Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: [Link]/locate/rser

Review on concentrating solar power plants and new developments

in high temperature thermal energy storage technologies
Ming Liu a,n, N.H. Steven Tay a, Stuart Bell b, Martin Belusko a, Rhys Jacob a, Geoffrey Will b,
Wasim Saman a, Frank Bruno a
a
Barbara Hardy Institute, School of Engineering, University of South Australia, Mawson Lakes Boulevard, Mawson Lakes, SA 5095, Australia
b
Energy and Process Engineering, Chemistry, Physics, Mechanical Engineering, Science and Engineering Faculty, Queensland University of Technology,
Brisbane 4001, Qld., Australia

art ic l e i nf o a b s t r a c t

Article history: A concentrating solar power (CSP) system converts sunlight into a heat source which can be used to drive
Received 24 June 2015 a conventional power plant. Thermal energy storage (TES) improves the dispatchability of a CSP plant.
Received in revised form Heat can be stored in either sensible, latent or thermochemical storage. Commercial deployment of CSP
2 September 2015
systems have been achieved in recent years with the two-tank sensible storage system using molten salt
Accepted 17 September 2015
as the storage medium. Considerable research effort has been conducted to improve the efficiency of the
CSP system and make the cost of electricity comparable to that of the conventional fossil-fuel power
Keywords: plant. This paper provides a comprehensive summary of CSP plants both in operation and under con-
Concentrating solar power (CSP) struction. It covers the available technologies for the receiver, thermal storage, power block and heat
Thermal energy storage
transfer fluid. This paper also reviews developments in high temperature TES over the past decade with a
Sensible heat storage
focus on sensible and latent heat storage. High temperature corrosion and economic aspects of these
Latent heat storage
Phase change material systems are also discussed.
& 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1412
2. Concentrating solar power plants and their thermal energy storage systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1412
2.1. Description of operational and under-constructed CSP plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1412
2.2. Thermal energy storage (TES) technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1414
2.2.1. Sensible heat storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1414
2.2.2. Latent heat storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1416
2.2.3. Thermochemical storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1417
3. Recent developments in thermal energy storage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1418
3.1. Development of molten salts and other liquid storage media/HTFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1418
3.2. Development of solid sensible storage media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1420
3.3. Development of latent heat thermal storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1421
3.3.1. Encapsulated PCMs (EPCMs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1422
3.3.2. Heat pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1422
3.3.3. Mobile PCM storage systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1425
3.3.4. Cascaded latent heat storage system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1425
4. Compatibility of containment materials with storage media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1426
4.1. Common corrosion mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1426
4.1.1. Nitrates, carbonates, sulphates and hydroxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1426
4.1.2. Chlorides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1427
4.2. Methods of evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1427
4.3. Corrosion mitigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1427

n
Corresponding author. Tel.: þ 61 8 83025132; fax: þ 61 8 83023380.
E-mail address: [Link]@[Link] (M. Liu).

[Link]
1364-0321/& 2015 Elsevier Ltd. All rights reserved.
1412 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

5. Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1427
6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1428
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1429
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1429

1. Introduction Firstly, a summary is provided on the current status of the CSP

technology as well as CSP plants that are currently operating or
Due to the worldwide population growth and industrialization, under construction. Current work on sensible heat storage
the demand of energy has been dramatically increasing. Renewable emphasizes the development of low cost liquid or solid storage
energy, which can be obtained from natural resources such as wind, media. Latent heat storage using PCMs are subject to the thermal
solar, biomass and geothermal, has been encouraged by many conductivity limitation particularly during discharge. The technol-
countries due to the advantages of being sustainable and not con- ogies which allow for the heat transfer improvement are presented,
tributing to the world's CO2 greenhouse gas emissions. However, including increasing the heat transfer area by encapsulation, adding
there is a mismatch between most renewable energy supplies and heat pipes, using mobile systems and cascaded systems. Thermo-
user demand. The easy integration of thermal energy storage (TES) chemical storage is a very promising technique for long-term sto-
makes concentrating solar power (CSP) dispatchable and unique rage, but at this stage it is still far from large-scale applications. Only
among all other renewable energy generating alternatives. potential chemical reactions and candidate reactants applicable to
CSP technologies generate electricity by concentrating the solar CSP systems are discussed. High-temperature corrosion of the
radiation beam onto a small area, where a heat transfer fluid (HTF) containment material presents major challenges in all TES types,
is heated up and this energy is ultimately transferred to the steam. and is also addressed in this work. The corrosion mechanism of
Electricity is then generated by an electric generator which is driven some inorganic salts is reviewed.
by a steam turbine with the efficiency limited by the Carnot cycle. A
TES system offers a highly efficient heat storage method as an
intermediate step for electricity generation. Depending on the size 2. Concentrating solar power plants and their thermal energy
of the TES system, it is able to mitigate the short load fluctuations storage systems
and shift or extend the energy supply. CSP with TES presents an
economic advantage if it supplies electricity at times of greatest 2.1. Description of operational and under-constructed CSP plants
need which generally has the highest electricity tariff [1].
One of the main barriers of wider adoption of CSP technology is The CSP technology can be classified into parabolic trough,
the high cost of solar generated electricity. The CSP generating cost linear Fresnel, tower and parabolic dish, according to the way that
is expected to reduce by 50–60% through technical improvement they focus the sun's ray and whether the receiver is fixed or
and scaling up and volume production [5]. Considerable research mobile. In parabolic trough and linear Fresnel systems, the mirror
effort has been continuously funded by various countries to make tracks the sun along one axis (line focus) and in tower and dish
CSP more cost effective, for example: systems, the mirror tracks the sun along two axes (point focus).
The receiver is fixed in linear Fresnel and tower systems and it is
 The European DISTOR project funded by the European Com- mobile in parabolic trough and dish systems. For each technology,
mission to demonstrate high temperature latent heat storage various options exist for the HTF, TES technology and power cycle.
systems for integration in CSP plants using absorbers with The CSP plants that are currently operating and being con-
direct steam generation [2]. structed have been reviewed. Details of their solar collector con-
 The SunShot Initiative launched by the United States Depart- figuration, solar field operating conditions, TES systems and cooling
ment of Energy in 2011. The target of the SunShot Initiative for methods have been summarized in Table 1. Figs. 1 and 2 present the
CSP is to reduce the levelized cost of CSP-generated electricity CSP capacity in various countries and for the different types of CSP
to less than USD$0.06/kW h by 2020 with the cost of thermal technologies, for solar plants that are both operational and under
storage less than USD$15/kW ht and the exergetic efficiency construction. The information has been obtained from a project
greater than 95% [3]. listing hosted by the National Renewable Energy Laboratory (NREL)
 The Australian Solar Thermal Research Initiative (ASTRI) funded ([Link] As of March 2015, the CSP
by the Australian Government. The goal is to lower the cost of market has a total capacity of 5840 MWe worldwide, among which
solar thermal power to AUD$0.12/kW h by 2020. 4800 MWe is operational and 1040 MWe is under construction.
Spain had a total operational capacity of 2405 MW and 100 MW is
Thermal energy for CSP systems can be stored in sensible heat under construction, making it the world's leading country in CSP.
storage by using either solid or liquid storage media, latent heat USA followed Spain, having a total capacity of 1795 MW. Interest
storage by using phase change materials (PCMs) and thermo- has grown in India, Chile, South Africa, Australia, China and a few
chemical storage through reversible chemical reactions. When Middle East countries. Parabolic trough systems are currently the
designing a TES system for a CSP plant, the following factors need most proven CSP technology and dominate the global market, being
to be considered: (1) nominal temperature and specific enthalpy installed in more than 80% of the CSP plants in operation and under
drop in load; (2) maximum load; (3) operational strategy and construction. Tower systems have just started to be introduced in to
(4) integration into the power plant [4]. A number of studies have commercial applications and linear Fresnel plants are currently
investigated the cost of thermal storage systems, considering the making the transition to commercial applications. Parabolic dishes
various elements involved which include the storage media, tank are at the early demonstration stage.
and other costs. Slightly less than half of the installed CSP capacity is integrated
The aim of this paper is to provide a state-of-the-art review of with thermal storage. With the maturity of molten salt storage
the development of TES technologies applicable to CSP systems over technology, over 80% of the capacity under construction has energy
the last decade, with the focus on sensible and latent heat storage. storage. This percentage increases to 88% in trough and tower
Table 1
Representative features of the different CSP technologies for current and future CSP plants.

Current trough Current tower Current linear Fresnel Current dish 2020 SunShot Target

Maturity High, commercially proven Medium, recently commercially Medium, pilot plants, commercial pro- Low, demonstration projects –
proven jects under construction

M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

Typical plant capacity 100 (MW) 50–100 (MW) 50 (MW) 3–30 each (kW) 200
Operating temperature of 290–390 290–565 250–390 550–750 Receiver outlet4 650 °C
solar field (°C)
Plant peak efficiency (%) 14–20 23–35 18 31.25
Annual average conversion 13–15 14–18 9–13 22–24
efficiency (%)
Collector concentration 70–80 41000 460 (depends on secondary reflector) 4 1300 –
(suns)
Power block cycle and fluid Superheated steam Rankine, steam Superheated steam Rankine, Saturated steam Rankine (steam @ Stirling/Brayton Supercritical steam Rankine, supercritical
conditions @380 °C/100 bar steam @ 540 °C/100–160 bar 270 °C/55 bar), superheated steam Ran- CO2 Brayton cycle (600 °C/200 bar), air
kine (steam @ 380 °C/50 bar) Brayton cycle
Power cycle efficiency (%) 37.7 41.6 – – 50–60
Heat transfer fluid Synthetic oil, water/steam (DSG), mol- Water/steam, molten salt, air Water/steam Air, hydrogen, helium Molten salt with lower melting points, air,
ten salt (demonstration), air (demonstration) steam, supercritical CO2
(demonstration)
Annual capacity factor (%) 20–25 (no TES) 40–45 (6–7.5 h TES) 22–24  25 66.6
40–53 (6 h TES) 65–80 (12–15 h TES)
Storage system Indirect two-tank molten salt storage Direct two-tank molten salt Short-term pressurized water storage No storage for Stirling dish, 14 h TES
(293–393 °C) storage (290–565 °C), (Ruths tank) chemical storage under
development
Capital cost (USD/kW) 4700–7300 (no TES, OECD countries) 6400–10,700 (with TES) – – 3770 Solar fieldo $75/m2; receiver and
3100–4050 (no TES, non-OECD HTF o $150/kW; power block o $1200/
countries) kW; TESo $15/kW h
6400–10,700 (with TES)
LCOE (USD/kW h) 0.26–0.37 (no TES) 0.2–0.29 (6–7.5 h TES) 0.19–0.38 (no TES) – 0.06
0.22–0.34 (with TES) 0.17–0.24 (12–15 h TES) 0.17–0.37 (6 h TES)
Cooling method Wet Wet, dry Dry Dry Dry
Suitable for air cooling Low to good Good Good Best –
Water requirement (m3/ 3 (wet) 1.8–2.8 (wet) 3.8 (wet)  0.08 (mirror washing)
MW h) 0.4–1.7 (hybrid) 0.3 (dry) 0.3–1 (hybrid)
0.3 (dry)

OECD: Organization of Economic Co-operation and Development.

1413
1414 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

salt is separately stored in the hot and cold tanks. It was reported
that the roundtrip energy efficiency can achieve up to 98% for the
storage system at the Solar Two power tower demonstration [8].
The operating temperature range of the storage system is
dependent on the solar field technologies. The current parabolic
trough and power tower technology can provide HTF at tem-
peratures of 393 °C and 565 °C, respectively, which result in a
storage temperature range of 292–385 °C and 290–565 °C,
respectively [9]. Extensive studies have been carried out to
increase the operating temperature of the solar field. For example,
over 650 °C was targeted in the SunShot program. The newly
developed volumetric receiver facilitates the operation at tem-
peratures over 1000 °C [10]. Higher operating temperature will
Fig. 1. Worldwide CSP capacity by December 2014 as in [Link] enable the possibility to increase the overall solar-to-electricity
solarpaces/. efficiency, reduce thermal storage volume and decrease the leve-
lized cost of electricity (LCOE) [11]. Table 1 outlines the repre-
sentative features of different CSP technologies for current and
next-generation CSP technologies [12–18].

2.2. Thermal energy storage (TES) technologies

TES technology solves the time mismatch between solar energy

supply and electricity demand, which provides a distinct advan-
tage to CSP plants compared to other renewable energies, such as
wind and photovoltaic. Also, electrical energy storage using bat-
tery has not proved to be economically viable [19]. Depending on
the daily and yearly variation of solar radiation and the electricity
demand profile, CSP plants integrated with thermal storage sys-
tems can have various operational strategies and the storage sys-
tem may offer the following functions [4,12]: (1) mitigating short
Fig. 2. CSP capacity categorized by receiver technologies and with/without storage.
fluctuations during transient weather conditions, e.g. cloudy per-
iods; (2) shifting the generation period from peak hours of solar
systems. While the current thermal storage technology used in insolation to peak hours of power demand; (3) extending the
linear Fresnel plants is short-term pressurized steam storage generation period when solar is not available, which improves the
(o1 h), molten salt is promising for long-term storage. Novatec and annual capacity factor and requires a larger solar field than a
BASF SE have started to demonstrate the linear Fresnel collector system without storage.
system (550 m absorber) integrated with direct molten salt storage
s
[6] with a commercial product – HITEC . Initially, they focused on 2.2.1. Sensible heat storage
salt-specific challenges such as salt freezing and re-melting, TES can be classified as sensible heat storage, latent heat sto-
degradation and corrosion issues and so far no major point of rage and thermochemical storage. In sensible TES systems, the
concern has been found [6]. energy is stored/released by raising/lowing the temperature of the
The annual solar to electricity efficiency is a significant para- storage medium in either a solid or liquid. The amount of energy
meter to evaluate a CSP system. The efficiency improvement stored is dependent on the quantity of the storage material (m),
results in a cost reduction for CSP plants. Fig. 3 presents the the specific heat capacity of the material (Cp) and the temperature
anticipated annual efficiency for various CSP technical options as change (ΔT) and is determined by Eq. (1). All TES systems cur-
well as the maturity of the technology [7]. In the current com- rently installed in CSP plants are using sensible storage materials,
mercial CSP plants, tower systems with molten salt as both the e.g. oil, molten salt, steam, ceramic and graphite.
HTF and the storage material are the most efficient option. The
Q ¼ mC p ΔT ð1Þ
annual efficiency of linear Fresnel systems with saturated/super-
heated steam is 9–13%, which is the lowest among those technical The state-of-the-art storage material is molten salts, for
options. For tower systems, the emerging options, which are example, Hitec/Hitec XL and Solar Salt have been commercially
expected to increase the efficiency from the current 18% to above deployed in parabolic trough and tower systems. Their key prop-
23%, include (1) using supercritical steam or carbon dioxide as the erties and costs are compared with those of a synthetic oil –
HTF and (2) using pressurized air as the HTF to drive a combined Therminol VP-1 and they are listed in Table 2. Synthetic oil was a
cycle (upper Brayton cycle and lower Rankine cycle) plant. How- common working fluid in trough plants before molten salts
ever, they are at a very early stage of development [7]. emerged. Solar Salt is relatively cheap and has the highest max-
The current state-of-the-art TES technology integrated into the imum operating temperature, however it freezes at 220 °C, which
parabolic trough and power tower plants is the two-tank sensible requires quite expensive anti-freeze systems. Hitec and Hitec XL
energy storage using a molten salt comprising of sodium nitrate have a more desirable freezing temperature of 120–140 °C and
and potassium nitrate (60–40 wt%). Fig. 4 presents the schematic they are able to withstand a temperature limit of over 500 °C.
diagram of the two-tank molten salt storage system. The cold HTF Abengoa Solar [20] analysed the feasibility, cost and performance
(e.g. molten salt, water/steam and synthetic oil) is heated up from of a parabolic trough plant with 6 h of storage. The modelling
the solar field and the thermal storage unit is charged by the hot results showed that the molten salt plant can reduce the storage
HTF either directly or indirectly through heat exchangers. When cost by up to 43.2%, solar field cost by up to 14.8% and LCOE by
the stored energy is needed, the system operates in reverse to 9.8–14.5% relative to an oil plant with Therminol VP-1. By repla-
generate steam to run the power plant. The hot and cold molten cing the synthetic oil with the molten salt as the working fluid in
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1415

Fig. 3. Annual solar-to-electricity efficiency as a function of development level [7].

