OVERVIEW OF MOLTEN SALT STORAGE SYSTEMS AND
MATERIAL DEVELOPMENT FOR SOLAR THERMAL POWER PLANTS
Thomas Bauer
Institute of Technical Thermodynamics
German Aerospace Center (DLR)
Linder Höhe, 51147 Köln, Germany
e-mail: [email protected]
ABSTRACT
Nils Breidenbach
Nicole Pfleger
Doerte Laing
Markus Eck
Institute of Technical Thermodynamics
German Aerospace Center (DLR)
Pfaffenwaldring 38-40, 70569 Stuttgart, Germany
e-mail: [email protected]
This paper gives an overview of thermal energy storage
(TES) systems based on molten salts. It summarizes stateof-the-art molten salt TES systems. Storage systems in a
research stage, such as the thermocline and floating barrier
concepts, are also discussed. The paper summarizes
previous work on molten salt material investigations at
DLR. Experimental results of novel salt formulations with
melting temperatures well below 100 °C are presented. An
overview of temperature dependent thermophysical property
values of Solar Salt is given. The thermal stability of
nitrate/nitrite mixtures was examined by dynamic
thermogravimetry and static measurements at constant
temperature. We discuss in this paper the impact of the
atmospheric conditions (partial oxygen pressure) and salt
composition on the stability of nitrate salts. At DLR, a
molten salt test loop is in the design stage. A scheme of the
test loop and the options for experiments are presented.
plants. At these temperatures molten salts are attractive
candidates because they have advantages in terms of a high
heat capacity, high density, high thermal stability, relatively
low cost and low vapor pressure. The low vapor pressure
results in storage designs without pressurized vessels. In
general there is experience with molten salts from a number
of industrial applications related to heat treatment,
electrochemical reactions and heat transfer. The application
of salts requires the consideration of the lower temperature
limit defined by the melting temperature. One major
difficulty with molten salts is unwanted freezing during
operation. Freezing must be usually prevented in the piping,
the heat exchanger and in the storage tanks using auxiliary
heating systems. Hence, salt mixtures with a low melting
temperature are developed. At high temperatures, salt
stabilities and corrosion aspects play a major role. For solar
thermal power plants, a non-eutectic salt mixture of 60 wt%
sodium nitrate and 40 wt% potassium nitrate is typically
utilized. This mixture is commonly called Solar Salt.
1. INTRODUCTION
2. MOLTEN SALT STORAGE SYSTEM OVERVIEW
Solar thermal power plants are a key technology for
electricity generation from renewable energy resources.
Thermal energy storage (TES) makes it possible to meet the
intermediate load profile with dispatchable power, a benefit
that has a high value to power utilities. Most commonly
three types of TES systems are distinguished. These are
sensible heat, latent heat and thermochemical heat storage.
The two groups of sensible heat storage are solids (e.g.
concrete, rock, ceramics) and liquids (e.g. water, molten
salt) [1].
The presented paper focuses on high-temperature (> 250 °C)
molten salt storage for concentrated solar power (CSP)
2.1. State of the art storage systems
At present, the two-tank molten salt storage is the only
commercially available concept for CSP plants with a large
thermal capacity requirement. This storage system
essentially consists of two tanks filled with molten salt at
different temperature and fill levels. The two major types
are direct and indirect TES systems. Figure 1 shows the
scheme of a direct system. In this system, the salt is at the
same time the heat transfer fluid (HTF) and the storage
medium. The first larger two-tank molten salt storage
system was demonstrated in the Solar Two Project at Sandia
1
National Laboratories, commissioned in 1996 with a tower
power plant. The operation time was 14 months and the salt
mass was about 1400 tones (Figure 1) [2]. In 2011 the
commercial solar tower power plant Gemasolar in
Andalusia was commissioned. Table 1 shows the main
characteristics of this storage system.
In an indirect storage system, the thermal storage is decoupled from the HTF loop of the solar receiver via a heat
exchanger (Figure 2). The first commercial system with
indirect heat exchange has been in operation in the solar
thermal power plant Andasol 1 since 2009 [3]. Table 1
compares commercial direct and indirect large-scale twotank molten salt TES systems.
