Progress in Energy and Combustion Science 53 (2016) 140

Contents lists available at ScienceDirect

Progress in Energy and Combustion Science

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p e c s

Thermal energy storage: Recent developments and practical aspects

Huili Zhang a, Jan Baeyens b, Gustavo Cceres c, Jan Degrve a, Yongqin Lv d,*
a Bio- & Chemical Reactor Engineering and Safety Section, Department of Chemical Engineering, KU Leuven, 3001 Leuven, Belgium
b
European Powder and Process Technology, Park Tremeland 9, 3120 Tremelo, Belgium
c
Facultad de Ingeniera y Ciencias, Universidad Adolfo Ibez, Diagonal Las Torres, 2640, 7941169 Peanoln, Santiago, Chile
d Beijing Key Lab of Bioprocess, College of Life Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China

A R T I C L E I N F O A B S T R A C T

Article history: Thermal energy storage (TES) transfers heat to storage media during the charging period, and releases
Received 31 July 2015 it at a later stage during the discharging step. It can be usefully applied in solar plants, or in industrial
Accepted 24 October 2015 processes, such as metallurgical transformations. Sensible, latent and thermo-chemical media store heat
Available online 15 December 2015
in materials which change temperature, phase or chemical composition, respectively. Sensible heat storage
is well-documented. Latent heat storage, using phase change materials (PCMs), mainly using liquid
Keywords:
solid transition to store latent heat, allows a more compact, ecient and therefore economical system
Thermal energy storage
to operate. Thermo-chemical heat storage (TCS) is still at an early stage of laboratory and pilot research
Sensible
Latent despite its attractive application for long term energy storage.
Thermo-chemical The present review will assess previous research, while also adding novel treatments of the subject.
Encapsulation TES systems are of growing importance within the energy awareness: TES can reduce the LCOE (levelized
Material properties cost of electricity) of renewable energy processes, with the temperature of the storage medium being
Improvements the most important parameter. Sensible heat storage is well-documented in literature and applied at large
Future R&D scale, hence limited in the content of the present review paper. Latent heat storage using PCMs is dealt
with, specically towards high temperature applications, where inorganic substances offer a high po-
tential. Finally, the use of energy storage through reversible chemical reactions (thermo-chemical storage,
TCS) is assessed. Since PCM and TCS storage media need to be contained in a capsule (sphere, tube, sand-
wich plates) of appropriate materials, potential containment materials are examined. A heat transfer uid
(HTF) is required to convey the heat from capture, to storage and ultimate re-use. Particle suspensions
offer a valid alternative to common HTF, and a preliminary assessment conrms the advantages of the
upow bubbling uidized bed and demonstrates that particulate suspensions enable major savings in
investment and operating costs.
Novel treatments of the TES subject in the review involve the required encapsulation of the latent
and chemical storage media, the novel development of powder circulation loops as heat transfer media,
the conductivity enhancement of PCMs, the use of lithium salts, among others.
2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction ............................................................................................................................................................................................................................................................. 2
1.1. Global energy and the required CO2 reduction .............................................................................................................................................................................. 2
1.2. Maturity of different energy storage systems and cost effects ................................................................................................................................................. 3
1.3. Energy storage in general, and thermal energy storage specically ...................................................................................................................................... 5
1.4. Heat balances in thermal energy storage ......................................................................................................................................................................................... 5
1.5. Classications of thermal energy storage ........................................................................................................................................................................................ 6
1.5.1. Classication according to temperature range and associated re-use technology ........................................................................................... 6
1.5.2. Classication according to energy storage mechanism ............................................................................................................................................. 8
1.5.3. Classication according to the storage concept ............................................................................................................................................................ 8
1.6. Objectives and structure of the paper ............................................................................................................................................................................................... 8

* Corresponding author. Beijing Key Lab of Bioprocess, College of Life Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China. Tel.: +86
10 64454356; Fax: +86 10 6471 5443.
E-mail address: lvyq@mail.buct.edu.cn (Y.Q. Lv).

http://dx.doi.org/10.1016/j.pecs.2015.10.003
0360-1285/ 2015 Elsevier Ltd. All rights reserved.
2 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

2. Sensible heat storage .......................................................................................................................................................................................................................................... 10

3. Latent heat storage ............................................................................................................................................................................................................................................. 10
3.1. Main characteristics .............................................................................................................................................................................................................................. 10
3.2. PCMs classication ................................................................................................................................................................................................................................ 10
3.3. Moderate or high temperature PCMs ............................................................................................................................................................................................. 11
3.4. Specic PCM applications ................................................................................................................................................................................................................... 13
3.4.1. Latent heat storage with cryogenics .............................................................................................................................................................................. 13
3.4.2. Latent heat storage with steam accumulators ........................................................................................................................................................... 14
3.4.3. Moderate to high temperature PCMs ............................................................................................................................................................................ 16
3.4.4. The growing importance of lithium salts ..................................................................................................................................................................... 16
3.5. Mechanisms to improve phase change material applications ............................................................................................................................................... 18
3.5.1. General principles ................................................................................................................................................................................................................. 18
3.5.2. Review of applied enhancement methods .................................................................................................................................................................. 19
3.6. Nanoparticle-encapsulated PCMs .................................................................................................................................................................................................... 21
3.7. Cascades of PCM systems ................................................................................................................................................................................................................... 22
4. Thermo-chemical heat storage ....................................................................................................................................................................................................................... 23
4.1. Principles .................................................................................................................................................................................................................................................. 23
4.2. Thermodynamic assessment ............................................................................................................................................................................................................. 23
4.2.1. Gibbs free energy and work recovery ........................................................................................................................................................................... 23
4.2.2. Calculation results ................................................................................................................................................................................................................ 24
4.3. Kinetics from thermogravimetric analysis .................................................................................................................................................................................... 24
5. Containment of LHS and TCS materials ....................................................................................................................................................................................................... 26
5.1. Preliminary selection concerns and criteria ................................................................................................................................................................................. 26
5.1.1. Pressure upon phase change or chemical reaction .................................................................................................................................................... 26
5.1.2. Design of the encapsulation .............................................................................................................................................................................................. 27
5.1.3. Preliminary selection criteria ............................................................................................................................................................................................ 28
5.2. Relevant properties of E-materials .................................................................................................................................................................................................. 28
5.2.1. INCONEL alloy 600 [187] .................................................................................................................................................................................................... 28
5.2.2. Incoloy 825 [191] ................................................................................................................................................................................................................. 28
5.2.3. Extruded tubular ceramics [192196] .......................................................................................................................................................................... 29
5.3. Experimental design data ................................................................................................................................................................................................................... 29
6. PCM/TCS integration in TES requires heat carriers .................................................................................................................................................................................. 30
6.1. Introduction ............................................................................................................................................................................................................................................. 30
6.2. Powders as novel heat carriers in renewable energy systems ............................................................................................................................................... 30
6.2.1. Dense up-ow of particles ................................................................................................................................................................................................ 32
6.2.2. Particle loops in SPT and CSP applications .................................................................................................................................................................. 32
6.2.3. Recent experimental results ............................................................................................................................................................................................. 32
7. Conclusion and recommendation for priority research ......................................................................................................................................................................... 35
Acknowledgements ............................................................................................................................................................................................................................................. 36
References .............................................................................................................................................................................................................................................................. 36

1. Introduction Because of the use of fossil fuels, current global CO2 emissions
are at 30.6 109 tpa of CO2 (against 28.2 109 tpa in 2005). Without
1.1. Global energy and the required CO2 reduction effective measures, it is however expected that world energy-
related emissions of CO2 will further increase to 33.5 109 tpa in
Energy supply is a vital issue, with special concerns of the public 2015 and 43.2 109 tpa in 2035 [5].
regarding the emission of greenhouse gases and the need to reduce The imposed reduction in CO2 emissions will require a combi-
the use of fossil fuels [1]. The worldwide economic crisis since 2008 nation of detailed strategies and tactics, including (i) a mix of energy
added additional challenges [2], leading worldwide governments generation technologies; (ii) a reduction in energy usage through
to enact new policies and nancial incentives in support of renew- the use of incentives, technologies, taxes and quotas; (iii) maxi-
able energies, enhancing their implementation and development, mizing CO2 absorption, through carbon sequestration by both natural
while simultaneously creating valuable new business opportuni- means and by technical developments; and (iv) the development
ties for companies involved in this energy sector [3,4]. One of the of highly-ecient energy capture, storage and re-use methods [3,7].
hot topics in the energy strategy is the capture and storage of thermal There is indeed still a considerable scope for improving the energy
energy as applicable to renewable energy concepts and in waste heat eciency. In the short or medium term, waste heat recovery and
recovery: these advanced energy utilization schemes call for the de- high temperature thermal energy storage are crucial concepts to im-
velopment and new usage of existing and/or new materials. Based plement such solutions, and even current power plants can use high
upon the current statistics and predictions of energy consump- temperature thermal energy storage to improve the energy balance
tion, the U.S. Energy Information Administration predicated an of their operations, since they increase the exibility and availabil-
increase in the total world energy use from 0.15 1012 MWh in 2008 ity of heat and electricity in traditional or sustainable power plants.
to 0.18 1012 MWh in 2020, and to 0.23 1012 MWh in 2035 [5]. Since The re-use of low grade heat, typically between ambient tem-
crude oil fuels remain an important source of energy, albeit with perature and 200 C, is not widespread since it is a technical and
depleting traditional reserves but with signicant current fracking economic challenge to obtain useful exergy and energy from low
operations, their prices are expected to remain around 60 US$/ grade heat. A large amount of low grade heat is available in the
barrel for the near future, while 2040 US$/barrel were valid from process industry, e.g. water from cooling towers with exhaust gas
1985 till 2005, but prices of 120140 US$/barrel are predicted from temperatures between 35 C and 55 C, and stack exhausts with a
2020 onward [5,6]. broader temperature range, between 30 C and ~180 C [8,9]. Highly
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
ecient heat exchangers or heat pumps might help in recycling this
low-grade heat. The barriers associated with low grade heat utili-
zation [7] include (i) the need to match the sources with potential
end-users around the process plant; (ii) the diculty in harness-
ing the potential of the waste water streams available at lower
temperatures (between 35 and 55 C), while achieving economic
benets; (iii) the economic and geographic limits towards dis-
tances of pumping recovered heat (steam or hot water); (iv) the heat
losses; and (v) the total cost for pipeline installation and pumping
cost.
In high grade heat applications, electricity generation and solar
energy are the key topics. We need energy electrical or thermal
but in most cases where and when it is not available. Energy
storage technologies are a strategic and necessary component for
the ecient utilization of renewable energy sources and energy con-
servation, since the addition of short and long term energy storage
will enable an extensive and more ecient use of the uctuating
renewable energy sources by matching the energy supply with
demand [9]. High grade waste heat is already commonly re-
used in chemical and mineral processes such as using the hot exhaust
gas of cement kilns to preheat the raw clinker meal, or through steam
generation in chemical reactors, such as FCC cracking regenera-
tors, production of acrylonitrile, and other synthesis reactions.
Low cost, fossil fuel-based electricity has always served as a major
cost competitor for electrical power generation and traditional power
plants store their energy resource on site in the form of a stock of
e.g. coal, oil, nuclear fuel or even water behind a dam. The alter-
nating current electricity grid requires that supply and demand is
always strictly matched, and electricity network and transmission
system operators have arrangements in place to ensure that this re-
quirement is always met. The traditional electricity market is
therefore largely based on fuels sold and traded as commodities,
and used to generate electricity to instantaneously match supply
with demand. A future electricity system predominantly or solely
supplied by renewable energy sources will therefore nd it di-
cult to meet the fundamental stability requirement of alternating
current networks to constantly match supply and demand unless
storage facilities are integrated to balance both [9].
1.2. Maturity of different energy storage systems and cost effects
The LCOE is calculated as the systems expected lifetime costs
(including construction, nancing, fuel, maintenance, taxes, insur-
ance and incentives), divided by the systems lifetime expected power
output (kWh). All cost and benet estimates are adjusted for in-
ation and discounted to account for the time-value of money. The
LCOE is very useful to compare various generation options, and a
relatively low LCOE means that electricity is produced at a low cost,
with higher returns for the investor.
system s expected lifetime costs ( )
LCOE =
system s lifetime expected power output (kWh)
For a conventional plant, the effects of ination on future plant
maintenance must be considered and the price of fuel for the plant
lifetime into the future needs to be estimated. A renewable energy
plant is initially more expensive to build, but has very low main-
tenance costs, and no fuel cost over its 2030 year life. The cost
structure of solar power plants will vary depending on the tech-
nology they use. Fig. 1 shows the capital cost structures for solar
power tower and parabolic trough collector, respectively. Clearly,
major cost items are the solar eld and the receiver, together with
the TES component and the power block. To reduce the capital cost
and associated LCOE, these are the major items to be considered.
Both improved TES-systems (reduction of the storage volume) and
an increased operation temperature of the HTF (reduction of solar
3
Fig. 1. Capital cost structure of solar power plants (adapted from Ref. 10).
receiver capacity, heliostat eld size and storage volumes) will enable
this reduction to be achieved. Novel HTF systems using powders and
latent/thermo-chemical TES will be important in this respect, and
are subsequently dealt with in the paper. LCOE estimates for con-
ventional sources of power depend on very volatile fuel cost
estimates, as shown in Fig. 2. These uncertainties must be fac-
tored into LCOE comparisons between different technologies. LCOE
estimates may or may not include the environmental costs and/or
benets associated with energy production. Governments have
worldwide begun to quantify these costs.
Fig. 2 shows some LCOEs for different renewable energies and
for different MWel produced. The shaded area between both hori-
zontal lines at 0.05 and 0.14 $/kWh indicates the range for electricity
generation based upon fossil fuel. In general the LCOE decreases
when more MW is produced and over the years, due mostly to tech-
nology improvements. This decrease in LCOE with time is very clear
for wind energy, for solar PV and for CSP (concentrated solar power).
Since the energy storage part of the CSP represents 918% of the
required investment in PTC or SPT, respectively, reductions in TES
costs will favourably impact the LCOE.
In order to achieve the European climate energy objectives as
dened in the European Unions 202020 targets and in the Eu-
ropean Commissions Energy Roadmap 2050, energy storage
technologies are necessary. The energy objectives contain impor-
tant assumptions on long term views for the energy policy based
on different energy mix assumptions. The text moreover does not
claim to determine the future.
The percentage of renewable energy sources in the gross nal
energy consumption should attain at least 55% in 2050, and this in
all scenarios of the Roadmap. This shift towards renewable energy
sources will surely lead to a situation in which, from time to time,
generation will considerably exceed demand or vice versa, with ex-
plicit concern to transmission and distribution networks. To respond
to this challenge and to provide a continued security of energy supply
at any time, energy storage is notably well suited. This is why the
crucial role of energy storage technologies for an increasingly de-
carbonized European energy system is recognized by the European
Commission Energy Roadmap 2050. A growing use of renewable
energy sources, especially considering e.g. wind and photovoltaic
(PV) generation, will boost the demand for exibility in the energy
system.
Energy storage is indeed very appropriate to respond to this chal-
lenge and to assure a continued security of energy supply at any
time.
4 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Fig. 2. LCOE for different renewable energies and MW electricity produced (adapted from Ref. 11).

As mentioned before, there are four main goals met by energy
storage.
Firstly, energy storage will ensure the balance between demand
and supply, especially in view of the expanding variability at the
generation side. Since energy storage allows a timely and geograph-
ical displacement between consumption and generation sites, it
promotes the integration of renewable energy sources generation.
Secondly, it is possible to manage transmission and distribu-
tion grids with energy storage, conrmed to become a fundamental
element of the future electricity infrastructure to stabilize the grid,
even with ideal expansion of transmission capacities. Moreover, some
storage technologies could be realized much faster than grid up-
grades. With the current trend of electricity production and
consumption being decentralized and uctuating, storage can be
used to improve generation, transportation and distribution: energy
storage will allow to size grids closer to average energy ows instead
of peak power requirements, whilst also resulting in reduced trans-
mission losses, and improving the stability and reliability of grids
with the use of storage.
Thirdly, energy storage technologies will have a major role in the
transition to a more ecient and sustainable energy use. This will
be mainly seen in the transportation sector with the use of (hybrid)
electric vehicles, but also in the emergence of intelligent build-
ings and smart grids. With energy storage, it is also possible to
manage electricity/heat generation and consumption on a local level,
in an optimal way for the whole power system.
Finally, energy storage will also contribute to a competitive and
secure electricity supply. Since energy storage can provide an eco-
nomically interesting alternative to grid expansion and load shedding,
it will be very important in new market designs. Market mecha-
nisms for exibility and security of supply and specic storage
regulation will help to establish a competitive energy storage market.
Within the different types of energy storage such as electrical,
thermal, mechanical, electrochemical and chemical storage, the tech-
nologies recommended by the European Association for Storage of
Energy (EASE) are listed in Table 1 with their actual state-of-the-
art repartition: thermal, electrochemical and chemical energy storage
seem most suitable or possible in different technology sections.
Sensible heat storage is one of the most developed technolo-
gies for thermal storage and has been used for many years in both
the domestic and the industrial sector, e.g. in the form of hot water
and ice storage systems, or using thermal uid or molten salts in
concentrated solar tower technology. Sensible heat storage is indeed
most commonly used for CSP plants using a two-tank storage of
molten salts. For low temperature applications (<40 C), under-
ground thermal energy storage is recently gaining popularity in the
European markets. For high temperature applications very few ex-
amples are currently used in the process industry, despite numerous
publications and research.
Latent heat storage, using PCMs, is in full development. By 2015,
the specic investment costs of latent heat storage, storage of in-
dustrial waste heat, and improved thermal management need to
be reduced below 100 /kWh. By 2020 the specic investment cost
for compact latent heat storage should be below 50 /kWh. Towards
2030 the intention is to have industrial process heat applications
with thermal energy storage [12].
Thermalchemical energy storage applies both thermal and
chemical storages, using the sensible heat of reactants and the re-
action enthalpy of reversible thermo-chemical reactions. According
to EASE [12], the niche applications for thermo-chemical storage

Table 1
EASE recommended technologies [12].

Technologies in focus Conventional generation Renewable generation Transmission Distribution Customers services

Pumped hydro energy storage Suitable Possible Suitable Possible Unsuitable

Compressed air energy storage Suitable Possible Suitable Possible Unsuitable
Electrochemical Possible Possible Suitable Suitable Suitable
Chemical Possible Possible Possible Unsuitable Possible
Electromagnetic energy storage, ywheels Unsuitable Possible Suitable Suitable Unsuitable
Thermal energy storage Suitable Possible Possible Possible Suitable
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 5

Fig. 3. Energy storage systems according to form of energy stored according to power rating and discharge period (adapted from Ref. 13).

still need to be identied. By 2020, the goal is to have a specic in-
vestment cost for thermo-chemical storage below 50 /kWh. Towards
2030 the intention is to have thermo-chemical storage tanks for solar
thermal power plants and industrial process heat applications. To
take advantage of the high energy storage density, the operating tem-
perature should exceed 400 C [12].
1.3. Energy storage in general, and thermal energy storage
specically
Energy storage, as suggested by its name, is to store a certain
form of energy, which can thus be used later when necessary: the
storage device is generally called an accumulator. Various forms of
energy, commonly including kinetic energy, potential energy (gravi-
tational), chemical energy, electrical energy or thermal energy, can
be stored by using appropriate methods, as illustrated in Fig. 3.
The different systems can be tentatively assessed in view of
energy density and prices per unit energy-stored, as illustrated in
Table 2.
High temperature thermal energy storage is the main objective
of this paper, although cryogenics and moderate temperature storage
will also be briey dealt with.
TES works by heating storage media during a charging period
and then releasing the heat when the energy is needed [14], thus
increasing the energy eciency of a system and being commonly
used to compensate the mismatch between energy supply and
demand [15]. Current applications include thermal building pro-
cesses [16]; solar applications such as cooking, solar water boilers,
air heating systems, greenhouses and concentrated solar power
plants [1721]; mining and metallurgy [22,23]; and solar tower
power plants [3,17].
1.4. Heat balances in thermal energy storage
Sensible heat storage (SHS) stores thermal energy by raising the
temperature of a solid or liquid without the occurrence of a phase
change. The following equation expresses the SHS potential:

Table 2
Characteristics of different energy storage systems (adapted from Ref. 14).

Storage mechanism Storage period Energy density Price per unit energy-
(kWh/m3) stored (/kWh)

Pumped hydro energy storage (PHES) Mechanical Daymonth 0.51.5 1070

Compressed air energy storage (CAES) Mechanical Day 36 2140
Flywheel energy storage system Mechanical Hour 2080 1054 105
Superconducting energy storage system (SCESS) Electrical Hour N/A 7 1042.3 105
Superconducting magnetic energy storage (SMES) Electrical Hour 0.22.5 6.3 1047.5 105
Lead-acid battery Conventional battery Daymonth 5080 100830
Nickelcadmium (NiCd) Conventional battery Day 60150 4501800
Lithium-ion (Li-ion) Conventional battery Daymonth 200500 5002000
Sodiumsulphur (NaS) battery Molten salt battery Daymonth 156255 280700
Sodium nickel chloride batteries (ZEBRA) Molten salt battery Daymonth 150290 75-150
Zincbromine ow battery (ZBB) Flow battery Daymonth 3060 110750
Polysulphidebromine ow battery (PSB) Flow battery Daymonth 1660 1201000
Vanadium redox battery (VRB) Flow battery Daymonth 1633 110750
Hydrogen-based energy storage system (HESS) Chemical Daymonth 2.7160 at 1700 bar 215
Combined powder loop + PCM Thermal Hourweek <150 120
Thermo-chemical storage (TCS) Thermal Hourweek 1120250 8100
6 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

TH
Q = m C p dT = mC p (TH TL ) when Cp ~ constant (1)
TL

where m is the mass of storage material, Cp is specic heat at con-

stant pressure of the SHS materials, and TL and TH are the low and
high temperatures of SHS use. Cp is assumed independent of tem-
perature within the narrow temperature range of the SHS application.
In applications below 100 C, water is the best SHS liquid, because
of its wide availability, low cost and relatively high specic heat.
Above 100 C, synthetic oils, molten salts, liquid metals, or powders
can be used. For air heating applications, rock bed type storage ma-
terials are more suitable. Besides physical properties, such as density
and specic heat of the storage materials, there are other proper-
ties that are important for the eciency of SHS, including operational
temperatures, thermal conductivity and diffusivity, vapour pres-
sure, compatibility among materials and their stability, heat losses
as a function of the surface area to volume ratio, and cost of the
materials and systems. A SHS system usually consists of a storage
medium, a container tank, and inlet/outlet connections. To prevent
any potential loss of thermal energy, thermal insulation is re-
quired. The storage medium can be solid or liquid. Solid media, such
as concrete and cast ceramics, are commonly used in the form of
packed beds, which require a uid for effective heat exchanging and
as heat carrier. If the uid is a liquid, the system is called a dual
storage system, because the heat capacity of the solid in the packed
bed cannot be neglected. One of the most signicant advantages
of the dual system is the use of inexpensive solids, such as rock, sand,
or concrete as storage materials. Liquid media mainly include molten
salts or synthetic oils. They are naturally stratied in the storage
vessel due to the difference in density between hot and cold layers:
the hot uid is supplied to the upper part of storage during charg-
ing, whereas the cold uid is extracted from the bottom part during
discharging.
Latent heat storage (LHS) makes use of the heat absorbed or re-
leased by the storage material when it experiences a phase change
between solid and liquid or liquid and gas. The storage capacity of
a LHS system with phase change materials is given by:
Tm TH
Q = m C ps dT + m H m + m C pl dT
TL Tm
where Tm is the melting point of the PCM, Cps and Cpl are the spe-
cic heat of the PCM in solid and liquid state respectively, and Hm
is the phase change enthalpy. If Cps and Cpl are not a function of tem-
perature, the heat storage capability can be calculated by integration
into the following equation:
Q = m [C ps (Tm TL ) + H m + C pl (TH Tm )]
LHS with PCMs is at present very promising, because it exhib-
its high energy storage density and can store heat at constant
temperatures, i.e. the phase transition temperatures of the mate-
rials [2226]. Phase transition may occur in forms of solidsolid,
solidliquid, solidgas, and liquidgas. Among these phase changes,
only the solidliquid phase change has been used for LHS. Liquid
gas phase changes are not practical for use as thermal storage due
to the large volumes involved or high pressures required to store
the energy when the compounds are in their gaseous or vapour
phase, although liquidgas transitions have a higher latent heat of
transformation than solidliquid transitions. Solidsolid phase
changes are usually too slow and have a very low value of latent
heat of phase change. Although PCMs are expected to have a po-
tential advantage towards energy storage in comparison with sole
sensible heat storage, Fig. 4 illustrates the temperature range where
the ecacy of sensible heat storage of the heat carrier (e.g. sand)
can exceed the latent heat storage capacity, as illustrated for e.g.
Fig. 4. Total heat stored vs. temperature for sensible and latent heat storage mate-
rials. 1: KNO3=NaNO3, 2: NaOH=NaCl, 3: SiO2, 4: SiC, 5: Sb2O3 [27]. HTF: Heat transfer
uid.
Sb2O3PCM: the total heat stored is lower than the sensible heat
stored in SiC and/or SiO 2 . Molten salts KNO 3 NaNO 3 and
NaOHNaCl, on the contrary, offer a signicant advantage. Due to
the latent heat of fusion, there is a steep jump in energy stored at
the temperature of the PCM melting point.
Thermo-chemical storage (TCS) uses the heat absorbed and re-
leased by a reversible chemical reaction. The amount of heat stored
is proportional to the amount of storage material (m), the endo-
thermic heat of the reaction (Hr) and the conversion (), given by:
Q = mH r with 1 (4)
It should be remembered that the external heat supply will also
need to cater for the sensible heat of the initial reactant between
the starting (ambient) and the reaction temperature.
For an operation between 750 C and 250 C, with specic heat
(2) value around 1 kJ/kg, the heat storage capacity of sensible heat
storage is ~500 kJ/kg, between 700 and 1000 kJ/kg in latent heat
storage due to the contribution of the heat of phase transition, and
in excess of 1000 kJ/kg for thermo-chemical storage, respectively.
1.5. Classications of thermal energy storage
TES systems can be classied according to different param-
(3) eters, being the temperature range of application, the mechanism
of energy storage, and the integration within the energy storage
concept.
1.5.1. Classication according to temperature range and associated
re-use technology
Essential to be able to capture, store and re-use thermal energy
is the use of a heat transfer uid (HTF): the ranges of applicable
thermal energy storage are hence a function of the HTF selected.
According to the temperature range at which the system works, TES
can be classied as low temperature thermal energy storage and
high temperature thermal energy storage. In this study the divi-
sion temperature is considered at about 200 C. The classication
according to the working temperatures of the HTF, TES and thermal
block is illustrated in Fig. 5.
The increased technological risk reects the impact of safety of
operations at extremely high temperature and the use of high value
construction materials. Traditional HTF include gas, water/steam,
thermal uids and molten salts. The use of particle suspension as
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 7

Fig. 5. Classication of energy storage systems according to the working temperature [28].