Fig. 4. Thermal storage system integrated in the CSP plant with solar field and power block [1]: (a) direct heat storage and (b) indirect heat storage.
1416 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

Table 2
Properties and cost for three different molten salts and Therminol VP1 [20,22].

Salt Composition (wt%) Melting point Maximum operation temperature Cost, (USD Cost, (USD$/kW h) Supplier
(°C) (°C) $/kg)

Solar Salt NaNO3–KNO3(60–40) 220 585 0.49 5.8 (ΔT ¼200 K) Chilean Nitrate or Coastal
Chemical
Hitec NaNO3–KNO3–NaNO2 (7–53– 142 450-538 0.93 10.7 (ΔT ¼ 200 K) Coastal Chemical
40) 610–710 under different atmo-
spheres [23]
Hitec XL NaNO3–KNO3–Ca(NO3)2 (7– 120 480–505 1.43 18.2 (ΔT ¼ 200 K) Coastal Chemical
45–48)
Therminol VP-1 – – 400 3.96 57.5 (ΔT ¼ 100 K) Solutia

the trough system, it allows for a higher solar field outlet tem-
perature, which will enable a higher Rankine power block effi-
ciency and a lower cost energy storage system [21].
Two-tank molten salt sensible storage is the most commonly
used storage technique in utility-scale CSP plants. The same fluid
operates as the HTF and the storage medium in the direct storage
system (Fig. 4a). In the indirect storage system (Fig. 4b), the HTF
operating in the solar field is coupled by means of an intermediate
heat exchanger to a different storage medium. Generally, parabolic
trough plants use the indirect storage approach, where the cold
molten salt is heated up in the heat exchanger by the oil HTF
delivered from the solar field. When the solar field cannot provide
heat, the hot salt can be used to heat the oil. This technique has
been demonstrated in the Andasol 1 plants in Spain and has been
widely applied [23]. Some parabolic trough and most of the tower
plants, which employ molten salt as the HTF, use the direct storage
approach. Direct storage removes the need for a heat exchanger
and hence reduces the cost and increases the overall efficiency.
The feasibility of this approach has been demonstrated in the
5 MW Archimede parabolic trough plant in Italy [9] and the Solar
Two tower plant in California [23].
A single tank thermocline storage, which eliminates one tank, Fig. 5. An ideal thermocline storage tank with a movable thermal insulation baffle.
enables a potential cost reduction of 35% compared to the two-
tank storage [24,25]. In thermocline systems, the hot and cold
fluids are stored in the same tank and they are separated due to
the thermal stratification [26]. An insulation baffle can be used to
keep the hot fluid and the cold fluid insulated from each other and
the baffle moves horizontally inside the tank, holding the hot fluid
above it and the cold fluid below (Fig. 5). This concept has been
patented and the first plant employing this technique will be
implemented in Spain [7]. Also, research has been conducted to
use low-cost filler materials to reduce the overall required amount
of the relatively higher cost molten salt storage medium [24,27–
30]. Thermoclines of this type as presented in Fig. 6 are named
dual-media or multi-media storage systems. The major obstacles
in the design and operation of dual-media thermoclines are:
(1) flow distribution [31]; (2) the stability of the filler material in
the hot fluid (molten salt or oil) environment and (3) thermo-
mechanically induced ratcheting [32]. It requires the appropriate
Fig. 6. Schematic illustration of a molten-salt thermocline tank, including the
selection of compatible filler material for the hot fluid [24,29,33–
porous quartzite rock bed and the liquid heel [28].
35] and also it is significant that both thermal and mechanical
considerations are investigated to avoid the potential ratcheting
[32,36]. The studied filler materials are introduced in Section 4.2.
energy converted during the phase change, latent storage system
2.2.2. Latent heat storage offers an isothermal heat storage/releasing process and a higher
Latent TES using phase change materials (PCMs), which take storage density compared to the sensible storage system with the
advantage of the large latent heat that can be stored/released by same temperature change. It potentially enables a smaller and
changing a material from one phase to another (generally solid to lower cost storage system compared to the sensible storage sys-
liquid), has attracted considerable attention for CSP applications tem. Currently, no commercial PCM storage system has been used
over the past decade [37–43]. The amount of the energy stored is in utility-scale CSP applications.
governed by both the specific heat and the phase change enthalpy  
and can be calculated by Eq. (2). Due to the significant quantity of Q ¼ m C p;s ðT m  T s Þ þ h þC p;l ðT l –T m Þ ð2Þ
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1417

Where Cp,s and Cp,l are the average specific heats of the PCM in 2.2.3. Thermochemical storage
solid and liquid phases, respectively, Tm is the melting tempera- In thermochemical storage, the energy is stored by means of
ture, h is the phase change enthalpy, and Ts and Tl are tempera- the reversible sorption process or chemical reaction:
tures of the solid and liquid PCM, respectively. AB þ Heat 3 A þB ð3Þ
Due to the high temperature requirements for CSP systems,
inorganic salts/salt eutectics and metals/metal alloys are potential In the endothermic reaction, the chemical reactant AB absorbs
PCMs [9,44–47]. Salts have been the most studied PCMs to reduce the heat supplied from the solar field and is dissociated into two
the cost of thermal storage. However, salts have low thermal products A and B which can be stored separately. In the reverse
conductivity, which limits the heat transfer between the HTF and process (exothermic reaction), chemicals A and B are put in touch
the PCM, particularly during discharging. Considerable efforts have and the initial reactant AB is formed with a heat release. The
been made to enhance the heat transfer within PCM storage sys- amount of heat stored in a thermochemical process is given by Eq.
tems, including extending the heat transfer surface by encapsu- (4) [12].
lating the PCM [48–50] or adding fins to the wall of the heat Q ¼ ar mΔH ð4Þ
exchanger tubes [42,51,52], adding heat pipes or thermosiphons in
the system [53,54] and composing high thermal conducting where, ar is the fraction reacted, m is the mass of the reactant AB
materials into the PCM, e.g. magnesium oxide [55,56] and graphite and ΔH is the heat of reaction per unit mass.
[57–59]. Less expensive metals and metal alloys containing zinc, Thermochemical storage possesses high energy density and
aluminum, copper, magnesium and silicon, have also been inves- negligible heat loss, which potentially offers a long-term storage
tigated [44,60–63]. Since they are high thermal conducting option with relatively small storage volume. The sorption process
materials, heat transfer is not a concern. Recently, a 1 MW h pro- is usually able to store low and medium grade heat at a tem-
totype storage system has been constructed and tested by the perature of less than 400 °C and the heat of reaction is typically
German Aerospace Center [64,65]. It is a three-part storage system between 20 and 70 kJ/mol [68]. The investigated processes include
for direct steam generation up to 500 °C, combining concrete metal salts with water, ammonia, methanol or methyl-ammonia
storage to preheat water, a PCM module for water evaporation and and metal alloys with hydrogen [68]. Generally, chemical reaction
concrete storage to superheat the steam. Encapsulated PCMs were allows the energy to be stored at over 400 °C with a higher heat of
investigated as the filler materials in the dual-media storage sys- reaction from 80 to 180 kJ/mol [68]. Various chemical reactions
tem. However, so far investigations are limited to numerical have been studied for potential thermochemical storage plants
modelling due to the difficulties and high cost associated with and they are summarized by Pardo et al. [69] and Pardo et al. [70]
high temperature experiments [49,66,67]. The encapsulated PCM in their work. Fig. 7 illustrates the investigated thermochemical
storage system is introduced in Section 3.3.1. storage systems and the candidate reactants that are promising for
CSP applications [69]. The volumetric energy density and mass
energy density of the studied reactants versus their turning tem-
peratures have been plotted in Fig. 8 and it enables the quick
screening of the candidate reactants in the desired temperature
and the energy density ranges [69]. The volumetric energy density
of thermochemical TES is 5–10 times higher than latent and sen-
sible TES respectively [69].
Compared to sensible and latent heat storage, thermochemical
storage is still at a very early stage of development and so far most
of the systems were only tested on laboratory scale for short dis-
charging period [68,71]. Several challenges need to be addressed
in future research on thermochemical storage: (1) the most sui-
table chemical reaction should be selected and the criteria sug-
gested by Wentworth and Chen [72] can be used as the selection
Fig. 7. Six candidate thermochemical storage systems at temperatures of 300– criteria [69]; (2) effective heat exchangers/reactors need to be
1100 °C [69]. investigated to enhance the mass and heat transfer as they are
poor [73,74]; and (3) advanced composite materials with high

Fig. 8. Energy density versus turning temperature for the reversible reactions: (a) volumetric energy density and (b) mass energy density [69].
1418 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

storage density, high thermal conductivity, good cyclic ability and Peng et al. [91] measured the thermo-physical properties of
low cost are desired [71]. Large-scale experiments are required to NaNO3–NaCl–NaNO2 systems with various compositions. The
prove the efficiency and economic feasibility of the thermo- eutectic mixture has a melting point of 228.13 °C and a stable/
chemical storage system for both short and long-term storage. decomposition temperature of 553.85 °C. Peng et al. [92] added
5 wt% chloride salt into Hitec and experimentally proved that the
new salt had better thermal stability and a lower freezing point
3. Recent developments in thermal energy storage and the optimum working temperature was increased from 500 °C
to 550 °C.
To achieve more cost-effective TES systems, there are many Ionic liquids (ILs), composed of a family of salts and having lower
promising technical developments, including: (1) development of freezing temperatures (o100 °C), are promising to replace the
new nitrate/nitrite molten salts which can operate over a wider conventional synthetic oil/molten salts as the working fluid for the
range of temperatures; (2) development of cheap solid heat sto- parabolic trough CSP plants [93,94]. Other advantages of ILs include
rage media with good physical properties; (3) development of new high heat capacity, high density, high thermal and chemical stabi-
PCMs and highly efficient latent energy storage system and lity, low vapor pressure and non-harmfulness [95]. A number of ILs
(4) identification of thermo-chemical processes suitable for the have been investigated by Wu et al. [96], Paul et al. [93], Fredlake
temperature range of CSP plant [7]. This work only covers the et al. [97] and Every et al. [98]. Moens et al. [95] summarized the
sensible and latent heat storage development . potential candidate ILs for parabolic trough systems, their thermal
stability and key issues for the development of new ILs, e.g. avail-
3.1. Development of molten salts and other liquid storage media/ ability, cost, material compatibility and environmental safety and
HTFs health. They pointed out that ILs containing halide as an anion are
not suitable for CSP technology because of thermal cleavage reac-
For most sensible TES systems, the cost will be reduced with a tions [95]. With an assumed price of ILs (USD$4.7/kg), the estimated
wider operational temperature range. At a given storage capacity, cost of a thermocline storage system using ILs is comparative with
the amount of salt required for a storage system in tower plants is that using molten salts and is more cost effective than using Ther-
approximately one third of that in parabolic trough plants as the minol VP-1 as the working fluid and the storage material [94].
tower can operate at higher temperatures. The desirable features Perissi et al. [99] investigated the corrosion behavior of several
of the storage media/HTFs include: (1) low freezing temperature containment materials (copper, nickel, AISI 1018 steel, brass and
and/or high maximum operational temperature; (2) high heat Inconel 600) to a typical IL-[C4mim][Tf2N] at 325 °C for 48 h and
capacity, thermal conductivity and density; (3) good thermal sta- pitting corrosion was detected on all of the metal/metal alloy
bility; (4) low corrosion to the containment material and (5) low samples, while at room temperature (25 °C), all materials under
cost [75]. investigation are corrosion resistant to [C6mim][PF6], [C8mim][PF6]
Extensive research has been carried out in this field. and [C4mim][Tf2N] [100]. Therefore, further containment material
Table 3 provides the material tests that have been conducted for investigation is required to use those ILs at high temperatures.
the molten salt development. LiNO3 has been added into NaNO3– Recently, research has been conducted on utilizing nano-
KNO3 to decrease the freezing temperature as well as to improve particles to enhance the specific heat capacity, thermal con-
the thermal stability [76]. A quaternary nitrate salt (LiNO3– ductivity and density of conventional oil HTFs, molten salts and
NaNO3–KNO3–Ca(NO3)2) was studied [77,78] and it is able to ILs. The effect of improvement is dependent on the material, shape
extend the operational temperature range from 130 to 550 °C. The and size of the particle as well as the type of the base fluid.
corrosion effects of the salt were studied at 390 °C on two low Cr However, the mechanism for the enhanced specific heat capacity
alloy steels (T11 and T22) and a carbon steel without Cr (A516 – a is still under investigation and has not been clearly explained
construction material in CSP plants). The 1–2% of Cr in T11 and T22 [101]. Four different stages of digestive ripening techniques were
reduces the corrosion behavior in this quaternary nitrate salt and studied by Singh et al. [102] to disperse the copper nano-particles
the increased composition of Cr in T22 helps to form a protective into two synthetic HTFs. They found that the combined tempera-
film. Sixteen quaternary nitrate salts with various compositions of ture and ultrasonic ripening processes should be used to accel-
KNO3, NaNO3, LiNO3 and Ca(NO3)2  4H2O were prepared and their erate the particle homogenization process and the stability of the
thermo-physical properties were measured [79]. It was found that suspension improves as the particle size decreases. Dudda and
the mixed salt has the most favorable properties when their mass Shin [101,103] investigated the effect of silica nano-particle size on
ratio is 6:1:2:1. Fernández et al. [80] investigated different addi- the specific heat of the Solar Salt and results showed that the
tions of Ca(NO3)2 and LiNO3 to the binary salt NaNO3/KNO3 and enhancement is increased with the larger particle size. The
analyzed three candidate ternary eutectic nitrate salt mixtures. enhancement can reach 27% for the 1 wt% 60 nm silica material.
The thermal stability of the nitrate salt mixture containing Ca With the same nano-material of  20 to 30 nm (1 wt%), the spe-
(NO3)2 decreases with an increasing ratio of Ca(NO3)2 [81]. The cific heat capacity of a eutectic alkali chloride salt (BaCl2, NaCl,
addition of Ca(NO3)2 reduces the corrosion rate on both A1 and CaCl2 and LiCl) is improved by 14.5% [104]. Silica-particles with a
T22 steels compared to the Solar Salt [82]. Wang et al. [83] added a nominal size of 10 nm is able to increase the specific heat of a
congruently melting compound (2KNO3  Mg(NO3)2[84]) into the eutectic Li2CO3 and K2CO3 salt by 22.37% with a concentration of
LiNO3–NaNO3–KNO3 ternary system to make a eutectic salt. This 1.0 wt% [105] and by 118–124% with a concentration of 1.5 wt% in
new salt has a melting temperature of 100.8 °C, which is roughly liquid state [106]. Also, microscopy results showed that the
17 °C lower than that of the LiNO3–NaNO3–KNO3 ternary system. nanofluids were fairly stable with minimal agglomeration of the
By using a thermodynamic modelling technique followed by silica nano-particles after nine repeated thermal cycles [107].
experimental verification, the University of Alabama [83–88] Multi walled carbon nano-tubes (nominal size 1.5 μm length and
developed two low melting temperature quaternary nitrate/nitrite 10–30 nm diameter) was also examined as an additive in the same
salts and one ternary nitrate salt and the thermodynamic model eutectic salt and it was found that the augmentation reaches a
was validated experimentally using the LiNO3–NaNO3–KNO3 maximum (17.91%) with 0.5 wt% of nano-tubes [105]. The specific
ternary system [89]. It can be expected that the quaternary reci- heat capacity of the eutectic Li2CO3 and K2CO3 salt was improved
procal system K, Li, Na//NO2, NO3 with a melting temperature by 23% with the addition of 1 wt% of TiO2-particles (  20 to
lower than 80 °C is feasible [90]. 30 nm) [108]. The Al2O3-particles having sizes below 50 nm was
Table 3
Summary of molten salt development.