Fig. 1: Scheme with two-tank direct storage system [2].
Table 1 shows that both systems have similar thermal
capacities. On the other hand, the salt mass of the direct
storage system is considerably lower compared to the
indirect system. This is due to the fact that the capacity of a
molten salt storage is proportional to the temperature
difference between the hot and the cold tank. In other
words, the low temperature difference between the cold and
hot tank of the described indirect system leads to a large
sized TES system.
2.2. Storage systems in a research stage
Several concepts for molten salt storage try to minimize
costs by using a single-tank system. A single tank requires
stratification with defined hot and cold temperature zones
and free convection is undesired. Free convection can be
suppressed by a floating barrier within the volume or
additional filler materials. The floating barrier concept was
examined by the company Sener [7]. The filler material
approach is commonly called thermocline concept. An
additional major advantage of the thermocline system would
be that molten salt volume is replaced by inexpensive filler
materials. Sandia National Laboratories performed material
investigations of the compatibility of filler materials and
molten nitrate salt, as well as modeling and experiments on
a system level [4,8,9]. Yang et al. modeled heat transfer
aspect of the thermocline concept [10,11]. Another paper
discusses further details of commercial and noncommercial molten salt TES systems [12].
3. MATERIAL REVIEW AND EXAMINATIONS
3.1. Thermophysical Properties of Solar Salt
Fig. 2: Scheme with two-tank indirect storage system [4].
TABLE 1 : COMMERCIAL TWO-TANK SYSTEMS
Direct storage [5,6] Indirect storage [3]
Gemasolar
Andasol 1
~ 1000 MWh #
1010 MWh
8500 tones
28500 tones
290 °C
292 °C
~565 °C
386 °C
#
Estimated from ΔT = 275 K, 8500 tones, 1.55 kJ kg-1 K-1
System name
Thermal capacity
Inventory
Cold tank temp.
Hot tank temp.
Previous work at DLR focused on temperature dependent
thermophysical properties of NaNO 3 [13]. In the present
paper, literature values of the density ρ, heat capacity c p ,
thermal diffusivity a and thermal conductivity k of Solar
Salt are compared. Reliable thermophysical properties are
important for the modeling and dimensioning of molten salt
storage systems. Previous authors focused mainly on single
thermophysical property values. We verified several
literature values using the correlation k = a · ρ · c p . Another
paper discusses details of the methodology, literature
sources and correlations between single salt and Solar Salt
properties [14]. Figure 3 includes consistent correlations of
thermophysical properties of Solar Salt. Results show that
these properties differ in terms of the uncertainty. Density
values show the lowest scattering. Heat capacity data vary
slightly more than ± 5 % compared to the average value.
Thermal diffusivity and thermal conductivity values showed
the strongest variations (± 15 %).
2
1.95
Murgulescu 1959 (Interpolation)
ρ in g·cm-3
1.9
Murgulescu 1969 (Interpolation)
Polyakov 1955 (Interpolation)
1.85
Pacheco 1995, Zavoico 2001 (SolarTwo)
Janz recom. 1972 (with +/-1.5 % error bar)
1.8
Janz extrapolated
Bradshaw 2009 (Molar volume addition)
1.75
ρ (g·cm-3) = 2.1060 - 6.6795E-04· T(°C)
cp in J·g-1·K-1
1.7
1.7
Rogers 1982 (Calc. from single salts)
Zavoico 2001 (SolarTwo)
Takahashi 1988 (Calc. from single salts)
Carling 1983 (Calc. from single salts)