HTF is considered as a novelty and offers major benets to enhance
the solar power tower commercial potential [7,21,28]. The use of
particle suspensions as heat carrier to transfer solar heat from the
receiver to the energy conversion process offers major advantages
in comparison with water/steam, thermal uids or molten salts. Since
the particle suspension has a heat capacity similar to that of molten
salts, without temperature limitation except for the maximum al-
Fig. 6. Evolution of standard reheat Rankine-cycle congurations (numbers refer to
Table 3) [28].
lowable wall temperature of the receiver tube, suspension
temperatures of up to 800 C can be tolerated for refractory steel
tubes (even higher when using ceramic or glass tubes), thus offer-
ing new opportunities for highly ecient thermodynamic cycles such
as obtained when using supercritical steam or CO2. With higher tem-
perature HTFs, a cascade of effects is noteworthy, with additional
high eciency thermodynamic systems being viable, as illus-
trated in Fig. 6 and assessed by e.g. Dunham and Iverson [29]. A
higher temperature operation will increase the power cycle e-
ciency, whilst increasing the temperature range over which the
storage operates, thus enhancing the storage density. The in-
creased eciency of the power cycle also reduces the thermal power
demand for a constant receiver eciency, which allows more elec-
tricity to be generated per unit of stored thermal energy. The
increased capacity of the storage moreover increases the power plant
capacity factor, thus reducing investments, despite the use of more
expensive construction materials. If a new high-temperature HTF
is to improve the economics of e.g. SPT applications, the cost re-
duction of the solar eld and storage must more than outweigh cost
increases of the receiver and power block. A high temperature
powder circulation loop allows advanced power cycle congura-
tions to be used, albeit not yet exploited in current solar power tower
applications. Such suspension receivers with outlet temperature
above 700 C are being developed in the Concentrated Solar Power
in Particles European project (CSP2) [30].
Supercritical Rankine cycles and associated steam turbines are
commonly designed for large power outputs (~800 MWel) [31], and
will need to be scaled-down for smaller plants by a redesign in order
to cope with the low volumetric ow rates and the resulting small

Table 3
Typical operating conditions reported by Spelling et al. [33].

Nbr. Power cycle Steam conditions Cycle eciency (%)

C Bar

1 High-tech. PTC plants 375 100 ~35

2 High-tech. molten salt SPT plants 535 115 ~40
3 Old subcritical fossil fuel plants 535 165 ~42
4 High-tech. subcritical fossil fuel plants 565 165 ~43
5 Old supercritical fossil fuel plants 565 255 ~44
6 High-tech. supercritical fossil fuel plants 600/610 270 ~45
7 Advanced supercritical fossil fuel plants 600/620 285 ~46
8 Ultra-supercritical fossil fuel plants 700/720 350 ~48
8 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Fig. 7. Classication of average storage systems according to applied concepts (adapted from Kuravi et al. [34]).

blade sizes, although solutions using radial turbines or operating
a high pressure turbine at higher speeds have been suggested [32].
In order to use solar power tower plants in arid locations, the power
block will use dry, indirect cooling for lack of sucient cooling water.
1.5.2. Classication according to energy storage mechanism
If the material is not subjected to a phase change or a chemical
reaction, due to temperature variations, it is referred to as SHS. LHS
requires the material to absorb or release enough energy to trans-
form its phase structure during melting/solidication. Finally, a TES
system is classied as TCS, if the storage media undergo revers-
ible endothermic/exothermic chemical reactions for heat storage.
1.5.3. Classication according to the storage concept
The different TES concepts are also classied regarding the lo-
cation of the storage media within the TES and on how the energy
reaches it. This classication involves active, passive and hybrid
storage, as illustrated in Fig. 7 and dealt with in detail by Kuravi et al.
[34].
In an active storage system, the storage medium circulates
through a heat exchanger (which can also be a solar receiver or a
steam generator) and it is charged or discharged by forced convec-
tion heat transfer. Active storage systems can be subdivided into
direct and indirect systems, where the former uses the same ma-
terial as heat transfer uid and storage medium, whereas the latter
uses a secondary storage medium.
Contrary to an active storage, the passive storage medium does
not circulate through the system, but always stays in the same lo-
cation. The heat transfer uid is responsible of carrying energy
through the system: in the charging process the HTF carries the re-
ceived energy from the energy source to the storage medium and
receives energy from the storage medium when discharging. This
type of system mainly uses a solid heat storage medium such as
concrete, casts, and contained-PCMs.
The hybrid storage combines active and passive concepts in trying
to improve the characteristics of the storage systems.
1.6. Objectives and structure of the paper
Energy capture and storage are hot research topics, as illus-
trated by the literature review of Fig. 8: both PCMs (and nanoparticle-
PCM) and TCS are within the focus of increasing current research,
whereas the use of powders as sensible HTF has only recently at-
tracted a lot of interest.

Fig. 8. Literature (19702015), as total number of papers per 5 years, related to energy capture and storage in SCOPUS with keywords: NP-PCM; PCM; particulate
HTFs; thermo-chemical storage.
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
Prior to delineating the objectives of the present review paper,
two previous review papers are important to cite, since both already
deal with specic aspects of TES.
The assessment of Yu et al. [35] evaluates TES by sorption process
using silica gels, zeolites, alumino-phosphates, silico-alumino-
phosphates and metal organic frameworks. These systems are
recognized for their high thermal energy storage densities and long
term applications. Intensive studies are however additionally re-
quired to demonstrate the technical feasibility of the concept.
Sorption energy storage is however not further discussed in the
current review.
A more comprehensive review of TES systems is provided by
Kuravi et al. [31]. This review discusses the thermal energy storage
design and practise for application and integration into concen-
trating solar power plants. This integration is studied at plant,
component and system level. Storage systems using active two-
tank, steam accumulation, extended/embedded heat transfer
structures, and packed bed systems were assessed. Energy and exergy
eciencies of the TES-systems were calculated to dene the eco-
nomic optimization. A life cycle assessment of TES systems and a
case study conclude the research. The present review will exten-
sively refer to Kuravi et al. [34] as far as these specic topics are
concerned. It will however add additional and up-to-date ndings
in various fundamental and practical aspects of TES, all integrated
into a design logic as illustrated in Fig. 9.
The structure comprises 7 sections. Section 1 positioned TES
systems and their growing importance within the energy aware-
ness, with aspects of TES classication and its main impacts.
Section 2 will deal with sensible heat storage, well-documented
in literature and hence limited in the content of this review.
Section 3 will deal with latent heat storage, specically towards
high temperature application, where inorganic substances offer a
Fig. 9. Layout of the review paper.
9
high potential. Experimental and modelling data will be com-
pared. Since inorganic solid/liquid phase change materials suffer from
a low thermal conductivity, conductive inserts offer a high poten-
tial, reducing the charging/discharging time by a factor of 2.
Specic applications of latent heat storage with cryogenics and
with steam accumulations will be discussed, both however within
very special ranges of application. Finally, recent developments in
enhanced high temperature PCMs and the important impact of using
lithium salts will be assessed: enhancement techniques appear to
be moderately to highly effective.
Section 4 will review the fundamentals of using reversible chem-
ical reactions to store heat for short to long term periods. Within
the selected reaction pairs, Ca(OH)2, CaCO3, Co3O4 and Mn2O3 offer
a high potential, both towards reaction temperature and revers-
ibility. All reactions can be treated as rst order chemical reactions,
with activation energy between 50 and 250 kJ/mol and reaction rates
of 1 to 2 103 s1 at reaction equilibrium temperature.
Since latent and thermo-chemical storage medium needs to be
contained in a capsule (sphere, tube, sandwich plates) of appro-
priate materials, Section 5 assesses the mechanical strength of the
containment from both a theoretical and experimental evalua-
tion. Alloy-capsules of 12 mm thickness meet the requirements of
long term stability, even after multiple charging/discharging cycles.
Section 6 deals with the current development of a novel heat
transfer uid, involving particulate suspensions. Literature and ex-
perimental data conrm the high heat transfer coecient achieved
over 400 W/m2 in circulating uidized bed, to 1100 W/m2 K in a
dense upow uidized bed. A tentative assessment selects to a po-
tential design and demonstrates that particulate suspensions lead
to major savings in investment and operating costs. Finally, Section
7 will summarizes the ndings and highlight required priority
research.
10 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

2. Sensible heat storage
Sensible heat storage materials undergo no phase change within
the temperature range required for the storage application [14].
Table 4 summarizes the main characteristics of the most common
solid and liquid sensible heat storage materials.
Within the indicated solids, concrete and cast ceramics have been
extensively studied due to their low costs and good thermal con-
ductivities, despite their moderate specic heats. In terms of liquids,
molten salts [17] and mineral oils [38] are widely used in solar towers
and parabolic trough collectors, respectively, being liquid at ambient
pressure, providing an ecient and low cost medium, and having
their operating temperatures compatible with current high-
pressure and high-temperature turbines (temperature range over
120600 C). From Table 4, it is clear that the main candidates for
liquid high temperature sensible heat storage are either a solar salt,
i.e. a binary salt consisting of 60% of NaNO3 and 40% of KNO3, that
melts at 221 C and is kept liquid at 288 C in an insulated storage
tank; or HitecXL, a ternary salt consisting of 48% Ca(NO3)2, 7% NaNO3,
and 45% of KNO3 operating beyond its melting point of 130 C.
Common mineral particles, such as silicon carbide, crystobalite
silica, exhibit high values of Cp, and are hence subject of increas-
ing research as SHS media, as described in Section 6 of this paper.
Sensible liquid heat materials have been widely studied and are
currently applied in solar thermal plant applications, despite im-
portant disadvantages that can affect the storage system design and
stability. The low solidication point may be a problem for solar
power plants because of the required heat tracing during non-
functioning periods. Furthermore, the sensible fraction of thermal
energy is seldom fully recovered due to the required temperature
difference for the heat transfer driving force. Another important dis-
advantage consists in the low energy storage density of sensible heat
materials, leading to the requirement of large volumes or quanti-
ties in order to deliver the amount of energy storage necessary for
high temperature thermal energy storage applications. The above
mentioned problems can imply signicant increments in the costs
of sensible heat storage systems. For solid SHS media, the number
of available materials is extensive and current handbooks and da-
tabases give no comparison or possible relationships between
properties of materials that enable their selection [39]. Materials
with the highest Cp are natural and polymeric materials, such as
natural rubber or the thermoplastic copolymers (acrylonitrile bu-
tadiene styrene) with a Cp value of around 2 kJ/kg K. Other composite
materials such as glass bre reinforced epoxy concretes have Cp
values close to 1 kJ/kg K. The specic heat capacity should be con-
sidered in conjunction with the price of the SHS material. Materials
for sensible thermal energy storage in the range of 15200 C were
considered and presented by Fernandez et al. [39]. Commercial prices
vary from 0.02 to 0.08 /kg for standard concrete, from 0.15 to 0.25 /
kg for high density concrete, from 0.1 to 0.3 /kg for common mineral
powders, and from 0.5 to 3 /kg for steel and alloys.
3. Latent heat storage
3.1. Main characteristics
Latent storage systems based on PCMs with solidliquid tran-
sition are considered to be very ecient in comparison to liquid
vapour and solidsolid transitions [40]. Liquidgas transition requires
a large volume recipient for the PCM and the solidsolid transi-
tion presents a low value of latent heat, making both alternatives
not considered as appropriate choices. A large number of materi-
als are known to melt with a moderate to high heat of fusion within
different ranges of temperature. No material yet studied has all the
optimum characteristics required for a PCM, and the selection of a
PCM for a given application requires careful consideration of the
properties of the various substances and/or mixtures [34,41]. The
main disadvantage of a PCM is its low range of thermal conductiv-
ity between 0.2 and 0.8 W/mK, whereas sensible heat materials score
better. Therefore, improving the PCM thermal conductivity will
enhance the TES eciency by improving its charging/discharging
processes as dealt with below.
3.2. PCMs classication
PCMs are generally classied in different categories consider-
ing their melting temperature and their material composition. The
melting temperature can be divided in two main groups: low tem-
perature (<200 C) and high temperature (>200 C). The categories
for material composition are mainly threefold, being organic, in-
organic and eutectic compounds [34], as illustrated in Fig. 10.
Organic materials can repeatedly freeze and melt without phase
segregations and are usually non-corrosive. Both paran and non-
paran compounds are limited in use to low and moderate
temperatures, hence outside the scope of the present review. Some
of the features of this group are their high heat of fusion, low thermal
conductivity and instability at high temperatures. Properties are re-
viewed in literature by e.g. Zalba et al. [36], Kuravi et al. [34],
Jegadheeswaran et al. [37] and other authors.

Table 4
Main characteristics of sensible heat storage solid and liquid materials (adapted from Refs. 34, 36, and 37).

Storage medium Temperature Average density Average heat Average heat

(kg/m3) conductivity (W/mK) capacity (kJ/kg K)
Cold (C) Hot (C)

Sand-rock-mineral oil 200 300 1700 1 1.3

Reinforced concrete 200 400 2200 1.5 0.85
NaCl (solid) 200 500 2160 7 0.85
Cast iron 200 400 7200 37 0.56
Silica re bricks 200 700 1820 1.5 1
Magnesia re bricks 200 1200 3000 1 1.15
HITEC solar salt 120 133 1990 0.60
Mineral oil 200 300 770 0.12 2.6
Synthetic oil 250 350 900 0.11 2.3
Silicon oil 300 400 900 0.1 2.1
Nitrite salts 250 450 1825 0.57 1.5
Nitrate salts 265 565 1870 0.52 1.6
Carbonate salts 450 850 2100 2 1.8
Liquid sodium 270 53 850 71 1.3
Silicon carbide 200 1400 3210 3.6 1.06
SiO2 (crystobalite) 200 1200 2350 0.92 1.13
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
Fig. 10. PCM classication considering material composition [24,36,37].
Inorganics are divided in two main groups called salt hy-
drates and metallics.
Salt hydrates are mixtures of inorganic salts and water forming
a typical crystalline solid of general formula ABnH2O. At the
melting point the hydrate crystals decompose into anhydrous
salt and water, or into a lower hydrate and water. This type of
material is used as PCM due to its high latent heat of fusion per
unit volume, relatively high thermal conductivity (almost twice
the thermal conductivity of paran) and small volume changes
at melting. The major problem of salt hydrates is that most of
them have incongruent melting, which means that the salt is not
entirely soluble in its water of hydration at the melting point.
This results in an irreversible meltingfreezing and a reduction
of the available salt hydrate with every chargedischarge cycle.
Metallics include low melting metals and eutectic alloys. Despite
their advantages, they have not yet been considered as a PCM
due to weight penalties. They are used when volume is a con-
sideration because of their high heat of fusion per unit volume.
Other features of these materials are their low heat of fusion per
unit weight, high heat of fusion per unit volume, high thermal
conductivity, low specic heat and relatively low vapour pressure.
Eutectics are melting compositions of two or more compo-
nents, which melt and freeze congruently forming a mixture of the
component crystals during crystallization. Due to this crystalliza-
tion, the materials have little chance of separation. Depending on
the nature of the material composition, they can be divided into
organicorganic, inorganicinorganic and inorganicorganic. Towards
CSP application, the most used heat storage medium is a eutectic
salt known as solar salt. This is a combination of 60% NaNO3 and
40% KNO3.
The low-temperature PCMs (organic, inorganic or eutectic) are
mainly used in waste heat recovery systems and buildings, while
high temperature PCMs (inorganic or eutectic) can be used in solar
power plants and other high temperature applications [25].
When developing a TES system, the selection of a PCM is an im-
portant task, and the selected PCM must t the requirements of the
system. Abhat [26], Jegadheeswaran et al. [37] and Zalba et al. [36]
described the main criteria that govern this selection, which can
be classied in thermal, physical, chemical and economical
properties.
Thermal properties should include a melting point in the desired
operating temperature range (temperature range of application);
a high latent heat of fusion per unit mass, so that a smaller amount
11
of material stores a given amount of energy; a high specic heat
to provide additional signicant sensible heat storage effects; and
a high thermal conductivity, so that the temperature gradients for
charging and discharging the storage material are small.
Physical properties should account for small volume changes
during phase transition, so that a simple container and heat ex-
changer geometry can be used; a high density; exhibiting little or
no sub-cooling during freezing; and a low vapour pressure in order
to avoid stresses and problems with the container and the needed
heat exchangers.
Chemical properties of the PCMs are focused upon their chem-
ical stability without decomposition; whilst avoiding corrosion of
construction materials. They must moreover contain non-poisonous,
non-ammable and non-explosive elements/compounds. Phase seg-
regation should be avoided.
Towards process economy, PCM materials should be available in
large quantities at low cost.
Since organic PCMs have applications at low to moderate tem-
perature (110260 C) only, Table 5 only lists physical properties of
non-organic PCMs.
To the extent of our current knowledge, there is no commer-
cial high temperature thermal energy storage module using PCM
storage, and only experimental modules have been constructed and
assessed. According to a recent study of the International Renew-
able Energy Agency (IRENA) [43], the status of the market for high
temperature thermal energy storage modules is still low. All the in-
vestment in this area has been focused on research and development.
One PCM test module recorded in literature was developed within
the DISTOR project, developed by the Institute of Technical Ther-
modynamics of the German Aerospace Center (DLR), which focuses
on the development of thermal storage systems using PCM in the
temperature range from 230 C to 330 C for systems using steam
between 30 and 100 bar. In this context, this project has devel-
oped innovative PCMs and heat transfer concepts to overcome PCM
disadvantages. From their research, they could conclude that the
nned tube design is the most promising and technically feasible
concept, therefore, it was decided to build a test 200 kW PCM storage
module based on the nned tube module design [44].
3.3. Moderate or high temperature PCMs
Although different PCMs can be used within a temperature range
of 300700/800 C, as illustrated in Table 5, their application at these
temperatures is still subject of widespread investigation. PCM
usage in high temperature thermal energy storage has been
12 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Table 5
Inorganic substances/compounds for potential use as PCM [34,36,41,42].

Inorganic compounds Melting Heat of fusion Density Specic heat at the Thermal
point (C) (kJ/kg) (kg/m3) melting point (kJ/kg K) conductivity (W/mK)

Mg(NO3)2.2H2O 130 275 NA NA NA

Hitec XL: 48% Ca(NO3)245%KNO37%NaNO3 140 NA 1992 1.44 0.519
Hitec: KNO3NaNO2NaNO3 142 84 1990 1.34 0.6
LiNO3NaNO3 195 NA NA NA NA
KNO3/NaNO3 eutetic 223 105 NA NA NA
NaNO3 307/308 74 2260/2257 NA 0.5
65.2%NaOH20%NaCl14.8%Na2CO3 318 290 2000 1.85 1.0
KNO3 333/336 266 2110 NA 0.5
22.9% KCl60.6% MnCl216.5% NaCl 350 215 2250 0.96 0.95
KOH 380 150 2044 NA 0.5
MgCl2/KCl/NaCl 380 400 1800 0.96 NA
Na2CO3BaCO3/MgO 500850 NA 2600 NA 5
Li2CO3 618 NA 2091 2.07 NA
Sb2O3 652 387 5670 0.43 NA
MgCl2 714 542 2140 NA NA
80.5% LiF19.5% CaF2 eutetic 767 790 2100/2670 1.97/1.84 1.7/5.9
LiF 850 811 NA NA NA
Na2CO3 854 276 2533 NA 2
K2CO3 897 236 2290 NA 2

NA, not available.

experimented for solar plant applications, but it has not yet been
commercially used [45,46]. The development of high temperature
thermal energy storage using PCMs is of increasing interest since
they are fairly cheap, have a high energy density, can be available
in large quantities, and are able to store and release thermal energy
at a constant temperature. In addition to adapting PCM melting
points to a high temperature range, their heat conduction charac-
teristics need to be improved in order to increase the eciency of
the charging and discharging processes. Although PCMs are ex-
pected to have a potential advantage towards energy storage in
comparison with sole sensible heat storage, Fig. 4 illustrated that
the eciency of sensible heat storage can in some cases exceed the
contained sensible and latent heat of the PCM. Molten salts gen-
erally offer signicant advantages. To counteract the slow charging/
discharging rates of PCMs, the integration of high conductivity
structures has been investigated. Figs. 11 and 12 illustrate some of
the experimental results [27].
The experimental results illustrate the effect of different param-
eters. The temperature of phase change should be a constant 230 C,
as obtained when a pure substance solidies. With mixtures, the
solidication takes place over a range of temperatures, sometimes
referred to as the mushy region between the solid and liquid zones
[36]. The effect of inserted metallic foam or metallic sponge is
evident, with a signicantly shorter discharge rate as a result. The
curve with forced water cooling (hence with a far higher external
heat transfer coecient) is very steep, and the temperature de-
creases from 270 C to 160 C in less than 700 s (against 2500
3700 s when air cooling is applied). This external heat transfer
coecient is an important parameter of the Biot-number, the di-
mensionless group that rates convection and conduction.
The date treatment can be performed by different methods:
(i) Using an analytical, unsteady-state conduction approach, in-
volving a material temperature to be dependent on both time
and spacial co-ordinates, with solutions provided by Carslaw
and Jaeger [47] and Smith [48]. The solution of the equa-
tions results in the prediction of the temperature at a given
point in the body at time t measured from the start of the
cooling or heating operation, for given values of

Fig. 11. Temperature vs. time cooling of the E-PCM (nitrates) [27]. Air-cooling of liquid
PCM (1) no inserts; (2) metallic sponge; (3) metallic foam. Air-cooling of solid PCM
(4) no inserts; (5) metallic sponge; (6) metallic foam; water-cooling of (7) liquid
PCM + foam; and (8) solid PCM + foam. Fig. 12. T vs. cooling time for E-PCMs [27].
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 13

Table 6
The effective conductivity of the PCM and insert-composites [27,5256].

