)
process, read the Sept. 2016 CEP Back to Basics article
“Introduction to Ammonia Production,” pp. 69–75).
The Haber-Bosch ammonia synthesis reactors are exo-
thermic. The reaction heat is usually captured as superheated
steam that is used for cogeneration of electricity or other
plant processes. This approach could, instead, be employed
to operate a standalone CSP plant. The feedstock of hydro-
gen and nitrogen could be produced by the decomposition
of ammonia, driven by solar energy from a concentrating
solar thermal (CST) system, with the reactants continuously
recycled.
Ammonia has been shown to be a viable solar TCES
medium over three decades of research at Australian
National Univ. (ANU). Researchers at ANU demonstrated
24-hr operation of a complete system, with a dish concentra-
tor, endothermic and exothermic reactors, and a storage ves-
sel (5). The Univ. of California, Los Angeles (UCLA), has
continued this work (6, 7). UCLA researchers are investigat-
ing cost-effective gas containment technologies, the use of
ammonia synthesis to enable production of 650°C steam for
a supercritical steam Rankine cycle, and integration of the
endothermic reactor within a CSP tower receiver (8).
How it works
In an ammonia TCES process (Figure 1), ammonia is
dissociated endothermically (NH3 + 66.8 kJ/mol ↔ 1/2 N2 +
3/2 H2) as it absorbs solar energy during the daytime. When
energy is required, the reverse reaction releases energy to
heat a working fluid, such as steam, to produce electricity.
A key feature of the process is the recuperative heat
exchangers paired with
the endothermic and exo­ Ammonia Dissociation
Reactor
thermic reactors, which lower (Endothermic Reactor/Receiver)
the endothermic reaction
products’ temperature to
near ambient and then raise
their temperature prior to
the synthesis reaction. As a
result, transport and storage
of the reactants and prod-
ucts is conducted at ambient
temperature.
In addition to eliminating
Solar Concentrators
piping networks and their (Dishes or Heliostats)
associated thermal losses and
the ability to store energy
u Figure 1. In an ammonia thermochemical storage system, high-
temperature solar concentrators decompose ammonia. The products
of the decomposition, hydrogen and nitrogen, are stored underground
at ambient temperature. When energy is needed, the gases are fed to
the ammonia synthesis reactors, which give off heat that is used to
create high-pressure steam for power generation.
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
indefinitely, ammonia TCES has several other advantages:
• The simple reaction has no side reactions, so there is
no need for elaborate control systems or nonstoichiometric
mixtures to maintain complete reaction reversibility.
• The constituents are inexpensive and remain stable at
the proposed operating temperatures.
• The large differences between the densities of ammonia
(liquid at the operating pressure) and nitrogen and hydro-
gen gases enable their storage in the same tank with natural
phase separation.
• Because of the natural phase separation in the storage
tank, the availability of the feedstocks for the two reactors is
independent of the degree of completion in the other reactor.
• The vast industrial experience in ammonia synthesis
can be tapped.
Conventional ammonia synthesis reactions are run at
high pressures (up to 300 bar). Because high pressure favors
the synthesis reaction, the reaction can be conducted at
higher temperatures, which increases reaction rates. To adapt
the conventional method to a closed-loop ammonia TCES
system, high pressure is applied to both the endothermic
reactors and the reactant storage volume to avoid a large
loss of work during gas compression. Designing the entire
system for high-pressure operation is a major engineering
challenge of developing the technology.
Integrating reactors and solar receivers
Under pressure, the endothermic ammonia dissociation
step requires a temperature of around 700°C to achieve near-
complete dissociation at reasonable reaction rates. A central
Power Block
(Steam)
Ammonia Synthesis
Reactor
(Exothermic Reactor)
Heat Heat
Exchanger Exchanger
N2/H2
Gas
Liquid
NH3
• Storage of Reactants
at Ambient Temperature
• High-Pressure Underground
Storage of Gases
CEP  July 2017  www.aiche.org/cep  37
SPECIAL SECTION: SOLAR ENERGY
receiver with heliostats or parabolic dish CST systems are
best suited to achieving this temperature. Trough or linear
Fresnel concentrators may also be considered, although
temperatures may reach only around 500°C and the level of
dissociation will be lower. To increase dissociation, however,
the phase separation of ammonia from the reactant mix in
storage could enable recirculation of unreacted ammonia.