Gustafsson 1968 (Calc. from single salts)
Jriri 1995 (Calc. from single salts)
Bradshaw 1987
Average without Zavoico
1.65
1.6
1.55
±5 % error bar of average
1.5
k in W·m-1·K-1
a in mm²·s-1
cp (J·g-1·K-1) = 1.5404 + 3.0924E-5· T(°C)
1.45
0.22
0.21
0.2
0.19
0.18
0.17
0.16
0.15
0.14
0.65
Gustafsson 1968 (Interpolation)
Ohta 1990 (Interpolation)
Kato 1983 (Interpolation)
Knothe 1985 (Interpolation)
Calculated using data by Janz and
average values
a = k / (cp·ρ)
± 15 % error bar of average
k (W·m-1·K-1) = 0.3804 + 3.452E-04 · T (°C)
±15 % error bar of average
0.6
Foosnæs 1982 (Interpolation)
McDonald 1970 (interpolation)
Omotani 1982 (Interpolation)
Tufeu 1985 (Interpolation)
Bloom 1965 (Interpolation)
White 1967 (Interpolation)
Zavoico 2001
Santini 1984 (Interpolation)
Kitade 1989 (Interpolation)
Turnbull 1961 (Interpolation)
Average without Zavoico
0.55
0.5
0.45
0.4
0
50
100 150 200 250 300 350 400 450 500 550 600
Temperature in °C
Fig. 3: Temperature dependent thermophysical data of Solar Salt
3
3.2. Literature review of alkali nitrate/nitrite salt mixtures
3.3. Development of low melting temperature mixtures
Table 2 presents a systematic list of the melting
temperatures of single salts and the minimum melting
temperature of salt systems with the cations calcium (Ca),
potassium (K), lithium (Li) and sodium (Na) (Table 2
vertical) and the anions nitrate (NO 3 ) and nitrite (NO 2 )
(Table 2 horizontal) using various literature sources [15].
The melting temperatures of the listed single salts range
from 220 °C (LiNO 2 ) to 561 °C (Ca(NO 3 ) 2 ). Salt mixtures,
rather than single salts, have the advantage of a lower
melting temperature. These mixtures can have similar
thermal stability limits as the single salts. Hence, salt
mixtures, such as eutectics, can have a wider temperature
range compared to single salts. Table 2 shows commonly
considered salt systems for solar applications in grey
background. These are the binary K,Na//NO 3 system with a
common anion (containing Solar Salt), the ternary
reciprocal systems K,Na//NO 3 ,NO 2 (containing a mixture
called Hitec) and the two ternary additive systems with a
common anion Ca,K,Na//NO 3 (HitecXL) and K,Li,Na//NO 3 .
Table 2 clearly shows the tendency of melting temperature
depression from the left to right hand side, as well as from
the top to the bottom of the table. Hence, it can be expected
that systems with a liquidus temperature lower than 80 °C
are feasible.
Bradshaw et al. published data of the quaternary reciprocal
system K,Li,Na//NO 2 ,NO 3 with a minimum melting
temperature of about 75 °C (Table 1) with the composition
33.5 mol% Li+, 47.9 mol% K+, 18.6 mol% Na+ and a
nitrate/nitrite ratio of 0.56 [16]. Own work focuses on phase
diagram determinations of this system using a heat flux type
differential scanning calorimeter (DSC) and a melting point
apparatus (MPA). A previous paper discusses details of the
experimental method [15]. Figure 4 shows DSC and MPA
measurement results of the liquidus temperature of the
system K,Li,Na//NO 2 ,NO 3 with a fixed NO 3 /NO 2 ratio of
about 0.56. The measurement points are marked by an
asterisk symbol. The mixture with the lowest liquidus
temperature had the composition: Li+ 33 mol%, K+ 48 mol%
and Na+ 19 mol% with a ratio NO 3 /NO 2 = 0.56. Figure 5
shows DSC measurements of this mixture. It can be seen
that a minimum melting mixture rather than a mixture with
a melting range was successfully identified [15].