Ref. keff
(1 )
52 k PCM k INS

53 k + 2k PCM (1 ) (k PCM k INS )

k PCM INS
k INS + 2k PCM + (1 ) (k PCM k INS )

k PCM
1 + (1 ) F k
1
1 1 3
with F = 1 + PCM 1 f i
k
54 k PCM 3 i =1 k INS
INS

1 (1 ) (F 1)

fi are the semi-principal axes of ellipsoids, proposed as
f1 = f2 = 0.125 f3 = 0.75

1 3 1 + 1 + k PCM 3 1 1
( ) ( ) ( ) ( )
k INS
55 k PCM
k
1 3 (1 ) + PCM 3
(1 )
k IN S

3 3
k INS + 1 k INS 1 (1 ) (k PCM k INS )
56 k PCM
3
k INS + 1 k INS + (1 ) (k PCM k INS )

kt t
= = 2 (Fourier-number )
C p R 2 R
1 k
= (inverse Biot-number )
Bi hR
with h = external heat transfer coecient (W/m2 K) and n = r/R di-
mensionless distance in the direction of the heat conduction.
(ii) By using Comsol-Multiphysics, as previously used by Lopez
et al. [49] and later extended by Piti et al. [50] and Parrado
et al. [51], presenting the conned melting in composite ma-
terials made of graphite-nitrates, ceramic-nitrates, and copper-
nitrates, respectively. Comsol-Multiphysics offers additional
calculation procedures such as internal pressure build-up due
to the thermal expansion of the PCM. The procedure Thermal
Stress contains Heat Transfer and Solid Mechanics phenom-
ena needs to be used. In order to provide results with a
minimum error range, it uses an extra-ne mesh. The shell
wall is considered to be homogeneous, isotropic and present
at a known and uniform temperature. At the interfaces, an
equality of phase temperature and heat ux continuity at the
melting front, and an equality of temperature and pressure
at the shell/salt interface are accepted.
To predict the effective thermal conductivities of the PCM with
inserts (foam, sponge), a series of empirical correlations was used,
as listed in Table 6. The symbols keff, kPCM and kINS represent respec-
tively the effective thermal conductivities of the composite mix, the
thermal conductivity of the PCM itself, and the thermal conduc-
tivity of the insert. The porosity, , represents the void space (PCM-
lled) of the composites. The sphericity, , is estimated at 0.95 for
the composites. The predicted values for the composite systems are
given in Table 7, with 10% deviation noticed between the differ-
ent predicted and average values. The inserts clearly increase the
Table 7
Calculated effective thermal conductivity of the PCM-insert composite.
System Reference of Equation: keff (W/mK)
52 53 54
Nitrate melt + foam 0.565 0.496 0.435
Nitrate solid + foam 0.965 0.869 0.763
Nitrate melt + sponge 0.710 0.547 0.492
Nitrate solid + sponge 1.178 0.955 0.766
Fig. 13. Comparison of experimental temperature vs. time curves and predictions
by solving the unsteady-state conduction equation and as predicted by Comsol
Multiphysics [27].
(5)
thermal conductivity of the PCM, with sponge (porosity 0.9) having
a slightly higher effect that the metallic foam.
(6)
Fig. 13 illustrates the comparison of calculated and experimen-
tal discharging rates of both liquid and solid PCM [27,49]: within
the temperature range under consideration, cooling rates can be fairly
well represented by a linear relationship of temperature and time.
This results in respective cooling rates, T/t (C/s), which provide
a tentative comparison as presented in Fig. 12.
The theoretical calculation of the cooling rate of the liquid PCM
underestimates the experimental temperature gradient, due to the
fact that only conduction is taken into consideration, whilst natural
convection within the liquid phase also contributes to the heat trans-
fer. When calculating the Rayleigh-number within the liquid salt,
values exceeding 107 are obtained, signicantly above the thresh-
old 1700 normally given for the onset of natural convection. Since
it is dicult to predict the contribution of natural convection, Zalba
et al. [36] propose to characterize the melt by an effective thermal
conductivity, exceeding the thermal conductivity as such. An ex-
pression of this effective thermal conductivity could not be obtained
yet, but it should indeed be expressed as: k total = k melt f (Ra ) with
Ra being the Rayleigh number.
To t the experimental results, the value of the f(Ra) ranges from
1.2 to 1.4, but further rening is required. The Comsol predictions
of the temperature evolution are very promising, again with an un-
derestimation of the cooling rate in the liquid phase, but a close
prediction of the periods of phase transformation and solids cooling.
3.4. Specic PCM applications
3.4.1. Latent heat storage with cryogenics
Cryogenic fuels, such as liqueed natural gas, have been deemed
to be promising alternative fuels due to their high energy density
[57]. They are stored at extremely low temperatures to maintain
the liquid state. At the consumers end, the cryogenic liquids need
to be re-gasied to about ambient temperature, and plenty of cryo-
genic cooling capacity would be released during this vapourization
process. It is therefore quite important and imperative to maxi-
55 56 Average mize the recovery of the cold energy either through the use of other
0.485 0.522 0.501 cryogens or PCMs [55,56]. Cryogens normally refer to as liquid media
0.851 0.914 0.872 at a temperature below approximately 150 C and are physical
0.532 0.602 0.567 energy carriers [58]. The use of cryogen as an energy carrier is dif-
0.932 1.050 0.978
ferent from normal heat storage in that the energy is stored through
14 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Table 8
Comparison of specic heat, latent heat and exergy density of cryogens and some commonly used heat storage materials [69].

Media Storage methoda Specic heat (kJ/kg K) Phase change temperature (C) Fusion/latent heat (kJ/kg) Exergy density (kJ/kg)

N2 (liquid cryogens) S+L 1.01.1 196 199 762

CH4 (liquid cryogens) S+L 2.2 161 511 1081
H2 (liquid cryogens) S+L 11.314.3 253 449 11987
a
In this description S indicates that thermal energy is stored in the form of sensible heat while L stands for latent heat.

decreasing its internal energy while increasing its exergy. The use
of liquid air/nitrogen as an energy carrier is not commonly known
until fairly recently [5961]. Cryogenic energy storages can have a
relatively high energy density (100200 W h/kg), low capital cost
per unit energy, are benign to the environment and have a rela-
tively long storage period. However, it has a relatively low eciency
(4050%) according to the current energy consumption for air liq-
uefaction [61]. Cryogenic energy storage is still under development
by Mitsubishi (Japan) [62], the University of Leeds (UK) and the
Chinese Academy of Sciences [63]. Specic topics of research involve
hydrogen releases [64], the phase change itself [65,66], and inte-
grated applications [67,68]. Table 8 and Table 9 illustrate that
although the specic heat and phase change heat of the cryogens
are of a similar order of magnitude than those of common heat
storage materials, the exergy density of cryogens is much higher.
Once produced and stored in insulated containers, the cryogen
is ready to be delivered. No extra energy is required except for the
pumping power consumption which is negligible. The only energy
loss of the cryogen is the heat dissipation of the cryogenic tank,
which can be less than 1% per day using conventional insulation
technologies [7072].
The exergy eciency for a cryogenic engine is about 40% as a
considerable portion of the exergy is in the form of cold, which is
wasted during the heating process. In order to improve the e-
ciency, the recovery of the waste cold energy can be integrated with
the direct expansion of the cryogen using cycles such as the Rankine
cycle as shown in Fig. 15. Many power cycles, based on Rankine and
Brayton cycles, were proposed to recover the cooling capacity of cryo-
genic liquids [7477]. Cryogenic liquids vapourizers are commonly
used as low temperature thermal sinks for the bottom Rankine cycle
(see Fig. 14), with waste heat or low-grade thermal energy (to
enhance exergy eciency) as thermal sources. Its alternative heat
source could also be using the atmosphere [60], unlike a conven-
tional heat engine that uses the atmosphere as a cold reservoir or
heat sink. The combined methods for cryogenic energy extraction
have been discussed in detail by Li et al. [61,68]. It has been con-
cluded that the combined cycle cryogenic heat engines could have
an exergy eciency of 6578% under ideal conditions [78].
Chen et al. [79] claimed that the exergy eciency could be im-
proved from ~78 to 117% if the working uid were superheated to
~100 C, and the exergy eciency could be doubled if the waste heat
temperature is as high as 300 C.
3.4.2. Latent heat storage with steam accumulators
The accumulator is an insulated steel pressure tank containing
hot water and steam under pressure, thus storing energy in the form
of latent heat stored in the liquid phase, i.e. water. It releases steam
by extracting latent heat when needed and to accept steam when
the demand is low. The volume specic thermal energy densities
are typically in the range of 2030 kW h/m3. The released steam
powers the steam turbine for electricity production. Storing steam
under pressure in saturated or superheated conditions in pressure
vessels is generally not economic due to the low volumetric energy
density [73]; it is nevertheless applied in specic cases as buffer
between e.g. a boiler and the steam-driven process of a paper mills
[80].

Table 9
Comparison of physical and chemical exergies of the cryogens [61].

Cryogen Thermal exergy (kJ/kg) Chemical exergy (kJ/kg) Gas density (kg/m3) Liquid density (kg/m3)

Liquid H2 11,897 116,528 0.0824 70.85

Liquid N2 762 0 1.1452 806.08
Liquid CH4 1081 51,759 0.6569 422.36

Fig. 14. Direct expansion-Rankine hybrid cycle [78].

 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 15

Fig. 15. Pressure drop accumulator [73].

There are three types of steam accumulators, as described and
illustrated in Steinmann and Eck [73] and DOE [81]:
Water lled accumulator:
1. Constant pressure accumulator: usually arranged in a verti-
cal position to allow for both hot and cold water to be stored
in the same vessel via thermal stratication, where the hot
water is in the top part and the cold water in the bottom part.
2. Pressure drop accumulator: also called feed water accumu-
lators, where pressure and temperature are allowed to vary
(see Fig. 15). The accumulator is usually arranged horizon-
tally and when fully charged, it is usually at 90% full of water.
Supercritical steam accumulator: this requires a storage condi-
tion of a minimum temperature of 647 K at pressure of 220 bar.
characteristic volume-specic storage capacity of PCMs is also de-
sirable, which is in the range of about 100 kW h/m3, signicantly
exceeding water (2030 kW h/m3) [73].
The design of accumulators should fulll important criteria, as
listed by DOE [81] and summarized below:
Provide sucient time from the end of one overload period to
the beginning of the next, to fully recharge the accumulator.
Sucient surplus boiler capacity should be available to re-
charge the water stored in the accumulator during off-peak times.
The accumulator volume should be large enough to contain
enough water to provide the required amount of ash steam
during the discharge period.

During the thermal storage process the temperature of the liquid

water can be increased by condensation of superheated steam, with
little variation of the liquid storage mass; or the mass in the volume
is increased by feeding saturated liquid water into the system, where
the pressure remains constant, charging indirectly where a heat ex-
changer is integrated into the liquid volume. The medium owing
in the heat exchanger needs not be water; hence heat from a heat
source at lower pressure can be used.
Steam is charged beneath the surface of the water by a distri-
bution manifold, which is tted with a series of steam injectors, until
the entire water content is at the required pressure and tempera-
ture. If the steam accumulator is charged using saturated (or wet)
steam, there may be a small gain in water due to the radiation losses
from the vessel.
To avoid a pressure drop in a constant pressure water lled ac-
cumulator, the application of a separate ash evaporator is required,
where the saturated liquid water is taken from the steam accumu-
lator, which is later depressurized externally. Cold water is fed into
the bottom of the storage vessel to keep the water level constant,
where mixing of hot and cold water must be minimized. Thermal
stress resulting from lling the pressure vessel with cold water must
be considered [73]. An alternative option is to maintain constant
pressure by integrating PCM into the storage vessel partly replac-
ing the liquid water (Fig. 16). PCMs usually exhibit a low thermal
conductivity so layers of this material must be thin to ensure a suf-
cient heat transfer rate or via the encapsulation of PCM in small
containers placed inside the liquid volume. PCM is not only attrac-
tive regarding the avoidance of thermo-mechanical stress, the Fig. 16. Constant pressure accumulator with PCM [73].
16 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
Higher steam release rates will produce wet steam, prevented
by ensuring a large enough water surface area, and hence again
determining the accumulator size.
The evaporation capacity must be sucient. This is a function
of the pressure at which the water is stored when fully charged
(the boiler pressure) and the minimum pressure at which the
accumulator will operate at the end of the discharge period (the
accumulator design pressure). More ash steam is produced when
the differential between these two pressures increases.
The accumulator design pressure must be higher than the down-
stream distribution pressure, to create a pressure differential and
allow for the required ow.
The steam inlet pipes must feed well below the water surface
level to ensure the maximum possible liquid head above them.
The steam injectors should be installed at a slight angle to avoid
erosion of the vessel, and discharge at a very high velocity to
provide turbulence and mixing of the stored water mass.
Both the design of the inlet pipe and the manifold system, to-
gether with the location of the steam injectors, must provide even
injection of steam throughout the length of the accumulator re-
gardless of the actual steam ow-rate.
The steam injector capacity will reduce as the pressure in the
vessel increases because the differential pressure between the in-
jected steam and the vessel pressure is reduced. At very low ow
rates, the steam feed will not be uniformly distributed and will tend
to issue from the injectors closest to the steam inlet pipe.
Supercritical water is water that exists above its thermody-
namic critical point (374 C, 22.1 MPa). Using water as the working
uid at supercritical conditions, the respective eciencies can be
raised from about 39% for subcritical operation to about 45% using
current technology. The supercritical uid is a single non-condensable
phase which has physical properties that are intermediary between
its liquid phase and gaseous phase, with a density typically being
closer to that of water but with transport properties being gas-
like. Supercritical water at low density behaves like a non-polar
solvent unlike its liquid counterpart due to loss of hydrogen bonding
[82]. It has high thermal eciency, as it combines high tempera-
ture, high pressure and its existence as a single phase which benets
from relative simplicity in system design due to not having to handle
multi-phase ow.
The problems associated with supercritical steam is that the pres-
ence of ionic species can induce corrosion. At low ionic density (low
pressure), corrosion is typically more severe in subcritical than
supercritical water, due to the fact that ionic species are insoluble
in supercritical water at low density. This is the reason that com-
ponents and/or piping used to preheat or cool uids in a supercritical
Fig. 17. Potential applications for TES using E-PCMs.
water system are more susceptible to corrosion than the reactor itself
[8385]. At high ionic density (high pressure), most salts will be
soluble. Therefore, pure distilled water should be used in supercritical
water/steam accumulator to eliminate salt-induced corrosion.
Pressure vessels are designed to have a thickness proportional
to the radius and operating pressure of the tank, and inversely pro-
portional to the maximum allowed normal stress of the particular
material used in the walls of the container. Theoretically, any ma-
terial with good tensile properties that is chemically stable in the
chosen application could be employed. No matter the shape, the
minimum mass of a pressure vessel scales with the pressure and
volume it contains and is inversely proportional to the strength to
weight ratio of the construction material (minimum mass de-
creases as strength increases). Godall [86] has suggested that the
best ratio of diameter to total length of a steam accumulator should
be between 1:4 and 1:6 in order to obtain optimal condition in terms
of thermal storage capacity and material cost.
3.4.3. Moderate to high temperature PCMs
Within the high temperature thermal energy storage option using
PCMs, various alternatives have been investigated, as illustrated in
Fig. 17. Each of the parallel topics is subsequently dealt with. To il-
lustrate the practical use of the results reported in previous sections,
design approaches for thermal energy storage installations are further
elaborated. This will be demonstrated by describing potential ap-
plications of high temperature thermal energy storage in three typical
case studies:
The encapsulated PCM is mixed within a high temperature con-
crete for passive heat energy storage.
The encapsulated PCM is mixed in a liquid HTF.
The encapsulated PCM is used in a solidgas conveying system
as heat carrier.
In these case studies, described in Table 10, the coated PCM in-
teracts with a different heat carrier, being solid, liquid or gas.
3.4.4. The growing importance of lithium salts
3.4.4.1. Lithium for sensible heat storage. Lithium has been used to
improve the properties of molten salts used in CSP by adding it to
their composition. Fig. 18 shows new compositions with lithium and
in comparison with the commercial solar salt (34 wt% KNO3 and
66 wt% NaNO3): the salts have nearly similar heat capacities, there-
fore they have similar energy density. The benets of adding lithium
to solar salts are noticed in the melting temperature of the com-
position, which is lowered. The range of working temperature is
extended, whilst improving the thermal stability [91]. Due to the
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 17

Table 10
Case studies of high temperature thermal storage applications.

Case study Description

High temperature A three component thermal storage system, combining sensible and latent heat storage was tested by Laing et al. [44] for use in solar
concrete as passive power plants. Investigations and experiments supported the hypothesis that high temperature concrete could serve as sensible heat
storage media storage up to 500 C. Within the limitations of the solid concrete structure, the incorporation of phase-change materials is benecial since
capable of absorbing a high quantity of energy (sensible and latent heat), while keeping the temperature of the matrix close to the
melting point of the phase change.
This use of encapsulated PCM mixed with concrete was further demonstrated for applications at ambient temperatures [16,24,87],
achieving high energy savings in cooling power.
Micro-encapsulated A micro-encapsulated PCM slurry is a suspension where the PCM is dispersed at 1020 wt % without signicantly altering the physical
PCM in slurry properties of the liquid (density, viscosity) [88]: PCM is microencapsulated using a polymeric capsule and dispersed in water. A slurry
with a 10 wt % concentration of paran was conducted by Delgado et al. [89] to investigate the effectiveness of its properties as a thermal
storage material and as a heat transfer uid. Results demonstrated an increase of approximately 25% in the convective heat transfer
coecient when compared to water, as the result of improving the overall heat capacity of the material through applying the latent heat
properties of the PCM.
The feasibility of using ionic liquids as liquid thermal storage media and heat transfer uids in a solar thermal power plant has been the
subject of additional investigations. According to Bridges et al. [90], nanoparticle enhanced ionic liquids have been shown to increase the
heat capacity of the ionic liquid without adverse secondary effects on the ionic liquids thermal stability. The ionic liquid physical
properties were shown to be only affected in cases of high nanoparticle loading. Those similarities between high and low temperature
slurries properties tend to indicate that a high temperature equivalent of an encapsulated PCM slurry is possible.
Encapsulated PCM in Heat storage units of new concentrated solar plants could benet from a heat transfer uid which incorporates encapsulated particles of
TES heat transfer PCM. The heat transfer capacities would be improved by the high conductivity and the high energy density of the PCM. In comparison
uids with a conventional thermal uid, e.g. Santotherm 350, the specic heat will be increased by a fraction of 3 whilst the energy density will
increase by about 5.
The main issues for developing salt-PCM modications would be the loss of uidity of the high temperature slurry. Further studies have
to prove if the addition of particles increases the thermal properties of the HTF without affecting the viscosity and associated pumping
power needed. Existing CSPs would then see their global eciency improved by the higher energy density of the HTF.
Fluidized bed Storage High heat transfer coecients can be achieved between a uidized bed of coated PCM particles and a heat exchanging surface [7]: heat
concept can be captured by the particulate gassolid owing suspension. Further research is required, since it was impossible to coat the PCM
particles by a non-porous metallic layer (chemical vapour deposition or electrolysis)

lower temperatures, it is expected that operational and mainte-
nance costs of CSPs are reduced, therefore savings should be obtained
[91]. In terms of costs, the U.S. Department of Energy performed a
comparison between salts used for CSP, including solar salt (40%
KNO3 and 60% NaNO3), and the cost of storing energy with each of
them in a two tank system [92].
Results are shown in Table 11: despite the higher price of the
salts with lithium nitrate, a lower cost per stored kWth is achieved,
which can be attributed to the higher energy densities and the
savings in heat tracing to keep the salts molten.
3.4.4.2. Lithium compounds as phase change materials. PCMs in-
cluding lithium have been a potential for building applications and
for high temperature TES. There are different lithium compounds
to be used in building applications which are compared to the well-
known octadecane and Glauber salt in Fig. 19. It can be seen that

Fig. 18. Heat capacity and melting temperature comparison of lithium compositions used in solar power plants (adapted from Ref. 91).
18 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Table 11
Costs of candidate salts for CSP [92].

Composition (wt %) Melting Heat capacity Energy density Salt price Two tank system cost/stored
point (C) (kJ/kg K) (MJ/m3) (US$/kg) energy (US$/kWhth)

40% KNO3 + 60% NaNO3 222 1.539 756 1.080 31.21

KNO3 + LiNO3 + NaNO3 117 2.32 1524 2.206 14.66
KNO3 + NaNO2 + NaNO3 99 1.462 1080 1.266 15.87
KNO3 + NaNO2 + LiNO2 + NaNO3 79 1.505 1073 1.928 19.11
KNO3 + LiNO3 + NaNO3 + MgK 101 1.579 1181 1.537 16.15
LiNO3 + NaNO2 + NaNO3 + KNO3 99 1.557 1114 1.809 18.27
LiNO3 + NaNO2 + NaNO3 + KNO2 + KNO3 95.7 1.546 1110 1.797 18.23

the melting temperatures of some lithium compounds are around
10 C, 3035 C and 7080 C which are useful for cooling, compa-
rable with Glauber salt and useful for domestic hot water respectively
[91]. Regarding their heat of fusion, lithium composites at 100 kJ/
kg are considered the lowest to be commercially feasible and those
with a latent heat in excess of 300 kJ/kg are considered of high po-
tential towards different applications [9194]. Lithium compositions
for high temperature TES are illustrated in Fig. 20.
The advantage of using Li-based PCMs is illustrated in Table 12,
where important properties are compared with a common NaNO3/
KNO3 molten salt.
Table 12 shows that a lithium salt mixture has the advantage of
a lower point of fusion and twice the sensible heat compared with
the current solar salt. The density values are quite similar, but the
price of the Li mixture is more than two times that of solar salt. Re-
garding physical properties as point of fusion, sensible heat and
density, a reduction in the size of the molten salt tanks, a reduc-
tion of heat tracing parasitics and a reduced amount of material can
be obtained. These improvements will reduce the nal prices of in-
vestment and cost of energy.

3.4.4.3. Lithium market. As far as the availability of the metal is con-

Fig. 19. Lithium compositions used in building applications (adapted from Ref. 91).
cerned, the worldwide reserves of lithium are estimated in 36.72
million tonnes, with Bolivia, Chile, Argentina and China having es-
timated reserves of 8.9, 8.04, 7.09 and 5.15 million tonnes
respectively. Other countries in Europe, Australia, Africa, Russia and
Canada represent reserves of 0.71.5 million tonnes each. As illus-
trated in Table 12, the cost of Li salts is about 23 times the cost
of common alkali-nitrates.

3.5. Mechanisms to improve phase change material applications

3.5.1. General principles

Fig. 20. Potential lithium compositions for solar energy applications (adapted from
Refs. 9194).
To improve the eciency of the charging and discharging pro-
cesses of PCMs, the most relevant parameter to be studied is their
thermal conductivity. In order to increase the thermal conductiv-
ity of PCMs, several heat transfer enhancement techniques have been
studied [14,27,34], such as the use of metal carrier structures made
of steel or stainless steel; a dispersion of high conductivity mate-
rial i.e. copper, silver or aluminium particles, within the PCM; the
impregnation of high conductivity porous materials, either as a metal
foam (copper, steel or aluminium) or as porous material like graph-
ite; the use of high conductivity, low density materials such as carbon
bres and paran composites; and the micro-encapsulation of PCMs
using graphite [48], polymers, or the nickel lm coating of PCM
copper spheres [95]. The general review of these techniques was
undertaken by Agyenim et al. [96], and the encapsulation of PCM
in shells was determined as being one of the most promising and
suitable approaches at the present stage of the investigations. The
reasons behind this selection are the result of the detailed assess-
ment, which summarizes as given below.

Table 12
Comparison of Li-based PCM and solar salt [93,94].