Material composition (wt%) Key properties Maximum opera- Stability test Material compatibility Ref.
tion temp. (°C)

LiNO3–NaNO3–KNO3–Ca(NO3)2 m.p.¼ 132 °C; Cp ¼ 1.518 J/g K; 553 The weight loss of this quaternary nitrate salt is 3% at a Gravimetric test, SEM and XRD examination on steel spe- [77,78,78]

M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

(10–20–60–10) μ¼ 0.013 Pa s @300 °C. temperature of 552.68 °C. cimens. T22 is more favourable to contain the salt than T11
and A516.
LiNO3–NaNO3–KNO3–Ca m.p.¼ 88.6 °C; 4500 °C Decomposition temperature 4600 °C. Underway. [79]
(NO3)2  4H2O (20–10–60–10) Cp ¼1.50 J/g K;
μo 0.005 Pa s;
ρ ¼1.75–1.95 g/m3;
k ¼0.530 W/m K.
NaNO3–KNO3–Ca(NO3)2 (15–43– m.p.¼ 130 °C; 550 The weight loss of this nitrate salt mixture is 3% at a tem- [80]
42) Cp ¼1.272 J/g K; perature of 554.39 °C.
μ¼ 0.014 Pa s @300 °C.
NaNO3–KNO3–LiNO3 (28–52–20) m.p.¼ 130 °C; 568 The weight loss of this nitrate salt mixture is 3% at a tem- Corrosion behaviour was studied at 390 °C for 2000 h on [76,80]
Cp ¼1.091 J/g K; perature of 600.05 °C. T22 and A1 steels and the results showed that T22 is better
μ¼ 0.03 Pa s @300 °C. than A1.
NaNO3–KNO3–LiNO3 (20.06–54.1– m.p.¼ 118 °C; 430 The weight loss was investigated from 200 to 500 °C and it [88,110]
25.9) Cp ¼1.63 J/g K @500 °C; starts to decompose at 435 °C. LiNO3 is unstable at high
temperatures and it decomposes to Li2O2 and Li2O.
LiNO3–KNO3–Ca(NO3)2 (30–60– m.p.¼ 132 °C; 560 The weight loss of this nitrate salt mixture is 3% at a tem- [80]
10) Cp ¼1.395 J/g K; perature of 567.18 °C.
μ¼ 0.03 Pa s @300 °C.
NaNO3–KNO3–Ca(NO3)2 (16–48– m.p.¼ 102.3 °C; 4500 The salt was tested at 500 °C during 500 h under air and the [81]
36) μ E 0.09 Pa s @200 °C. weight loss is less than 2.5%.
LiNO3–NaNO3–KNO3–2KNO3  Mg m.p.¼ 102.3 °C; [83]
(NO3)2 (17.77–15.28–35.97– Cp ¼1.67 J/g K @500 °C;
30.98)
LiNO3–NaNO3–KNO3– NaNO2 m.p.¼ 94.3 °C; 425 The weight loss was investigated from 200 to 500 °C and it The SS316L sample was immersed into the molten salt at [85–87]
(17.5–14.18–50.53–17.78) ΔH¼ 16.79 J/g; was found that the salt starts to decompose at 430 °C. 420 °C for 748 h. The formation of lithium iron oxide pre-
Cp ¼1.66 J/g K @500 °C; vents the SS316L from severe corrosion and SS316L is
ρ ¼1.715 g/m3 @500 °C. considered as a compatible containment material.
Hitec þ 5% chloride salt m.p.¼ 140 °C; 550 The weight loss of Hitec is 10% at 550 °C after 30 h and weight [92]
Cp ¼2.32 J/g K; loss of Hitec þ 5% chloride salt is 7% at 550 °C after 30 h.
μ¼ 0.02 Pa s @300 °C;
k ¼1.3061 W/m K @300 °C.

1419
1420 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

Low cost recycled shaped ceramics, which can be produced

from industrial vitrification of asbestos containing wastes, have
been studied by many researchers. Ceramics can be directly
shaped at the exit of the furnace without any additional energy
consumption [34] and the material is obtained as ceramics and
glass forms. The study indicated that the ceramic form is more
appropriate for CSP applications [118–120] as it presents com-
parable storage capacity (2.5–3.2 MJ/(m3 K)) to synthetic high
temperature ceramics [119]. With this filler material, the molten
salt in the conventional plants (e.g. ANDASOL 1) can be potentially
reduced by 8 times with the same storage capacity [34]. Calvet
et al. [34] studied the compatibility of a commercial form ceramic,
s
Cofalit , with Solar Salt and HITEC XL at 500 °C for 500 h. Fig. 10
presents pictures of the surface of (a) the reference sample that
did not undergo any test; (b) the sample tested in the HITEC XL
Fig. 9. Packed bed thermal storage tank (diameter: 263 mm; height: 550 mm) with and (c) the sample tested in the Solar Salt. The isothermal testing
s
ceramic particles [35]. results showed that the Cofalit ceramic is thermo-chemically
stable in Solar Salt and the authors recommended that the
able to increase the specific heat of two ionic liquids ([C4mim] dynamic testing with a flowing Solar Salt must be conducted to
[NTf2] and [C4mpyrr][NTf2]) to roughly 23% from 25 °C to 275 °C verify this filler material. Motte et al. [121] investigated the fea-
[109]. sibility of six different industrial waste ceramics in a direct molten
salt storage configuration and they found that the Cofalit, coal fly
3.2. Development of solid sensible storage media ashes and laboratory blast furnace slags are compatible with the
conventional Solar Salt. Another post-industrial waste, electric arc
Sensible energy storage using solid media is especially suitable furnace slags from the steelmaking industry (also called “black
when air is used as the working fluid in the solar receiver because slags”), was shaped in the form of plates and thermally cycled in
air can be employed directly as the working fluid in the storage hot air [122]. This material has a slightly lower storage capacity
system and the temperature constraint due to the chemical (2.3 MJ/(m3 K)) than ceramic form [122].
instability between the air and the storage media is eliminated Pacheco et al. [29] evaluated 17 minerals through the com-
[111]. However, due to the lower volumetric heat capacity and patibility test (isothermal test) with Hitec XL and Solar Salt and the
thermal conductivity of the air compared to the liquid HTF, this thermal cycling test up to 553 cycles in Hitec XL. The test facility
system will result in higher pressure drops and larger sized storage for thermal cycling test was shown in Fig. 11. The minerals and
tank [111]. Solid media can also be employed as filler materials molten salts were contained in a chamber with a diameter of
into the nitrate salts – thermocline storage system to prevent 10 cm and a height of 30 cm and this chamber was placed between
convective mixing and reduce the amount of salts required. The the hot and cold salt tanks (400 °C and 290 °C, respectively), each
filler material should be inexpensive, widely available, compatible having 15 cm in diameter and 2 m in height. Quartzite and silica
with the nitrate salts and non-hazardous and also have a high heat filter sand were selected as the most wear-resistant filler materials
capacity and a low void fraction [29]. The commonly investigated based on their performance, availability and cost. To verify the
system is the packed bed/dual-media thermocline system using operation and performance of a molten salt thermocline system
rocks or other ceramic materials [35]. Fig. 9 shows a packed bed with quartzite rock and silica sand as filler materials, a 2.3 MW h
sensible storage unit with ceramic particles. Most of the research pilot-scale test equipment was designed and constructed, which is
work in this area is on the development of numerical models [112] present in Fig. 12 [123]. This storage tank is made from carbon
and only a few include experimental work. steel and has a diameter of 3.0 m and a height of 6.1 m. In total,
In the packed bed/dual-media thermocline storage system, the 49.9 t of rock and 21.9 t of sand were loaded into this thermocline
solid material is loosely packed and the heated fluid (air or molten tank. The charging and discharging tests showed that the ther-
salt) is circulated through the storage material. During the char- mocline storage system is a feasible option for thermal storage in a
ging process, the heated fluid flows from the solar field into the parabolic trough plant.
top of the tank (i.e. upper plenum) and then passes evenly down Following Pacheco et al. work [29], Brosseau et al. [123] con-
through the porous medium (heating up the storage material) and ducted the long-term cycling test in the 285–450 °C temperature
passes out through the lower plenum. Cold fluid is drawn off at the range to simulate a 30-year plant life. Hitec XL, quartzite rock and
bottom and returned to the solar field. When the bed is fully silica sand were used in the test and it took 14 months to complete
charged, its temperature becomes uniform. When the stored the 10,000 thermal cycles. The results revealed that: (1) there was
energy is needed, cold fluid enters from the bottom and flows to no significant deterioration of the stainless steel basket material;
the top of the tank. Then the fluid is heated and delivered to the (2) the majority of the rock and sand were still tightly packed;
heat exchanger to heat up the steam or directly to the air Brayton (3) the deterioration of rock and sand filler materials in both iso-
cycle turbine if air is used. thermal and thermal cycling tests was considered minimal.
Concrete and cast ceramics have been extensively investigated Thermal ratcheting is one of the critical problems when
because of their low cost, good thermal conductivity and moderate designing a thermocline system. Thermal ratcheting is a process by
specific heat [113,114]. Rocks, metals, concrete, sand and brick which the tank wall of the thermocline is progressively expanded
were recommended by Hasnain [115] as high temperature storage from cyclic operation leading to catastrophic failure of the wall
media. Rock or pebble bed storage, having a typical characteristic material [124]. If the thermal expansion coefficient of the tank wall
size of 1–5 cm, can be used for temperatures up to 1000 °C. The is greater than that of the filler particles, a radial gap is created by
use of metal media can be advantageous where high thermal the thermal expansion during the charge cycle and the filler par-
conductivity is required and the cost is of secondary importance. ticles settle down to fill this gap. When the tank is then cooled
Graphite [116] and copper powder [117] were studied as additives during the discharge cycle, the tank wall cannot shrink to its ori-
to increase the thermal conduction of the concrete. ginal diameter due to the gravity of the particle. It will result in a
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1421

Fig. 10. Pictures of (a) reference ceramic sample (b) sample tested in HITEC XL and (c) sample tested in Solar Salt.

Fig. 11. Schematic and photograph of thermal cycling test rig to test selected candidate minerals [29].

mechanical/hoop stress proportional to the strain of the initial

expansion. If this stress exceeds the yield strength of the wall
material, plastic deformation occurs and the wall effectively
“ratchets” to a new diameter. These ratchets may accumulate with
each cycle until the ultimate strength of the material is exceeded,
causing the tank to rupture [125]. To avoid thermal ratcheting,
temperature oscillations in the tank wall should be minimized by
inclusion of insulation layer and the wall material should be highly
resilient to accommodate the stress [124]. So far, thermal ratcheting
has been investigated via computational fluid dynamics (CFD)
simulation by Flueckiger et al. [32,126] and Yang and Garimella
[127,128]. Another approach to avoid this problem is to replace the
packed bed with structured materials, e.g. concrete [129] and
ceramic [130]. One of the studies compared the cost and efficiency
of the structured concrete and packed-bed thermocline and the
simulation result suggested that the structured concrete thermo-
cline is slightly costly ($34 vs. $30/kW h) and less efficient (84% vs.
Fig. 12. Photograph of thermocline test equipment [123].
92.37%) [129].