TABLE 2 : MATRIX OF (MINIMUM) MELTING
TEMPERATURE OF SUBSYSTEMS OF THE QUINARY
RECIPROCAL SYSTEM Ca,K,Li,Na//NO 2 ,NO 3 [15]
#
Decomposition at the melting temperature
Fig. 4: Part of liquidus phase diagram K,Li,Na //
NO 2 ,NO 3 ; the minimum is marked red
Liquidus temperature
[°C]
7
Heating rate 5 K/min
Heating rate 2 K/min
Heating rate 1 K/min
6
DSC signal [μV/mg]
NO 2
NO 3
NO 2 , NO 3
Single salts and binary systems with common cation
Ca
398 °C#
561 °C#
393 °C
K
440 °C
334 °C
316-323°C
Li
220 °C
254 °C
196 °C
Na
275 °C
306 °C
226-233 °C
Binary systems with common anion and ternary reciprocal
Ca,K
185 °C
145-174 °C 130 °C
Ca,Li
205-235 °C 235 °C
178 °C
Ca,Na
200-223 °C 226-230 °C 154 °C
K,Li
98 °C
126 °C
94 °C
K,Na
225 °C
222 °C
142 °C
Li,Na
151 °C
196 °C
126 °C
Ternary additive common anion and quaternary reciprocal
Ca,K,Li
N/A
117 °C
N/A
Ca,K,Na
N/A
130 °C
N/A
Ca,Li,Na
N/A
170 °C
N/A
K,Li,Na
N/A
119 °C
75 °C
Quaternary additive common anion and quinary reciprocal
Ca,K,Li,Na
N/A
109 °C
N/A
5
4
3
2
85
84
83
82
81
0
1 2 3 4 5 6
Heating rate [K/min]
1
0
-1
-50
0
50
100
150
200
Temperature [°C]
Fig. 5: DSC measurement results of the mixture Li +
33 mol%, K + 48 mol% and Na + 19 mol%
4
Mass losses with gas evolution of alkali metal nitrate salts
may occur due to three mechanisms: nitrite formation in the
melt and oxygen release (1), alkali metal oxide formation in
the melt and nitrogen/nitrogen oxide release (2) and
vaporization of the nitrate salts (3) [17,18]. The
decomposition temperature depends on various aspects.
They include the definition itself, the experimental method
(e.g. heating rate), crucible material and atmosphere. A
comparison of results from different authors is usually
difficult, because these aspects differ among the authors.
Hence, own measurements of single salts and selected salt
mixtures were performed. Measurements parameters were
atmospheric pressure, dry nitrogen and a heating rate of
10 K min-1. Another paper discusses further details of the
experimental method [15]. Figure 6 presents results of
thermogravimetry (TG) measurement of single nitrate and
nitrite salts, as well as four mixtures with a low melting
temperature.
Stern reports the decomposition pressure, or alternatively
the equilibrium constant, in increasing order as follows
KNO 3 < NaNO 3 < Ca(NO 3 ) 2 < LiNO 3 (highest stability for
KNO3). Nitrites are considered to be less stable than nitrates
[18]. Own measurements show that mass losses occur
mostly in a temperature range from 500 °C to 600 °C. Less
stable salts were LiNO 3 and Ca(NO 3 ) 2 , as well as the
mixture Ca,K,Na//NO 3 . The stabilities of the salt mixtures
K,Na//NO 3 (Eutectic) and K,Li,Na//NO 2 ,NO 3 (80 °C) were
similar to the single salt NaNO 3 . Hence, the novel low
melting salt mixture K,Li,Na//NO 2 ,NO 3 is a promising
candidate as HTF due to its wide operation range.
101
Mass loss - TG signal [%]
100
99
KNO3
NaNO3
Ca(NO3)2
LiNO3
NaNO2
KNO2
K,Li,Na // NO2,NO3 (Tm=75°C)
Ca,K,Na // NO3 (Hitec XL)
K,Na // NO3 (Eutectic)
K,Na // NO2,NO3 (Hitec)
98
97
96
95
stability tests in an open-system in oxygen-nitrogen
atmospheres at ambient pressure.
Static measurements at constant temperatures in synthetic
air atmosphere were performed. It is well known that molten
nitrates decompose to nitrites. The thermal dissociation is
reversible and for NaNO 3 the equilibrium reaction can be
written as shown in Equation 1.
NaNO3( l ) ⇔ NaNO2 ( l ) + 12 O2 ( g )
(1)
We examined the kinetics of the nitrite formation in the
NaNO 3 melt in the temperature range 450 to 550 °C. A full
discussion of these measurements is presented in [15].