Mixture Point of fusion (C) Thermal stability (C) Sensible heat (kJ/kg K) Density (kg/m3) Price (2014) (US$/tonne)

Solar salt (60% NaNO3 + 40%KNO3) 222 588.51 1.54 2192 716.9
25.92%LiNO3 + 20.01%NaNO3 + 54.07%KNO3 117 Not dened 2.32 1720 1684.4
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
3.5.2. Review of applied enhancement methods
3.5.2.1. Fins. Fins are being used in order to increase the heat trans-
fer surface area with the storage media [97], therefore enhancing
the rate at which thermal energy is distributed within it. They can
be axial or radial and are usually attached to tubes [98]. Depend-
ing on their shape and their conguration inside the storage tank
with the storage media, this mechanism has different results. For
example, according to Zhang and Faghri [99], for internally nned
tubes, the melting fraction volume can be signicantly increased
by adding more ns and increasing their thickness and height.
Castell et al. [98] experimented adding longitudinal graphite ns
to PCMs (sodium acetate trihydrate whose melting point is 58 C)
and tested if these ns could decrease the solidication time of PCMs.
They concluded that despite ns altered natural convection within
the PCM and they did not enhance the heat transfer coecient of
the material, it was possible to decrease the solidication time of
the PCM. Two cases where analysed: PCMs with small ns and PCM
with longer ns. Both were compared with a base case module which
had no ns. In the rst case, a lower temperature difference was
necessary to achieve the same heat transfer coecient as with no
ns. Therefore, the needed time to solidify the PCM decreased. In
the second case, a lower heat transfer coecient was obtained in
comparison to the PCM with no ns. This was attributed to the width
of the ns that should be interfering with natural convection of the
PCM. Despite that the heat transfer coecient decreased, it is im-
portant to notice that the solidication time of the PCM was reduced.
Lamberg [100] analysed the two phase solidication problem in
a low temperature nned PCM storage system (it used aluminium
ns and paran as PCM) by using an approximate analytical model.
It is important to note that the geometry of the storage module had
a big inuence in the accuracy of the analytical model. Hence,
through this analysis, it was concluded that the solidication was
dominated by conduction while natural convection existed only
during the beginning of the solidication process. In terms of com-
parison, the natural convection is negligible compared to conduction.
Erek et al. [97] numerically and experimentally investigated latent
heat thermal energy storage with phase change around a radially
nned tube. The results showed that the stored energy increases
with increasing n radius and decreasing n space.
Velraj et al. [101] investigated the solidication of PCM in a cy-
lindrical vertical tube with radial internal ns arrangement. It was
concluded that this conguration, which forms a V-shaped enclo-
sure for the PCM, gives maximum benet to the n arrangement.
With an imposed constant heat ux as a boundary condition,
Talati et al. [102], and Mostaffa et al. [103] studied solidication of
a nned rectangular PCM storage analytically and numerically, with
the salt hydrate ClimSel C23 as the PCM and aluminium for the ns.
Results showed that when the length of n to the height of storage
ratio is smaller than unity, the PCM solidied more quickly in the
storage.
Within the framework of the DISTOR project [104], a prototype
of high temperature thermal energy storage module with graph-
ite ns was tested: to increase the transfer area, expanded graphite
was adhered to tubes through which owed a bi-phase medium.
It was concluded that for the same amount of PCM, area and metal
material, heat transfer was enhanced by using ns [104].
3.5.2.2. Fluidized bed. The term uidized bed is unavoidably con-
nected to the term particulate solid material [105110]. The
geometrical, physical and aerodynamical properties of the par-
ticles affect the onset of uidization, the characteristics, behaviour
and the main parameters of uidized beds. The most important solid
properties are particle density, diameter, and porosity.
Recently, this technology has been used to store energy through
PCM. Izquierdo-Barrientos et al. [111] studied the performance of
an air-uidized bed of micro-encapsulated PCM as a thermal storage
19
system and compared it with a uidized bed of sand and a xed
bed with sand and PCM. They concluded that PCM is an alterna-
tive material that can be used to increase the eciency of storing
thermal energy in the form of latent heat in uidized bed systems.
In fact, under the experimental conditions tested in this work, greater
charging eciencies are observed for the PCM than with just sand
in xed and uidized beds. Furthermore, when the charging e-
ciency is stabilized, the PCM shows similar results for both beds.
The PCM also presents a better recovery eciency for both beds after
the solidication time. Moreover, the inuence of the bed height
was tested. This study showed that it takes longer to achieve a certain
temperature when more mass is used, but higher eciency values
are obtained. The analysis of the inuence of the ow rate showed
that a higher ow rate allows a faster achievement of the set tem-
perature. At the end of the charging process, similar eciencies are
measured at the different studied ow rates. The cycling study re-
vealed that the PCM suffers attrition during the uidization process,
although no loss of PCM is observed under the experimental con-
ditions tested in this work (75 h of continuous operation with 15
charging/discharging cycles).
Piti et al. [7] introduced PCMs and their potential application
in high temperature energy capture and storage, using a circulat-
ing uidized bed (CFB) as transfer/storage mode. Thermal
considerations determined the optimum size range for the applied
particles (<400 m) when convection heat transfer dominates. In
the design of a CFB heat collector, the heat transfer coecient
between the riser wall and the owing suspension is an impor-
tant design parameter to determine the required heat exchange
surface area.
Peng et al. [112] predicted the thermal behaviour of the system
with a concentric-dispersion model for a packed bed latent heat
thermal energy storage using PCM capsules and a molten salt HTF.
The radial heat transfer and wall heat losses were considered. The
effects of PCM capsule diameter, uid inlet velocity and storage tank
height on the temperature proles and eciency of a packed bed
were investigated for the charging process. The heat transfer between
PCM capsules and molten salt was shown to be signicantly inu-
enced by the capsule diameter, with ner sizes shortening the
effective charging time, thus increases the charging eciency. The
numerical concentric-dispersion model established by the authors
enables a fair prediction of the thermal behaviour in a packed bed
latent heat thermal energy storage. The uid inlet velocity has a
smaller inuence on the charging process, whereas a higher thermal
storage tank increases the charging eciency.
3.5.2.3. Microencapsulation. Microencapsulation is dened as a
process in which tiny particles or droplets are surrounded by a
coating, or embedded in a homogeneous or heterogeneous matrix,
to produce small capsules with many useful properties [113116].
Unless the matrix encapsulating the PCM has a high thermal con-
ductivity, the microencapsulation system suffers from low heat
transfer rate. The rigidity of the matrix prevents convection cur-
rents and all heat transfer needs to occur by conduction. This can
reduce seriously the heat transfer rates, especially in the charging
process. The most important parameters in microencapsulation are
the thickness of the shell; the geometry of the encapsulation; and
the encapsulation size [117].
Microencapsulation of phase change materials refers to a tech-
nique in which a large number of small PCM particles are contained
within a sealed, continuous matrix which ranges in size from less
than 1 mm to more than 300 mm [118]. Nowadays the cost of the
microencapsulation is high compared to other storage methods, and
it is used only in thermal control applications [116]. Microencap-
sulation of PCMs is an effective way of enhancing their thermal
conductivity and preventing possible interaction with the surround-
ings and leakage during the melting process. This method is classied
20 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
into three categories: physical, physico-chemical, and chemical
methods.
Li [119] studied the improvement in thermal conductivity of PCMs
(paran) with nano-graphite. It was demonstrated that nano-
graphite can be prepared easily from exfoliated graphite and
expanded graphite. Moreover, the price of nano-graphite is low, while
the improvement effect of nano-graphite on the thermal conduc-
tivity is signicant. Results showed that the thermal conductivity
of the composite PCMs with 1% and 10% nano-graphite is 2.89 times
and 7.41 times of that of paran, respectively. The phase change
temperature of the nano-graphite/paran composite PCM was
slightly less than that of pure paran. As the nano-graphite content
increased, the thermal conductivity of the nano-graphite/paran
composite increased while the latent heat gradually decreased.
Parrado et al. [51] conducted a numerical study of a deform-
able spherical copper shell containing solar salt at high temperatures.
The thermo-mechanical analysis by Comsol Multiphysics was per-
formed during charging and discharging, and showed that the
volume expansion and pressure rise are below the acceptable me-
chanical properties of copper and are not expected to crack the shell.
3.5.2.4. Macroencapsulation. A signicant quantity of heat storage
media is encapsulated, where the mass per unit may range from
grams to kilograms [40]: by careful selection of the capsule geom-
etry and the capsule material, the macroencapsulation can be used
for a wide variety of energy storage needs. The shape of
macrocapsules varies from rectangular panels to spheres or pouches
without a dened shape. The macrocapsule shape selection will
depend on the needs and design of each intended application. Ac-
cording to Regin et al. [40] the most cost-effective containers are
plastic bottles (high density and low density polyethylene bottles,
polypropylene bottles), tin-plated metal cans and mild steel cans.
In the DISTOR project using PCM and expanded graphite, a
macroencapsulation of phase change materials was tested [104]. The
extra volume required in this process, based on the thermal ex-
pansion of PCM due to temperature changes, drastically reduced the
volumetric storage capacity [101]. This design was compared with
the graphite n concept, mentioned above. Results showed that for
the same amount of PCM, the same area and material for heat ex-
change, heat transfer is more effective using ns.
Cascades of macroencapsulated PCMs, as will be further de-
tailed in Section 3.7, can store 50% more energy per unit volume
than the conventional system using two tanks.
3.5.2.5. Metal matrix. A metal matrix, such as a porous matrix, has
been used to enhance the heat transfer rate in TES. A porous matrix
is characterized by two parameters, i.e. the porosity and the cell size
which refer respectively to the percentage of volume that will be
occupied by PCM and the size of the pore expressed in pores per
inch (PPI). For the rst parameter, Mesalhy et al. [120] indicated that
the performance improvement depends on both the porosity of the
matrix and its conductivity. Low values of porosity lead to a higher
effective thermal conductivity, resulting in an increased perfor-
mance enhancement. However, a decrease in the porosity has the
opposite effect because the low porosity of the matrix hampers the
movement of the liquid PCM and the natural convection. For the
second parameter, Elgafy et al. [121] showed how the mean pore
size of the additives could be of critical importance on the system
performance. If it is too small, the PCM molecular motion will be
hindered, thereby it will be very dicult to impregnate the porous
media with the PCM, which will adversely affect the latent heat
storage capacity.
Metals with the highest thermal conductivities are silver, copper,
gold and aluminium [122], which turns them into a desirable option
for metal matrix production to latent heat thermal energy storage
systems. Their characteristics are given in Table 13. Due to their lower
Table 13
Metals with high thermal conductivities [42].
Metal Thermal Density Specic Melting
conductivity (kg/m3) heat point (C)
(W/mK) (J/kgK)
Silver 429 10490 237 961.9
Copper 387.6 8978 381 1084.6 C
Gold 401 19300 129 1064.4
Aluminium 237 2800 910 660.2 C
Nickel 90.3 8910 440 1450 C
prices in comparison with gold and silver, copper and aluminium
are most commonly used. Also, nickel has been studied, as re-
ported in Table 14.
3.5.2.6. Graphite composites. Another alternative to improve thermal
energy storage materials is producing a new compound with the
storage media. To improve the PCM thermal conductivity, the use
of graphite to enhance the heat transfer rate has been investi-
gated by numerous researchers.
Mehling et al. [130] demonstrated that PCM/graphite compos-
ites show a signicant improvement in the heat transfer within latent
heat storage materials. They studied two PCMs: the eutectic salt
MgNO3/LiNO36H2O and paran RT50. In fact, the thermal conduc-
tivity in the composite material is up to 100 times higher than in
the pure PCM, leading to the phase front moving between 10 and
30 times faster.
Py et al. [131] experimented a new supported PCM made of par-
an impregnated by capillary forces in compressed expanded
natural graphite (CENG) matrix. High loads of paran were ob-
tained: from 65% to 95% weight depending upon the bulk graphite
matrix density. Composite PCM/CENG thermal conductivities were
found to be equivalent to those of the sole graphite matrix from 4
to 70 W/mK instead of the 0.24 W/mK of the pure paran. Thermal
power and capacity of the composite are theoretically compared to
those of conventional systems in the case of two usual external ge-
ometries: tubes and spherical hollow nodules. The compressed
expanded natural graphite induced a decrease in overall solidi-
cation time and a stabilization of the thermal storage power. An
optimization procedure of the composite composition was pro-
posed according to the antagonistic behaviours of the thermal power
and the thermal capacity in respect to the compressed expanded
natural graphite content. Within the usual external heat transfer
coecient range, the estimated compressed expanded natural graph-
ite matrix optimized densities fell within the practicable range.
Cabeza et al. [132] investigated three heat enhancement methods
to improve the low conductivity of the PCM water/ice. These were
the addition of stainless steel pieces, copper pieces, and a graph-
ite matrix impregnated with the PCM. The use of graphite composite
allows an even larger increase in the heat transfer than with copper.
The heat ux is about four times larger on heating and three times
larger on cooling as compared to using pure ice.
Lopez et al. [49] studied the behaviour of graphite with closed
and open pores, made with salt as a PCM. A pore-thermo-elastic
analysis was performed from which was concluded that varia-
tions in expansion volumes of the salt due to fusion were limited
by the matrix of graphite, which causes an increase of pressure in
the pores. Higher pressure causes a progressive increase in the
melting temperature of the salt and a gradual decrease of the latent
heat. This implies that a signicant part of the energy supplied to
the material is used to heat it (sensible heat instead of latent heat).
Zhao et al. [133] studied the addition of the highly thermal con-
ductive ENG-TSA (expanded natural graphite treated with sulphuric
acid) to solar salt (60% sodium nitrate and 40% potassium nitrate),
in order to enhance the thermal conductivity of nitrates when used
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 21

Table 14
Summary of metal matrix studies.

Author Studied material Main conclusions

Copper studies Zhao et al. [123] Copper foam embedded in paran
RT58 (low temperature storage)
Vadwala [124] Paran wax and open cell copper
foam
(low temperature storage)
Li et al. [125] Copper matrix and sodium nitrate
(high temperature storage)
Aluminium Bauer and Wirtz Composite developed by plate-like
studies [126] thermally conductive aluminium
containing pentaglycerine as PCM
Tong et al. [127] Water/aluminium foam composite
(low temperature storage)
Nickel studies Xiao et al. [128] Paran/nickel foam composite
(low temperature storage)
The addition of metal foams with a cell size of 10 PPI and 95% porosity can increase
the overall heat transfer rate by 310 times. Also, smaller porosities and pore densities
produced better heat transfer performance, depending on the metal foam parameters.
The equivalent thermal conductivity of a foam-wax composite was found to be 3.8 W/
mK, which was 18 times higher than that of pure paran wax, i.e. 0.21 W/mK.
Moreover, copper foam reduced the time required to melt approximately the same
amount of wax to 36% of that without the use of metal foam.
Heat transfer coecient of the thermal energy storage with copper matrix can be
signicantly increased up to 28.1 times by heat conduction when the PCM is in the
solid phase, and up to 3.1 times by the combination of natural convection and heat
conduction when the PCM is in the liquid phase. Hence, both the melting and
solidication times are substantially shortened. In addition, it was concluded that
porosity plays a more important role to enhancing the heat transfer eciency in
thermal energy storage system than pore density.
An increase in aluminium foam metal fraction results in an increase in the effective
thermal conductivity of the composite because only about 2% of the heat ow is
through the PCM, and the interfacial bond resistance decreases due to increased
contact area. The trade-off is a reduced charging time versus a reduced volumetric
latent heat storage due to the increasing aluminium foam metal fraction.
An increase in the heat transfer rate during melting and freezing of the PCM (water).
It was demonstrated that thermal conductivity of the paran/nickel foam composite
was nearly three times larger than that of pure paran, 0.4 and 1.2 W/mK
respectively.

Fiedler et al. [129] compared a copper and aluminium metal matrix and found that copper matrices achieve approximately an increase of 80% in the effective thermal con-
ductivity compared to aluminium matrices. It was found that a further increase is possible using diamond-coated copper matrices. Xiao et al. [128] and Fiedler et al. [129]
concluded that copper is an acceptable matrix material.

as PCM. Different samples of this composite were evaluated. The
results showed that the highest effective thermal conductivity of
composite PCM was 50.78 W/mK, which was 110 times higher than
that of salt powder. Also, it was concluded that the addition of ENG-
TSA causes a decrease in latent heat, but there is no remarkable
variation in phase change temperature.
3.6. Nanoparticle-encapsulated PCMs
ductivity than aluminium, carbon/copper nanoparticle dispersed
paran composite in the same condition showed in Fig. 21.
Additional recent work on nanoparticle PCMs is summarized in
Table 15. The review by Kibria et al. [134] illustrates how dispers-
ing nanoparticles in PCMs enhance the overall heat capacity, whilst
also demonstrating the effects upon the PCM viscosity and
subcooling.
Major areas of concern remain to be claried.

As illustrated in Table 15, most of the research work in the lit-
erature was conducted on improving the thermal properties of
organic PCMs such as parans and fatty acids. Paran draws a great
attraction to researchers for its desirable characteristics i.e. good heat
storage density, melting or solidication compatibly with little or
no sub-cooling, being inert to most common chemical reagents and
of low cost. The main drawback of these PCMs is their low thermal
conductivity, hence the growing interest in dispersing high con-
ductive nanoparticles within the PCMs [134]. Wu et al. [146] tested
different metal nanoparticles in order to enhance the thermal con-
ductivity of paran wax. The results concluded that Cu/paran
composite had a better cooling/heating rate i.e., better thermal con-
Although the thermal conductivity of the PCM increases with
increase in concentration of nanoparticles, surface modica-
tion of the nanoparticles might be required. The diameter of the
nanoparticles was shown to have a limited effect only. Mostly
carbon nano-tubes and bres, graphite and graphene have been
investigated.
Metal and metal oxide nanoparticles show an enhancement in
thermal conductivity that varies with their shape, size and con-
centrations. The enhancement is also temperature dependent.
The presence of nanoparticles reduces the solidication and
melting time of the composition. Too high a concentration of
nanoparticles could negatively modify other properties of PCM.

Table 15
Recent relevant publications concerning nanoparticle-PCM (>2011).

Authors Topics

Kibria et al. [134] Review of thermophysical properties of nanoparticle-dispersed PCM

Warzoha et al. [135] Temperature-dependent thermal properties of a paran PCM embedded with herringbone style graphite nanobres
Sahan et al. [136] Improving thermal conductivity PCM: a study of paran nanomagnetite composites
Yu et al. 2014 [137] Inuence of nanoparticles and graphite foam on the supercooling of acetamide
Jourabian et al. [138] The expedited melting of PCM through dispersion of nanoparticles in the thermal storage unit
Motahar et al. [139] A novel PCM containing mesoporous silica nanoparticles for thermal storage: thermal conductivity and viscosity
Cingarapu et al. [140] Nanouids with encapsulated tin nanoparticles for advanced heat transfer and thermal energy storage
Sciacovelli et al. [141] Melting of PCM in a thermal energy storage unit: Numerical investigation and effect of nanoparticle enhancement
Kashani et al. [142] Numerical study of solidication of a nano-enhanced PCM in a thermal storage system
Khodadadi et al. [143] Thermal conductivity enhancement of nanostructure-based colloidal suspension utilized as PCM for thermal energy storage
Abolghasemi et al. [144] Heat transfer enhancement of a thermal storage unit consisting of a PCM and nanoparticles
Guo and Wang [145] Numerical investigation of nanoparticles-enhanced high temperature PCM for solar energy storage
22 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
Fig. 21. Heating (a) and cooling (b) curves of PCMs with different metal nanoparticles
[146]. Charging cycle (left); discharging cycle (right).
Nanoparticles generally reduce the latent heat capacity of PCM.
The determination of the optimum addition of nanoparticles has
to be explored.
Predicting the thermal conductivity and latent heat capacity of
a nanoparticle-PCM is very dicult.
Nanoparticles present affect the nucleation and decrease the
super-cooling of the PCM.
The presence of nanoparticles will moreover signicantly in-
crease the viscosity of the PCM-melt.
As far as now, mostly organic nanoparticles have been used. It
is however not excluded to use mineral or metal nanoparticles in
suspension with the HTF. The segregation of these denser par-
ticles from the lighter HTF needs however to be cautiously
investigated.
3.7. Cascades of PCM systems
The use of multiple component PCMs in PCM-TES systems has
been shown as an effective way for the enhancement in their per-
formances [134]. It means that in a system more than one PCM are
used. These PCMs have different melting points. The differences
between the temperatures of HTF and the melting points of the PCMs
have signicant effects on the heat transfer rate of the unit and thus
enhance the performance of the system during the charging
(melting) and discharging (solidication) process. In the case of a
single PCM, this temperature difference decreases in the ow di-
rection of HTF, thus leading to a decrease in heat transfer rate and
poor performance of the unit. If multiple PCMs with different melting
temperatures are packed in the unit, in decreasing order of their
melting points, nearly a constant temperature difference can be
maintained during the melting process, even though the temper-
ature of HTF decreases in the ow direction. This leads to an almost
constant heat ux to the PCMs. During discharging, if the ow di-
rection of the HTF is reversed, the PCMs remain in increasing order
of their melting points. A nearly constant heat ux from the PCM
to HTF is maintained [34,147149].
An example of experimental multiple PCM is shown by Kuravi
et al. [34] the shell side is loaded with different PCMs and synthet-
ic oil is allowed to ow through the inner tube, all PCMs being
arranged in the ow direction (axial direction) in contact with the
HTF.
In solar thermal applications, the PCMs are mostly kept in the
storage tank, which contains the HTF, used as the heat carrier
[150,151]. Since the PCM is surrounded by the HTF, there is no axial
temperature variation within the HTF. In such congurations, mul-
tiple PCMs should be arranged in the radial direction instead of in
the axial direction. With multiple PCM arrangement, it is possible
to have lowest melting point PCM at the centre and other PCMs ar-
ranged in the increasing order of their melting points from the centre
of the unit to the outer wall. This may lead to more or less simul-
taneous melting at all points, because a nearly constant temperature
difference can be maintained in the radial direction.
A similar work, with three coaxial cylindrical tubes to store three
PCMs and the entire unit being kept in water bath, has been re-
ported by Wang et al. [152,153]. The PCMs are arranged in a similar
way as discussed above. The experiments for the charging process
with different HTF temperatures indicate that the melting time in
three PCMs system is decreased by 3742%, as compared with that
of the single PCM unit, which has been attributed to the fact that
the three PCM unit terminates the melting process earlier by 15
25%. With multiple phase PCMs, it is possible to produce variable
distribution of phase change temperature [151,152]. In the systems
with constant temperature boundary conditions, the distribution
of a linear phase change temperature can be realized by properly
arranging those PCMs with different melting points. Simulation and
experimental results indicated that a 25% reduction in phase change
time can be achieved by using multiple phase PCMs. It is also found
that the eciency in the enhancement is dependent on the number
of the PCMs. For practical applications, the number of the PCMs
should be in the range of 510 [147,154].
Peir et al. [155] performed an experimental evaluation of the
advantages of using the multiple PCM conguration instead of the
single PCM conguration in thermal energy storage (TES) systems
at pilot plant to ll the gap of experimental and high scale studies
on this concept in the literature. For the evaluation were selected
two PCM with melting temperatures in the range of 150200 C due
to their high value of heat of fusion and compared: d-mannitol and
hydroquinone. Three congurations were evaluated: single PCM
with hydroquinone; single PCM with d-mannitol; and multiple
PCM with hydroquinone and d-mannitol. Results showed that the
multiple PCM conguration introduced an effectiveness enhance-
ment of 19.36% if compared with single PCM conguration as well
as a higher uniformity on the HTF temperature difference between
the inlet and outlet.
In order to improve cascade thermal energy storage, Tian et al.
[156] experimented adding metal foam to the PCM. Exergy e-
ciency of STES cannot be signicantly improved by CTES (15% to ca.
30%), nor by MF-CTES; exergy transfer rate of STES is increased by
CTES (up to 23%), and is further increased by MF-CTES (by 27 times).
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 23

Fig. 22. Schematics of the thermal energy storage by reversible chemical reactions.

4. Thermo-chemical heat storage a complete analysis. It is however important that an early guide as
to the viability of each system is obtained. The present section hence
4.1. Principles examines thermodynamic and engineering criteria to provide a sci-
entic comparison among the systems listed before [169,170]. The
One of the most novel approaches to the problem of storing analysis will moreover provide guidance as to the important as-
thermal energy from a solar collection eld, or from transforming sessment criteria of new candidate reaction systems.
excess renewable power (wind turbines, photovoltaics) into heat
[14], is the use of reversible chemical reactions. Unlike sensible or 4.2. Thermodynamic assessment
latent thermal energy storage, mostly limited in time due to heat
losses, chemical energy storage enables to bridge long duration 4.2.1. Gibbs free energy and work recovery
periods between supply and demand, hence making it particular- A reversible reaction system suitable for high temperature energy
ly suitable for a large scale electricity generation. In such a system, storage, and subsequent steam turbine or Stirling engine use for work
as outlined in Fig. 22, excess energy is converted into high grade recovery, should absorb heat at temperatures below 1000 C in order
heat at the absorber containing the reactant during the charging to lessen problems of material limitation and heat losses, and should
cycle: the high grade heat is converted into chemical energy by pro- furthermore be capable of delivering heat to the turbine inlet at typ-
moting an endothermic reaction whereby reaction products are ically 500600 C [168]. As initially presented by Williams and Carden
either stored for re-use in the reverse reaction, or exhausted. At the [171], the thermodynamic criteria for the preliminary screening of
time of energy demand by the central power generation plant or candidate reactions are assessed from available thermodynamic equi-
other users, the high grade heat is regenerated through the reverse librium data. Williams and Carden [171] proposed to use two
exothermic reaction cycle and is converted into electricity and low parameters for the initial screening of reversible reactions. The rst
grade heat (CHP concept) through e.g. a standard steam-driven turbo- is the equilibrium temperature, T*, i.e the temperature at which
alternator or through a heat carrier powered Stirling engine. neither forward nor reverse reactiions are thermodynamically fa-
A large number of reversible chemical reactions can be used for voured. T* is derived by applying the denition of Gibbs free energy
thermo-chemical energy storage, each within a given range of equi- to the condition of T*.
librium temperatures and heats of reaction [157163]. Table 16
selects some of the possible reaction pairs, within the target tem- G ( T*, P ) = 0 (7)
perature range of 300 C to 1000 C. Numerous low temperature
reversible reactions are not included. Clearly, the Mn2O3MnO pair where G is Gibbs free energy change for the reaction, i.e.
does not match the proposed temperature range, but can similar- G = H TS, in this case
ly equivalent to oxides, be used in the reaction of MnO and H2O to
H (T *, P )
produce H2 [164168]. T*= (8)
The selection of the optimum system for high temperature ap- S (T *, P )
plications is governed by both the overall thermodynamic
with H and S respectively the reaction enthalpy and entropy at
performance, and by chemical engineering criteria (kinetics, design
pressure P and temperature T*.
factors, etc.) which determine practicability and cost of operation.
The second screening parameter is the maximum work recov-
In a complete analysis, these factors are normally combined in terms
ery eciency, wmax , dened as the ratio of the maximum available
of global cost-effectiveness, expressed as the levelized cost of elec-
or ideal work to the actual work, and obtained from standard chem-
tricity (LCOE), and previously discussed in Section 1. At the present
ical thermodynamics [42].
stage of development, there is a lack of comparative data for such
The lost work is only zero, when a process is completely re-
versible. For irreversible process, the energy that becomes unavailable
for work is positive, and ~TsS. The engineering signicance is clear:
Table 16
the greater the irreversibility of a process, with a greater the rate
Possible reaction pairs [157163].
of entropy generation corresponding with a greater amount of energy
Reaction Teq (P = 1 atm) (C) Hr at Teq (kJ/kg)
that becomes unavailable for work.
Mg(OH)2 MgO + H2O 259 1396 The maximum work recovery eciency is hence:
MgCO3 MgO + CO2 303 1126
Ca(OH)2 CaO + H2O 479 1288 Ws G (Ts , P )
CaMg(CO3)2 MgO + CaO + 2CO2 490 868 wmax = = (9)
CaCO3 + H2O 573 137 W ideal H (Ts , P )
Ca(OH)2 + CO2
CaCO3 CaO + CO2 839 1703 where Ts represents the ambient or the sink temperature. The at-
2Co3O4 6CoO + O2 870 844 tainment of the maximum work implies that the endothermic heat
5Mn2O3 5Mn3O4 + O2 90 6 185 exchange operation must be reversible and complete, whilst also
Mn2O3 2MnO + 1/2 O2 1586 1237
the exothermic reaction must proceed reversibly. The reaction must
24 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Table 17
Screening parameters for a selection of reaction pairs.