and discharging processes. One of the major obstacles of using

3.3. Development of latent heat thermal storage
latent heat storage is the low thermal conductivity of the inorganic
Compared to sensible heat storage, latent heat thermal storage salt PCMs, which obstructs the heat transfer between the PCM and
exhibits desirable characters for CSP applications due to its high the HTF. For CSP applications, the energy to charge the latent heat
energy density and the isothermal behaviour during the charging storage system is delivered from the solar collector and then the
1422 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

working fluid for the power cycle is heated up by the storage sys- melt cycles. A more recent study showed that the capsule has
tem in the discharge process. For both processes, a small tem- stood the test of 2000 cycles [138]. The preliminary cost estima-
perature difference is desirable to minimise the exergy losses and to tion indicated that the TES system based on the developed PCM
improve the efficiency of the power cycle. Reducing the thermal capsules can produce energy at a cost of USD$14/kW hth[137].
resistance between the PCM and the HTF is a critical issue and the By using an electroplating technique, copper balls with a size of
techniques under development include: increasing the area for heat 2–3 mm in diameter with a melting temperature of 1083 °C was
transfer by packaging or utilizing extended surfaces, compositing coated by nickel [139,140]. To avoid the mixing between the
high conducting materials into the PCM, embedding heat pipes as copper and nickel, carbon, ruthenium and chromium were studied
thermal conduits, employing mobile systems and cascaded PCMs. as the inhibition layer [139,140]. The inhibition material and the
PCM/coating material should be immiscible or their solubility
3.3.1. Encapsulated PCMs (EPCMs) should be very minimal. The thermal cycling test demonstrated
In most cases, the PCM needs to be encapsulated in order to hold that the thick layer of carbon and ruthenium (4940 μm) was
the liquid phase PCM and also to avoid the PCM contacting with the effective against the copper volume expansion [139]. The thickness
environment [131]. In building applications where the temperature of the film can be controlled by the operation time [139]. An in-
is generally below 30 °C, the PCM can be packaged in tubes, pou- house equipment (shown in Fig. 15) was made by Zhang et al.
ches, spheres, panels or other receptacles [132] and they are com- [140] to enable the electroplating of chromium layer on copper
mercially available [133]. For high temperature applications, steel, and a commercial electroplating equipment was used to coat the
nickel, sodium silicate, silicon dioxide, calcium carbonate and tita- nickel film after applying the chromium layer. As a result, the
nium dioxide have been identified as potential shell materials [134]. copper ball was uniformly encapsulated with chromium and
Selection of the containment/shell material is critical as most of the nickel layers with a thickness of 100 μm and 400 μm, respectively.
high temperature PCMs are inorganic salts and they are corrosive to The encapsulated capsule has been subject to 1000 freeze-melt
common containment materials [9]. The criteria for containment cycles between 1050 °C and 1150 °C and no leakage was found
material selection have been discussed in [135,136]. The spherical from the outside. The scanning electron microscope (SEM) image
capsule is the mostly investigated encapsulation for high tem- demonstrated that the chromium layer was cracked after repeated
perature applications as it offers a high ratio of area for heat transfer cycles due to its poor ductility. However, it still behaved as an
to volume and it is easy to be packed into the storage tank with effective barrier to restrain the mixing between copper and nickel.
good bed porosity. The EPCM can be manufactured by batch Eutectic aluminium–silicon (88–12 wt%) alloy, which has a very
assemble and individual assemble [1]. Large numbers of EPCMs can high phase change enthalpy of 560 kJ/kg and a melting tempera-
be made in a single process in batch assemble and they are typically ture of 576 °C [62], is a very promising PCM. Mobley and Rapp
smaller in size. Individual assemble is a much more labor-intensive [141] proposed a number of methods to apply silicon coating on
manufacturing process and the size of the EPCM produced is this eutectic alloy to form the self-encapsulation. Bennett and
typically large. Braunstein [142] investigated the stability of the silicon on the
Core-to-shell ratio determines the mechanical strength of the eutectic alloy under conditions of thermal cycling with an upper
EPCM and also it affects the heat transfer performance of the PCM temperature of 600 °C, 650 °C and 700 °C. A section of cycled
[136]. Pendyala [135] from the University of South Florida suc- samples was examined by optical micrography and electron
cessfully encapsulated sodium nitrate with silicon dioxide as microscopy techniques. As a result, this self-encapsulated Al-Si
shown in Fig. 13. The spheres were made by using a pelletizer. PCM is not feasible at/above 700 °C due to the large penetration of
Before coating the salt with silicon dioxide, the PCM pellet was eutectic into the silicon [142]. A kind of clay and ceramic container
coated with a layer of high temperature polymer and cured, so that was chosen to encapsulate this alloy and no corrosion resulted
the pellet becomes insoluble in water and several organic solvents between the container material and the alloy [62].
[135]. The silicon dioxide is then coated using self-assembly,
hydrolysis and simultaneous chemical oxidation at various tem- 3.3.2. Heat pipes
peratures [135]. Researchers from the same institute developed an An effective approach to increase the heat transfer rate
automated industrial encapsulation process by using a laboratory between the PCM and the HTF is to use embedded heat pipes/
development model of an industrial coating system (LDCS) [137]. It thermosiphons, which can efficiently transfer heat by means of
is reported that they succeeded in preparing PCM capsules with evaporation and condensation using an internal working fluid
metal coating. The capsules have undergone 200 freeze-melt [53,143,144]. At the hot side of the heat pipe (evaporator), the
thermal cycles between 250 °C and 325 °C and there is no leak- liquid internal working fluid absorbs heat and turns into vapour
ing observed and the discolouring of the metal coat is due to the and creates a high pressure. Due to the pressure difference, the
oxidation of the outer metal layer. Fig. 14 shows (a) LDCS coater; vapour travels along the heat pipe to the cold side (condenser) and
(b) a prepared PCM capsule and (c) the capsule after 200 freeze- condenses back into the liquid. Then the liquid returns to the hot

Fig. 13. (a) Schematics of PCM encapsulation by silicon oxide; (b) polymer coated spherical pellets [138].
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1423

Fig. 14. (a) A picture of LDCS coater; (b) a PCM capsule with metal coating and (c) a PCM capsule after 200 thermal cycles [137].

Fig. 16. The schematic diagram of a heat pipe latent heat thermal storage system
[148]: (1) heat pipe; (2) longitudinal fins; (3) storage chamber; (4) PCM; (5) cold
fluid flow passage (condenser) and (6) hot fluid flow passage (evaporator).

variable energies, e.g. solar energy. By expanding the studies of

Horbaniuc et al. and Liu et al., Jung and Boo [151] developed a
Fig. 15. Schematic diagram of the chromium periodic-barrel electroplating equip- thermal resistance network model to enable the design of a real
ment [140]. CSP application. The model was based on a staggered tube bank
configuration and the PCM temperature and the heat transfer rate
side by means of capillary action, centrifugal force or gravity and were obtained by a row-to-row heat transfer analysis. Only con-
the cycle continues. The heat transfer behaviour of the heat pipe duction was considered in the model and the discrepancies were
has been investigated by using numerical models with varying found to be less than 8% when comparing the experimental and
complexity [54,145–147]. However, only limited experimental theoretical PCM temperatures. Considering the natural convection
work has been conducted and most of the work has been focused in the PCM, Shabgard et al. [152] modelled the heat transfer per-
on low-temperature, small-scale storage systems. formance of a large-scale high temperature heat pipe PCM storage
Horbaniuc et al. [148] studied the effect of the number of system in both charging and discharging processes. The size and
longitudinal fins attached to the heat pipe on the total discharge the capital cost of a PCM storage system with staggered heat pipes
time of the storage system by using an analytical model. The was estimated for a 50 MWe parabolic trough CSP plant [38] and
proposed heat pipe PCM storage system is shown in Fig. 16. Liu was compared with a two-tank sensible storage system of the
et al. [149,150] extended the work of Horbaniuc et al. and con- same capacity developed by Herrmann et al. [153]. The result
structed a prototype storage system with five heat pipes and a low demonstrated that the heat pipe PCM storage system potentially
melting temperature PCM. They operated the system in three allows for a capital cost reduction by at least 15%.
modes: charging only (only hot fluid flows through the passage), The National Aeronautics and Space Administration (NASA) of
discharging only (only cold fluid flows through the passage) and USA proposed a novel concept of a heat pipe receiver with high
simultaneous charging and discharging (both hot and cold fluids temperature PCMs for an advanced solar dynamic power genera-
flow through their corresponding passages). The testing result tion system [154], shown in Fig. 17. The heat pipe, which uses a
indicated that this proposed device can perform the designed liquid metal (e.g. sodium) as the internal working fluid, transfers
functions and the flexible operation mode allowed for the con- the solar flux to the PCM and circulates fluid in the receiver. This
tinuous operation of systems using time dependent and intensity apparatus comprises three portions: the receiver, the thermal
1424 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

storage device (TSD) and the heat source heat exchanger (HSHX). A shell-and-tube PCM storage system integrated with heat
Twenty four heat pipes were lined along the side wall of the pipes has been investigated by Shabgard et al. [157] and Nithya-
cylindrical container and passed through the three portions. Each nandam and Pitchumani [39,158]. The two configurations of the
heat pipe was rounded by 24 PCM annular canisters, which were storage system are presented in Fig. 18. In configuration 1
separated by a ceramic fibre spacer and axisymmetric to the heat (Fig. 18a), the HTF passes through tubes with PCM in the shell side
and in configuration 2 (Fig. 18b) the HTF passes over tubes that
pipe axis. The thermal performance of the heat pipe receiver and
contain the PCM. A unit cell can be defined and referred to Module
the influence of void cavity inside the PCM canister were studied
1 (Fig. 18c) for configuration 1 and Module 2 (Fig. 18d) for con-
by Gui et al. [155,156] using a CFD software – Fluent. The simu-
figuration 2, respectively. The heat transfer in the two storage
lation results showed that the heat pipe receiver alleviates thermal
configurations in both charging and discharging has been mod-
spots and thermal ratcheting and the void cavity reduces the elled by Shabgard et al. [157] using a thermal resistance network.
usability and thermal storage ability of PCM storage. s
In the analysis, potassium nitrate and Therminol were employed
as the PCM and the HTF, respectively. The effect of the number of
heat pipes in each Module (¼ 1, 2, 3, 4) as well as their orientation
(horizontal and vertical) was also studied. This work quantified the
augmentation associated with two terms: heat pipe effectiveness (ε)
and generalized heat pipe effectiveness (~ε ). ε is defined as the ratio
of the amount of energy stored/extracted from the PCM system
with heat pipes to that without heat pipes. ε~ is defined as (ε  1)/
number of heat pipes and it indicated the enhancement effect of
the associated heat pipes. The analysis suggested that for Module
1, ε is slightly affected by the orientation of heat pipes and ε~ is
nearly independent of both the number and orientation of heat
pipes. For Module 2, ε is significantly influenced by the heat pipe
orientation and the thermal performance of Module 2 can exceed
that of Module 1 with use of less numbers of heat pipes.
Following Shabgard et al. modelling work [157], Nithyanandam
and Pitchumani [39] optimised the design of the system with four
heat pipes (two vertical and two horizontal) for maximizing
energy transferred, effectiveness and energy transfer rate during
charging and discharging processes and they presented the opti-
mum design conditions for the two modules. The optimum con-
ditions showed that the heat pipe effectiveness (ε) is significantly
higher for Module 1 and the heat transfer rate during charging and
discharging of the PCM is higher for Module 2. Nithyanandam and
Pitchumani [158] conducted a detailed heat transfer analysis on
Fig. 17. The configuration of (a) the heat pipe receiver and (b) a unit [155].
the two modules based on the three-dimensional computational

Fig. 18. Two configurations of heat pipe PCM storage systems: (a) the PCM surrounds the HTF tubes; (b) the HTF passes over tubes containing PCM; (c) Module 1 and
(d) Module 2 [157].
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1425

modelling using FLUENT and the different arrangements of heat et al. [159] experimentally investigated a dynamic melting process
pipes under investigation are shown in Fig. 19a–e. As a result, of a tube-in-tank PCM system. A U-tube system with a pre-melt
Module 1 with 4 heat pipes, Module 2 with 4 heat pipes and tube was embedded in the frozen PCM to allow for an initial melt
Module 2 with 2 vertical heat pipes arrangement is preferred to path in the PCM. The melted PCM was circulated using another
maximize the energy transfer capability, the power density and pump to create dynamic melting of the PCM. By applying this
the power density per unit heat pipe, respectively. In another dynamic melting concept, the heat transfer rate between the PCM
work, the dynamic performance of a PCM storage system with two and the HTF can be doubled without any reduction in compactness
horizontal heat pipes (Module 1) was simulated subjected to four factor. Tay et al. [160] further studied this dynamic system using a
cyclic charging and discharging process and it was found that CFD model and the analysis showed the dynamic melting can
equal durations of charging and discharging allows for a more maintain the storage system at a higher local effectiveness for
efficient operation [144]. majority of the phase change process. Pointner et al. [161,162]
suggested a PCM Flux concept. The PCM is encapsulated in the flat
3.3.3. Mobile PCM storage systems containers that are arranged as parallel layers as shown in Fig. 20
A new concept was proposed recently to enhance the heat and the containers are moved slowly in one direction for charging
transfer between the PCM and the HTF by actively transporting the and in the other one for discharging. A screw heat exchanger (SHE)
presented in Fig. 21 was employed by Zipf et al. [163] to transport
PCM along the heat transfer area and the flow of the liquid PCM
the PCM during the phase change and the concept was proved by
enables forced convection and hence increases the overall heat
conducting experiments on two prototypes. In the prototype, the
transfer coefficient. In conventional stationary PCM systems, when
PCM used was the eutectic mixture of NaNO3 and KNO3 having a
discharging the layer of the solidified PCM on the heat exchanger
melting temperature of 221 °C. Thermal oil and steam were used
surfaces grows with time and the thermal resistance between the
as heat transfer media for the two prototypes respectively. The
PCM and the HTF increases, resulting in a decreased heat transfer
PCM was transported along the heat transfer area from left to right
rate. If the PCM in the system is made mobile then this problem inside the SHE and it was simultaneously heated or cooled by the
can be overcome. The flow rate of the PCM could be controlled to heat transfer medium. The molten PCM and solid PCM were stored
reach the required nominal power output from CSP plants. Tay separately in two tanks.