Figure 7 plots static NaNO 3 measurements at the three
temperatures 450, 500 and 550 °C. Synthetic air with a rate
of 100 and 600 ml min-1 was purged through the melt. The
figure shows the nitrite formation depending on the time
with the salt temperature as a parameter. A simple empirical
exponential growth model was fitted to the measurements
with 100 ml min-1 using non-linear regression techniques.
The model uses two parameters and can approximately
describe the reaction kinetics. The three model curves show
that the temperature level not only affects the amount of
NO 2 - in equilibrium but also the decomposition rate. The
equilibrium at 550 °C was quickly reached after several tens
of hours. On the other hand, at 450 °C the time constant was
much longer (several hundreds of hours).
Molar ratio NO2-/NO3- ,pO2=0.21(air)
3.4. Thermal stability examination of nitrate/nitrite salts
0.045
550
min-1 -1
550°C
°C100
100mlml·min
500
min-1 -1(Experiment
1) 1)
500°C
°C600
600mlml·min
(Experiment
500
min-1 -1(Experiment
1) 1)
500°C
°C100
100mlml·min
(Experiment
-1
500°C
°C100
100mlml·min
(Experiment
500
min-1 (Experiment
2) 2)
-1
450°C
°C100
100mlml·min
450
min-1
0.040
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
0
100
200
300
400
500
600
700
800
900 1000
Time t in hours
Fig. 7: Kinetics of nitrite formation in NaNO3 in synthetic
air atmosphere in an open type system
94
0
100
200
300
400
Temperature [°C]
500
600
700
Fig. 6: Thermogravimetric measurement results of single
alkali nitrate salts (filled symbols) and alkali metal nitrite
salts (asterisk symbols) and salt mixtures (open symbols)
More detailed stability examination focused on NaNO 3 and
Solar Salt. The thermal stability of these nitrate salts was
examined by two methods. These were dynamic and static
For Solar Salt, we examined not only the kinetics of nitrite
formation but also the kinetics of oxide formation in the
melt by an indirect method. The examined oxides form an
alkaline solution if dissolved in water. Subsequently, an
acid-base titration with HCl solution was performed [14].
As discussed in the previous paragraph, the primary
decomposition reaction with nitrite formation in the melt
and oxygen release is well examined. The secondary
decomposition reaction with alkali metal oxide formation
5
Fig. 8: Kinetics of oxide formation in Solar Salt in synthetic
air atmosphere in an open type system at 550 °C
oxygen pressure leads to a high decomposition temperature.
In other word, the stability of NaNO 3 and Solar Salt
improve with increased partial oxygen pressures. Results
were extrapolated to equilibrium conditions (0 K min-1
heating rate). For NaNO 3 and equilibrium conditions,
results show a decomposition temperature of 525 °C in
synthetic air and 542 °C in oxygen at atmospheric pressures.
For Solar Salt and equilibrium conditions, the values are
529 °C (synthetic air) and 562 °C (oxygen). This value is
lower compared to the previously reported value of 565 °C
[8]. Own measurements refer to a mass loss of 3wt%. It
should be considered that different mass loss definitions will
result in other stability limits. Also, static and dynamic test
may result in different decomposition temperatures. Our
measurements show that the thermal stability of Solar Salt is
higher compared to NaNO 3 . It can be also seen that the
stability of Solar Salt (Figure 10) depends more strongly on
the partial oxygen pressure than NaNO 3 (Figure 9). For
NaNO 3 the stability difference between synthetic air and
oxygen is 17 K (542 °C minus 525 °C), whereas the Solar
Salt value is 33 K.
Thermal decomposition temperature Tz
of NaNO3 3wt% in °C
and nitrogen/nitrogen oxide is less understood and oxides
within the melt could be corrosive. The chemistry of these
oxygen species is complex and some of the evidence is in
conflict. In literature the formation of oxide, superoxide and
peroxide is discussed [18,19]. Results in Figure 8 refer to
measurements of Solar Salt at 550 °C in synthetic air
atmosphere. The figure plots not only the NO 2 -/NO 3 - ratio
(left hand axis) but also the equivalence point of the titration
(right hand axis). It can be seen that the NO 2 -/NO 3 - ratio
reaches equilibrium after a few ten hours as opposed to the
volume of titration which increases steadily. It can be
concluded that although the NO 2 -/NO 3 - ratio reaches
equilibrium, the alkali metal oxide level with nitrogen
release reaches no equilibrium within the experimental time
frame. Hence, further work focuses on the long-term
thermal stability limit of Solar Salt to examine the different
decomposition mechanisms.