Reaction Teq (P = 1 atm) (C) wmax (30120 C) )

Mg(OH)2 MgO + H2O 259 0.280.43

MgCO3 MgO + CO2 303 0.340.48
Ca(OH)2 CaO + H2O 479 0.470.60
CaMg(CO3)2 MgO + CaO + 2CO2 490 0.490.61
CaCO3 + H2O 573 0.550.64
Ca(OH)2 + CO2
CaCO3 CaO + CO2 839 0.640.73
2Co3O4 6CoO + O2 870 0.730.82
6Mn2O3 4Mn3O4 + O2 906 0.640.74

separate gas carrier is used to exhaust the gas/vapour reaction prod-

ucts. For Mn2O3 the temperature is a more outspoken function of
pressure for all pressures. The work eciency is rather low for car-
bonates and hydroxides. Oxides provide the higher value of work
Fig. 23. Equilibrium temperature versus pressure for selected reaction pairs. eciency in the selected range of sink temperatures.

4.3. Kinetics from thermogravimetric analysis

hence proceed so that chemical equilibrium is maintained at all tem-
peratures. All the reaction heat and all heat imbalances in the heat
exchanger must be converted into maximum work by use of a Carnot
cycle available throughout the temperature range. The advantage
of both T* and wmax is that they may be estimated to a fair approx-
imation for a given reaction by use of standard reference data alone.
4.2.2. Calculation results
From the calculations, Figs. 23 and 24 illustrate both screening
parameters, T* and wmax, for the selected reaction pairs. Fig. 23 plots
the equilibrium temperature versus the equilibrium pressure for the
different reactions in Table 17. The maximum work eciency versus
the sink temperature is plotted in Fig. 24.
To make a selection between the reactions, different aspects need
to be taken into consideration. Firstly, the temperature should pref-
erably be as close as possible to 500900 C. Secondly a higher
maximum work recovery eciency is preferred. But also the prac-
tical aspects need to be considered, because adding water vapour
is easier than collecting, storing and injecting carbon dioxide, whilst
O2 reactions are most favourable since air can be the reactant of the
reverse reaction. Oxides and hydroxides are thus favoured above
carbonates.
Fig. 24 shows that the temperature is not very dependent on the
pressure for most of the systems selected except at low partial pres-
sures of the gas/vapour phase, as commonly encountered when a
Thermogravimetry (TGA) measures the weight of the sample
every second, recording M0, Mt and M, at the initial measure-
ment time t0, time t and the time for completing the reaction t.
The conversion is calculated as
M 0 Mt
= (10)
M0 M
The reaction rate constant can be derived within the complete
time and/or temperature scale, as described in detail for other chem-
ical process by Brems et al. [172,173].
For example, typical TGA plots for Mg(OH)2 are illustrated in
Fig. 25. The decomposition rate is slightly dependent on the tem-
perature ramp, . Similar weight loss curves were obtained for all
reaction pairs tested.
Additional TGA plots are illustrated in Figs. 26 and 27 for the re-
versible reactions of Mn2O3 Mn3O4 and Co3O4 CoO, respectively.
Clearly, the oxidation reaction is slower than the reduction and
incomplete, assumed to be the results of compaction of the reduced
reaction compound, hence hampering diffusion of O2 during the
reverse reaction [170,174]. This is also illustrated in the BET ad-
sorption curves of Mn2O3 and Mn3O4 (Fig. 28) [170].
A fair approximation of the reaction kinetics is obtained by trans-
forming the TGA data, using the Arrhenius equation of the reaction
rate constant, k:

E
k = Aexp A (11)
RT

Plotting ln(k) versus 1/T provides EA/R as the slope of the linear
t, and ln(A) as ln(k) at 1/T = 0.
In this equation k is the reaction rate constant; A the pre-
exponential factor with same units as k; Ea the activation energy,
J/mol; R the universal gas constant, 8.314 J/mol K; and T the tem-
perature, K.
The rate of reaction, d/dt, is a function of the rate constant k
(dependent on temperature) and a function of conversion f() (in-
dependent on temperature). This results in:

d
= k (T ) f ( ) (12)
dt
Since k is described by the Arrhenius expression, the overall ex-
Fig. 24. Maximum work eciency vs. sink temperature for the selected reaction pairs. pression of the reaction rate is:
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 25

Fig. 25. TGA curves of Mg(OH)2 at different heating rates [168].

Fig. 26. Weight vs. time graph for charging and discharging cycles of Mn2O3 Mn3O4 at = 5 K/min [170].

d E linearity of the conversion rates expression versus 1/T for various

= k (T ) f ( ) = Aexp A f ( )
dt RT
Different methods were tested to t the experimental data, i.e.
(i) iso-conversional methods (Friedman, FlynnWallOzawa and
KissingerAkahiraSusone); and (ii) the mechanistic methods based
upon the CoatsRedfern approach, using different expressions for
f(). The iso-conversional methods do not provide the expected
(13) -values, resulting in widely different values of the activation
energy with degree of conversion. Within the different f() ap-
proach (nucleation, diffusion, reaction order), the expression as a
rst order reaction [f() = ln(1 )] provides a very fair t for all
of the reaction pairs tested. The activation energies and pre-
exponential factors (A) were hence determined by solving
Eq. 13 for the experimental values of , and using a rst order

Fig. 27. Weight vs. time graph for charging and discharging cycles of Co3O4 CoO at = 5 K/min [170].
26 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Fig. 28. The BET isotherm linear plots of Mn2O3 (a) and Mn3O4 (b), respectively.

expression. Values are given in Table 18, with A determined for
T > Teq + 100 K:
For the given reversible reactions, the maximum conversion is
given by Eq. 13 above and the reaction stoichiometry. To achieve
99% of the maximum conversion in a rst order reaction, Eq. 11 de-
termines the required reaction time above Teq below 300 s for all
of the reactions studied. This is however tentative only, pilot/large
scale experimentation is certainly required.
Recently, the German Aerospace Center (DLR) had been also
testing TCS heat storage materials at laboratory scale. In order to
investigate TCS at a larger scale, a test bench as well as a reactor
containing around 20 kg of reaction material was built and brought
into operation. Due to the good availability at low cost and its
favourable temperature range for CSP plants, previous work at DLR
focused on the reversible dissociation reaction of calcium hydrox-
ide [175177]:
CaO(s) + H2O(g) Ca (OH)2(s) + Hreaction
The dehydration was performed at 450 C and the rehydration
at about 550 C. Reversible conversion of 77% of the material with
no degradation effects after 10 cycles was determined.
Within their results they evidenced losses of chemically stored
thermal energy, but they concluded that these could be estimated
by implementing a loss term into their simulation and accounted
for around 2 kW during the rst minutes of the reaction. Addition-
ally, the good agreement between experimental and simulated results
showed that the reaction kinetics determined in micro-scale mea-
surements as well as the known bed properties offered a sucient
representation of the effective conditions within the bed. Anyhow,
research is still needed in order to develop TCS at an industrial level.
The selection of the optimum reaction system for high temper-
ature applications is governed by both the overall thermodynamic
performance, and by chemical engineering criteria (kinetics, design
factors, etc.) which determine practicability and cost of operation.
In a complete analysis, these factors are normally combined in terms
of global cost-effectiveness, expressed as the levelized cost of elec-
tricity (LCOE), as explained in Section 1. At the present stage of
development, there is a lack of comparative data for such a com-
plete analysis. It is however important that an early guide as to the
viability of each system is obtained, and this is provided by ther-
modynamic calculations and kinetic measurements.
From the previous and preliminary assessment, TCS certainly
offers a potential for medium to high temperature heat storage. The
reactions are fast and generally reversible (except Mn2O3 and Co3O4).
The reaction rate itself can be considered as moderate (300 s), and
is thereafter not a limiting factor. The heat penetration inside the
chemicals packed bed is governed by conduction, where the con-
duction approach of PCM storage, as developed in Section 3.3 (with
or without inserts), will determine the rate of heat penetration, and
hence the overall charging/discharging rate of the TCS system.
Additional pilot-scale experiments with Co3O4 and doped Co3O4
were examined by Tescari et al. [178] and the reaction concepts were
numerically modelled by commercial CFD software. Adding up to
10 wt% of non-reactive oxides to the Co3O4 is shown to prevent con-
solidation and enhance the reversibility of the Co3O4/CoO redox
reaction.
5. Containment of LHS and TCS materials
5.1. Preliminary selection concerns and criteria

5.1.1. Pressure upon phase change or chemical reaction

Table 18 Phase change materials and thermochemical storage (TCS)
First-order reaction results for different reactions.
systems certainly offer a signicant potential towards thermal energy
Reaction EA Teq (P = 1 A (s1) storage. Of course, the materials used need to be properly con-
(kJ/mol) atm) (K)
tained in e.g. a spherical capsule, heat exchanger tube or sandwich
Mg(OH)2 MgO + H2O 198 532 4.26 1016 plates, or in any other container that meets the required need of
MgCO3 MgO + CO2 52 576 86.3 being closed and able to withstand the high temperatures envis-
Ca(OH)2 CaO + H2O 131 752 2.09 106
CaMg(CO3)2 MgO + CaO + 2CO2 219 763 1.15 1012
aged, to resist possible chemical attacks by the contained materials
CaCO3 + H2O Ca(OH)2 + CO2 185 846 2.89 108 [179184], and to resist the pressure build up during the phase
CaCO3 CaO + CO2 228 1112 7.03 107 change (expansion for solid to liquid phase) or during the release
2Co3O4 6CoO + O2 189 1115 1.41 106 of gases and vapours during the thermochemical transformations.
6Mn2O3 5Mn3O4 + O2 254 1179 2.60 108
This containment is hereafter referred to as encapsulation.
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
In most PCM systems, the density of the liquid is 35% lower than
the density of the corresponding solid. In thermochemical storage,
gas (CO2, O2) or H2O-vapour are formed, again with a correspond-
ing pressure increase in the encapsulation. This pressure increase
can be signicantly higher than in PCMs, and only special mea-
sures such as a pressure relief valve on the encapsulation will be
able to cope with the P.
For PCM or TCS encapsulation, mostly tubular containment will
be applied. A spherical encapsulation, applied by e.g. vapour de-
position or electrolysis is also possible, but with additional problems
of very high pressure build-up on the capsule [51].
For PCM encapsulations, the pressure to be considered can be
calculated as e.g. for a capsule of volume V, lled for ~95% with
molten PCM. After cooling and solidication, the lling volume will
reduce (through contraction) by e.g. 3.1% for Sb2O3 and 4.4% for KNO3
respectively. The tube is then closed by welding a lid onto the cyl-
inder with material-appropriate welding rods, or by a ceramic lid
in the event of using ceramic tubes. The enclosed air will expand
upon heating and re-melting of the PCM.
Since the starting pressure is 1 bar, the internal pressure upon
melting will increase (reducing free volume and increasing tem-
perature) to P2.
For Sb2O3
P2 =
[0.05 + (1 0.05) % ]V (273.15 + 750) P 5.6P
(273.15 + 20)
1 1
0.05V
For KNO3
0.05 + 0.95 0.044 (273.15 + 400)
P2 = P1 4.2P1
0.05 (273.15 + 20)
The internal pressure on the encapsulation will hence increase
to approximately 5.3 bar for Sb2O3 and 4.2 bar for KNO3. These values
are slightly pessimistic since the expansion of the capsule should
be taken into account (V increases as T increases). In TCS-systems,
such pressure increases are not expected, since a tubular encapsu-
lation will be vented to the atmosphere at 1 bar during the
endothermic reaction, whilst pressures will be set at 13 bar during
the exothermic reverse reaction. The mechanical analysis of E-PCMs
is hence also valid for the reversible reaction TCS case.
In the event of a complete encapsulation such as in coated PCM
spheres [51], no air buffer is present, and the capsule material will
need to fully withstand the expansion of the phase change.
5.1.2. Design of the encapsulation
Calculations should focus upon normal design parameters for
straight pipes under internal pressure, to determine:
minimum required wall thickness
comparison with Youngs modulus (limit of plastic deforma-
tion) [51]
pressure to rupture [185]
These calculations are available on Internet calculators, provid-
ed by ASME and engineering companies.
5.1.2.1. Wall thickness. When the wall thickness to withstand an in-
ternal pressure is used as a design parameter, it requires that the
relief-valve setting may not be higher than the design pressure. For
piping, the chosen design pressure and temperature are the
maximum intended operating pressure and the coincident tem-
perature combination. This results in the maximum wall thickness.
The allowable stresses for possible capsule materials will be given
in Section 5.2.
27
The formula for the minimum required wall thickness tm is given
in Eq. 14. For straight metal pipes under internal pressure, this
formula is valid for D0/t ratios larger than 6. The factor Y in this
formula varies with material and temperature. It is used to take the
redistribution of circumferential stress into account: this redistri-
bution occurs under steady-state creep at high temperatures and
permits a slightly lesser thickness at this range.
PD 0
tm = +C (14)
2(SE + PY )
where
tm = minimum required wall thickness, m;
P = design pressure, Pa;
D0 = outside diameter of pipe, m;
C = sum of allowances for erosion, corrosion and any thread or
groove depth, m;
SE = allowable stress, Pa;
S = basic allowable stress for materials, excluding casting, joints,
or structural-grade quality factors, Pa;
E = quality factor as the product of several parameters. Typical
quality factors are: casting quality factor E0, joint quality factor
Ej, and structural-grade quality factor Es.
Y = coecient: 0.4 for ductile nonferrous materials, 0 for brittle
materials such as cast iron.
5.1.2.2. Pressure to deformation and to burst (rupture). The Barlow
formula [183] for the pressure to deformation or to burst is given
in Eq. 15:
2St w
P= (15)
D 0
with:
P = pressure, Pa;
S = material strength, Pa, where the time-dependent creep
strength at temperature is commonly used
tw = wall thickness, m;
D0 = outside pipe diameter, m;
= safety factor: calculations commonly use 1 as safety factor
5.1.2.3. Youngs modulus. Youngs modulus (tensile modulus or elastic
modulus) characterizes the material and is the stiffness of an elastic
material. The denition of Youngs modulus is the ratio between
the stress along an axis, and the strain along that axis in the range
of stress in which Hookes law is valid. The strain is dened as the
ratio of deformation over the initial length. Eq. 16 shows the def-
inition of Youngs modulus.
tensile stress F A0 FL 0
E= = = = (16)
extensional strain L L 0 A 0 L
When a material has high Youngs modulus, it is rigid. It is im-
portant to not confuse rigidity with strength, stiffness, hardness or
toughness. The difference between rigidity and stiffness is that ri-
gidity characterizes the material (intensive property), the stiffness
looks to products and construction (extensive property). The hard-
ness of a material is the relative resistance of its surface to the
penetration of a harder body. Toughness is the amount of energy
that can be absorbed by the material before fracturing when sub-
jected to strain. Youngs modulus indicates e.g. how much a sample
extends under tension or shortens under compression. Youngs
modulus is directly applicable to uni-axial stresses, these are tensile
28 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Table 19
Selection of encapsulating materials ( poor; + acceptable; ++ excellent) [186].

Required property AISI 316 AISI 304 AISI 310 AISI 321 Incoloy 825 Inconel 600 Ceramics

Resistance to corrosive environments + + + ++ ++ ++ ++

Resistance to Cl stress corrosion + ++ + ++
Resistance to reducing environments + + + ++ ++ ++ ++
Resistance to oxidizing environments ++ ++ ++ ++ ++ ++ ++
Resistance to NO3 + ++ + ++
Resistance to localized corrosion + ++ + ++
Application 500700 C + + ++ ++ ++ ++ ++
Application 700900 C + ++ ++ ++ ++
Application <900 C ++ ++ ++
Possible lid-welding ++ ++ ++ ++ ++ ++

It is clear that only AISI 310 (to 700 C), AISI 321, Incoloy 825, Inconel 600 and ceramic tubes (to 900 C) meet the required general properties at temperatures in excess of
600 C. In addition to these properties, several thermal and mechanical properties will determine the nal selection, together with the mechanical calculations of the ge-
ometry of the tubular encapsulation.

or compressive stresses in one direction without stress in the other INCONEL alloy 600 with a minimum nickel content of 72% is vir-
directions. tually immune to chloride-ion stress-corrosion cracking. Stress-
corrosion problems occur in INCONEL alloy 600 in high-temperature,
5.1.3. Preliminary selection criteria high-strength caustic alkalis. In this case the material should be
In view of the concerns of operating temperature, thermal cycling stress-relieved at 900 C/1 hr or 790 C/4 hr prior to use and oper-
between charging and discharging, possible oxidation by air or chem- ating stresses should be kept to a minimum. Fig. 31 shows the alloys
ical attacks by the PCM or TCS reactants, the selection is rather resistance to oxidation and scaling at 980 C, with other potential
limited and illustrated in Table 19. alloys included. This gure was obtained from weight-loss deter-
minations which indicate the ability of a material to retain a
5.2. Relevant properties of E-materials protective oxide coating under conditions of cyclic exposure to the
temperature. INCONEL alloy 600 offers the best long term oxida-
Since extensive properties are provided in standard handbooks tion resistance.
or websites of suppliers [187189], or by the American Iron and Steel When the machinability of INCONEL alloy 600 is compared with
Corporation [190], only essential properties are summarized here- stainless steel, it is slightly more machinable than Type 304 and
after; with an example more detailed description for INCONEL alloy slightly less machinable than Type 303. The joining of INCONEL alloy
600. 600 is done by conventional welding processes, using specic elec-
trodes and either shielded metal-arc, gas tungsten-arc and gas metal-
5.2.1. INCONEL alloy 600 [187] arc, or submerged-arc welding.
INCONEL alloy 600 is an engineering material with excellent
mechanical properties, high strength and good workability. It is used
5.2.2. Incoloy 825 [191]
for applications which require resistance to corrosion and heat. The
Similar to Inconel alloy 600, Incoloy alloy 825 is a NiFeCr
high nickel content makes INCONEL alloy 600 resistant to corro-
alloy with excellent thermal properties and corrosion resistance. The
sion by organic and inorganic compounds and to chloride-ion stress-
presence of molybdenum and copper enhances the resistance to a
corrosion cracking. Chromium makes the alloy resistant to sulphur
reducing atmosphere and to pitting and crevice corrosion.
compounds and to oxidizing conditions at high temperatures or in
corrosive solutions. INCONEL alloy 600 is used in a variety of ap-
plications with temperatures from cryogenic to above 1095 C.
Mechanical properties are listed in Table 20 for tubes. The yield
strength is between 170 and 345 MPa, the yield strength together
with elongations of 3555% [187] makes the alloy easy to fabricate.
Fig. 29 shows the high-temperature tensile properties of an an-
nealed (870 C/1 hr) hot-rolled plate. Data for other materials are
also included.
Although both INCONEL alloy 600 and Incoloy 825 are applica-
ble at temperatures beyond ~700 C, the mechanical properties
decrease signicantly with increasing temperature.
Fig. 30 presents the rupture life of hot-rolled material at tem-
peratures of 5381093 C, and demonstrates that the applicable stress
decreases with temperature and duration of use.

Table 20
Typical mechanical properties for tubular INCOLOY alloy 600 [187].

Tensile Yield strength Hardness,

strength (0.2% offset) Rockwell ()
(MPa) (MPa)

Hot-nished 520690 170345

Annealed hot-nished 520690 170345
Cold drawn, cold-nished 990 915 34 Rockwell C
Fig. 29. High-temperature tensile properties of annealed, 870 C/1 hr, hot-rolled plate
Cold drawn, annealed 550690 170345 88 max. Rockwell B
of various alloys (adapted from Refs. 186 and 187).
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 29

Table 21
Typical room-temperature tensile properties of INCOLOY alloy 825 [188].

Form and condition Tensile Yield strength Elongation

strength (0.2% offset) (%)
(MPa) (MPa)

Tubing, annealed 772 441 36

Tubing, cold drawn 1000 889 15
Plate, annealed 662 338 45
Sheet, annealed 758 421 39

The addition of chromium confers resistance to different oxi-

Fig. 30. Rupture life of plate in the as-hot-rolled condition (538 C704 C data) and
in the hot-rolled and solution-annealed (1121 C/2 h condition (732 C1093 C data))
(adapted from Ref. 187).
dizing substances such as nitric acid, nitrates and oxidizing salt. The
addition of titanium, in combination with an appropriate heat treat-
ment, stabilizes the alloy against sensitization to intergranular
corrosion.
Physical and mechanical properties are discussed in literature
[190,191]. INCOLOY alloy 825 has good mechanical properties from
cryogenic temperatures to moderately high temperatures. When the
temperature exceeds 540 C, microstructural changes (phase for-
mation) can occur which lower ductility and impact strength. So
at temperatures where creep-rupture properties are design factors,
the alloy is not used. Some tensile properties of INCOLOY alloy 825
are given in Table 21: the alloy can be strengthened signicantly
by cold work.
Fig. 29 showed the high-temperature tensile properties and in-
dicated the typical usage range. The ultimate tensile strength of the
material was found to be 720 MPa. The alloy has good impact
strength at room temperature and retains its strength at high
temperatures.

5.2.3. Extruded tubular ceramics [192196]

Ceramic materials are far less documented than their steel coun-
terparts. In tubular ceramics, mostly alumina and its composites,
zirconia, SiC and SiN are used. The alumina ceramics are recog-
nized for their ease of production and moderate costs, although still
between 4 and 5 times the cost of their steel equivalents. Some es-
sential parameters are given in Table 22.
To seal the ceramic heat exchanger tube after having been lled
with PCM or TCS materials, different options are offered by ceramic
specialists, including tungsten carbide, or silicon carbide or graph-
ite seals [196198]. For less demanding applications, steel hardware
(clamps, clips, etc.) are proposed, e.g. Ref. 199.

Fig. 31. Results of cyclic oxidation tests at 980 C (cycles were 15 minutes of heating 5.3. Experimental design data
and 5 minutes of cooling in air) (adapted from Refs. 186 and 187).
Some initial experiments were carried out by the authors,
using a welded AISI 321 tube lled with PCM. The capsules were

Table 22
Mechanicalthermal properties of ceramics (adapted from Refs. 192196).

Material properties Unit Alumina Zirconia Silicon carbide Silicon nitride (10 bar)

Sintered Inltrated

Purity >99% ZrO2 (TZP) SSiC SiSiC ND-SN

Density kg/m3 3900 6000 3120 3100 3200

Four-point bending strength 4PB MPa 350 900 400 360 800
Comparison stress OV MPa 400 1060 500 470 925
Youngs modulus E GPa 380 210 400 380 320
Fracture toughness KIC MPa/m2 4 8 3 3.5 6.5
Hardness 10 HV GPa 20 12 25 20 15
Thermal expansion 106/K 8 11 4.5 4 3
Thermal conductivity W/mK 25 2 100 110 30
Max. working temperature Tmax K 1773 1273 1773 1653 1523
Chemical resistance Rated Good Good Very good Good Good
Raw material costs /kg 5 75 15 7 60
30 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Table 23 calculated required wall thickness is far below the standard thick-
Test results for AISI 321, Pdeform (pressure to initial deformation) and Prupt (pressure nesses of commercial tubes. Predicted and experimental pressures
to rupture).
to deformation and rupture are in very fair agreement. Further ex-
Number of KNO3 E-PCM, 39 mm O.D. Sb2O3 E-PCM, 27.6 mm O.D. perimental conrmation is however required.
charging/ 2.0 mm wall thickness 1.5 mm wall thickness
discharging
Pdeform Prupt Pdeform Prupt 6. PCM/TCS integration in TES requires heat carriers
cycles
(MPa) (MPa) (MPa) (MPa)

0 (reference) 14.8 41.6 16.8 46.2 6.1. Introduction

250 14.6 41.5 16.2
500 14.5 40.9 16.0
750 14.4 40.6 15.7
1000 14.4 39.8 15.9
1250 14.1 39.4 14.8
subjected to pull and press experiments, and pressures to defor-
mation and to rupture were measured and given in Table 23. Clearly,
the pressure required to deform or burst the tested encapsula-
tions signicantly exceeds the pressure generated during the phase
change (~5 bar = 0.5 MPa). Moreover, the KNO3 tube of larger di-
ameter and thickness performs better, with a lower reduction in
pressure after multiple cycling. These results are however in line
with the theoretical predictions, as shown below.
The minimum required wall thickness at the maximum oper-
ating temperature is calculated according to Eq. 14. The pressure
to deformation and rupture is calculated according to the Barlow
formula, Eq. 15. The maximum intended operating pressure P is
3 bar; the outside diameter for cylinder 1 and 2 is respectively 39 mm
and 27.6 mm. The allowable stress SE (yield strength) at 650 C for
AISI 321 is 250 MPa. The coecient Y is 0.4 because the material
is nonferrous and the factor C is taken as 0.01 mm, which is a pes-
simistic estimation. The minimum required wall thickness for the
two cylinders and the selected materials is summarized in Table 24.
A commercial wall thickness of 1.5 or 2 mm provides a large excess
wall thickness.
For the pressure to deformation, S is the yield strength at 650 C,
this is the same as SE in the calculation for the minimum required
wall thickness, and the safety factor is 1.5. To calculate the pres-
sure to rupture, S is the tensile strength and the safety factor is 1.
The results of the calculations for the pressure to deformation and
rupture are given in Table 25. Clearly, experimental values are about
1020% below the theoretical predictions, so the pressure equa-
tions can hence be used. For high numbers of cycles, the design
pressure should however be increased by 10 to conservatively 20%.
Form the above data and experimental treatment, it emerges that
the proper properties of the encapsulation materials as given in spe-
cialized references provide a very fair basis for design calculations.
Yield strength, tensile strength, and time-dependent creep strength
are the most important properties that enable to assess required
wall thickness and allowable pressures to deformation and rupture.
Thermal cycling of E-PCM sample tubes has demonstrated that the
45.0
44.6
43.9
TES generally requires the use of a heat transfer uid during the
43.5 charging and discharging processes. Although water/steam, gas, syn-
42.2 thetic oils and others can be used, current HTFs are commonly
molten salts, when large-scale heat storage is included, however with
drawbacks of temperature limitation required heat tracing of the
molten salt circuits, and high pumping power requirements. The
maximum temperature for the application of thermal uids is
between 350 and 400 C, whilst molten salt applications are limited
to ~560 C.
Molten salts are mostly nitrates, such as solar salt (NaNO3/
KNO3) or Hitec (NaNO2/NaNO3/KNO3), with an upper temperature
limit of ~600 C for solar salt, and a lower temperature limit through
solidication at 230 and 140 C respectively. The molten salts have
a low vapour pressure, but their moderate viscosity as a function
of temperature leads to considerable pumping energy. The insta-
bility of the molten salts should be carefully considered, with e.g.
a slow conversion at ~538 C (Hitec HTS) from NO2 NO3, or a
decomposition of NO3 salts to NO2 and O-radicals with possible
formation of CO32 plugs [200], according to:
NO3 NO2 + 1 2O2; 2NO2 O* + 3 2O2 + N2;
and CO2 + O* CO23
As a result of these instabilities, physical properties (, , etc.)
will change and environmental issues should be taken into con-
sideration (e.g. emissions through pressure venting of the formed
gases), whilst plugging problems might occur.
Recently, the use of particle suspensions as HTF has been ad-
vocated. Since the particle suspension has a heat capacity similar
to that of molten salts, without temperature limitation except for
the maximum allowable wall temperature of the receiver tube, sus-
pension temperatures of up to 800 C can be tolerated for refractory
steel tubes (even higher when using ceramic or glass tubes), thus
offering new opportunities for highly ecient thermodynamic cycles
such as obtained when using supercritical steam or CO2, as previ-
ously illustrated in Fig. 6. Despite the pre-cited advantages of using
circulating particles as HTF, multiple questions need to be investi-
gated concerning powder stability (attrition); equipment erosion;
or the overall economy of the system when the receiver tempera-
ture is increased, leading to increased power cycle eciency,
increased storage density, reduced thermal power requirements, and
increased plant capacity factor.