Tower receivers with heliostat concentrators are cur-
rently favored by the CSP industry. Parabolic dish concen-
trators offer the highest optical efficiencies, but they are the
least commercially mature because they have so far lacked
a proven thermal energy storage solution. The ammonia
TCES system, with its ambient-temperature energy transport
feature, could enable CSP plants composed of multiple dish
concentrators.
Researchers at ANU designed a receiver reactor for
ammonia dissociation on a 20-m2 parabolic dish concen-
Dish Concentrator
Receiver/Reactor
pu Figure 2. This
dish concentrator
prototype includes
a 15-kWsol solar
dissociation reactor
that is integrated
into the receiver.
38  www.aiche.org/cep  July 2017  CEP
trator (Figure 2). The 15-kWsol receiver/reactor consists
of 20 catalyst-filled Inconel 601 reactor tubes mounted to
manifolds that provide parallel flow of the reactants (9).
The tubes form a cone inside an insulated cavity in the
receiver structure.
The system successfully handled startup, shutdown,
and operation under intermittent clouds. It operated at up to
30 MPa and 700°C; the temperature was controlled by vary-
ing the flowrate of the ammonia feed. Flow was distributed
among the tubes by tuning their pressure drop characteristics
during manufacturing.
An ammonia dissociation reactor in a tower receiver
Because much of the growth in CSP deployment has
involved tower receivers, our research focuses on the techni-
cal feasibility of operating a 100-MWe tower system with
six or more hours of storage.
Ammonia dissociation within a tower receiver has yet
to be demonstrated. Our research considers the conceptual
design of such a receiver. We examined the molten salt
tower receiver in the well-known U.S. National Renew-
able Energy Laboratory’s (NREL) System Advisor Model
(SAM) as a starting point (10). This receiver is designed for
a 100-MWe, 10-hr storage plant with a surrounding heliostat
field (similar to the Crescent Dunes plant in Nevada).
Our goal was to develop a conceptual design for an
ammonia dissociation receiver that fits within the same size
cylindrical envelope as the SAM receiver (17.7 m wide,
20.4 m high) and that has the same thermal capacity under
the design conditions (670 MWth).
The conceptual design uses multiple tubular packed-bed
catalytic dissociation reactors, similar to the parabolic dish
receiver designed at ANU, that are 2.7 m long with 5-cm-
wide beds (Figure 3). We assumed the same iron-cobalt
catalyst used successfully in the ANU dish. We also assumed
q Figure 3. The central receiver (left) of this conceptual design would
include a dissociation receiver (center) composed of multiple tubular
packed-bed catalytic dissociation reactors arranged in skirts (right).
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
that these endothermic reactors will operate at 30 MPa, as
required by the exothermic heat-recovery process.
In addition to containing the pressure, the dissociation
reactor/receiver must be able to convert sufficient quantities
of ammonia to meet the thermal power requirement at a rea-
sonable operating temperature. We established that around
33,000 such tubes would be needed to absorb the 670 MWth
design input, and that the reactant temperature would need to
start at 450°C and increase to 700°C at the exit.
The tubes would be mounted at a downward angle
(7 deg. from horizontal) in a series of tight skirts (Figure 3).
The angle is intended to match the incident solar flux to the
requirements of the reaction. An angle equal to the rim angle
from the receiver to the outermost heliostat would allow
radiation from those heliostats to penetrate to the center of
the receiver, while radiation from closer heliostats would
land along the length of the tubes.
To enhance the receiver’s thermal efficiency, the tubes
are arranged so that the lower-temperature inlet condition
is at the outermost point and the higher temperatures are
toward the inside. The gaps between adjacent skirts serve as
cavities that shield the highest-temperature portions of the
reactor tubes from the surroundings, reducing radiation and
convection heat losses. The downward-angled tubes form a
stagnant, stably stratified air space inside the cavities, which
also reduces convection losses.
We predict this design will achieve a receiver thermal
efficiency of 94%.
Gas storage
For deployment, our receiver design would require the
support of a gas storage system. A TCES system capable
of driving a 100-MWe generator for six hours would need
a 1,300-MWth storage system. The nitrogen and hydrogen
gas mixture has an energy density of about 400 MJ/m3 at
the proposed storage conditions of 298 K and 30 MPa and
requires roughly 12,000 m3 of storage volume. To be conser-
vative, we assumed 20 MPa pressure and 33% cushion gas
(i.e., the amount of gas that must be left in situ to preserve
the integrity of the storage space), which increases the stor-
age volume to about 24,000 m3.