680
TZ = 612.5 + 3.474 ⋅ pO 2
660
TZ = 595.8 + 4.084 ⋅ pO 2
0.5
10 K∙min-1
640
5 K∙min-1
620
TZ = 569.1 + 2.700 ⋅ pO 2
600
0.5
2 K∙min-1
580
TZ = 525.3 + 3.069 ⋅ pO 2
560
0.5
0.5 K∙min-1
540
0 K∙min-1 (extrapolated
from 0.5 and 2 K∙min-1)
520
500
0
10
70
20
30
40
50
60
80
90
Relative partial oxygen pressure po2 in % (21 % = air)
100
Fig. 9: Experimental thermogravimetry results of the
thermal stability of NaNO 3 depending on the partial oxygen
pressure and the heating rate (0.5 K min-1 to 10 K min-1).
Thermal decomposition temperature Tz
of Solar Salt 3wt% in °C
In addition to static measurements, dynamic measurements
with a heating ramp were performed by thermogravimetry
(TG) [14]. The major parameters were the salt composition,
partial oxygen pressure and the heating rate. This approach
led to a large number of TG measurements. For all
measurements, the thermal decomposition temperature
refers to the temperature with a mass loss of 3 wt%
compared to the base line. The atmospheres were pure
nitrogen or an oxygen-nitrogen mixture with a total flow of
100 ml min-1. The heating rates ranged from 0.5 K min-1 to
10 K min-1.
Work at Sandia aims for oxygen-stabilized Solar Salt up to
650°C [20]. Own work focused on the impact of the partial
pressure of oxygen on the thermal stability. Figure 9 and 10
plot results of dynamic TG-measurements of NaNO 3 and
Solar Salt depending on the partial oxygen pressure. It can
be seen that the measured decomposition temperature
depends on the heating rate. Higher heating rates result in
higher decomposition temperatures due to the measurement
principle. All measurement series show that a higher partial
0.5
680
TZ = 603.2 + 6.635 ⋅ pO 2
660
5 K∙min-1
0.5
640
2 K∙min-1
620
TZ = 576.7 + 7.490 ⋅ pO 2
600
580
TZ = 520.9 + 6.439 ⋅ pO 2
0.5
0.5
0.5 K∙min-1
560
0 K∙min-1 (extrapolated
from 0.5 and 2 K∙min-1)
540
520
500
0
10
20
40
50
70
30
60
80
90
Relative partial oxygen pressure po2 in % (21 % = air)
100
Fig. 10: Experimental thermogravimetry results of the
thermal stability of Solar Salt depending on the partial
oxygen pressure and heating rate (0.5 K min-1 to 5 K min-1)
6
4. DESIGN OF A MOLTEN SALT LOOP AT DLR
Research on molten salt as heat transfer or storage medium
focuses currently on mainly two aspects:
•
•
Development of innovative salt mixtures with lower
melting temperature and/or higher thermal stability
Demonstration and optimization of molten salt
processes e.g. as heat transfer medium in parabolic
troughs or in innovative storage systems.
Especially the second focus has gained more importance.
The main advantage of using molten salt in parabolic trough
plants are the potential reduction of investment costs for the
heat transfer fluid, since nitrate salt is less expensive than
thermal oil and the simplification of the process technology
because the molten salt heat transfer fluid also acts as the
storage medium. Another advantage is the higher operation
temperature of the solar field. This enables a higher
conversion efficiency of the power block and a larger
temperature difference in the storage system and thus a
significantly increased specific storage capacity (Table 1).
However, to-date no commercial large-scale parabolic
trough plants applying this technology has been realized.