6.2. Powders as novel heat carriers in renewable energy systems

Table 24
Minimum required wall thickness (mm) for the selected materials. A novel application of powders in renewable energy relies on
AISI 321 their use as heat transfer medium for heat capture, conveying and
Test 1 0.033 storage. Gas, mostly air, but possibly inert gas such as N2, is used
Test 2 0.027 as powder carrying medium. According to the gas velocity applied,
different gassolid concepts are possible. With increasing gas ow
rate, the operation moves from a xed bed to a bubbling uidized
or ultimately circulating uidized bed and pneumatic conveying
Table 25
Pressure to deformation and rupture (MPa) for AISI 321. [201]. Without, or with a limited gas ow, falling particle reactors
or mechanical conveying could be used. Whereas powder reactors
Pdeform (MPa) Prupt (MPa)
in chemical applications mostly use Geldart A or B powders as bed
Test 1 17 45 material [202], HTF applications will mostly use A-type powders
Test 2 18 48
in order to limit the required gas ow rate and associated sensible
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 31

Fig. 32. Operation modes of powder-gas systems [201].

heat losses. Fig. 32 illustrates the different operation modes of solid

gas systems, with representative values of the major characteristic
transition velocities given in Table 26.
Compared to the moving bed concept, introducing relative par- Table 26
ticle movement improves the heat transfer and gas/solid mixing, Transition velocities associated with Fig. 34.
but abrasion, erosion and particle elutriation increase with increas- Utrans Equations Ref.
ing gas velocity. Bubbling uidized beds (BFB) and BFB with induced
2.07d p g0.06
solids circulation (UBFB), operating with A-powders and at low U, Umb U mb = exp (0.716 F45 ) , F45: ne fraction less 203
g0.347
provide an excellent balance between heat transfer rates and par- than 45 m

ticle stability. Umf Ar = 1823Remf1.07 + 21.27Remf2 110

Falling particle reactors have poor heat transfer and particle sta- Utf Retf = 1.24Ar0.45, 2 < Ar < 108 204
bility could be critical. The heat capturing capacity of the solids is UTR ReTR = 3.23 + 0.23Ar 201

low per unit reactor length. Circulating uidized bed (CFB) systems UCh 2gD (s Ch
4.7
1) 205
= 0.00872g0.77
U Ch U t
offer the advantage of operating at high solid circulation uxes,
(>>500 kg/m2 s), thus being capable of conveying signicant amounts p g
Usalt U salt = 4.43 dp 206
g
of heat. The somewhat lower heat transfer coecient than in BFB
and UBFB, and higher risk of abrasion and attrition are drawbacks. Ums U ms = U mf + 0.07 gD , D: diameter of the bubbling 207
uidized bed
The application of a higher value of U will moreover necessitate
32 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
appropriate de-dusting equipment and heat recovery on the exhaust
gas. The application of CFB seems hence limited to very large storage
and re-use capacities. The rotary drum and screw conveyer can offer
advantages for heat storage, but can be excluded as heat capture
option due to the low heat transfer rate achieved. Among the dif-
ferent concepts, the UBFB (low to moderate CSP capacities) and CFB
(large capacities) are recommended.
The BFB, CFB and moving bed concepts can however be applied
in the TES and thermal block of the plant, with some tentative il-
lustrations are given in Fig. 33 and Fig. 34 as example applications.
The excellent heat transfer characteristics of powder beds more-
over facilitate the integration of E-PCM, as temporary heat storage
medium. The loop of Fig. 33 is generally composed of a hot storage
silo fed from the discharge of the heat or energy receiver, and feeding
a uid bed heat exchanger, where the particles transmit their energy
to submerged tubes inside whose a working uid (for example
steam) is generated and further expanded in a turbine: the uid bed
heat exchanger is a common device in the electrical power indus-
try (mostly implemented for coal and biomass combustion in
uidized beds).
The cooled particles exit the boiler (continuous circulation) and
are returned towards the cold storage tank by mechanical convey-
ing. Finally, a conveying system raises the particles from the cold
silo to the heat receiver inlet. Particles are used as heat transfer uid
and heat storage medium, possibly complemented by using an
E-PCM.
The loop of Fig. 34 includes the same unit components. Heat is
supplied to the hot-side of the Stirling engine in a BFB of the par-
ticle heat carrier.
6.2.1. Dense up-ow of particles
Literature concerning dense up-ow systems, with forced up-
ow and bubble-induced mixing is scarce. Initial research mostly
covered moving packed beds. Fluidized bed up-ow reactors were
described by several researchers. The dense up-ow column is fed
at its bottom by a circulation ow of solids at an appropriate pres-
sure to overcome the pressure drop of the upwards-moving bed of
solids. The column itself is either aerated from the windbox [3,7]
or by a combined windbox aeration and additional aeration at the
bottom of the riser column itself [21,28].
6.2.2. Particle loops in SPT and CSP applications
Since the particle solar receiver is the key component, previ-
ous investigations in the eld of solar receivers using particles as
HTF are dealt with in Table 27. Tubular absorbers are generally pro-
posed for use in current solar thermal receivers. The rst studies
on direct absorption solar receivers started in the early 1980s with
three concepts, the uidized bed receiver [208], the free falling par-
ticle receiver [209] and the rotary kiln receiver [210,211].
In the free falling particles concept, the solids are dropped di-
rectly into the concentrated solar beam from the top of the receiver
and are heated during their pass through the concentrated solar ra-
diation. The principle was validated by on-sun experiments at pilot
scale [212]. A receiver prototype was tested at the National Solar
Thermal Test Facility (NSTTF) in Albuquerque NM, USA. Short batch
runs were performed for a total particle inventory of about 1800 kg.
The receiver eciency generally increased with the particle ow
rate and varied from about 35% to 52%, thus in good agreement with
simulated data. A review of the falling particle receiver was pro-
posed by Tan and Chen [213]. CNRS (FR) developed a sand heater
loop using sand particles as HTF [211]: it combined a solar rotary
kiln that delivered hot sand to a heat storage/heat recovery sub-
system consisting of a hot and a cold heat storage bin and a
multistage uidized heat exchanger.
The air jet ow concept of Rger et al. [214] used falling par-
ticles that line the inner wall of a cylinder closed at its top, whilst
the bottom part was facing the concentrated solar beam. Heat re-
covery from the hot storage is then possible using uidized bed heat
exchangers as described by Warerkar et al. [215], or by moving bed
particle-air heat exchangers tested by Al-Ansary et al. [216]. In the
former study, storage bins are integrated at the top of the tower.
Direct and indirect absorption particulate receivers were dis-
cussed in detail by Zhang et al. [28]. Indirect receivers operate at
lower ux density in the range 200400 kW/m2, but offer a better
control of particle circulation within the receiver and a possible man-
agement of operating pressure and heat losses.
The particle mass ow rate inside the solar receiver is one of the
main issues for a high power solar concentrating system using par-
ticles as HTF. In industry, BFB and CFB are widely used at large scale
in oil reneries and in combustors [217,218]: the dense phase u-
idized bed can be used in standpipes to provide an important and
steady downward ow of solid as shown by Bodin et al. [219], Smol-
ders et al. [220], and Chan et al. [221]. In this regime, the suspension
is uniform, it has a low voidage, and it circulates slowly (a few cm/
s), thus limiting both the energy consumption and its use as HTF.
Novel concepts use a dense suspension of small size solid par-
ticles, patented by Flamant and Hemati [222], and are currently
developed in the framework of both a French National and a Eu-
ropean project [30]: they involve a dense particle upow suspension
receiver (solid fraction in the range 3040%) in vertical absorbing
tubes submitted to concentrated solar energy.
6.2.3. Recent experimental results
Recent experimental investigations of using a particulate HTF
were presented by Flamant et al. [21] and Zhang et al. [28], where
the concept of using UBFB and BFB particle suspension carriers was
tested at the solar receiver of Font Romeu (F). The selected par-
ticles were 64 m silicon carbide (3120 kg/m3), conveyed in a single
tube at air velocities between 0.03 and 0.22 m/s, achieving powder
circulation uxes between 8 and 50 kg/m2s. The tube wall temper-
ature reached up to ~800 C.
The experimental results of the heat transfer coecient, HTC,
against solid ux and for all values of operating gas velocity are il-
lustrated in Fig. 35. The heat transfer coecient is a clear function
of G. Values of HTC range from >430 W/m2 K to 1120 W/m2 K. These
measured values include both the contribution of particle convec-
tion and radiation at higher temperatures. The radiation contribution
can be separately assessed from the StefanBoltzmann equation and
ranges from ~30 W/m2 K at low wall and bed temperature, to 200 W/
m2 K for high temperatures [28].
CFB experiments were also carried out and were partly re-
ported by Zhang et al. [231]. Solids circulation was achieved via a
downcomer and L-valve. Heat supply was by hot water or thermal
uid (Santotherm 350). The downcomer was water-cooled through
a 0.2 m long concentric cooler.
Temperatures were measured and recorded (wall, different lo-
cations in the CFB, feeding and overow lines of the uid). The HTC
was calculated from the known exposed surface area, Aex, the mea-
sured temperature difference, T, and heat input, Q. The bed material
used was rounded sand of 75 m, 2260 kg/m3, and Cps = 1.05 kJ/
kg K. Various combined (U,G) values were tested to scan the different
riser hydrodynamic regimes (dilute ow, core/annulus ow, dense
ow): G was varied from 8 to ~500 kg/m2 s and U UTR from 1 to
18 m/s (with U, the supercial gas velocity, and UTR, the required
onset supercial gas velocity for CFB riser ow).
The HTC varied from ~20 W/m 2 K (G = 0) to ~400 W/m 2 K
(G = 500 kg/m2 s). If a higher temperature would have been used (wall
temperature >600 C), the radiation contribution should have in-
creased the measured HTC. The required supercial gas velocity in
a CFB (>>3 m/s) signicantly exceeds the low value in the UBFB
(~0.1 m/s).
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 33

Fig. 33. Possible layout with heat capture/storage and powder heat carrier system [28].
34 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140

Fig. 34. Possible layout with BFB heating of the hot side Stirling heat exchanger.

Table 27
Literature review on solar receiver developments using powders as HTF.

Year Concept and Use Ref.

1980 Concept of a uidized bed receiver, no solids circulation, batch operation Flamant et al. [211]
1982 Fluidized bed solar receiver Flamant [208]
1982 Free falling particles receiver, model of particle selection and radiative heat transfer were proposed [220] Martin and Vitko [209]
1985 The use of small refractory pebbles or sand as the working media in a high temperature solar central receiver system Falcone et al. [223]
was proposed and examined. Particle selection and radiative heat transfer model of free fall receiver
1989 A two-part system, where a rotary kiln receiver connects to hot storage/heat recovery sub-system (storage bins and Bataille et al. [210]
multistage uidized heat exchanger)
2004 Two high-porosity materials, a double-layer silicon carbide foam and a screen-printed porous silicon carbide material Fend et al. [224]
were proposed to be used as volumetric receivers for concentrated solar radiation.
2010 Pilot scale prototype validation of high temperature solar thermal concentrating plant at National Solar Thermal Test Siegel et al. [212]
Facility (NSTTF). Cavity receiver with falling particle lms
2010 Effect of wind speed and parasitic air ows inside the cavity on particle aerodynamics Tan and Chen [213]
2010 An internally circulating uidized (draft tube) using uidized particles as water splitting reactor with solar simulated Gokon et al. [225]
light irradiation (non-CSP application)
2011 Face-down solid particle receiver concept and modelling Rger et al. [214]
2011 Fluidized bed heat exchanger model Warerkar et al. [215]
2011 A simulation analysis was applied to a porous media solar tower receiver. The established working diagram provides a Xu et al. [226]
reference for receiver design and reconstruction.
2011 Combined conveyor and heat transfer device which consist in a rotatable conveyor drum and heat transfer uid Jeter and Stephens [227]
conduits within the drum
2012 Particleair heat exchangers integrated with storage bins at the top of the tower was proposed Al-Ansary et al. [216]
2012 CFB particulate receiver Brems et al. [106].
2013 CFB with encapsulated PCM particles Piti et al. [7]
2013 Concept of screw conveyor heat exchanger, whereby PCMs can be transported within the shaft and height. Decouples Zipf et al. [228]
the heat transfer and storage and develops an economically viable latent heat storage.
2013 A dense particle suspension receiver (DPSR) was proposed and tested at pilot scale. Upward circulation of a dense Flamant et al. [18,21]
suspension of particles (SiC) as HTF to extend the working temperature limit to the bearing capacity of the receiver
tube.
2014 An inner-circulating uidized bed spouted by concentric gas streams with high and low velocities in the centre and Matsubara et al. [229]
outer annulus, respectively, with an irradiation source at the top of the receiver container.
2014 A thermal energy storage (TES) system was developed by NREL using solid particles as the storage medium for CSP Ma et al. [230]
plants. Based on their performance analysis, particle TES systems using low-cost, high T withstand able and stable
material can reach 10$/kWhth, half the cost of the current molten-salt based TES
2014 A heat transfer model to predict the heat transfer coecient for different operating conditions in CFB receiver was Zhang et al. [231]
proposed, with air ow rate and solids circulation ux as dominant parameters.
2014 Optimized optical particle properties for a high temperature solar receiver Ordez et al. [232]
2015 Design optimization of a small heat exchange receiver for solar tower power plants Fernndez and Miller [233]
2015 Design of a packed bed TES for high temperature Zanganeh et al. [234]
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140 35

At equivalent values of the specic heat of molten salts and

powders, and with a lower powder temperature after heat recov-
ery (~150 C against ~300 C for molten salts), the higher receiver
temperatures of the powders reduce the storage requirement on a
mass basis by ~50%, however only about 30% on a volumetric basis,
due to the lower bulk density of the powder bed. Parasitic power
consumption is also signicantly reduced due to the reduced
pumping energy, and lack of heat tracing of the circuits. The com-
bination of these technological advantages will considerably reduce
the LCOE of the SPT concept. The dense particle suspension is a
promising option, justifying additional testing (e.g. using multiple
receiver tubes), and allowing a more correct analysis of economic
advantages in both operating and investment costs, as tentatively
proposed in Table 28. A multi-tube receiver (16 parallel tubes) has
been studied within the CSP2 project [27] at the CNRS 1MW solar
furnace, and results are presented in a separate paper.

7. Conclusion and recommendation for priority research

Fig. 35. Experimental values of the heat transfer coecient, HTC, vs. G, for all ex-
perimental supercial gas velocities.
The studies conrmed that upowing particles can be used in
the novel design of a solar receiver, both in UBFB and CFB mode.
UBFB heat transfer coecients up to 1100 W/m2 K were obtained
for operation at low supercial gas velocities, thus limiting sensi-
ble heat losses by the exhaust air. Although CFBs could be interesting
for large scale operations, due to the higher solid ux achieved, the
comparison of Table 28 tentatively illustrates the main advan-
tages of using the UBFB principle, even for high receiver capacities.
This selection accounts for sizes, complexity and operating costs.
The high heat transfer coecient obtained within a given range
of supercial gas velocities, U and solid ux, G, and the possible op-
eration of the powder upow bed at high temperatures (>750 C)
moreover offers new opportunities for high eciency thermody-
namic cycles, with a conversion increase from the current 40% to
over 46% (leading to a reduction of the receiver and heliostat
requirements).
This work reviewed some recent developments and design rec-
ommendations for the high temperature heat storage, where
sensible/latent/reaction heat each offers a contribution with re-
versible reactions being the most effective (heat storage >>1000 kJ/
kg), followed by latent heat storage (5001000 kJ/kg) and sensible
heat storage materials (<500 kJ/kg). This is dealt with in Section 1
of the paper.
Section 2 of the paper specically dealt with sensible heat storage.
Sensible heat storage materials are well-documented, have the lowest
cost materials but at the same time the lowest storage capacity com-
pared to PCM and TCS. Most commonly used minerals costs range
from 0.05 /kg to 0.1 /kg, hence cheaper than latent heat storage
costs sensible heat materials have excellent thermal conductivi-
ties: 1.0 W/mK7.0 W/mK for sand-rock minerals, concrete and
rebricks, 37.0 W/mK40.0 W/mK for ferroalloys. Their disadvan-
tage is their heat capacity, ranging from 0.56 kJ/kg K to 1.3 kJ/kg K,
making storage unit sizes to be bigger than when using latent heat.
The energy stored by a SHS material towards variations of 1 C in
its temperature can be 100 times lower compared to the energy
storage capacity when it has a phase change, as can be noticed in

Table 28
Comparison of UBFB and CFB solar receiver (10 MWth) with SiC powders (50 m, 3100 kg/m3 and specic heat at 500 C of 1.2 kJ/kg K) [203].

Parameter Equation or reference UBFB CFB

Voidage, 0.60.7 [21,28] 0.950.99 [235,236]

Pressure drop, P/L (mbar/m) p(1 ) 124 6.2
Heat transfer coecient, HTC (W/m2 K) 950 600 for G > 400 kg/m2 s [231]
Riser diameter, D (m) 0.04 [21,28] 0.08 [231]
Solids circulation ux, G (kg/m2 s) 50 [21,28] >400 [231,235]
Solids ow in riser tube, Qs (kg/s) (D2/4) G 0.063 2.0
Temperature increase in riser passage (K) Inlet ~250 C 500 500
Outlet ~750 C
Average temperature in riser, Tr (K) (Tin + Tout)/2 500 500
Temperature of riser wall, Tw (K) 800850 800850
Driving T (K) T (Tw Tb) 300 300
Required heat exchanging surface, S (m2) (10 MWth 106)/(HTC T) 30 55.5
Height of riser, H (m) e.g. 3 2 m each Commonly 6 m
Number of parallel riser tubes () S/(HD) 3 40 36
Pressure drop, P (mbar) P/(LH) +disperser + separator (2 124) + 50 300 (6 6.2) + 50 90
Air ow to riser, Qa (N m3/h) UBFB: 0.05 Nm/s (i.e. 0.14 m/s at 750 C) ~30 1300
CFB: 2 Nm/s (i.e. 5.5 m/s at 750 C)
Heat loss by air ow (90% recycle), Hloss (kWth) [QaCpa(Tout 20)] 0.1 14564 kJ/h 123,370 kJ/h
0.8 kWth 35 kWth
Air pumping power, P (kWel) UBFB: blower <1 kWel <4 kWel
CFB: fan