The high pressure required to contain the gas could
necessitate a costly conventional steel pressure-containment
system. Therefore, we explored strategies to cost-effectively
contain the gas. We considered a variety of underground stor-
age solutions that exploit the surrounding geology to provide
the bulk of the pressure containment — including storage in
depleted oil or gas wells, aquifers, salt caverns, rock caverns,
and buried pre-stressed reinforced concrete pressure ves-
sels — as well as several technologies for boring and lining
underground holes. Because the liquid ammonia has a much
smaller volume than the gaseous phase (N2 + 3H2), it could
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
be stored in a connected ground-level containment vessel.
The pressure of the storage system will cycle between
maximum and minimum values as the system is charged by
ammonia decomposition and discharged by ammonia syn-
thesis. If the synthesis reaction is operated at a fixed design
pressure (or between narrow limits), feed gas will need to
be compressed much of the time. The gas storage would be
designed for an economically optimal storage pressure that
is independent of the reaction system.
The storage options were evaluated based on attributes
such as estimated cost, location flexibility, potential for
contamination of the surrounding environment, poten-
tial for contamination of the gas store, and ease of gas
injection/extraction. Based on the initial evaluation, we
identified mined salt caverns and large-diameter shafts as the
most promising storage options.
Gas storage: Salt caverns
Salt caverns can be mined in either salt domes or salt
beds (Figure 4). Mining a salt cavern is simple and employs
established processes, such as solution mining or leaching,
whereby fresh water is pumped into the salt dome or bed and
brine is extracted. Over 2,000 salt caverns are used for hydro-
carbon and gas storage in North America, including about
500 in Texas alone. Caverns are typically located at depths of
200–2,000 m and can be as large as 500,000 m3, and they are
capable of tolerating pressures up to about 50 MPa.
The rock salt of the cavern walls is chemically inert to
hydrogen, and it is one of the only rock formations with
permeability low enough to contain hydrogen gas. Pure
hydrogen or hydrogen-rich gas mixtures have been stored in
salt caverns at several sites around the world. Praxair oper-
Cavern Salt Deposit
p Figure 4. Salt caverns are a viable and economical option for gas
storage because the surrounding rock provides adequate pressure
containment, the salt deposits are inert to hydrogen, and the permeability
of the salt is low enough to contain hydrogen.
CEP  July 2017  www.aiche.org/cep  39
SPECIAL SECTION: SOLAR ENERGY

5.5 m Pressure Pipe bolts and then backfilled with earth and rock to complete the
pressure containment.
Backfill Shaft-drilling companies estimate a volume cost of
with Rock
around $400/m3 for the storage system, which translates to
12-mm an energy storage cost of $4.80/kWhth if no cushion gas is
Reinforced
Earth Steel Case
Concrete Cap
required, or $7.20/kWhth if 33% cushion gas is assumed.
1,000 m
Rock Bolts Driving power cycles
to Reinforce Cap
The overall ammonia production process is energy inten-
Steel Plate sive. However, ammonia synthesis reactors actually produce
Cement
energy as heat.
Grout Conventional ammonia synthesis reactors generally con-
sist of adiabatic reactor beds connected in series in one or
more pressure vessels. Heat is extracted from the reactants
p Figure 5. The surrounding rock provides the majority of the pressure between beds to drop the reactants’ temperature to below the
containment for gas storage in an underground shaft. Shafts are not subject chemical equilibrium state. Bayonet-type heat exchangers
to the location constraint of salt caverns.
are used to produce steam at around 450°C. The steam can
ates a 470,000-m3 (71 million-Nm3) salt cavern hydrogen be used in standard steam turbine power blocks that achieve
storage facility at 15 MPa in Texas (11). reasonable conversion efficiencies.
A 2008 U.S. Dept. of Energy (DOE) report on the The overall economic performance of a CSP system is
economic feasibility of syngas storage (12) estimated a salt strongly linked to the conversion efficiency of its power
cavern facility would cost around $50/m3. For a 24,000-m3 cycle, which can be improved by higher input temperatures.
facility, this translates to a very low thermal storage cost State-of-the-art steam turbine power cycles employ steam at
of about $1/kWh of storage capacity. Because size is not pressures above 22.1 MPa (i.e., supercritical steam). Super-
a limiting factor, a salt cavern could be suitable for longer critical steam temperatures are typically higher than 600°C.