Converting the design of parabolic trough collector is not a
straightforward issue. The operation temperature is more
than 150 °C above state of the art collectors and components
such as receivers, flexible tube connections and control
valves, have to be qualified and have to prove reliability
under realistic operating conditions. The challenging
chemical behavior, creepage and corrosion of molten salts
introduce new requirements on components that are in
contact with the salt.
Even though molten salt storage systems are commercially
available for indirect and direct storage systems and realized
in ANDASOL-like power plants [3] as well as in the
Gemasolar power plant [6], storage costs are a major issue in
the economic evaluation of concentrated solar power (CSP)
plants. Therefore, the development of new storage concepts
with reduced costs for the storage material, such as the
thermocline concept [8] or the floating barrier concept [7] is
crucial for improving the profitability of CSP. Additionally,
components such as pumps, valves and other instruments are
a major engineering issue in a molten salt system with
parabolic trough collectors or a central receiver and a molten
salt storage. Up to now, only a few suppliers are available
and thus experience with molten nitrate salt operation over
500 °C is limited.
In order to face the mentioned challenges a molten salt test
loop as shown in Figure 11 will be erected at DLR. The test
loop will be designed to be able to store the complete salt
inventory in the cold tank. By using the cold tank as a drain,
different components can be installed in the measurement
section. In addition, the hot tank can be drained in order to
test various thermocline setups. The system will have a
heater to achieve salt temperatures over 500 °C and a heat
sink for operating thermal cycles. For the testing and
examination of components and storage tanks, the test loop
is anticipated to have a scale which can provide real
operation conditions of a solar thermal power plant in terms
of mass flow rates and receiver diameters. For the first test
phase, Solar Salt will be used as the HTF and storage
medium.
Cooler
Measurement
Section
Heater
Cold
Tank
Hot Tank /
Thermocline
Fig. 11: Schematic of the molten salt test loop in a design
stage at DLR
5. SUMMARY AND CONCLUSION
This paper gave an overview of commercial molten salt TES
systems for CSP plants. These systems are the two-tank
direct storage system for power towers and two-tank
indirect storage systems for parabolic troughs. The
following research and development aspects were identified:
•
•
•
•
Identification and characterization of salt
formulations with a low liquidus temperature and
suitable thermal stability
Examination of nitrate salts at their thermal
stability limit under different atmospheric
conditions including corrosion aspects
Component tests under realistic operation
conditions (e.g. valves, pumps, solar collectors)
Development of new storage concepts with
improved economics compared to the two-tank
design (e.g. floating barrier, thermocline concept)
Phase diagram measurements demonstrated that novel salt
mixtures with a low melting temperature are feasible.
However, the identification of further suitable multi-
7
component mixtures requires a large number of
measurements due to their degrees of freedom. Thus,
advanced techniques to determine phase diagrams will be
required. They include phase diagram prediction by
modeling and high-throughput experimental methods.
The thermal stability of nitrate salts depends on several
aspects. They include the salt composition, the gas
composition with partial pressures and the release of
decomposition products (open and closed type system). The
paper described results on the primary decomposition
reaction with nitrite formation in the melt and oxygen
release and a secondary decomposition reaction with alkali
metal oxide formation in the melt and nitrogen/nitrogen
oxide release. Measurements indicate that the kinetic time
constants of these two decompositions are not the same
under the examined experimental conditions. Hence, further
work on the nitrate salt chemistry near the stability limit is
required.
In order to assess operational challenges on a system level, a
flexible molten salt test loop is in a design stage at DLR.
This test loop should allow for component tests with typical
mass flow rates and tube diameters in CSP plants. In addition,
thermal cycling operation and full drainage procedures are
anticipated. This should allow for the research and
development of new single-tank concepts in a sufficient
scale.
6. ACKNOWLEDGMENTS
We express our thanks especially to Ulrike Kröner and
Markus Braun for the experimental part and data analysis.