The UBFB application results in lower heat losses, but a slightly higher pressure drop. Towards the number of parallel receiver tubes, the CFB is more advantageous. The
conveying of the powders from the cold storage to the feeding system of the receiver is supposed to be operated by mechanical conveying (as a series of bucket elevators).
Further scale-up work is required to conrm the tentative economic and design assessment.
36 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
Table 29
Energy densities from SHS and LHS materials.
SHS
Material Specic heat (kJ/kg K) Energy density (kJ/m3 K)
Water (liquid) 4.2 4200
60% NaNO3 40% KNO3 1.5 3144
Paran wax 1.92 1593.6
Table 29, where the difference of energy storage capacity of three
different materials is illustrated when used as SHS (in their liquid
state) and LHS (when they change from solid to liquid).
Section 3 dealt with latent heat storage, specically towards high
temperature application, where inorganic substances offer a high
potential. Experimental and modeling data will be compared. Since
inorganic solid/liquid phase change materials suffer from a low
thermal conductivity, conductive inserts offer a high potential, re-
ducing the charging/discharging time by a factor of ~2.
Section 4 reviewed the fundamentals of using reversible chem-
ical reactions to store heat for short to long term storage. Within
the selected reaction pairs, Ca(OH)2, CaCO3, Co3O4 and Mn2O3 offer
a high potential, both towards achieved reaction temperature and
reversibility. All reactions can be treated as rst order chemical re-
actions, with activation energies between 50 and 250 kJ/mol and
reaction rate constants of 1 to 2 103 s1 at reaction equilibrium
temperature.
Since latent and thermal chemical storage medium needs to be
contained in a capsule (sphere, tube, sandwich plates) of appro-
priate materials, Section 5 assessed the mechanical strength of the
containment from both a theoretical and experimental evalua-
tion. Alloy-capsules of 12 mm thickness meet the requirements of
long term stability, even after charging/discharging cycles.
Section 6 specically dealt with a novel particulate HTF, con-
sidered for both heat capture and storage. Since particulate
suspensions deal with sensible heat storage, the paper reviews pre-
vious investigation and recent experimental results. Literature and
experimental data conrm the high heat transfer coecient achieved
(>400 W/m2 K in circulating uidized beds, to >1100 W/m2 K in a
dense upow uidized bed).
A tentative technical and economic assessment demonstrates that
particulate suspensions lead to major savings in investment and op-
erating costs.
Although sucient fundamental and practical information on
SHS/LHS and TCS is available in literature, the achievement of a high
eciency TES needs to be proven, preferably on pilot or large scale.
These requirements, subjects of additionally required research,
include (i) a high energy density in the storage material (storage
capacity); (ii) a high heat transfer between the HTF and the con-
tained storage media; (iii) a chemical and mechanical stability of
the storage materials used during extensive cycles; (iv) the com-
patibility between HTF, storage medium and heat exchanger; (v) the
complete reversibility of the charging/discharging cycles; (vi) working
at low thermal losses; (vi) ease of control, operation, and integra-
tion into the proposed heat storage concept; and (vii) operation at
as high a temperature as possible.
According to the analysis, specic priority research targets should
include (i) the further development and testing of PCMs and their
charging/discharging enhancement techniques; (ii) the proof of
concept of using nanoparticle-encapsulated PCMs; (iii) the further
development and large-scale conrmation of TCS storage tech-
niques, with special emphasis on the complete reversibility of the
endothermic/exothermic cycling; (iv) the large scale testing of ap-
propriate PCM/TCS containments; (v) the further development of
particle suspension HTFs, possibly even using TCS powders in a cycle
between a reducing (charging) and oxidation (discharging) atmo-
LHS
Melting temperature (C) Latent heat (kJ/kg) Energy density (GJ/m3)
0 335 307
223 105 230
32 251 208
sphere; and (vi) nally, exergy and energy analyses should be
conducted to optimize a storage module: entropy generation anal-
ysis is specically important in optimizing design and operation
parameters of the systems.
Acknowledgements
Yongqin Lv gratefully acknowledges the nancial support from
the special assistance of the National Natural Science Foundation
of China (21306006, and 21436002), 973 programs (2014CB745103,
2013CB733603), the Xiamen Scientic and Technological project
(3502Z20142012), the Public Hatching Platform for Recruited Talents
of Beijing University of Chemical Technology, the High-Level Faculty
Programs of Beijing University of Chemical Technology (20130805),
and the Fundamental Research Funds for the Central Universities
(YS1407). This work was also supported in Chile by the projects
CONICYT/FONDAP/15110019 (SERC-CHILE), CONICYT/FONDECYT/
1151061, by Center UAI Earth and the R&D project engineer
Macarena Montan. Huili Zhang would like to express her sincere
gratitude to the China Scholarship Council for sponsoring her Ph.D.
study at KU Leuven in Belgium (File No. 201206880024). Jan Baeyens
and Huili Zhang acknowledge EU funding under the CSP2- program
(Project No. 282932).
References
[1] Asif M, Muneer T. Energy supply, its demand and security issues for developed
and emerging economies. Renew Sustain Energy Rev 2007;11:1388413.
doi:10.1016/j.rser.2005.12.004.
[2] International Energy Agency. World energy outlook. 2011.
[3] Zhang HL, Baeyens J, Degrve J, Cacres G. Concentrated solar power plants:
review and design methodology. Renew Sustain Energy Rev 2013;22:46681.
doi:10.1016/j.rser.2013.01.032.
[4] Zhang HL, Van Gerven T, Baeyens J, Degrve J. Photovoltaics: reviewing the
European Feed-in-Tariffs and changing PV eciencies and costs. Sci World J
2014;2014:404913. doi:10.1155/2014/404913.
[5] Energy Information Administration. International energy outlook. Washington,
DC: 2011.
[6] International Energy Agency. World energy outlook 2012. Paris, France: 2012.
[7] Piti F, Zhao CY, Baeyens J, Degrve J, Zhang HL. Circulating uidized bed heat
recovery/storage and its potential to use coated phase-change-material (PCM)
particles. Appl Energy 2013;109:50513. doi:10.1016/j.apenergy.2012.12.048.
[8] Ammar Y, Joyce S, Norman R, Wang Y, Roskilly AP. Low grade thermal energy
sources and uses from the process industry in the UK. Appl Energy 2012;89:3
20. doi:10.1016/j.apenergy.2011.06.003.
[9] Naish C, McCubbin I, Edberg O, Harfoot M. Outlook of energy storage
technologies (Study No. IP/A/ITRE/FWC/2006-087/Lot 4/C1/SC2). 2008.
[10] Astrom Technical Advisors, S&L. CSP today. CSP benchmarking int., 2013.
[11] Bilgili M, Ozbek A, Sahin B, Kahraman A. Concentrated solar power plants:
review and design methodology. Renew Sustain Energy Rev 2013;22:46681.
doi:10.1016/j.rser.2013.01.032.
[12] Teller O, Nicolai J-P, Lafoz M, Laing D, Tamme R, Pedersen AS, et al. Joint
EASE/EERA recommendations for a European energy storage technology
development roadmap towards 2030. 2013.
[13] The Liquid Air Energy Network. Energy storage systems according to form of
energy stored. <http://www.liquidair.org.uk>; 2014. [accessed 05.05.15].
[14] Fernandes D, Piti F, Cceres G, Baeyens J. Thermal energy storage: How
previous ndings determine current research priorities.. Energy
2012;39:24657. doi:10.1016/j.energy.2012.01.024.
[15] Medrano M, Gil A, Martorell I, Potau X, Cabeza LF. State of the art on high-
temperature thermal energy storage for power generation. Part 2 case
studies. Renew Sustain Energy Rev 2010;14:5672. doi:10.1016/
j.rser.2009.07.036.
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
[16] Castelln C, Medrano M, Roca J, Nogus M, Castell A, Cabeza LF. Use of
microencapsulated phase change materials in building applications. Build X
2007.
[17] Torresol Energy. Gemasolar. <http://www.torresolenergy.com/TORRESOL/
gemasolar-plant/en>; 2015. [accessed 10.05.15].
[18] Flamant G, Gauthier D, Benoit H, Sans J-L, Boissire B, Ansart R, et al. A new
heat transfer uid for concentrating solar systems: particle ow in tubes.
Energy Procedia 2014;49:61726. doi:10.1016/j.egypro.2014.03.067.
[19] Zhang HL, Baeyens J, Degrve J, Piti F. Latent heat storage with phase change
materials (PCMs). J Technol Innov Renew Energy 2013;112.
[20] Benoit H, Perez I, Gauthier D, Sans J-L, Flamant G. On-sun demonstration of
a 750C heat transfer uid for concentrating solar systems: dense particle
suspension in tube. Sol Energy 2015;118:62233.
[21] Flamant G, Gauthier D, Benoit H, Sans J-L, Garcia R, Boissire B, et al. Dense
suspension of solid particles as a new heat transfer uid for concentrated solar
thermal plants: on-sun proof of concept. Chem Eng Sci 2013;102:56776.
doi:10.1016/j.ces.2013.08.051.
[22] Maruoka N, Akiyama T. Development of OCM for high temperature application
in the steelmaking industry. In: Annex 17, advanced thermal energy storage
techniques feasibility studies and demonstration projects. 3rd Expert. meet.
work. annex, vol. 17. Tokyo, Japan: 2002.
[23] Maruoka N, Sato K, Yagi J, Akiyama T. Development of PCM for recovering
high temperature waste heat and utilization for producing hydrogen by
reforming reaction of methane. ISIJ Int 2002;42:21519. doi:10.2355/
isijinternational.42.215.
[24] Sharma A, Tyagi VV, Chen CR, Buddhi D. Review on thermal energy storage
with phase change materials and applications. Renew Sustain Energy Rev
2009;13:31845. doi:10.1016/j.rser.2007.10.005.
[25] Kearney D, Herrmann U, Nava P, Kelly B, Mahoney R, Pacheco J, et al.
Assessment of a molten salt heat transfer uid in a parabolic trough solar eld.
J Sol Energy Eng 2003;125:170. doi:10.1115/1.1565087.
[26] Abhat A. Low temperature latent heat thermal energy storage: heat
storage materials. Sol Energy 1983;30:31332. doi:10.1016/0038-
092X(83)90186-X.
[27] Zhang HL, Baeyens J, Degrve J, Cceres G, Segal R, Piti F. Latent heat storage
with tubular-encapsulated phase change materials (PCMs). Energy
2014;76:6672. doi:10.1016/j.energy.2014.03.067.
[28] Zhang HL, Benoit H, Gauthier D, Degrve J, Baeyens J, Prez Lpez I, et al.
Particle circulation loops in solar energy capture and storage: gas-solid ow
and heat transfer considerations. Appl Energy 2016(161C):20624.
doi:10.1016/j.apenergy.2015.10.005.
[29] Dunham MT, Iverson BD. High-eciency thermodynamic power cycles for
concentrated solar power systems. Renew Sustain Energy Rev 2014;30:75870.
doi:10.1016/j.rser.2013.11.010.
[30] CSP2. CSP2 project. <http://www.csp2-project.eu/>; 2015 [accessed 28.01.15].
[31] Cziesla F, Kremer H, Much U, Riemschneider J-E, Quinkertz R. Advanced 800+
MW steam power plants and future CCS options. Siemens Ind Turbomach
2009;121.
[32] Singer C, Buck R, Pitz-Paal R, Mller-Steinhagen H. Assessment of solar power
tower driven ultrasupercritical steam cycles applying tubular central receivers
with varied heat transfer media. J Sol Energy Eng 2010;132:041010.
doi:10.1115/1.4002137.
[33] Spelling J, Gallo A, Romero M, Gonzalez-Aguilar J. A high eciency solar
thermal power plant using a dense particle suspension as heat transfer uid.
Int. Conf. Conc. Sol. Power Chem. Energy Syst., Beijing, China 2014.
[34] Kuravi S, Trahan J, Goswami DY, Rahman MM, Stefanakos EK. Thermal energy
storage technologies and systems for concentrating solar power plants. Prog
Energy Combust Sci 2013;39:285319. doi:10.1016/j.pecs.2013.02.001.
[35] Yu N, Wang RZ, Wang LW. Sorption thermal storage for solar energy. Prog
Energy Combust Sci 2013;39:489514. doi:10.1016/j.pecs.2013.05.004.
[36] Zalba B, Marin JM, Cabeza LF, Mehling H. Review on thermal energy storage
with phase change: materials, heat transfer analysis and applications. Appl
Therm Eng 2003;23:25183. doi:10.1016/S1359-4311(02)00192-8.
[37] Jegadheeswaran S, Pohekar SD. Performance enhancement in latent heat
thermal storage system: a review. Renew Sustain Energy Rev 2009;13:2225
44. doi:10.1016/j.rser.2009.06.024.
[38] Abengoa. Innovative technology solutions for sustainability. <http://www
.abengoa.es/web/es/> 2015 [accessed 01.28.15].
[39] Fernandez AI, Martnez M, Segarra M, Martorell I, Cabeza LF. Selection of
materials with potential in sensible thermal energy storage. Sol Energy Mater
Sol Cells 2010;94:17239. doi:10.1016/j.solmat.2010.05.035.
[40] Regin AF, Solanki SC, Saini JS. Heat transfer characteristics of thermal energy
storage system using PCM capsules: a review. Renew Sustain Energy Rev
2008;12:243858. doi:10.1016/j.rser.2007.06.009.
[41] Kenisarin MM. High-temperature phase change materials for thermal energy
storage. Renew Sustain Energy Rev 2010;14:95570. doi:10.1016/
j.rser.2009.11.011.
[42] Perry RH, Green DW, editors. Perrys chemical engineers handbook. 8th ed.
New York: McGraw-Hill; 2008.
[43] International Renewable Energy Agency. Concentrating solar power technology
brief. 2013.
[44] Laing D. Solar thermal energy storage technologies. Hannover; 2008.
[45] Gil A, Medrano M, Martorell I, Lzaro A, Dolado P, Zalba B, et al. State of the
art on high temperature thermal energy storage for power generation. Part
1 concepts, materials and modellization. Renew Sustain Energy Rev
2010;14:3155. doi:10.1016/j.rser.2009.07.035.
37
[46] Laing D, Bahl C, Bauer T, Lehmann D, Steinmann W-D. Thermal energy storage
for direct steam generation. Sol Energy 2011;85:62733. doi:10.1016/
j.solener.2010.08.015.
[47] Carslaw HS, Jaeger JC. Conduction of heat in solids. Oxford University Press;
1986.
[48] Smith GD. Numerical solution of partial differential equations. Finite difference
methods. Oxford, UK: Oxford University Press; 1985.
[49] Lopez J, Caceres G, Palomo Del Barrio E, Jomaa W. Conned melting in
deformable porous media: a rst attempt to explain the graphite/salt
composites behaviour. Int J Heat Mass Transf 2010;53:1195207. doi:10.1016/
j.ijheatmasstransfer.2009.10.025.
[50] Piti F, Zhao CY, Cceres G. Thermo-mechanical analysis of ceramic
encapsulated phase-change-material (PCM) particles. Energy Environ Sci
2011;4:2117. doi:10.1039/c0ee00672f.
[51] Parrado C, Cceres G, Bize F, Bubnovich V, Baeyens J, Degrve J, et al. Thermo-
mechanical analysis of copper-encapsulated NaNO3KNO3. Chem Eng Res Des
2015;93:22431. doi:10.1016/j.cherd.2014.07.007.
[52] Zakri T, Laurent J-P, Vauclin M. Theoretical evidence for Lichteneckers mixture
formulae based on the effective medium theory. J Phys D Appl Phys
1998;31:158994. doi:10.1088/0022-3727/31/13/013.
[53] Maxwell JC. A treatise on electricity and magnetism. Oxford University Press;
1954.
[54] Lal K, Parshad R. Test and utilization of the Fricke and Pearce equations for
dielectric correlation between powder and bulk. J Phys D Appl Phys
1974;7:45561. doi:10.1088/0022-3727/7/3/313.
[55] Geldart D, Abrahamsen AR. Fluidization of ne porous powders. Chem Eng
Prog Symp Ser 1981;77:160.
[56] Hamilton RL, Crosser OK. Thermal conductivity of heterogeneous two-
component systems. Ind Eng Chem Fundam 1962;1:18791. doi:10.1021/
i160003a005.
[57] Tan H, Li C, Li Y. Simulation research on PCM freezing process to recover and
store the cold energy of cryogenic gas. Int J Therm Sci 2011;50:22207.
doi:10.1016/j.ijthermalsci.2011.04.017.
[58] Ahern JE. Applications of the second law of thermodynamics to cryogenics
a review. Energy 1980;5:8917. doi:10.1016/0360-5442(80)90104-8.
[59] Ordonez C. Liquid nitrogen fueled, closed Brayton cycle cryogenic heat
engine. Energy Convers Manag 2000;41:33141. doi:10.1016/S0196-
8904(99)00117-X.
[60] Ordonez CA, Plummer MC. Cold thermal storage and cryogenic heat engines
for energy storage applications. Energy Sources 1997;19:38996. doi:10.1080/
00908319708908858.
[61] Li Y, Chen H, Ding Y. Fundamentals and applications of cryogen as a thermal
energy carrier: a critical assessment. Int J Therm Sci 2010;49:9419.
doi:10.1016/j.ijthermalsci.2009.12.012.
[62] Kishimoto K, Hasegawa K, Asano T. Development of generator of liquid air
storage energy system. 1998.
[63] Wen DS, Chen HS, Ding YL, Dearman P. Liquid nitrogen injection into water:
pressure build-up and heat transfer. Cryogenics (Guildf) 2006;46:7408.
doi:10.1016/j.cryogenics.2006.06.007.
[64] Markert F, Melideo D, Baraldi D. Numerical analysis of accidental hydrogen
releases from high pressure storage at low temperatures. Int J Hydrogen Energy
2014;39:735664. doi:10.1016/j.ijhydene.2014.02.166.
[65] Byon C. Numerical study on the phase change heat transfer of LNG in glass
wool based on the VOF method. Int J Heat Mass Transf 2015;88:207.
doi:10.1016/j.ijheatmasstransfer.2015.03.092.
[66] Konopka M, Behruzi P, Schmitt S, Dreyer EM. Phase change in cryogenic upper
stage tanks. 50th AIAA/ASME/SAE/ASEE Jt. Propuls. Conf., Cleveland, USA: 2014,
p. 16. doi:10.2514/6.2014-3994.
[67] Pacheco KA, Li Y, Wang M. Study of integration of cryogenic air energy storage
and coal oxy-fuel combustion through modelling and simulation, vol. 33.
Elsevier; 2014 doi:10.1016/B978-0-444-63455-9.50091-X.
[68] Li Y, Wang X, Ding Y. A cryogen-based peak-shaving technology: systematic
approach and techno-economic analysis. Int J Energy Res 2013;37:54757.
doi:10.1002/er.1942.
[69] Donatini F, Zamparelli C, Maccari A, Vignolini M. High ecency integration
of thermodynamic solar plant with natural gas combined cycle. 2007 Int Conf
Clean Electr Power, IEEE 2007;7706. doi:10.1109/ICCEP.2007.384301.
[70] Fu J, Sunden B, Chen X, Huang Y. Inuence of phase change on self-
pressurization in cryogenic tanks under microgravity. Appl Therm Eng
2015;87:22533. doi:10.1016/j.applthermaleng.2015.05.020.
[71] Yang JH, Yang GS. The temperature eld research for large LNG cryogenic
storage tank wall. Appl. Mech. Mater. 2014;668669:7336.
[72] Kotek M, Suma J, Chrz V. Holding time of horizontal cryogenic vacuum
insulated tanks. Refrig Sci Technol 2014;1319.
[73] Steinmann W-D, Eck M. Buffer storage for direct steam generation. Sol Energy
2006;80:127782. doi:10.1016/j.solener.2005.05.013.
[74] Zhang N, Lior N. A novel Brayton cycle with the integration of liquid hydrogen
cryogenic exergy utilization. Int J Hydrogen Energy 2008;21424.
[75] Wang Q, Li Y, Wang J. Analysis of power cycle based on cold energy of liqueed
natural gas and low-grade heat source. Appl Therm Eng 2004;24:53948.
doi:10.1016/j.applthermaleng.2003.09.010.
[76] Wang Q, Li Y, Chen X. Exergy analysis of liqueed natural gas cold energy
recovering cycles. Int J Energy Res 2005;29:6578. doi:10.1002/er.1040.
[77] Zhang N, Lior N. A novel near-zero CO2 emission thermal cycle with LNG
cryogenic exergy utilization. Energy 2006;31:166679. doi:10.1016/
j.energy.2005.05.006.
38 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
[78] Szargut J, Szczygiel I. Utilization of the cryogenic exergy of liquid natural gas
(LNG) for the production of electricity. Energy 2009;34:82737. doi:10.1016/
j.energy.2009.02.015.
[79] Chen H, Chen Y, Hsieh H-T, Siegel N. Computational uid dynamics modeling
of gas-particle ow within a solid-particle solar receiver. J Sol Energy Eng
2007;129:160. doi:10.1115/1.2716418.
[80] Van de Velden M, Baeyens J, Dougan B, Mc Murdo A. Investigation of
operational parameters for an industrial CFB combustor of coal, biomass and
sludge. China Particuol 2007;5:24754. doi:10.1016/j.cpart.2007.05.001.
[81] Department of Energy. Improving steam system performance, a sourcebook
for industry (108 pages). DOE/GO-12004-1868. 2012.
[82] Tassaing T, Danten Y, Besnard M. Infrared spectroscopic study of hydrogen-
bonding in water at high temperature and pressure. J Mol Liq 2002;101:149
58. doi:10.1016/S0167-7322(02)00089-2.
[83] Gupta G, Ampornrat P, Ren X, Sridharan K, Allen TR, Was GS. Role of grain
boundary engineering in the SCC behavior of ferriticmartensitic alloy HT-9.
J Nucl Mater 2007;361:16073. doi:10.1016/j.jnucmat.2006.12.006.
[84] Hirose T, Shiba K, Enoeda M, Akiba M. Corrosion and stress corrosion cracking
of ferritic/martensitic steel in super critical pressurized water. J Nucl Mater
2007;367370:11859. doi:10.1016/j.jnucmat.2007.03.212.
[85] Mitton DB, Zhang SH, Han EH, Hautanen KE, Latanision RM. Assessment of
corrosion and failure mechanisms in supercritical water oxidation systems.
Proc. 13th ICC, Melbourne, Australia: 1996, p. 259.
[86] Godall PM. The ecient use of steam. Surrey, England: IPC Science and
Technology Press Limited; 1980.
[87] Cabeza LF, Castell A, Barreneche C, de Gracia A, Fernndez AI. Materials used
as PCM in thermal energy storage in buildings: a review. Renew Sustain Energy
Rev 2011;15:167595. doi:10.1016/j.rser.2010.11.018.
[88] Chen B, Wang X, Zeng R, Zhang Y, Wang X, Niu J, et al. An experimental study
of convective heat transfer with microencapsulated phase change material
suspension: laminar ow in a circular tube under constant heat ux. Exp
Therm Fluid Sci 2008;32:163846. doi:10.1016/j.expthermusci.2008.05.008.
[89] Delgado M, Lzaro A, Mazo J, Marn JM, Zalba B. Experimental analysis of a
microencapsulated PCM slurry as thermal storage system and as heat transfer
uid in laminar ow. Appl Therm Eng 2012;36:3707. doi:10.1016/
j.applthermaleng.2011.10.050.
[90] Bridges NJ, Visser AE, Fox EB. Potential of nanoparticle-enhanced ionic liquids
(NEILs) as advanced heat-transfer uids. Energy Fuels 2011;25:48624.
doi:10.1021/ef2012084.
[91] Cabeza LF, Gutierrez A, Barreneche C, Ushak S, Fernndez G, Ins Ferndez
A, et al. Lithium in thermal energy storage: a state-of-the-art review. Renew
Sustain Energy Rev 2015;42:110612. doi:10.1016/j.rser.2014.10.096.
[92] U.S. Department of Energy. Solar energy technologies program. 2010.
[93] The University of Alabama. Novel molten salts thermal energy storage for
concentrating solar power generation 2010.
[94] Fernndez AG, Ushak S, Galleguillos H, Prez FJ. Development of new molten
salts with LiNO3 and Ca(NO3)2 for energy storage in CSP plants. Appl Energy
2014;119:13140. doi:10.1016/j.apenergy.2013.12.061.
[95] Maruoka N, Akiyama T. Thermal stress analysis of PCM encapsulation for heat
recovery of high temperature waste heat. J Chem Eng Japan 2003;36:7948.
doi:10.1252/jcej.36.794.
[96] Agyenim F, Hewitt N, Eames P, Smyth M. A review of materials, heat transfer
and phase change problem formulation for latent heat thermal energy storage
systems (LHTESS). Renew Sustain Energy Rev 2010;14:61528. doi:10.1016/
j.rser.2009.10.015.
[97] Erek A, Ylken Z, Acar MA. Experimental and numerical investigation of thermal
energy storage with a nned tube. Int J Energy Res 2005;29:283301.
doi:10.1002/er.1057.
[98] Castell A, Sol C, Medrano M, Roca J, Cabeza LF, Garca D. Natural convection
heat transfer coecients in phase change material (PCM) modules with
external vertical ns. Appl Therm Eng 2008;28:167686. doi:10.1016/
j.applthermaleng.2007.11.004.
[99] Zhang Y, Faghri A. Heat transfer enhancement in latent heat thermal energy
storage system by using an external radial nned tube. J Enhanc Heat Transf
1996;3:11927. doi:10.1615/JEnhHeatTransf.v3.i2.50.
[100] Lamberg P. Approximate analytical model for two-phase solidication problem
in a nned phase-change material storage. Appl Energy 2004;77:13152.
doi:10.1016/S0306-2619(03)00106-5.
[101] Velraj R, Seeniraj RV, Hafner B, Faber C, Schwarzer K. Experimental analysis
and numerical modelling of inward solidication on a nned vertical tube
for a latent heat storage unit. Sol Energy 1997;60:28190. doi:10.1016/S0038-
092X(96)00167-3.
[102] Talati F, Mosaffa AH, Rosen MA. Analytical approximation for solidication
processes in PCM storage with internal ns: imposed heat ux. Heat Mass
Transf 2010;47:36976. doi:10.1007/s00231-010-0729-9.
[103] Mosaffa A, Talati F, Rosen MA, Basirat TH. Phase change material solidication
in a nned cylindrical shell thermal energy storage: an approximate analytical
approach. Therm Sci 2013;17:40718.
[104] DISTOR project: D2.5 Report on material & geometry denition of
macroencapsulation 2005.
[105] Van Caneghem J, Brems A, Lievens P, Block C, Billen P, Vermeulen I, et al.
Fluidized bed waste incinerators: design, operational and environmental issues.
Prog Energy Combust Sci 2012;38:55182. doi:10.1016/j.pecs.2012.03.001.
[106] Brems A, Cceres G, Dewil R, Baeyens J, Piti F. Heat transfer to the riser-wall