periods of storage at plants that are larger than our concep- A primary goal of our research was to experimentally
tual design. demonstrate heating of supercritical steam by ammonia syn-
Suitable salt deposits are present on every continent, and thesis. Our aim was to heat the steam at 26 MPa from 350°C
there is sufficient overlap between areas with salt deposits to 650°C. We achieved this using a tubular packed-bed reac-
and areas of high solar radiation. The availability of salt tor with counterflow heat-transfer tubes for the steam. The
deposits is sufficient to allow ammonia CSP systems to reactants experienced a rapid temperature increase of around
provide a large fraction of electricity in countries with high 50°C near the inlet, and the steam exited at a temperature
incident solar radiation. Nevertheless, siting
CSP plants in areas with suitable salt deposits is Table 1. The cost of an ammonia-based concentrating solar power (CSP)
a significant constraint. plant with energy storage is in range of the DOE’s targets for 2020.
Fixed Costs $/kWth Cost for 6 hr, $/kWhth
Gas storage: Underground shafts
Reactor 22.11 3.69
The use of large-diameter shafts for gas
storage removes the site choice constraint Recuperating Heat Exchanger 17.07 2.85
(Figure 5). Drilling is an established technology, Compressor 27.86 4.64
although it has not yet been used for gas stor- Chilled Separator and 19.50 3.25
age. The mining industry drills shafts that are other Balance of Plant (BOP)
7.5 m wide and reach depths of 1,000 m. Offset for Boiler* –26.38 –4.40
For a gas storage application, the shaft Variable Costs Salt Cavern Shaft Drilling
would be lined with a relatively thin (about
Underground Storage Volume — 1.50 7.20
12 mm) steel lining and grouted with cement,
Ammonia Inventory — 1.50 1.00
and top and bottom plates would be welded in
place. The top plate would be located a suf- Total — 13.0 18.2
ficient distance below ground level to resist the * The ammonia TCES system heats steam in the synthesis reactor directly, so a separate
steam generating heat exchanger is not required, as it is with a molten-salt-based storage,
internal pressure and would be coupled to the for example. Thus, the equivalent cost of such a heat exchanger can be counted as a credit
surface by a pressure pipe. The top would be to the overall capital cost of ammonia-based TCES. The boiler heat exchanger cost has
been estimated.
reinforced with a suitable concrete cap and rock

40  www.aiche.org/cep  July 2017  CEP Copyright © 2017 American Institute of Chemical Engineers (AIChE)
 within about 15°C of the peak reactant temperature.
This approach and temperature characteristic could also
be applied to other advanced power cycles under develop-
ment, such as those based on supercritical CO2.
Economics
The goal of the DOE SunShot program, which funds our
research, is to reduce the total levelized cost of solar energy
to $0.06/kWh of electrical energy produced by 2020 (6).
Within the SunShot effort, the target for thermal energy stor-
age is an installed cost of $15/kWhth of capacity.
Table 1 presents the estimated cost of a 1,300-MWth
storage system. The overall cost of salt cavern storage and
shaft drilling would be $13/kWhth and $18/kWhth, respec-
tively. For 6-hr storage, the ammonia TCES system cost is
within the SunShot target range.
Figure 6 compares the specific cost of installed storage
capacity as a function of the storage capacity in hours. The
specific cost includes the variable cost of storage and the
fixed cost amortized over the number of storage hours.
At 10–15 hr of storage, the effective cost of installed
storage capacity drops well below the SunShot target for
both salt cavern and shaft storage. The marginal cost of
storage in a salt cavern is sufficiently low, which suggests
that the technology could be suited to meeting peak demands
through several days of continuous clouds.
Closing thoughts
Solar energy systems with built-in energy storage will
be key to achieving the decarbonization of the energy sector
needed to keep global temperatures from rising more than
2°C. CSP integrated with thermal storage could be a good
solution to this challenge. Ammonia TCES is especially
attractive because of the vast established industry knowl-
50
45
40
Specific Cost, $/kWhth
35
30
25
20 Drilled Shaft
15 (Average)
10
Salt Cavern
5 (Average)
0 11. Watme, R., “Large Scale Hydrogen Storage in Salt Caverns,”
0 5 10 15 20
Length of Time of Storage, hr
p Figure 6. The cost of an ammonia thermochemical energy storage
(TCES) system is a function of the storage capacity in hours. The broad line
widths represent the uncertainty in the estimates.
Copyright © 2017 American Institute of Chemical Engineers (AIChE)
edge and capabilities. Salt caverns and drilled shafts could
meet the needs of high-pressure storage at an economically
feasible cost.