3.3.4. Cascaded latent heat storage system

Utilizing more than one PCM in a cascaded configuration can
improve the heat transfer rate of the storage system during both
charging and discharging processes [164]. Fig. 22 shows a sche-
matic diagram of a cascaded shell-and-tube PCM storage system.
In this type of system, a few modules containing different PCMs
are connected to each other in series and the melting temperature
decreases in the charging process (PCM melts) and increases in the
discharging process (PCM solidifies). In a single PCM system, the
temperature of the HTF decreases in the direction of flow in the
charging process, resulting in a decreased heat flux to the PCM and
the heat transfer rate reduces. When the heat transfer rate drops
below a certain quantity, it is not able to produce the required
nominal power output from CSP plants. However, in a cascaded
configuration, although the temperature of the HTF reduces in the
flow direction in the charging process, the temperature difference
between the PCM and the HTF can be maintained almost constant
if the PCM is properly chosen. This leads to an almost constant
heat flux into the PCM and a stable power output from the plant.
The benefits of employing multiple PCMs include enhancing the
heat transfer rate, achieving constant HTF outlet temperature
[165,166], improving the exergy efficiency [152,167,168] and
increasing the effectiveness of extracting the stored energy
[165,169]. The exergy efficiency can be maximized through the
utilization of infinite numbers of PCMs. However, three or four

Fig. 19. Schematic diagram of the different arrangements of heat pipes: (a) 2 ver-
tical heat pipes (2-VHP); (b) 2 horizontal heat pipes (2-HHP); (c) 3 heat pipes (3-
*HP); (d) 3 heat pipes (3-YHP) and (e) 4 heat pipes (4-HP). Fig. 20. PCM Flux module and the structure of the stack [161].
1426 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

2OH  2H2O þO2  (8)

2
According to the Lux-Flood model, the O is the basic com-
ponent and considered analogous to the H þ ion in aqueous solu-
tion, whilst the corresponding oxides are the conjugate acids. The
basicity of the melt can be defined by pO2  , which is the con-
centration of the O2  ions in the melt according to:

pO2  ¼  log(O2  ) (9)

Dissolution of metal oxides formed by the oxide ion reducing

surface metal can occur via acidic or basic pathways. A nickel
oxide for example will dissociate in an acidic environment
Fig. 21. Configuration of a double screw heat exchanger [163]. according to:

NiO2Ni2 þ þO2  (10)

2
This occurs when the O activity in the melt is low, thus nickel
oxide will dissociate to form O2  , pushing reaction (10) to the
right. In a basic melt, the abundance of O2  ions causes the nickel
oxides to form anions such as; NiO22  , NiO22  or NiO32  (which
correspond to nickel oxide states þ2, þ 3 and þ4 respectively).
This can occur via one of the following reactions:

NiOþ R2O22R þ þ NiO22  (11)

Fig. 22. Physical model of a cascaded shell-and-tube PCM storage system.
þ 1
2NiOþ R2O þ½O222R þ 2NiO2 (12)
PCMs are preferred from a practical point considering the asso-
ciated cost increased with the need for a separate tank for each NiOþ R2O þ½O222R þ þNiO32  (13)
PCM used [1]. Cascading EPCMs with different melting tempera-
This acid/base dissolution regime allows the formation of phase
tures in a single tank allows for the high energy and exergy effi-
stability diagrams to describe the behaviour of the system under
ciency [51,170] without increasing the tank cost. To optimize the various conditions. These diagrams are equivalent to the Pourbaix
overall exergy efficiency of multiple PCMs, the optimum melting diagrams used in aqueous corrosion. A number of these diagrams
temperatures of PCMs were identified as an approximate geome- have been developed and discussed in the literature [176–186].
trical regression [171], i.e. Tm1/Tm2 ffi Tm2/Tm3 ffi Tm3/Tm4 ffi…… ffi A post mortem study of the Solar Two project developed by
Tmn  1/Tmn(n ¼number of PCMs, Tm is in Kelvin). Sandia National Laboratories evaluated the metal losses of cou-
pons in various components of the molten salt system. The Solar
Salt was found to be quite compatible with 316, 304 and 347
4. Compatibility of containment materials with storage media stainless steel, with metal losses less than 4 μm reported. The
lower alloy steels used were more susceptible, especially when at
Compatibility of containment materials with molten salt thermal high temperatures or under thermal cycling.
storage media is a significant technical challenge for CSP plants. Fernández et al. [187] used gravimetric analysis and electrical
Most operational plants use Solar Salt for storage, which is relatively impedance spectroscopy (EIS) to measure the corrosion of an
compatible with common stainless steel grades such as 304, 316 alumina forming stainless steel (AFA OC-4), relative to 304 stain-
and 347 at high temperatures [172]. The other molten salts which less and T22 steels at 390 °C, also in Solar Salt. They concluded that
have been used in a pilot scale plant were fluorides in the Molten the Al–Cr rich oxide layer on the surface of the OC-4 steel was
Salt Reactor Experiment at Oak Ridge National Laboratories [173– highly protective in the nitrates used with virtually no corrosion
175]. PCM candidates considered for CSP applications contain car- detected after immersion times up to 2000 h, although its per-
bonates, sulphates, chlorides, hydroxides, their eutectics or mix- formance did not appear to be significantly better than 304 grade.
tures [9,45]. This presents a materials challenge for identifying Another report to come out of Sandia National Laboratories
compatible containment materials with acceptable mechanical [188] investigated corrosion in Solar Salt at 600 °C and 680 °C, a
properties, and corrosion resistance over the temperature range higher temperature than any previous studies. Inconel 625 and
under which these systems will be operated. Hastelloy 230 tokens were tested in 316 stainless steel vessels for
periods of up to 4000 h. They found that whilst corrosion rates
4.1. Common corrosion mechanisms were below 25 μm/year at 600 °C for both alloys, at 680 °C the
corrosion rate increased to 594 μm/year and 688 μm/year for the
4.1.1. Nitrates, carbonates, sulphates and hydroxide Inconel 625 and Hastelloy 230 respectively. In addition to the
Nitrates, carbonates, sulphates and hydroxides all act in a corrosion data, it was found that at 680 °C the salt underwent
similar way with regards to corrosion. As the fluid in contact with significant decomposition to NO2 with an associated degrading of
the metal is non-aqueous, the pH scale for acidity/basicity is not its thermophysical properties and contained significant amounts
applicable. Instead the Lux-Flood model is used, which describes of chromium from the 316 container and test samples.
the acidity/basicity of the melt according to the following Lim et al. [189] investigated corrosion of 316L and 310S stain-
reactions: less steel in a 52 mol% Li2CO3–48 mol% Na2CO3 molten salt,
between 520 °C and 650 °C. They found corrosion and pitting was
NO3  2NO2 þ þO2  (5)
greatly reduced as temperature was increased from 580 °C to
CO3 2
2CO2 þO 2
(6) 600 °C as the surface scale transitioned from a porous LiFe5O8 film
to a dense LiFeO2 film. They also found higher corrosion rates as
SO42  2SO3 þ O2  (7) the partial pressure of CO2 in the atmosphere was increased, but
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1427

little effect from adding O2. They suggested that additional CO2 4.2. Methods of evaluation
could lead to acidic dissolution, or drive an equilibrium shift
towards the formation of LiFe2O8 scale. There are a number of different methods for determining cor-
Yan et al. [190] also investigated the corrosion behaviour of an rosion rates. By far the most common one is the gravimetric ana-
alumina forming austenitic stainless steel. Their tests were lysis i.e. mass loss. By comparing the initial and final weights of a
undertaken in sodium sulphate at 900 °C and compared to Ni- sample, and the time over which it was tested, a rate of mass loss
based superalloys; K438 and K417, and 316L stainless steel. It was can be determined. When combined with material characterization
found that chromium leeched through the pre-formed Al2O3 sur- methods such as microscopy, spectroscopy and diffractometry,
face oxide to form a dense Cr2O3 layer on the surface. They con- important information such as corrosion type, chemical reactions
cluded that the Al2O3/Cr2O3 layer was an effective barrier against and mechanisms, and material changes can be obtained.
sulphur penetration due to its high density, compared to the other Whilst gravimetric analysis is very common, one major dis-
alloys tested. advantage is that experimental tests take many weeks or even
months to complete. Electrochemical methods used in aqueous
4.1.2. Chlorides corrosion can also be applied to molten salt corrosion as the molten
Chloride salts are becoming more common in the area of salts are conductive. This means electrochemical methods such as
EIS and a myriad of voltametric methods can be used to assess and
molten salt thermal storage as they have high latent heats of
infer corrosion rates, mechanisms and reactions in molten salts.
fusion. However, they are considered to be highly corrosive to
Electrochemical studies and reports dating from the late 1980s
steel, especially in the presence of impurities and moisture
began to identify specific reaction pathways and mechanisms for
[191,192]. Grabke et al. [186] defined the corrosion mechanism in
corrosion. Early work by a Japanese group resulted in the review
chloride salts at temperatures below 850 °C to involve the pro-
paper proposing corrosion mechanisms for molten chlorides, car-
duction of chlorine gas from the salt reacting with metal oxides on
bonates and sulphates mentioned earlier [179]. They presented a
the surface of the metal, or with any moisture in the system. The
number of E-pO2  diagrams – the analogue of a Pourbaix diagram
chlorine gas is highly mobile and can penetrate through the oxide in aqueous corrosion – for various metal/salt systems. Then, by
scale and form metal chlorides with the metals at the metal/metal applying anodic and cathodic polarization they developed their
oxide interface. The metal chlorides are quite volatile at these proposed corrosion mechanisms. Another research group active in
temperatures and will sublimate. These gaseous metal chlorides the 1990s [196–199] also used polarization curves in addition to
will diffuse through the scale until they reach higher oxygen cyclic voltammetry to identify reactions and mechanisms for dif-
environment and undergo oxidation. This liberates the chlorine as ferent metals and alloys. Most corrosion studies in molten carbo-
a gas which can diffuse back towards the metal/metal oxide nates have used electrochemical methods as they have been aimed
interface and cause further corrosion. at materials for molten carbonate fuel cells [191].
This corrosion mechanism does not vary much with high alloy EIS is an analysis technique for materials in which ionic con-
steels as iron chloride is the favoured product at the metal/oxide duction is the major conduction mechanism. As molten salts are
interface. However, alloys which form thick oxide layers are more conductive due to their ionic species then EIS can be applied to
resistant to this form of corrosion as the rate limiting step is FeCl2 them. Impedance is influenced by five physical processes; bulk
evaporation and diffusion through the scale. resistive–capacitive effects, electrode reactions, adsorption at
A recent study tested three different alloys which had chro- electrodes, bulk generation–recombination effects, and diffusion
mium, nickel and iron as the major alloying element in NaCl at [200]. Either a detailed physiochemical model, or an equivalent
850 °C for 168 h [193]. It was found that the chromium based alloy circuit designed to represent various mass and charge transport
was the least resistant, followed by the nickel then iron based mechanisms, is used to define the system and fitted to immitance
alloys. It can be concluded that chromium loss was the major vs. frequency data. Aung and Liu's electrochemical study success-
contributor to total mass loss, which occurred predominantly at fully used EIS data to measure hot corrosion [201]. Other recent
grain boundaries. The nickel and iron based alloys formed a pro- studies which use EIS to study corrosion include Ni et al. [202],
tective scale, underneath of which formed a Cr2O3 layer which Frangini et al. [203] and Fernández et al. [187].
prevented chromium leeching from the alloy matrix.
Vignarooban et al. [194] tested commercial Hastelloy types 4.3. Corrosion mitigation
C-276, C-22, and N in three NaCl, KCl and ZnCl2 eutectics at both
Material selection and compatibility is a matter of matching the
250 °C and 500 °C. They conducted potentiodynamic testing and
structural metal to the media in which it is contact. The protective
verified it with an immersion, mass loss test. Results showed that
oxide layer must have a low solubility, or it must be coated with
the C-276 and C-22 alloy were significantly more resistant than
something that does. Plating a base metal with nickel [204], alu-
the N type alloy. In contrast to the previous study however, they
minium [202], and forming perovskite surface layers have all been
concluded that the higher chromium content in the C-276 and C-
reported [205,206]. Another method for mitigating corrosion is to
22 (16% and 22% respectively) protected those alloys more than
add a metal to the salt which will favourably react with the
the N type (7% Cr). They did mention that the N type alloy contains cations. The potential of the salt will stay more negative than the
1% silicon (compared to 0.08% in the other two alloys), which may redox potential of the structural providing the additive metal
have contributed to its poorer performance and provide an alter- remains available for oxidation. This cathodic protection ensures
native explanation. that the corrosion of the steel is less thermodynamically favour-
Liu et al.[195] tested a stainless steel (TP347H), a Ni–Cr–Mo able than oxidation of the additive [207].
alloy (C22) as both a metal and a laser cladded coating in chloride
salts produced from biomass combustion. They used gravimetric
analysis, microscopy and XRD and found severe intergranular 5. Cost
corrosion in the TP347H compared to the C22 alloy. They sug-
gested that this was due to higher solubility of Cr2O3 and MoO2 The economic implications of various TES systems in conjunc-
protective layers of the stainless steel compared to the NiO base tion with CSP plants has been studied in great detail [208]. The
metal oxide of the C22 Nickel alloy. total cost is dependent on the energy density of the storage and
1428 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

containment material as well as the practical limitations of the 6. Conclusions

storage mechanism, whether it be sensible, latent or thermo-
chemical. Cost reductions are achieved through optimisation of As a renewable energy resource, solar energy has the dis-
these two factors without compromising the thermal efficiency of advantage of supplying intermittent power for electricity genera-
the plant. tion. TES allows the mitigation of short fluctuations and extension
Due to the large volume of storage media involved, the material of electricity supply to more desirable periods, making CSP dis-
patchable. Over 80% of the CSP capacity under construction has
price essentially relates to the bulk price. Table 4 provides a
TES, while this number is less than 50% in the operational CSP
summary of costs for various materials used for different storage
capacity.
types. PCM systems can either be encapsulated or arranged in a
All three basic types of TES (sensible, latent and thermo-
coil-in-tank configuration. The cost of encapsulation can account chemical) have the potential to make the next generation CSP
for approximately 15% of the total EPCM system [209]. A coil-in- plants viable. However, more research is still required to develop
tank system containing PCM, is able to achieve a higher packing new storage media with improved thermal characteristics and to
density than the EPCM system, reducing the tank cost and HTF improve the thermal performance and economics of the TES sys-
cost. In order to achieve a high heat exchange effectiveness, a large tem. The target of the SunShot Initiative is to reduce the cost of the
coil surface area is required, which can be of similar cost to the TES system from the present 5C\  t\\vskip\  tˈ/kW h to
storage media [210]. 1C\  t\\vskip\  tˈ/kW h by 2020. The paper overviews the tech-
The cost of storage tank has been investigated by Kelly and nical developments in high temperature TES over the last decade
Kearney [211] and Herrmann et al. [153]. The impact of the cost of to achieve cost-effective TES for CSP. Current research efforts in the
tank is largely dependent on the system configuration (1 or area of sensible TES focus on (1) developing new molten salts with
lower freezing temperatures and higher decomposition tempera-
2 tanks) and the tank material. Generally the cost of the tank is
tures; (2) developing ionic liquids for next generation HTFs;
between 15% and 30% of the overall system cost. Piping and
(3) utilizing nano-based technology to improve the specific heat
instruments and controls have been estimated to be $642/MW h t capacity and thermal conductivity and (4) developing low-cost
and $151/MW h t respectively [212]. Heat tracing system, if solid storage media and evaluating them through compatibility
required, has been estimated to cost $50/m-cable [211]. testing with molten salts. So far, the long-term stability of the new
The total of these input costs is multiplied by 1.3 to deliver the molten salts/HTFs and their compatibility with containment and
constructed cost, and this value is multiplied by 1.1 and 1.07 to piping materials have not been fully investigated. Future research
consider engineering/inspection and contingency costs, respec- is required to address these issues to warrant the commercial
tively [212]. The balance of system costs, including heat exchan- application of the developed molten salts/HTFs.
gers and pumps, as well as operating and maintenance costs are Recent developments in latent TES aim to maximize the
largely overlooked in TES cost estimation and therefore it is diffi- extraction rate of the stored heat from the storage system by
cult to achieve a fair comparison between systems. An example of reducing the thermal resistance between the PCM and the HTF
through encapsulation, utilizing heat pipes and making PCMs
the capital cost of some high temperature TES systems is given in
mobile. Several useful laboratory-scale tests have been conducted
Table 5.
on cascaded PCM systems and the results indicate that they can
offer a high energy and exergy efficiency storage solution com-
pared to a system containing only one PCM. Future research
should focus on PCM characterization to facilitate accurate design
of the latent TES system. In addition, pilot and demonstration scale
Table 4 tests are needed to verify the deployment of latent TES in CSP
The cost of some common storage materials. applications.
High temperature corrosion of the containment material is a
Material Cost/kg ($US/kg) Material Cost/kg ($US/kg)
severe problem experienced by all types of TES. Corrosion in
Quartz/Sand 0.72 Matrices 0.33 [Vermiculite] [71] molten salts is a very complex issue, encompassing many aspects
Carbonate salts 0.23 [Na2CO3] 1.11 [Silica Gel] [71] of chemistry, metallurgy and thermodynamics. As yet, the
1.1 [K2CO3] 1.93 [Zeolite] [71] mechanisms of corrosion in various salts are not fully resolved and
7 [Li2CO3] 2.75 [Activated Carbon]
[71]
the effects of different alloying elements are ambiguous. Addi-
Molten salts 0.49 [Solar Salt] 2.15 [Expanded Natural tionally, how containment materials will be affected by the ther-
Graphite] [71] mal cycling and mechanical loading required for a PCM system is
1.43 [Hitec XL] 1.32 [Activated Alumina] almost totally uninvestigated and will be critical for providing safe
[71]
and cost effective solutions for thermal storage.
0.93 [Hitec] Chloride salts 0.35 [KCl]
1.4 [Halotechnics] 0.2 [MgCl2] The cost of thermal storage is a function of both the storage and
Ammonia 0.7 [Anhydrous] 0.12 [NaCl] containment materials and the configuration associated with the
7.8 [LiCl] storage system. Costs can only be minimized by considering both
issues in connection with the rest of a CSP plant.