7. REFERENCES
(1) Tamme, R., Bauer, T. and Hahne, E., Heat Storage
Media, Ullmann's Encyclopedia of Industrial Chemistry,
2000
(2) Pacheco, J., Final Test and Evaluation Results from the
Solar Two Project, SAND2002-0120, 2002
(3) Relloso, S. and Delgado, E., Experience with Molten
Salt Thermal Storage in a Commercial Parabolic Trough
Plant. ANDASOL-1 Commissioning and Operation, Solar
Paces, Berlin, Germany, 2009
(4) Kolb, G. J., Evaluation of Annual Performance of 2Tank and Thermocline Thermal Storage System for Trough
Plants, Solar Paces, Perpignan, France, 2010
(5) Concentrating Solar Energy of the future: Torresol
Energy's molten salt power tower CSP plant Gemasolar,
Online Article: www.solarserver.com (accessed 13.3.2012)
(6) Relloso, S. and Lata, J., Molten Salt Thermal Storage:
A Proven Solution to increase Plant Dispatchability.
Experience in Gemasolar Tower Plant, Solar Paces,
Granada, Spain, 2011
(7) Lata, J. and Blanco, J., Single Tank Thermal Storage
Design for Solar Thermal Power Plants, Solar Paces,
Perpignan, France, 2010
(8) Pacheco, J. E., Showalter, S. K. and Kolb, W. J.,
Development of a Molten-Salt Thermocline Thermal
Storage System for Parabolic Trough Plants, Journal of
Solar Energy Engineering 124, pp. 153, 2002
(9) Brosseau, D. A., Hlava, P. F. and Kelly, M. J., Testing
Thermocline Filler Materials and Molten-Salt Heat Transfer
Fluids for Thermal Energy Storage Systems Used in
Parabolic Trough Solar Power Plants, SAND2004-3207,
2004
(10) Yang, Z. and Garimella, S. V., Thermal analysis of
solar thermal energy storage in a molten-salt thermocline,
Solar Energy 84, pp. 974-985, 2010
(11) Yang, Z. and Garimella, S. V., Molten-salt thermal
energy storage in thermoclines under different
environmental boundary conditions, Applied Energy 87, pp.
3322-3329, 2010
(12) Breidenbach, N., Bauer, T., Laing, D. and Tamme, R.,
Assessment and development of molten salt storage and
heat transfer fluids for solar thermal power plants and
industrial processes, World Engineers' Convention, Geneva,
Italy, 2011
(13) Bauer, T., Laing, D. and Tamme, R., Characterization
of Sodium Nitrate as Phase Change Material, International
Journal of Thermophysics 33, pp. 91-104, 2012
(14) Bauer, T., Pfleger, N., Breidenbach, N., Eck, M.,
Laing, D. and Kaesche, S., Material Aspects of Solar Salt
for Sensible Heat Storage, Innostock 2012, The 12th
International Conference on Energy Storage, Lleida, Spain,
2012 (submitted for publication)
(15) Bauer, T., Laing, D. and Tamme, R., Recent Progress
in Alkali Nitrate/Nitrite Developments for Solar Thermal
Power Applications, Molten Salts Chemistry and
Technology, MS9, Trondheim, Norway, 2011
(16) Cordaro, J. G., Rubin, N. C., Sampson, M. D. and
Bradshaw, R. W., Multi-Component Molten Salt Mixtures
based on Nitrate/Nitrite Anions, Solar Paces, Perpignan,
France, 2010
(17) Carling, R.W., Kramer, C.M., Bradshaw, R.W.,
Nissen, D.A., Goods, S.H., Mar, R.W., Munford, J.W.,
Karnowsky, M.M., Biefeld, R.N., Norem, N.J., Molten
nitrate salt technology development - Development Status
Report, SAND80-8052, 1981
(18) Stern, H.K., High Temperature Properties and Thermal
Decomposition of Inorganic Salts with Oxyanions, CRCPress, 2001
(19) Plambeck, J.A., Fused salt systems, in "Encyclopedia
of electrochemistry of the elements Vol X", edited by A. J.
Bard, Marcel Dekker, 1976
(20) Bradshaw, R., Cordaro, J., Jayaraman, S., Sun, A.,
Siegel, N., Gill, D., Hachman, J., Zimmerman, M. and
Sartor, G., Advanced Heat Transfer Fluid Development
(CSP Agreement 16749), Presentation at DOE CSP
Program Annual Review, US, 2010
8