Recent work has demonstrated that ammonia synthesis
can heat supercritical steam from 350°C to 650°C, improv-
ing the potential efficiency of conversion. And, it has also
been shown that endothermic dissociation reactors can be
integrated into dish or tower receivers with high thermal
efficiency. Together, these advances have moved ammonia-
based TCES closer to becoming a reality for CSP. CEP
Acknowledgments
The information, data, and work presented in this article was funded by
the Office of Energy Efficiency and Renewable Energy, U.S. Dept. of Energy
(DOE), Award No. DE-EE0006536. The authors gratefully acknowledge the
support.
Literature Cited
1. Pardo, P., et al., “A Review on High Temperature Thermo­
chemical Heat Energy Storage,” Renewable and Sustainable
Energy Reviews, 32, pp. 591–610 (Apr. 2014).
2. Romero-Paredes, H., et al., “Thermochemical Storage of Solar
Energy by Means of Sulfates: A Review,” in “Proceedings of the
13th SolarPACES Conference,” Seville, Spain (2006).
3. Levy, M., et al., “Solar Energy Storage via a Closed-Loop
Chemical Heat Pipe,” Solar Energy, 50 (2), pp. 179–189 (1993).
4. Brown, T., “The Ammonia Monitor,” https://ammoniaindustry.
com/ammonia-monitor (accessed Apr. 18, 2017).
5. Dunn, R., et al., “A Review of Ammonia-Based Thermochemical
Energy Storage for Concentrating Solar Power” Proceedings of
the IEEE, 100 (2), pp. 391–400 (Feb. 2012).
6. Office of Energy Efficiency and Renewable Energy, “SunShot
Initiative Goals,” U.S. Dept. of Energy, Washington, DC,
http://energy.gov/eere/sunshot/mission (accessed Apr. 20, 2017).
7. Office of Energy Efficiency and Renewable Energy, “Con-
centrating Solar Power: Efficiently Leveraging Equilibrium
Mechanisms for Engineering New Thermochemical Storage,”
U.S. Dept. of Energy, Washington, DC, http://energy.gov/eere/
sunshot/concentrating-solar-power-efficiently-leveraging-equilib-
rium-mechanisms-engineering-new (accessed Apr. 20, 2017).
8. Lavine, A., et al., “Thermochemical Energy Storage with Ammo-
nia: Aiming for the SunShot Cost Target,” in “Proceedings of
Molten Salt SolarPACES 2015 Conference,” Cape Town, South Africa
(Two Tanks) (Oct. 13–16, 2015).
9. Lovegrove, K., et al., “A Solar-Driven Ammonia-Based
Thermoc­hemical Energy Storage System,” Solar Energy, 67 (4),
pp. 309–316 (1999).
10. National Renewable Energy Laboratory, “System Advisor
Model (SAM),” U.S. Dept. of Energy, Washington, DC,
https://sam.nrel.gov (accessed Apr. 20, 2017).
25 30
Renewable Resources for Fuel Cells — Workshop, http://cafcp.
org/sites/files/3_Watwe_LargeScaleHydrogenStorageCaverns.pdf
(Oct. 2010).
12. Apt, J., et al., “An Engineering-Economic Analysis of Syngas
Storage,” DOE/NETL-2008/1331 (July 31, 2008).
CEP  July 2017  www.aiche.org/cep  41
SPECIAL SECTION: SOLAR ENERGY
MEET THE AUTHORS
HAMARZ ARYAFAR NICK AUYEUNG CHEN CHEN
PETER KREIDER
HAMARZ ARYAFAR, PhD, is the Chief Technical Officer of Element 16
Technologies, Inc., and an assistant professor in the Electromechanical
Engineering Technology Dept. at California State Polytechnic Univ., Pomona
(Email: hamarz@ucla.edu). He received his PhD in aerospace engineering
from the Univ. of California, Los Angeles (UCLA). He previously served as lead
engineer for Marvel Energy, Inc. — an LA-based energy storage start-up. As a
postdoctoral researcher at UCLA, he investigated ammonia thermochemical
energy storage technologies. He has secured over $3.1 million in grants and
government funding related to his research, and has three pending patent
applications for energy storage innovations.