Table 5
Example TES capital costs.

Type Configuration Storage material Heat transfer fluid ΔT (°C) Capital cost ($/kW ht)

Sensible Two-tank Molten salt (Hitec XL) Same as storage material 300 28.21
Sensible Two-tank Liquid sodium Same as storage material 300 48.48
Sensible Thermocline Quartz/sand Molten salt (Solar Salt) 300 20.26
Latent EPCM Chloride salt [Alumina Shell] Air 300 19.74
Latent Coil-in-tank Carbonate salt [stainless tubing] SCO2 or liquid sodium 300 21
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1429

Acknowledgements [26] Gil A, Medrano M, Martorell I, Lázaro 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
This research was performed as part of the Australian Solar 2010;14:31–55.
Thermal Research Initiative (ASTRI), a project supported by the [27] Xu C, Li X, Wang Z, He Y, Bai F. Effects of solid particle properties on the
thermal performance of a packed-bed molten-salt thermocline thermal
Australian Government, through the Australian Renewable Energy
storage system. Appl Therm Eng 2013;57:69–80.
Agency (ARENA). [28] Flueckiger SM, Iverson BD, Garimella SV. Economic optimization of a con-
centrating solar power plant with molten-salt thermocline storage. J Sol
Energy Eng 2013;136:011015.
[29] Pacheco JE, Showalter SK, Kolb WJ. Development of a molten-salt thermo-
cline thermal storage system for parabolic trough plants. J Sol Energy Eng
References 2002;124:153–9.
[30] Van Lew JT, Li P, Chan CL, Karaki W, Stephens J. Analysis of heat storage and
delivery of a thermocline tank having solid filler material. J Sol Energy Eng
[1] Stekli J, Irwin L, Pitchumani R. Technical challenges and opportunities for
2011;133:021003.
concentrating solar power with thermal energy storage. J Therm Sci Eng Appl
[31] Wang L, Yang Z, Duan Y. Influence of flow distribution on the thermal per-
2013;5:021011.
formance of dual-media thermocline energy storage systems. Appl Energy
[2] Pitz-Paal R, Amin A, Oliver Bettzuge M, Eames P, Flamant G, Fabrizi F, et al.
2015;142:283–92.
Concentrating solar power in Europe, the Middle East and North Africa: a
[32] Flueckiger S, Yang Z, Garimella SV. An integrated thermal and mechanical
review of development issues and potential to 2050. J Sol Energy Eng
investigation of molten-salt thermocline energy storage. Appl Energy
2012;134:024501.
[3] Steinmann W-D, Laing D, Tamme R. Latent heat storage systems for solar 2011;88:2098–105.
thermal power plants and process heat applications. J Sol Energy Eng [33] Yang X, Yang X, Ding J, Shao Y, Qin FGF, Jiang R. Criteria for performance
2010;132:021003. improvement of a molten salt thermocline storage system. Appl Therm Eng
[4] U.S. Department of Energy. 2014 SunShot initiative portfolio book: con- 2012;48:24–31.
centrating solar power. Washington, DC: Office of Energy Efficiency & [34] Calvet N, Gomez JC, Faik A, Roddatis VV, Meffre A, Glatzmaier GC, et al.
Renewable Energy, U.S. Department of Energy; 2014. Compatibility of a post-industrial ceramic with nitrate molten salts for use as
[5] Herrmann U, Kearney DW. Survey of thermal energy storage for parabolic filler material in a thermocline storage system. Appl Energy 2013;109:387–
trough power plants. J Sol Energy Eng 2002;124:145–52. 93.
[6] Morina G, Karla M, Mertinsa M, Seliga M. Molten salt as a heat transfer fluid [35] Yang X, Yang X, Qin FGF, Jiang R. Experimental investigation of a molten salt
in a linear fresnel collector – commercial application backed by demon- thermocline storage tank. Int J Sustain Energy 2014:1–9.
stration. In: Proceedings of international conference on concentrating solar [36] Brown BM, Strasser MN, Selvam RP. Development of a structured thermo-
power and chemical energy systems, SolarPACES 2014. Beijing (China); 2014. cline thermal energy storage system. In: Proceedings of World Renewable
[7] European Academies Science Advisoty Council. Concentrating solar power: Energy Forum, WREF 2012, Including World Renewable Energy Congress XII
its potential contribution to a sustainable energy future. EASAC policy report. and Colorado Renewable Energy Society (CRES) annual conference. Denver,
European Academies Science Advisoty Council; 16 November 2011. Colorado (USA); 2012.
[8] Pacheco J. Final test and evaluation results from the solar two project. [37] Fernandes D, Pitié F, Cáceres G, Baeyens J. Thermal energy storage: “how pre-
Albuquerque, NM: Sandia National Laboratories; 2002 SAND2002-0120. vious findings determine current research priorities”. Energy 2012;39:246–57.
[9] Liu M, Saman W, Bruno F. Review on storage materials and thermal perfor- [38] Robak CW, Bergman TL, Faghri A. Economic evaluation of latent heat thermal
mance enhancement techniques for high temperature phase change thermal energy storage using embedded thermosyphons for concentrating solar
storage systems. Renew Sustain Energy Rev 2012;16:2118–32. power applications. Sol Energy 2011;85:2461–73.
[10] Romero M, González-Aguilar J. Solar thermal CSP technology. Wiley Inter- [39] Nithyanandam K, Pitchumani R. Analysis and optimization of a latent ther-
discip Rev: Energy Environ 2014;3:42–59. mal energy storage system with embedded heat pipes. Int J Heat Mass Transf
[11] Kutscher C, Mehos M, Turchi C, Glatzmaier G, Moss T. Line-focus solar power 2011;54:4596–610.
plant cost reduction plan. National Renewable Energy Laboratory; 2010 [40] Ren N, Wu Y-t, Wang T, Ma C-f. Experimental study on optimized compo-
Golden, Colorado, NREL/TP-5500-48175. sition of mixed carbonate for phase change thermal storage in solar thermal
[12] Kuravi S, Trahan J, Goswami DY, Rahman MM, Stefanakos EK. Thermal energy power plant. J Therm Anal Calorim 2011;104:1201–8.
storage technologies and systems for concentrating solar power plants. Prog [41] Steinmann W-D, Laing D, Tamme R. Development of PCM storage for process
Energy Combust Sci 2013;39:285–319. heat and power generation. J Sol Energy Eng 2009;131:041009.
[13] U.S. Department of Energy. The SunShot Vision study-concentrating solar [42] Steinmann W-D, Tamme R. Latent heat storage for solar steam systems. J Sol
power: technologies, cost, and performance. U.S. Department of Energy; Energy Eng 2008;130:011004.
2012. [43] Tamme R, Laing D, Steinmann W-D. Advanced thermal energy storage
[14] Hoffschmidt B, Alexopoulos S, Gottsche J, Sauerborn M, Kaufhold O. High technology for parabolic trough. J Sol Energy Eng 2004;126:794–800.
concentration solar collectors. In: Sayigh A, editor. Comprehensive renew- [44] Adinberg R, Zvegilsky D, Epstein M. Heat transfer efficient thermal energy
able energy. Elsevier Ltd; 2012. storage for steam generation. Energy Convers Manag 2010;51:9–15.
[15] Turchi CS, Wagner MJ, Kutscher CF. Water use in parabolic trough power [45] Kenisarin MM. High-temperature phase change materials for thermal energy
plants: summary results from WorleyParsons' analyses. National Renewable storage. Renew Sustain Energy Rev 2010;14:955–70.
Energy Laboratory, Golden, Colorado; 2010. [46] Kotzé JP, von Backström TW, Erens PJ. High temperature thermal energy
[16] U.S. Department of Energy. Concentrating solar power commercial applica-
storage utilizing metallic phase change materials and metallic heat transfer
tion study: reducing water consumption of concentrating solar power elec-
fluids. J Sol Energy Eng 2013;135:035001.
tricity generation. U.S. Department of Energy; 2008.
[47] Liu M, Gomez JC, Turchi CS, Tay NHS, Saman W, Bruno F. Determination of
[17] International Renewable Energy Agency. Renewable power generation costs
thermo-physical properties and stability testing of high-temperature phase-
in 2012: An overview. International Renewable Energy Agency; 2013.
change materials for CSP applications. Sol Energy Mater Sol Cells
[18] International Renewable Energy Agency. Renewable energy technologies:
2015;139:81–7.
cost analysis series. Volume 1: Power sector-concentrating solar power.
[48] Elmozughi AF, Solomon L, Oztekin A, Neti S. Encapsulated phase change
International Renewable Energy Agency; 2012.
material for high temperature thermal energy storage – heat transfer ana-
[19] Divya KC, Østergaard J. Battery energy storage technology for power systems
—an overview. Electr Power Syst Res 2009;79:511–20. lysis. Int J Heat Mass Transf 2014;78:1135–44.
[20] Abengoa Solar. Development of molten-salt heat transfer fluid technology [49] Bellan S, Gonzalez-Aguilar J, Romero M, Rahman MM, Goswami DY, Stefa-
for parabolic trough solar power plants. Abengoa Solar, LLC; 2013. nakos EK, et al. Numerical analysis of charging and discharging performance
[21] Kolb GJ, Diver RB. Conceptual design of a 2  trough for use within salt and of a thermal energy storage system with encapsulated phase change mate-
oil-based parabolic trough power plants. J Sol Energy Eng 2010;132:041003. rial. Appl Therm Eng 2014;71:481–500.
[22] Kearney D, Herrmann U, Nava P, Kelly B, Mahoney R, Pacheco J, et al. Over- [50] Zhao W, Elmozughi AF, Oztekin A, Neti S. Heat transfer analysis of encap-
view on use of a molten salt htf in a trough solar field. In: Proceedings of sulated phase change material for thermal energy storage. Int J Heat Mass
NREL parabolic trough thermal energy storage workshop. Golden, CO; 2003. Transf 2013;63:323–35.
[23] Medrano M, Gil A, Martorell I, Potau X, Cabeza LF. State of the art on high- [51] Kurnia JC, Sasmito AP, Jangam SV, Mujumdar AS. Improved design for heat
temperature thermal energy storage for power generation. Part 2 – case transfer performance of a novel phase change material (PCM) thermal
studies. Renew Sustain Energy Rev 2010;14:56–72. energy storage (TES). Appl Therm Eng 2013;50:896–907.
[24] Brosseau D, Kelton JW, Ray D, Edgar M, Chisman K, Emms B. Testing of [52] Liu Z, Sun X, Ma C. Experimental investigations on the characteristics of
thermocline filler materials and molten-salt heat transfer fluids for thermal melting processes of stearic acid in an annulus and its thermal conductivity
energy storage systems in parabolic trough power plants. Sol Energy Eng- enhancement by fins. Energy Convers Manag 2005;46:959–69.
Trans ASME 2005;127:109–16. [53] Robak CW, Bergman TL, Faghri A. Enhancement of latent heat energy storage
[25] Kearney D, Herrmann U, Nava P, Kelly B, Mahoney R, Pacheco J, et al. using embedded heat pipes. Int J Heat Mass Transf 2011;54:3476–84.
Assessment of a molten salt heat transfer fluid in a parabolic trough solar [54] Shabgard H, Faghri A. Performance characteristics of cylindrical heat pipes
field. J Solar Energy Eng 2003;125:170–6. with multiple heat sources. Appl Therm Eng 2011;31:3410–9.
1430 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