of a circulating uidised bed (CFB). Energy 2013;50:493500. doi:10.1016/
j.energy.2012.10.037.
[107] Van de Velden M, Baeyens J, Seville JPK, Fan X. The solids ow in the riser of
a circulating uidised bed (CFB) viewed by positron emission particle
tracking (PEPT). Powder Technol 2008;183:2906. doi:10.1016/
j.powtec.2007.07.027.
[108] Smolders K, Baeyens J. Thermal degradation of PMMA in uidised beds. Waste
Manag 2004;24:84957. doi:10.1016/j.wasman.2004.06.002.
[109] Smolders K, Baeyens J. Overall solids movement and solids residence time
distribution in a CFB-riser. Chem Eng Sci 2000;55:410116. doi:10.1016/
S0009-2509(00)00084-1.
[110] Wu SY, Baeyens J. Effect of operating temperature on minimum
uidization velocity. Powder Technol 1991;67:21720. doi:10.1016/0032-
5910(91)80158-F.
[111] Izquierdo-Barrientos MA, Sobrino C, Almendros-Ibez JA. Energy storage with
PCM in uidized beds: modeling and experiments. Chem Eng J 2015;264:497
505. doi:10.1016/j.cej.2014.11.107.
[112] Peng H, Dong H, Ling X. Thermal investigation of PCM-based high temperature
thermal energy storage in packed bed. Energy Convers Manag 2014;81:4207.
doi:10.1016/j.enconman.2014.02.052.
[113] Zhang HZ, Wang XD. Synthesis and properties of microencapsulated
n-octadecane with polyurea shells containing different soft segments for heat
energy storage and thermal regulation. Sol Energy Mater Sol Cells
2009;93:136676. doi:10.1016/j.solmat.2009.02.021.
[114] Salunkhe PB, Shembekar PS. A review on effect of phase change material
encapsulation on the thermal performance of a system. Renew Sustain Energy
Rev 2012;16:560316. doi:10.1016/j.rser.2012.05.037.
[115] Hawlader MNA, Uddin MS, Khin MM. Microencapsulated PCM thermal-energy
storage system. Appl Energy 2003;74:195202. doi:10.1016/S0306-
2619(02)00146-0.
[116] Hawlader MNA, Uddin MS, Zhu HJ. Preparation and evaluation of a novel solar
storage material: microencapsulated paran. Int J Sol Energy 2000;20:22738.
doi:10.1080/01425910008914357.
[117] Zhao CY, Zhang GH. Review on microencapsulated phase change materials
(MEPCMs): fabrication, characterization and applications. Renew Sustain
Energy Rev 2011;15:381332. doi:10.1016/j.rser.2011.07.019.
[118] Jamekhorshid A, Sadrameli SM, Farid M. A review of microencapsulation
methods of phase change materials (PCMs) as a thermal energy storage (TES)
medium. Renew Sustain Energy Rev 2014;31:53142. doi:10.1016/
j.rser.2013.12.033.
[119] Li M. A nano-graphite/paran phase change material with high
thermal conductivity. Appl Energy 2013;106:2530. doi:10.1016/
j.apenergy.2013.01.031.
[120] Mesalhy O, Lafdi K, Elgafy A, Bowman K. Numerical study for enhancing the
thermal conductivity of phase change material (PCM) storage using high
thermal conductivity porous matrix. Energy Convers Manag 2005;46:84767.
doi:10.1016/j.enconman.2004.06.010.
[121] Elgafy A, Lafdi K. Effect of carbon nanober additives on thermal behavior of
phase change materials. Carbon N Y 2005;43:306774. doi:10.1016/
j.carbon.2005.06.042.
[122] Goodfellow. Metal thermal catalogue 2015.
[123] Zhao CY, Lu W, Tian Y. Heat transfer enhancement for thermal energy storage
using metal foams embedded within phase change materials (PCMs). Sol
Energy 2010;84:140212. doi:10.1016/j.solener.2010.04.022.
[124] Vadwala P. Thermal energy storage in metal foams lled with paran wax
2011.
[125] Li Z, Wu Z-G. Numerical study on the thermal behavior of phase change
materials (PCMs) embedded in porous metal matrix. Sol Energy 2014;99:172
84. doi:10.1016/j.solener.2013.11.017.
[126] Bauer CA, Wirtz RA. Thermal characteristics of a compact, passive thermal
energy storage device. Proc. 2000 ASME IMECE, Orlando, Florida, USA: 2000,
p. 17.
[127] Tong X, Khan JA, RuhulAmin M. Enhancement of heat transfer by inserting a
metal matrix into a phase change material. Numer Heat Transf Part A Appl
1996;30:12541. doi:10.1080/10407789608913832.
[128] Xiao X, Zhang P, Li M. Preparation and thermal characterization of paran/
metal foam composite phase change material. Appl Energy 2013;112:135766.
doi:10.1016/j.apenergy.2013.04.050.
[129] Fiedler T, chsner A, Belova IV, Murch GE. Thermal conductivity enhancement
of compact heat sinks using cellular metals. Defect Diffus. Forum 2008;273-
276:2226.
[130] Mehling H, Hiebler S, Ziegler F. Latent heat storage using a PCM-graphite
composite material. Proc. Terrastock 2000-8th int. conf. therm. energy storage,
Stuttgart, Germany: 2000, p. 37580.
[131] Py X, Olives R, Mauran S. Paran/porous-graphite-matrix composite as a high
and constant power thermal storage material. Int J Heat Mass Transf
2001;44:272737. doi:10.1016/S0017-9310(00)00309-4.
[132] Cabeza LF, Mehling H, Hiebler S, Ziegler F. Heat transfer enhancement in water
when used as PCM in thermal energy storage. Appl Therm Eng 2002;22:1141
51. doi:10.1016/S1359-4311(02)00035-2.
[133] Zhao YJ, Wang RZ, Wang LW, Yu N. Development of highly conductive
KNO3/NaNO3 composite for TES (thermal energy storage). Energy
2014;70:2727. doi:10.1016/j.energy.2014.03.127.
[134] Kibria MA, Anisur MR, Mahfuz MH, Saidur R, Metselaar IHSC. A review on
thermophysical properties of nanoparticle dispersed phase change materials.
Energy Convers Manag 2015;95:6989. doi:10.1016/j.enconman.2015.02.028.
[135] Warzoha RJ, Weigand RM, Fleischer AS. Temperature-dependent thermal
properties of a paran phase change material embedded with herringbone
 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
style graphite nanobers. Appl Energy 2015;137:71625. doi:10.1016/
j.apenergy.2014.03.091.
[136] ahan N, Fois M, Paksoy H. Improving thermal conductivity phase change
materials a study of paran nanomagnetite composites. Sol Energy Mater
Sol Cells 2015;137:617. doi:10.1016/j.solmat.2015.01.027.
[137] Yu J, Chen X, Ma XL, Song QF, Zhao YK, Cao JH. Inuence of nanoparticles and
graphite foam on the supercooling of acetamide. J Nanomater 2014;2014:110.
doi:10.1155/2014/313674.
[138] Jourabian M, Farhadi M, Sedighi K. On the expedited melting of phase change
material (PCM) through dispersion of nanoparticles in the thermal storage
unit. Comput Math Appl 2014;67:135872. doi:10.1016/j.camwa.2014.02.004.
[139] Motahar S, Nikkam N, Alemrajabi AA, Khodabandeh R, Toprak MS, Muhammed
M. A novel phase change material containing mesoporous silica nanoparticles
for thermal storage: a study on thermal conductivity and viscosity.
Int Commun Heat Mass Transf 2014;56:11420. doi:10.1016/
j.icheatmasstransfer.2014.06.005.
[140] Cingarapu S, Singh D, Timofeeva EV, Moravek MR. Nanouids with
encapsulated tin nanoparticles for advanced heat transfer and thermal energy
storage. Int J Energy Res 2014;38:519. doi:10.1002/er.3041.
[141] Sciacovelli A, Colella F, Verda V. Melting of PCM in a thermal energy storage
unit: numerical investigation and effect of nanoparticle enhancement. Int J
Energy Res 2013;37:161023. doi:10.1002/er.2974.
[142] Kashani S, Ranjbar AA, Madani MM, Mastiani M, Jalaly H. Numerical study
of solidication of a nano-enhanced phase change material (NEPCM) in a
thermal storage system. J Appl Mech Tech Phys 2013;54:70212. doi:10.1134/
S0021894413050027.
[143] Khodadadi JM, Fan L, Babaei H. Thermal conductivity enhancement of
nanostructure-based colloidal suspensions utilized as phase change materials
for thermal energy storage: a review. Renew Sustain Energy Rev 2013;24:418
44. doi:10.1016/j.rser.2013.03.031.
[144] Abolghasemi M, Keshavarz A, Mehrabian MA. Thermodynamic analysis of a
thermal storage unit under the inuence of nano-particles added to the phase
change material and/or the working uid. Heat Mass Transf 2012;48:196170.
doi:10.1007/s00231-012-1039-1.
[145] Guo CX, Wang YH. Numerical investigation of nanoparticle-enhanced high
temperature phase change material for solar energy storage. Adv. Mater. Res.
2012;512-515:9614.
[146] Wu S, Zhu D, Zhang X, Huang J. Preparation and melting/freezing characteristics
of Cu/paran nanouid as phase-change material (PCM). Energy and Fuels
2010;24:18948. doi:10.1021/ef9013967.
[147] Gong Z-X, Mujumdar AS. Cyclic heat transfer in a novel storage unit of multiple
phase change materials. Appl Therm Eng 1996;16:80715. doi:10.1016/1359-
4311(95)00088-7.
[148] Gong Z-X, Mujumdar AS. A new solar receiver thermal store for space-based
activities using multiple composite phase-change materials. J Sol Energy Eng
1995;117:215. doi:10.1115/1.2847798.
[149] Michels H, Pitz-Paal R. Cascaded latent heat storage for parabolic trough solar
power plants. Sol Energy 2007;81:82937. doi:10.1016/j.solener.2006.09.008.
[150] Seeniraj RV, Velraj R, Lakshmi Narasimhan N. Heat transfer enhancement study
of a LHTS unit containing dispersed high conductivity particles. J Sol Energy
Eng 2002;124:243. doi:10.1115/1.1488669.
[151] Seeniraj RV, Lakshmi Narasimhan N. Performance enhancement of a solar
dynamic LHTS module having both ns and multiple PCMs. Sol Energy
2008;82:53542. doi:10.1016/j.solener.2007.11.001.
[152] Wang J, Ouyang Y, Chen G. Experimental study on charging processes of a
cylindrical heat storage capsule employing multiple-phase-change materials.
Int J Energy Res 2001;25:43947. doi:10.1002/er.695.
[153] Wang J, Chen G, Zheng F. Study on phase change temperature distributions
of composite PCMs in thermal energy storage systems. Int J Energy Res
1999;23:27785.
[154] Cui H, Yuan X, Hou X. Thermal performance analysis for a heat receiver using
multiple phase change materials. Appl Therm Eng 2003;23:235361.
doi:10.1016/S1359-4311(03)00210-2.
[155] Peir G, Gasia J, Mir L, Cabeza LF. Experimental evaluation at pilot plant
scale of multiple PCMs (cascaded) vs. single PCM conguration for
thermal energy storage. Renew Energy 2015;83:72936. doi:10.1016/
j.renene.2015.05.029.
[156] Tian Y, Zhao CY. Thermal and exergetic analysis of metal foam-enhanced
cascaded thermal energy storage (MF-CTES). Int J Heat Mass Transf
2013;58:8696. doi:10.1016/j.ijheatmasstransfer.2012.11.034.
[157] Agraotis C, Roeb M, Sattler C. Cobalt oxide-based structured thermochemical
reactors/heat exchangers for solar thermal energy storage in Concentrated Solar
Power plants. Proc ASME 2014 8th int conf energy sustain collocated with
ASME 2014 12th int conf fuel cell sci eng technol 2014. doi:10.1115/ES2014-
6336. ISBN 978-0-7918-4586-8.
[158] Aydin D, Casey SP, Riffat S. The latest advancements on thermochemical heat
storage systems. Renew Sustain Energy Rev 2015;41:35667. doi:10.1016/
j.rser.2014.08.054.
[159] Carrillo AJ, Serrano DP, Pizarro P, Coronado JM. Thermochemical heat storage
based on the Mn2O3/Mn3O4 redox couple: inuence of the initial particle
size on the morphological evolution and cyclability. J Mater Chem A
2014;2:1943543. doi:10.1039/C4TA03409K.
[160] Carrillo AJ, Moya J, Bayn A, Jana P, de la Pea OShea VA, Romero M, et al.
Thermochemical energy storage at high temperature via redox cycles of Mn
and Co oxides: pure oxides versus mixed ones. Sol Energy Mater Sol Cells
2014;123:4757. doi:10.1016/j.solmat.2013.12.018.
39
[161] Michel B, Mazet N, Neveu P. Experimental investigation of an innovative
thermochemical process operating with a hydrate salt and moist air for thermal
storage of solar energy: global performance. Appl Energy 2014;129:17786.
doi:10.1016/j.apenergy.2014.04.073.
[162] Neveu P, Tescari S, Aussel D, Mazet N. Combined constructal and exergy
optimization of thermochemical reactors for high temperature heat
storage. Energy Convers Manag 2013;71:18698. doi:10.1016/
j.enconman.2013.03.035.
[163] Yan T, Wang RZ, Li TX, Wang LW, Fred IT. A review of promising candidate
reactions for chemical heat storage. Renew Sustain Energy Rev 2015;43:1331.
doi:10.1016/j.rser.2014.11.015.
[164] Meng Q-L, Lee C, Kaneko H, Tamaura Y. Solar thermochemical process for
hydrogen production via two-step water splitting cycle based on Ce1xPrxO2
redox reaction. Thermochim Acta 2012;532:1348. doi:10.1016/
j.tca.2011.01.028.
[165] Le Gal A, Abanades S. Catalytic investigation of ceria-zirconia solid solutions
for solar hydrogen production. Int J Hydrogen Energy 2011;36:473948.
doi:10.1016/j.ijhydene.2011.01.078.
[166] Fresno F, Fernndez-Saavedra R, Beln Gmez-Mancebo M, Vidal A, Snchez
M, Isabel Rucandio M, et al. Solar hydrogen production by two-step
thermochemical cycles: evaluation of the activity of commercial ferrites. Int
J Hydrogen Energy 2009;34:291824. doi:10.1016/j.ijhydene.2009.02.020.
[167] Meier A, Steinfeld A. Solar thermochemical production of fuels. Adv Sci Technol
2010;74:30312. doi:10.4028/www.scientic.net/AST.74.303.
[168] Charvin P, Stphane A, Florent L, Gilles F. Analysis of solar chemical processes
for hydrogen production from water splitting thermochemical cycles. Energy
Convers Manag 2008;49:154756. doi:10.1016/j.enconman.2007.12.011.
[169] Sol A, Fontanet X, Barreneche C, Fernndez AI, Martorell I, Cabeza LF.
Requirements to consider when choosing a thermochemical material
for solar energy storage. Sol Energy 2013;97:398404. doi:10.1016/
j.solener.2013.08.038.
[170] Zhang HL, Baeyens J, Degrve J, Huys K, Kong WB, Lv YQ. Thermo-chemical
heat storage for dispatching power generation on demand. Proceeding 13th
int. conf. energy storage, Beijing, China: 2015, p. 62.
[171] Williams OM, Carden PO. Energy storage eciency for the ammonia/
hydrogen-nitrogen thermochemical energy transfer. Int J Energy Res
1979;3(1):2940. doi:10.1002/er.4440030105.
[172] Brems A, Baeyens J, Beerlandt J, Dewil R. Thermogravimetric pyrolysis of waste
polyethylene-terephthalate and polystyrene: a critical assessment of kinetics
modelling. Resour Conserv Recycl 2011;55:77281. doi:10.1016/
j.resconrec.2011.03.003.
[173] Brems A, Baeyens J, Vandecasteele C, Dewil R. Polymeric cracking of waste
polyethylene terephthalate to chemicals and energy. J Air Waste Manage Assoc
2011;61:72131. doi:10.3155/1047-3289.61.7.721.
[174] Alonso E, Hutter C, Romero M, Steinfeld A, Gonzalez-Aguilar J. Kinetics of
Mn2O3-Mn3O4 and Mn3O4-MnO redox reactions performed under
concentrated thermal radiative ux. Energy and Fuels 2013;27:488490.
doi:10.1021/ef400892j.
[175] Pardo P, Anxionnaz-Minvielle Z, Roug S, Cognet P, Cabassud M. Ca(OH)2/CaO
reversible reaction in a uidized bed reactor for thermochemical heat storage.
Sol Energy 2014;107:60516. doi:10.1016/j.solener.2014.06.010.
[176] Schmidt M, Szczukowski C, Rokopf C, Linder M, Wrner A. Experimental
results of a 10 kW high temperature thermochemical storage reactor based
on calcium hydroxide. Appl Therm Eng 2014;62:5539. doi:10.1016/
j.applthermaleng.2013.09.020.
[177] Linder M, Rokopf C, Schmidt M, Wrner A. Thermochemical energy storage
in kW-scale based on CaO/Ca(OH)2. Energy Procedia 2014;49:88897.
doi:10.1016/j.egypro.2014.03.096.
[178] Tescari S, Agraotis C, Breuer S, de Oliveira L, Puttkamer MN, Roeb M, et al.
Thermochemical solar energy storage via redox oxides: materials and reactor/
heat exchanger concepts. Energy Procedia 2014;49:103443. doi:10.1016/
j.egypro.2014.03.111.
[179] Porisini FC. Salt hydrates used for latent heat storage: corrosion of metals and
reliability of thermal performance. Sol Energy 1988;41:1937. doi:10.1016/
0038-092X(88)90136-3.
[180] Singh IB, Sen U. Inuence of temperature and sulphate ion on corrosion of
mild steel in molten NaNo3. Br Corros J 1992;27:299304. doi:10.1179/
bcj.1992.27.4.299.
[181] Bradshaw RW, Goods SH. Corrosion of alloys and metals by molten nitrates,
SANDIA report. 2000.
[182] Baraka A, Abdel-Rohman AI, El Hosary AA. Corrosion of mild steel in molten
sodium nitratepotassium nitrate eutectic. Br Corros J 1976;11:446.
doi:10.1179/000705976798320313.
[183] Cabeza LF, Illa J, Roca J, Badia F, Mehling H, Hiebler S, et al. Immersion corrosion
tests on metal-salt hydrate pairs used for latent heat storage in the 32 to 36C
temperature range. Mater Corros 2001;52:1406. doi:10.1002/1521-
4176(200102)52:2<140::AID-MACO140>3.0.CO;2-R.
[184] Bradshaw RW, Goods SH. Corrosion resistance of stainless steels during thermal
cycling in alkali nitrate molten salts, SANDIA Report. 2001.
[185] Barlow P. Barlows formula. <http://en.wikipedia.org/wiki/Barlow%27s
_formula>2015 [accessed 05.05.15].
[186] Perry RH, Green DW. Perrys chemical engineers handbook. [Chapter 23]. 6th
ed. New York: McGraw-Hill Book Company; 1985.
[187] Specialmetals. INCONEL alloy 600 2008.
[188] ATI. Allegheny technologies incorporated. <http://www.atimetals.com>; 2015.
[189] Global Metals Pty Ltd. Global metals 2015. <http://www.globalmetals.com.au/>.
40 H. Zhang et al./Progress in Energy and Combustion Science 53 (2016) 140
[190] American Iron and Steel Institute. High temperature characteristics of stainless
steel, vol. 240. 2010 doi:10.1088/1742-6596/240/1/012071.
[191] Special metals. INCOLOY alloy 825 2004.
[192] Singh JP, Bansal NP, Goto T, Lamon J, Choi SR, Mahmoud MM, editors.
Processing and properties of advanced ceramics and composites iv: ceramic
transactions. Hoboken, New Jersey: John Wiley & Sons; 2012.
[193] Ananthakumar S, Manohar P, Warrier KGK. Microstructural features and
mechanical properties of Al2O3Al2TiO 5 composite processed by gel
assisted ceramic extrusion. Br Ceram Trans 2002;101:3843. doi:10.1179/
096797801125000393.
[194] Maddrell ER. Pressureless sintering of silicon carbide. J Mater Sci Lett
1987;6:4868. doi:10.1007/BF01756807.
[195] Guo XM, Yan YJ, Chen J, Huang ZR, Liu XJ. Mechanical properties and
microstructure of extruded SiC ceramics. J Inorg Mater 2009;24:11558.
doi:10.3724/SP.J.1077.2009.01155.
[196] RHI AG. RHI. <http://rhi-ag.com/internet_en/>; 2015.
[197] Ortech Advanced Ceramics. Ceramic seal. <http://www.ortechceramics.com/
ceramic_seal.htm>; 2015.
[198] GraphiteStore.com. Technical ceramics. <https://www.graphitestore.com/
cat.asp/spcat_id/4>; 2015.
[199] McDanel Ceramics. Ceramics. <http://www.mcdanelceramics.com/
products.html>; 2015.
[200] Bradshaw RW, Siegel NP. Molten nitrate salt development for thermal energy
storage in parabolic trough solar power systems. In: Es2008 Proc 2nd int conf
energy sustain, vol. 2. 2009. p. 6317.
[201] Zhang HL, Degrve J, Dewil R, Baeyens J. Operation diagram of circulating
uidized beds (CFBs). Procedia Eng 2015;102:1092103. doi:10.1016/
j.proeng.2015.01.232.
[202] Geldart D. Types of gas uidization. Powder Technol 1973;7:28592.
doi:10.1016/0032-5910(73)80037-3.
[203] Abrahamsen AR, Geldart D. Behaviour of gas-uidized beds of ne powders
part I. Homogeneous expansion. Powder Technol 1980;26:3546. doi:10.1016/
0032-5910(80)85005-4.
[204] Bi HT, Grace JR. Flow regime diagrams for gas-solid uidization and upward
transport. Int J Multiph Flow 1995;21:122936. doi:10.1016/0301-
9322(95)00037-X.
[205] Punwani DV, Modi MV, Tarman PB. Generalized correlation for estimating
choking velocity in vertical solids transport. Conf. int. powder bulk solids handl.
process. conf., Chicago, USA 1976.
[206] Brook N. Fluid transport of coarse solids. Min Sci Technol 1987;5:197217.
doi:10.1016/S0167-9031(87)90415-4.
[207] Baeyens J, Geldart D. An investigation into slugging uidized beds. Chem Eng
Sci 1974;29:25565. doi:10.1016/0009-2509(74)85051-7.
[208] Flamant G. Theoretical and experimental study of radiant heat transfer in a
solar uidized-bed receiver. AIChE J 1982;28:52935. doi:10.1002/
aic.690280402.
[209] Martin J, Vitko J. ASCUAS: a solar central receiver utilizing a solid thermal
carrier. 1982.
[210] Bataillie D, Laguerie C, Royere C, Gauthier D. Echangeurs de chaleur gaz-solide
lits uidiss multi-tages dans le domaine des moyennes et hautes
tempratures. Entropie 1989;25:11326.
[211] Flamant G, Hernandez D, Bonet C, Traverse J-P. Experimental aspects of the
thermochemical conversion of solar energy; decarbonation of CaCO3. Sol
Energy 1980;24:38595. doi:10.1016/0038-092X(80)90301-1.
[212] Siegel NP, Ho CK, Khalsa SS, Kolb GJ. Development and evaluation of a
prototype solid particle receiver: on-sun testing and model validation. J Sol
Energy Eng 2010;132:021008. doi:10.1115/1.4001146.
[213] Tan T, Chen Y. Review of study on solid particle solar receivers. Renew Sustain
Energy Rev 2010;14:26576. doi:10.1016/j.rser.2009.05.012.
[214] Rger M, Amsbeck L, Gobereit B, Buck R. Face-down solid particle receiver
using recirculation. J Sol Energy Eng 2011;133:031009. doi:10.1115/1.4004269.
[215] Warerkar S, Schmitz S, Goettsche J, Hoffschmidt B, Reiel M, Tamme R.
Air-sand heat exchanger for high-temperature storage. J Sol Energy Eng
2011;133:021010. doi:10.1115/1.4003583.
[216] Al-Ansary H, El-Leathy A, Al-Suhaibani Z, Jeter S, Sadowski D, Alrished A, et al.
Experimental study of a sandair heat exchanger for use with a high-
temperature solar gas turbine system. J Sol Energy Eng 2012;134:041017.
doi:10.1115/1.4007585.
[217] Mahmoudi S, Chan CW, Brems A, Seville J, Baeyens J. Solids ow diagram of
a CFB riser using Geldart B-type powders. Particuology 2012;10:5161.
doi:10.1016/j.partic.2011.09.002.
[218] Smolders K, Baeyens J. Elutriation of nes from gas uidized beds: mechanisms
of elutriation and effect of freeboard geometry. Powder Technol 1997;92:35
46. doi:10.1016/S0032-5910(97)03214-2.
[219] Bodin S, Briens C, Bergougnou M, Patureaux T. Standpipe ow modeling,
experimental validation and design recommendations. Powder Technol
2002;124:817. doi:10.1016/S0032-5910(01)00473-9.
[220] Smolders K, Geldart D, Baeyens J. The physical models of cyclone diplegs in
uidized beds. Chinese J Chem Eng 2001;9:33747.
[221] Chan CW, Seville JPK, Fan X, Baeyens J. Solid particle motion in a standpipe
as observed by Positron Emission Particle Tracking. Powder Technol
2009;194:5866.
[222] Flamant G, Hemati M. Dispositif collecteur dnergie solaire. French Patent
No. 1058565, 2010.
[223] Falcone PK, Noring JE, Hruby JM. Assessment of a solid particle receiver for a
high temperature solar central receiver system. 1985.
[224] Fend T, Pitz-Paal R, Reutter O, Bauer J, Hoffschmidt B. Two novel high-porosity
materials as volumetric receivers for concentrated solar radiation. Sol Energy
Mater Sol Cells 2004;84:291304. doi:10.1016/j.solmat.2004.01.039.
[225] Gokon N, Yamamoto H, Kondo N, Kodama T. Internally circulating uidized
bed reactor using m-ZrO[sub 2] supported NiFe[sub 2]O[sub 4] particles for
thermochemical two-step water splitting. J Sol Energy Eng 2010;132:021102.
doi:10.1115/1.4001154.
[226] Xu C, Song Z, Chen L, Zhen Y. Numerical investigation on porous media heat
transfer in a solar tower receiver. Renew Energy 2011;36:113844.
doi:10.1016/j.renene.2010.09.017.
[227] Jeter SM, Stephens JH. System and method of thermal energy storage. US2012/
0132398A1, 2011.
[228] Zipf V, Neuhuser A, Willert D, Nitz P, Gschwander S, Platzer W.
High temperature latent heat storage with a screw heat exchanger: design
of prototype. Appl Energy 2013;109:4629. doi:10.1016/j.apenergy.2012.
11.044.
[229] Matsubara K, Kazuma Y, Sakurai A, Suzuki S, Soon-Jae L, Kodama T, et al.
High-temperature uidized receiver for concentrated solar radiation by a
beam-down reector system. Energy Procedia 2014;49:44756. doi:10.1016/
j.egypro.2014.03.048.
[230] Ma Z, Glatzmaier GC, Mehos M. Development of solid particle thermal energy
storage for concentrating solar power plants that use uidized bed technology.
Energy Procedia 2014;49:898907. doi:10.1016/j.egypro.2014.03.097.
[231] Zhang HL, Baeyens J, Degrve J, Brems A, Dewil R. The convection heat transfer
coecient in a circulating uidized bed (CFB). Adv Powder Technol
2014;25:71015. doi:10.1016/j.apt.2013.10.018.
[232] Ordez F, Caliot C, Bataille F, Lauriat G. Optimization of the optical particle
properties for a high temperature solar particle receiver. Sol Energy
2014;99:299311. doi:10.1016/j.solener.2013.11.014.
[233] Fernndez P, Miller FJ. Performance analysis and preliminary design
optimization of a small particle heat exchange receiver for solar tower
power plants. Sol Energy 2015;112:45868. doi:10.1016/j.solener.2014.
11.012.
[234] Zanganeh G, Pedretti A, Haselbacher A, Steinfeld A. Design of packed bed
thermal energy storage systems for high-temperature industrial process heat.
Appl Energy 2015;137:81222. doi:10.1016/j.apenergy.2014.07.110.
[235] Chan CW, Seville JPK, Parker DJ, Baeyens J. Particle velocities and their residence
time distribution in the riser of a CFB. Powder Technol 2010;203:18797.
doi:10.1016/j.powtec.2010.05.008.
[236] Zhang H, Degrve J, Baeyens J, Dewil R. The voidage in a CFB riser as function
of solids ux and gas velocity. Procedia Eng 2015;102:111222. doi:10.1016/
j.proeng.2015.01.234.