NICK AUYEUNG, PhD, is an assistant professor at Oregon State Univ.’s
School of Chemical, Biological, and Environmental Engineering (Email:
nick.auyeung@oregonstate.edu). Prior to joining the faculty at Oregon State,
he worked as a postdoctoral associate focusing on solar thermochemical
conversion processes for fuels and energy storage in the Dept. of Mechani-
cal and Aerospace Engineering at the Univ. of Florida. He earned his BS from
the Univ. of Connecticut and PhD from Oregon State Univ., both in chemical
engineering. He completed his PhD dissertation on thermochemical water
splitting. For the past three years, AuYeung has been an active leader of ses-
sions related to solar energy for power generation and chemical processing
at AIChE Annual Meetings.
CHEN CHEN is a doctoral student in the Mechanical and Aerospace Engineer-
ing Dept. at the Univ. of California, Los Angeles (UCLA) (Box 951597, Univ. of
California, Los Angeles, Los Angeles, CA 90095). His emphasis is on heat and
mass transfer. He received his master’s degree in mechanical engineering
from Columbia Univ. He has published two research papers on ammonia solar
thermochemical energy storage systems.
JEFF CUMPSTON, PhD, is a postdoctoral researcher at Rheinisch-­Westfälische
Technische Hochschule (RWTH) Aachen Univ. His research focuses include
optimization of CSP power plant layouts and optimization of energy storage
scheduling to maximize revenue in variable electricity markets. He has
conducted atmospheric monitoring at Australia’s Davis Station in Antarctica.
Cumpston received a BS in Physics, a Bachelor of Photonics, and a PhD
in solar engineering from Australian National Univ. He was awarded the
Australian Solar Energy Society’s Wal Read Memorial Prize and the Australian
Institute of Energy’s Post Graduate Research Award.
42  www.aiche.org/cep  July 2017  CEP
JEFF CUMPSTON
ADRIENNE LAVINE KEITH LOVEGROVE ALEXANDER MITSOS
PETER KREIDER, PhD, is a research fellow in the solar thermal group at the
Australian National Univ. and is currently managing a solar thermochemical
energy storage project for the Australian Renewable Energy Agency (ARENA)
(Email: peter.kreider@anu.edu.au). His research interests focus on reaction
engineering for solar fuels and functional materials. He received his BS from
the Univ. of Colorado Boulder and PhD from Oregon State Univ., both in
chemical engineering. He is an AIChE senior member and is active in sessions
related to solar energy for power generation and chemical processing at
AIChE Annual Meetings.
ADRIENNE LAVINE, PhD, is a professor and past department chair of the
Mechanical and Aerospace Engineering Dept. at the Univ. of California, Los
Angeles (UCLA) (Email: lavine@seas.ucla.edu). Her research addresses vari-
ous aspects of heat transfer, including solar thermal energy storage, thermal
energy harvesting, and thermal aspects of nanoscale and macroscale manu-
facturing processes. She earned a PhD in mechanical engineering from the
Univ. of California, Berkeley. She received the Presidential Young Investigator
Award from the National Science Foundation (NSF) and the F.W. Taylor Medal
of CIRP from the International Academy of Production Engineering. She is
a Fellow of the American Society of Mechanical Engineers (ASME) and has
served as an associate editor of the Journal of Heat Transfer and the Journal
of Engineering for Industry.
KEITH LOVEGROVE, PhD, is the head of solar thermal for the ITP Energised
group and managing director of ITP Thermal Pty Ltd. (Email: keith.lovegrove@
itpau.com.au). He is also a board member of the Australian Solar Thermal
Energy Association. He received a PhD from the Australian National Univ.
(ANU) in philosophy, and was previously the leader of the solar thermal
group at ANU. He has 29 years of experience in the field of solar thermal
energy and 15 years of experience teaching undergraduate and postgraduate
courses in energy systems and systems engineering.
ALEXANDER MITSOS, PhD, is a full professor (W3) at Rheinisch-Westfälische
Technische Hochschule (RWTH) Aachen Univ. and Director of the Laboratory
for Process Systems Engineering (AVT.SVT) (Email: amitsos@alum.mit.edu).
His research focuses on optimization of energy and chemical systems and the
development of numerical algorithms to enable optimization. Mitsos received
a Diplom-Ingenieur from the Univ. of Karlsruhe and PhD from Massachusetts
Institute of Technology (MIT), both in chemical engineering. He has over
70 publications in peer-reviewed journals and has received several awards
including the 2016 AIChE Computing and Systems Technology (CAST) Div.
Outstanding Young Researcher award.
Copyright © 2017 American Institute of Chemical Engineers (AIChE)