[55] Petri RJ, Ong ET, Olszewski M. High temperature composite thermal storage [84] Reddy RG, Wang T, Mantha D. Thermodynamic properties of potassium
systems. In: Proceedings of the 19th intersociety energy conversion engi- nitrate–magnesium nitrate compound [2KNO3  Mg(NO3)2]. Thermochim
neering conference. San Francisco, CA; 1984. Acta 2012;531:6–11.
[56] Gokon N, Nakano D, Inuta S, Kodama T. High-temperature carbonate/MgO [85] Mantha D, Wang T, Reddy RG. Thermodynamic modeling of eutectic point in
composite materials as thermal storage media for double-walled solar the LiNO3–NaNO3–KNO3–NaNO2 quaternary system. Sol Energy Mater Sol
reformer tubes. Solar Energy 2008;82:1145–53. Cells 2013;118:18–21.
[57] Bauer T, Tamme R, Christ M, Őttinger O. PCM-graphite composites for high [86] Wang T, Mantha D, Reddy RG. Novel low melting point quaternary eutectic
temperature thermal energy storage. In: Proceedings of the tenth interna- system for solar thermal energy storage. Appl Energy 2013;102:1422–9.
tional conference on thermal energy storage, Atlantic; 2006. [87] Wang T, Reddy R.G. Novel LiNO3–NaNO3–KNO3–NaNO2 molten salts for solar
[58] Lopez J, Caceres G, Palomo Del Barrio E, Jomaa W. Confined melting in thermal energy storage applications. Energy Technology 2014. John Wiley &
deformable porous media: A first attempt to explain the graphite/salt Sons, Inc. pp. 141–148.
composites behaviour. Int J Heat Mass Transf 2010;53:1195–207. [88] Wang T, Mantha D, Reddy RG. Thermal stability of the eutectic composition
[59] Pincemin S, Olives R, Py X, Christ M. Highly conductive composites made of in LiNO3–NaNO3–KNO3 ternary system used for thermal energy storage. Sol
phase change materials and graphite for thermal storage. Sol Energy Mater Energy Mater Sol Cells 2012;100:162–8.
Solar Cells 2008;92:603–13. [89] Mantha D, Wang T, Reddy RG. Thermodynamic modelling of eutectic point in
[60] Sun JQ, Zhang RY, Liu ZP, Lu GH. Thermal reliability test of Al–34%Mg–6%Zn the LiNO3–NaNO3–KNO3 ternary system. J Phase Equilibia Diffus
alloy as latent heat storage material and corrosion of metal with respect to 2012;33:110–4.
thermal cycling. Energy Convers Manag 2007;48:619–24. [90] Bauer T, Laing D, Tamme R. Recent progress in alkali nitrate/nitrite devel-
[61] Blanco-Rodríguez P, Rodríguez-Aseguinolaza J, Risueño E, Tello M. Thermo- opments for solar thermal power applications. Molten salts chemistry and
physical characterization of Mg–51%Zn eutectic metal alloy: a phase change technology. John Wiley & Sons, Ltd; 2014. p. 543–53.
material for thermal energy storage in direct steam generation applications. [91] Peng Q, Yang X, Ding J, Wei X, Yang J. Thermodynamic performance of the
Energy 2014;72:414–20. NaNO3–NaCl–NaNO2 ternary system. J Therm Anal Calorim 2014;115:1753–8.
[62] Wang X, Liu J, Zhang Y, Di H, Jiang Y. Experimental research on a kind of [92] Peng Q, Ding J, Wei X, Yang J, Yang X. The preparation and properties of
novel high temperature phase change storage heater. J Sol Energy Convers multi-component molten salts. Appl Energy 2010;87:2812–7.
Manag 2006;47:2211–22. [93] Paul TC, AKMM Morshed, Fox EB, Visser AE, Bridges NJ, Khan JA. Thermal
[63] Wang Z, Wang H, Li X, Wang D, Zhang Q, Chen G, et al. Aluminum and performance of ionic liquids for solar thermal applications. Exp Therm Fluid
silicon based phase change materials for high capacity thermal energy sto- Sci 2014;59:88–95.
rage. Appl Therm Eng 2015;89:204–8. [94] Blake DM, Moens L, Hale MJ, Price H, Kearney D, Herrmann U. New heat
[64] Laing D, Bahl C, Bauer T, Lehmann D, Steinmann W-D. Thermal energy sto- transfer and storage fluids for parabolic trough solar thermal electric plants.
rage for direct steam generation. Sol Energy 2011;85:627–33. In: Proceedings of 11th SolarPACES international symposium on con-
[65] Laing D, Bauer T, Lehmann D, Bahl C. Development of a thermal energy centrating solar power and chemical energy technologies. Zurich (Switzer-
storage system for parabolic trough power plants with direct steam gen- land); 2002.
[95] Moens L, Blake DM, Rudnicki DL, Hale MJ. Advanced thermal storage fluids
eration. Energy Eng 2010;132:021011.
[66] Galione PA, Pérez-Segarra CD, Rodríguez I, Oliva A, Rigola J. Multi-layered for solar parabolic trough systems. J Sol Energy Eng 2003;125:112–6.
[96] Wu B, Reddy RG, Rogers RD. Novel ionic liquid thermal storage for solar
solid-PCM thermocline thermal storage concept for CSP plants. Numerical
thermal electric power systems. In: Proceedings of solar forum 2001 solar
analysis and perspectives. Appl Energy 2015;142:337–51.
energy: the power to choose. Washington, DC; 2001.
[67] Galione PA, Pérez-Segarra CD, Rodríguez I, Torras S, Rigola J. Multi-layered
[97] Fredlake CP, Crosthwaite JM, Hert DG, SNVK Aki, Brennecke JF. Thermo-
solid-PCM thermocline thermal storage for CSP. Numerical evaluation of its
physical properties of imidazolium-based ionic liquids. J Chem Engineer Data
application in a 50 MWe plant. Sol Energy 2015;119:134–50.
2004;49:954–64.
[68] Cot-Gores J, Castell A, Cabeza LF. Thermochemical energy storage and con-
[98] Every HA, Bishop AG, MacFarlane DR, Oradd G, Forsyth M. Transport prop-
version: a-state-of-the-art review of the experimental research under
erties in a family of dialkylimidazolium ionic liquids. Phys Chem Chem Phys
practical conditions. Renew Sustain Energy Rev 2012;16:5207–24.
2004;6:1758–65.
[69] Pardo P, Deydier A, Anxionnaz-Minvielle Z, Rougé S, Cabassud M, Cognet P. A
[99] Perissi I, Bardi U, Caporali S, Lavacchi A. High temperature corrosion prop-
review on high temperature thermochemical heat energy storage. Renew
erties of ionic liquids. Corros Sci 2006;48:2349–62.
Sustain Energy Rev 2014;32:591–610.
[100] M.F. A, R.G. R. Corrosion of steel in ionic liquids. J Min Metall Sect B: Metall
[70] Pardo P, Anxionnaz-Minvielle Z, Rougé S, Cognet P, Cabassud M. Ca(OH)2/
2003;39:81–91.
CaO reversible reaction in a fluidized bed reactor for thermochemical heat
[101] Dudda B, Shin D. Effect of nanoparticle dispersion on specific heat capacity of
storage. Sol Energy 2014;107:605–16.
a binary nitrate salt eutectic for concentrated solar power applications. Int J
[71] Aydin D, Casey SP, Riffat S. The latest advancements on thermochemical heat
Therm Sci 2013;69:37–42.
storage systems. Renew Sustain Energy Rev 2015;41:356–67. [102] Singh D, Timofeeva EV, Moravek MR, Cingarapu S, Yu W, Fischer T, et al. Use
[72] Wentworth WE, Chen E. Simple thermal decomposition reactions for storage
of metallic nanoparticles to improve the thermophysical properties of
of solar thermal energy. Sol Energy 1976;18:205–14. organic heat transfer fluids used in concentrated solar power. Sol Energy
[73] Schaube F, Wörner A, Tamme R. High temperature thermochemical heat 2014;105:468–78.
storage for concentrated solar power using gas–solid reactions. J Sol Energy [103] Dudda B, Shin D. Investigation of molten salt nanomaterial as thermal energy
Eng 2011;133:031006. storage in concentrated solar power. In: Procedings of ASME 2012 interna-
[74] Neveu P, Tescari S, Aussel D, Mazet N. Combined constructal and exergy tional mechanical engineering congress & exposition. Houston Texas (USA);
optimization of thermochemical reactors for high temperature heat storage. 2012 .
Energy Convers Manag 2013;71:186–98. [104] Shin D, Banerjee D. Enhancement of specific heat capacity of high-
[75] Vignarooban K, Xu X, Arvay A, Hsu K, Kannan AM. Heat transfer fluids for temperature silica-nanofluids synthesized in alkali chloride salt eutectics
concentrating solar power systems – a review. Appl Energy 2015;146:383– for solar thermal-energy storage applications. Int J Heat Mass Transf
96. 2011;54:1064–70.
[76] Pérez FJ, Fernández A, Lasanta MI, García G. New molten salts as con- [105] Shin D, Jo B, Kwak H-E, Banerjee D. Investigation of high temperature
centrated fluids for energy storage in csp plants. In: Proceedings of Inter- nanofluids for solar thermal power conversion and storage applications. In:
national Conference on Concentrating Solar Power and Chemical Energy Proceedings of 14th international heat transfer conference. Washington, DC
Systems, SolarPACES 2014. Beijing (China); 2014. (USA); 2010.
[77] Fernández AG, Galleguillos H, Pérez FJ. Corrosion ability of a novel heat [106] Shin D, Banerjee D. Enhanced specific heat capacity of nanomaterials syn-
transfer fluid for energy storage in CSP plants. Oxid Met 2014:1–15. thesized by dispersing silica nanoparticles in eutectic mixtures. J Heat Transf
[78] Fernández AG, Ushak S, Galleguillos H, Pérez FJ. Thermal characterisation of 2013;135:032801.
an innovative quaternary molten nitrate mixture for energy storage in CSP [107] Shin D, Banerjee D. Enhanced specific heat of silica nanofluid. J Heat Transf
plants. Sol Energy Mater Sol Cells 2015;132:172–7. 2010;133:024501.
[79] Ren N, Wu Y-t, Ma C-f, Sang L-X. Preparation and thermal properties of [108] Shin D, Banerjee D. Experimental investigation of molten salt nanofluid for
quaternary mixed nitrate with low melting point. Sol Energy Mater Sol Cells solar thermal energy application. In: Proceedings of ASME/JSME 2011 8th
2014;127:6–13. thermal engineering joint conference. Honolulu, Hawaii (USA); 2011.
[80] Fernández AG, Ushak S, Galleguillos H, Pérez FJ. Development of new molten [109] Paul TC, Morshed AM, Khan JA. Nanoparticle Enhanced Ionic Liquids (NEILS)
salts with LiNO3 and Ca(NO3)2 for energy storage in CSP plants. Appl Energy as working fluid for the next generation solar collector. Procedia Eng
2014;119:131–40. 2013;56:631–6.
[81] Gomez J, Calvet N, Starace AK, Glatzmaier GC. Ca(NO3)2–NaNO3–KNO3 [110] Wang T, Reddy RG. Thermodynamic properties of novel low melting point
molten salt mixtures for direct thermal energy storage systems in parabolic LiNO3–NaNO3–KNO3 ternary molten salt for parabolic trough solar power
trough plants. J Sol Energy Eng 2013;135:021016. generation. Energy Technology 2013. John Wiley & Sons, Inc. pp. 1–16.
[82] Fernández AG, Galleguillos H, Fuentealba E, Pérez FJ. Corrosion of stainless [111] Hänchen M, Brückner S, Steinfeld A. High-temperature thermal storage using
steels and low-Cr steel in molten Ca(NO3)2–NaNO3–KNO3 eutectic salt for a packed bed of rocks e Heat transfer analysis and experimental validation.
direct energy storage in CSP plants. Sol Energy Mater Sol Cells 2015;141:7– Appl Therm Eng 2011;31:1798–806.
13. [112] Andreozzi A, Buonomo B, Manca O, Tamburrino S. Thermal energy storages
[83] Wang T, Mantha D, Reddy RG. Thermodynamic properties of LiNO3–NaNO3– analysis for high temperature in air solar systems. Appl Therm Eng
KNO3–2KNO3  Mg(NO3)2 system. Thermochim Acta 2013;551:92–8. 2014;71:130–41.
 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432 1431

[113] Skinner JE, Strasser MN, Brown BM, Panneer Selvam R. Testing of high- [143] Sharifi N, Wang S, Bergman TL, Faghri A. Heat pipe-assisted melting of a
performance concrete as a thermal energy storage medium at high tem- phase change material. Int J Heat Mass Transf 2012;55:3458–69.
peratures. J Sol Energy Eng 2013;136:021004. [144] Nithyanandam K, Pitchumani R. Computational modeling of dynamic
[114] John E, Hale M, Selvam P. Concrete as a thermal energy storage medium for response of a latent thermal energy storage system with embedded heat
thermocline solar energy storage systems. Sol Energy 2013;96:194–204. pipes. J Sol Energy Eng 2014;136:011010.
[115] Hasnain SM. Review on sustainable thermal energy storage technologies. Part 1: [145] Faghri A. Heat pipe science and technology. Washingto, DC: Taylor and
heat storage materials and techniques. Energy Convers Manag 1998;39:1127–38. Francis; 1995.
[116] Yuan H-W, Lu C-H, Xu Z-Z, Ni Y-R, Lan X-H. Mechanical and thermal prop- [146] Kaya T, Goldak J. Three-dimensional numerical analysis of heat and mass
erties of cement composite graphite for solar thermal storage materials. Sol transfer in heat pipes. Heat Mass Transf 2007;43:775–85.
Energy 2012;86:3227–33. [147] Tournier JM, El-Genk MS. A heat pipe transient analysis model. Int J Heat
[117] Yuan H, Shi Y, Lu C, Xu Z, Ni Y, Lan X. Enhanced performance of high tem- Mass Transf 1994;37:753–62.
perature aluminate cementitious materials incorporated with Cu powders for [148] Horbaniuc B, Dumitrascu G, Popescu A. Mathematical models for the study of
thermal energy storage. Cem Concr Compos 2015;55:139–44. solidification within a longitudinally finned heat pipe latent heat thermal
[118] Faik A, Guillot S, Lambert J, Véron E, Ory S, Bessada C, et al. Thermal storage storage system. Energy Convers Manag 1999;40:1765–74.
material from inertized wastes: evolution of structural and radiative prop- [149] Liu Z, Wang Z, Ma C. An experimental study on heat transfer characteristics
erties with temperature. Sol Energy 2012;86:139–46. of heat pipe heat exchanger with latent heat storage. Part I: charging only
[119] Py X, Calvet N, Olives R, Meffre A, Echegut P, Bessada C, et al. Recycled and discharging only modes. Energy Convers Manag 2006;47:944–66.
material for sensible heat based thermal energy storage to be used in con- [150] Liu Z, Wang Z, Ma C. An experimental study on the heat transfer character-
centrated solar thermal power plants. J Sol Energy Eng 2011;133:031008. istics of a heat pipe heat exchanger with latent heat storage. Part II: simul-
[120] Guillot S, Faik A, Rakhmatullin A, Lambert J, Veron E, Echegut P, et al. Cor- taneous charging/discharging modes. Energy Convers Manag 2006;47:967–
rosion effects between molten salts and thermal storage material for con- 91.
centrated solar power plants. Appl Energy 2012;94:174–81. [151] Jung EG, Boo JH. Thermal analytical model of latent thermal storage with heat
[121] Motte F, Falcoz Q, Veron E, Py X. Compatibility tests between solar salt and pipe heat exchanger for concentrated solar power. Sol Energy 2014;102:318–
thermal storage ceramics from inorganic industrial wastes. Appl Energy 32.
2015;155:14–22. [152] Shabgard H, Robak CW, Bergman TL, Faghri A. Heat transfer and exergy
[122] Calvet N, Dejean G, Unamunzaga L, Py X. Waste from metallurgic industry: a analysis of cascaded latent heat storage with gravity-assisted heat pipes for
sustainable high-temperature thermal energy storage material for con- concentrating solar power applications. Sol Energy 2012;86:816–30.
centrated solar power. In: Proceedings of the ASME 7th International Con- [153] Herrmann U, Kelly B, Price H. Two-tank molten salt storage for parabolic
ference on Energy Sustainability collocated with the ASME 2013 Heat trough solar power plants. Energy 2004;29:883–93.
Transfer Summer Conference and the ASME 2013 11th International Con- [154] Strumpf HJ. Advanced heat receiver conceptual design study. Washington,
ference on Fuel Cell Science, Engineering and Technology. Minneapolis, DC: National Aeronautics and Space Administration; 1987.
[155] Xiaohong G, Bin L, Yongxian G, Xiugan Y. Two-dimensional transient thermal
Minnesota (USA); 2013.
analysis of PCM canister of a heat pipe receiver under microgravity. Appl
[123] Brosseau DA, Hlava PF, Kelly MJ. Testing thermocline filler materials and
Therm Eng 2011;31:735–41.
moltensalt heat transfer fluids for thermal energy storage systems used in
[156] Gui X, Qu W, Lin B, Yuan X. Two-dimensional transient thermal analysis of a
parabolic trough solar power plants. Albuquerque, NM: Sandia National
phase-change-material canister of a heat-pipe receiver under gravity. J
Laboratories; 2004 SAND2004-3207.
Therm Sci 2010;19:160–6.
[124] Flueckiger SM, Yang Z, Garimella SV. Design of molten-salt thermocline tanks
[157] Shabgard H, Bergman TL, Sharifi N, Faghri A. High temperature latent heat
for solar thermal energy storage. Purdue University: Cooling Technologies
thermal energy storage using heat pipes. Int J Heat Mass Transf 2010;53:2979–
Research Center; 2013.
88.
[125] Flueckiger SM, Yang Z, Garimella SV. Review of molten-salt thermocline tank
[158] Nithyanandam K, Pitchumani R. Computational studies on a latent thermal
modeling for solar thermal energy storage. Heat Transf Eng 2012;34:787–
energy storage system with integral heat pipes for concentrating solar
800.
power. Appl Energy 2013;103:400–15.
[126] Flueckiger SM, Yang Z, Garimella SV. Thermomechanical simulation of the
[159] Tay NHS, Bruno F, Belusko M. Experimental investigation of dynamic melting
solar one thermocline storage tank. J Sol Energy Eng 2012;134:041014.
in a tube-in-tank PCM system. Appl Energy 2013;104:137–48.
[127] Yang Z, Garimella SV. Thermal analysis of solar thermal energy storage in a
[160] Tay NHS, Belusko M, Liu M, Bruno F. Investigation of the effect of dynamic
molten-salt thermocline. Sol Energy 2010;84:974–85.
melting in a tube-in-tank PCM system using a CFD model. Appl Energy
[128] Yang Z, Garimella SV. Molten-salt thermal energy storage in thermoclines
2015;137:738–47.
under different environmental boundary conditions. Appl Energy
[161] Pointner H, Steinmann W-D, Eck M. Introduction of the PCM flux concept for
2010;87:3322–9. latent heat storage. Energy Procedia 2014;57:643–52.
[129] Strasser MN, Selvam RP. A cost and performance comparison of packed bed [162] Pointner H, Steinmann W-D, Eck M. PCMflux as a fully controllable dynamic
and structured thermocline thermal energy storage systems. Sol Energy latent heat storage system. Eurotherm seminar no 99: advances in thermal
2014;108:390–402. energy storage. Lleida (Spain); 2014.
[130] Motte F, Bugler-Lamb SL, Falcoz Q, Py X. Numerical study of a structured [163] Zipf V, Neuhäuser A, Willert D, Nitz P, Gschwander S, Platzer W. High tem-
thermocline storage tank using vitrified waste as filler material. Energy perature latent heat storage with a screw heat exchanger: design of proto-
Procedia 2014;49:935–44. type. Appl Energy 2013;109:462–9.
[131] Mehling H, Cabeza LF. Heat and cold storage with PCM-an up to date intro- [164] Domanski R, Fellah G. Exergy analysis for the evaluation of a thermal storage
duction into basics and applications. Berlin, Germany: Springer; 2008. system employing PCMS with different melting temperatures. Appl Therm
[132] Zhang Y, Zhou G, Lin K, Zhang Q, Di H. Application of latent heat thermal Eng 1996;16:907–19.
energy storage in buildings: state-of-the-art and outlook. Build Environ [165] Michels H, Pitz-Paal R. Cascaded latent heat storage for parabolic trough solar
2007;42:2197–209. power plants. Sol Energy 2007;81:829–37.
[133] Waqas A, Ud Din Z. Phase change material (PCM) storage for free cooling of [166] Seeniraj RV, Lakshmi Narasimhan N. Performance enhancement of a solar
buildings—a review. Renew Sustain Energy Rev 2013;18:607–25. dynamic LHTS module having both fins and multiple PCMs. Sol Energy
[134] Jacob R, Bruno F. Review on shell materials used in the encapsulation of 2008;82:535–42.
phase change materials for high temperature thermal energy storage. Renew [167] Li Y-Q, He Y-L, Wang Z-F, Xu C, Wang W. Exergy analysis of two phase change
Sustain Energy Rev 2015;48:79–87. materials storage system for solar thermal power with finite-time thermo-
[135] Pendyala S. Macroencapsulation of phase change materials for thermal dynamics. Renew Energy 2012;39:447–54.
energy storage. Tampa, Florida: University of South Florida; 2012. [168] Jegadheeswaran S, Pohekar SD, Kousksou T. Exergy based performance
[136] Salunkhe PB, Shembekar PS. A review on effect of phase change material evaluation of latent heat thermal storage system: A review. Renew Sustain
encapsulation on the thermal performance of a system. Renew Sustain Energy Rev 2010;14:2580–95.
Energy Rev 2012;16:5603–16. [169] Liu M, Tay NHS, Belusko M, Bruno F. Investigation of cascaded shell and tube
[137] Goswami DY. Development and demonstration of an innovative thermal latent heat storage systems for solar tower power plants. In: Proceedings of
energy storage system for baseload power generation; 2012. international conference on concentrating solar power and chemical energy
[138] Alam TE, Dhau J, Goswami DY, Stefanakos E. Encapsulation of phase change systems, SolarPACES 2014. Beijing (China); 2014.
materials and use in thermal energy storage; 2014. [170] Yang L, Zhang X, Xu G. Thermal performance of a solar storage packed bed
[139] Maruoka N, Sato K, Yagi J-i, Akiyama T. Development of PCM for recovering using spherical capsules filled with PCM having different melting points.
high temperature waste heat and utilization for producing hydrogen by Energy Build 2014;68(Part B):639–46.
reforming reaction of methane. ISIJ Int 2002;42:215–9. [171] Gong Z-X, Mujumdar AS. Thermodynamic optimization of the thermal pro-
[140] Zhang G, Li J, Chen Y, Xiang H, Ma B, Xu Z, et al. Encapsulation of copper- cess in energy storage using multiple phase change materials. Appl Therm
based phase change materials for high temperature thermal energy storage. Eng 1997;17:1067–83.
Sol Energy Mater Sol Cells 2014;128:131–7. [172] Dawson D, Bradshaw B, Goods S. Final test and evaluation results from the
[141] Mobley CE, Rapp RA. Hypereutectic heat storage shot. Columbus, Ohio: The Solar Two project. Appendix H: coupon corrosion tests, salt chemistry and
Ohio State University Research Foundation; 1986. post mortem analysis. Kruizenga Sandia National Laboratories; 2002.
[142] Bennett MR, Braunstein J. Hypereutectic heat storage alloy: silicon shell [173] Kroger JW. Evaulation of Hastelloy N alloys after nine years exposure to both
integrity in molten Al–Si eutectic. Oak Ridge, Tennessee: Oak Ridge National a molten flouride salt and air at temperatures from 700 to 560 °C. Oak Ridge,
Lab; 1988. Tennessee: Oak Ridge National Laboratory; 1972 ORNL-TM-4189.
1432 M. Liu et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1411–1432

[174] McCoy HE, McNabb B. Intergranular cracking of INOR-8 in the MSRE. [195] Liu S, Liu Z, Wang Y, Tang J. A comparative study on the high temperature
Laboratory, Oak Ridge, Tennessee: Oak Ridge National; 1972. corrosion of TP347H stainless steel, C22 alloy and laser-cladding C22 coating
[175] Keiser JR. Status of tellurium–Hastelloy N studies in molten fluoride salts. in molten chloride salts. Corros Sci 2014;83:396–408.
Oak Ridge, Tennessee: Oak Ridge National Laboratory; 1977 ORNL/TM-6002. [196] Vossen JPT, Janssen AHH, de Wit JHW. Corrosion behavior of nickel–iron
[176] Gupta DK, Rapp RA. The solubilities of NiO, Co3O4, and ternary oxides in fused alloys in molten carbonate. J Electrochem Soc 1996;143:58–66.
Na2SO4 at 1200 °K. J Electrochem Soc 1980;127:2194–202. [197] Vossen JPT, Makkus RC, de Wit JHW. Corrosion behavior of chromium in
[177] Gupta DK, Rapp RA, Erratum. The solubilities of NiO, Co3O4, and ternary molten carbonate. J Electrochem Soc 1996;143:66–73.
oxides in fused Na2SO4 at 1200 °K. J Electrochem Soc 1980;127:2656; [198] Vossen JPT, Plomp L, de Wit JHW. Corrosion of nickel in molten carbonate. J
1980;127:2656: 1980;127:2656. 1980;127. Electrochem Soc 1994;141:3040–9.
[178] Jose PD, Gupta DK, Rapp RA. Solubility of α-Al2O3 in Fused Na2SO4 at 1200 K. J [199] Vossen JPT, Plomp L, de Wit JHW, Rietveld G. Corrosion behavior of stainless
Electrochem Soc 1985;132:735–7. steel and nickel‐base alloys in molten carbonate. J Electrochem Soc
[179] Nishikata A, Numata H, Tsuru T. Electrochemistry of molten salt corrosion. 1995;142:3327–35.
Mater Sci Eng: A 1991;146:15–31. [200] Macdonald JR. Impedance spectroscopy. Ann Biomed Eng 1992;20:289–305.
[180] Orfield ML, Shores DA. Solubility of NiO in molten Li2CO3, Na2CO3, K2CO3, and [201] Aung NN, Liu X. High temperature electrochemical sensor for in situ mon-
Rb2CO3 at 910 °C. J Electrochem Soc 1988;135:1662–8. itoring of hot corrosion. Corros Sci 2012;65:1–4.
[181] Orfield ML, Shores DA. The solubility of NiO in binary mixtures of molten [202] Ni CS, Lu LY, Zeng CL, Niu Y. Evaluation of corrosion resistance of aluminium
carbonates. J Electrochem Soc 1989;136:2862–6. coating with and without annealing against molten carbonate using elec-
[182] Shi DZ, Rapp RA. The solubility of SiO2 in fused Na2SO4 at 900 °C. J Electro- trochemical impedance spectroscopy. J Power Sources 2014;261:162–9.
chem Soc 1986;133:849–50. [203] Frangini S, Loreti S. The role of temperature on the corrosion and passivation
[183] Tzvetkoff T, Girginov A, Bojinov M. Corrosion of nickel, iron, cobalt and their of type 310S stainless steel in eutectic carbonate melt. J Power Sources
alloys in molten salt electrolytes. J Mater Sci 1995;30:5561–75. 2006;160:800–4.
[184] Zhang YS. Solubilities of Cr2O3 in fused Na2SO4 at 1200 K. J Electrochem Soc [204] Olson L, Sridharan K, Anderson M, Allen T. Nickel-plating for active metal
1986;133:655–7. dissolution resistance in molten fluoride salts. J Nucl Mater 2011;411:51–9.
[185] Zhang YS, Rapp RA. Solubilities of α-Fe2O3 and Fe3O4 in Fused Na2SO4 at [205] Frangini S, Masci A, Zaza F. Molten salt synthesis of perovskite conversion
1200 K. J Electrochem Soc 1985;132:2498–501. coatings: a novel approach for corrosion protection of stainless steels in
[186] Grabke HJ, Reese E, Spiegel M. The effects of chlorides, hydrogen chloride, molten carbonate fuel cells. Corros Sci 2011;53:2539–48.
and sulfur dioxide in the oxidation of steels below deposits. Corros Sci [206] Frangini S, Zaza F, Masci A. Molten carbonate corrosion of a 13-Cr ferritic
1995;37:1023–43. stainless steel protected by a perovskite conversion treatment: relationship
[187] Fernández AG, Rey A, Lasanta I, Mato S, Brady MP, Pérez FJ. Corrosion of with the coating microstructure and formation mechanism. Corros Sci
alumina-forming austenitic steel in molten nitrate salts by gravimetric ana- 2012;62:136–46.
lysis and impedance spectroscopy. Mater Corros 2014;65:267–75. [207] Laurenty BF, Fukuda G, Damba DD, Peterson PF. Inhibiting corrosion by
[188] McConohy G, Kruizenga A. Molten nitrate salts at 600 and 680 °C: thermo- molten fluoride salts: investigations on Flinak. Berkeley: University of Cali-
physical property changes and corrosion of high-temperature nickel alloys. fornia; 2005.
Sol Energy 2014;103:242–52. [208] Wagner SJ, Rubin ES. Economic implications of thermal energy storage for
[189] Lim T-H, Hwang ER, Ha HY, Nam SW, Oh I-H, Hong S-A. Effects of tem- concentrated solar thermal power. Renew Energy 2014;61:81–95.
perature and partial pressure of CO2/O2 on corrosion behaviour of stainless- [209] Nithyanandam K, Pitchumani R. Optimization of an encapsulated phase
steel in molten Li/Na carbonate salt. J Power Sources 2000;89:1–6. change material thermal energy storage system. Sol Energy 2014;107:770–
[190] Yan YF, Xu XQ, Zhou DQ, Wang H, Wu Y, Liu XJ, et al. Hot corrosion behaviour 88.
and its mechanism of a new alumina-forming austenitic stainless steel in [210] Bai F, Wang Y, Wang Z, Sun Y. Economic evaluation of shell-and-tube latent
molten sodium sulphate. Corros Sci 2013;77:202–9. heat thermal energy storage for concentrating solar power applications. In:
[191] Kruizenga AM. Corrosion mechanisms in chloride and carbonate salts. Proceedings of Solar PACES 2014. Beijing (China); 2014.
Albuquerque, New Mexico: Sandia National Laboratories; 2012 SAND2012- [211] Kelly B, Kearney D. Thermal storage commercial plant design for a 2-tank
7594. indirect molten salt system. Golden, Colorado: National Renewable Energy
[192] Sridharan K, Allen TR. Chapter 12 – corrosion in molten salts. In: Lantelme F, Laboratory; 2006 NREL/SR-550040166.
Groult H, editors. Molten salts chemistry. Elsevier; 2013. [212] Turchi C, Heath G. Molten salt power tower cost model for the system advisor
[193] Yang L, Bo C, Junwei W, Zhiping W, Wensheng L. Corrosion behavior of Cr, Fe model (SAM). Golden, Colorado: National Renewable Energy Laboratory;
and Ni based superalloy in molten NaCl. Rare Met Mater Eng 2014;43:17–23. 2013 NREL/TP-5500-57625; 2013.
[194] Vignarooban K, Pugazhendhi P, Tucker C, Gervasio D, Kannan AM. Corrosion
resistance of Hastelloys in molten metal-chloride heat-transfer fluids for
concentrating solar power applications. Sol Energy 2014;103:62–9.