SolarPACES 2012 Commercial thermal storage.

Molten salts vs
Steam accumulators
1

C. Prieto , A. Jove F. Ruiz , A. Rodriguez , E. Gonzalez

1

Master of Chemical Engineering, Abengoa, c.prieto@solar.abengoa.com

Master of Mechanical Engineering, Abengoa, a.jove@solar.abengoa.com

3

Physical degree, Abengoa, fj.ruiz@solar.abengoa.com

Master of Mechanical Engineering, Abengoa, e.gonzalez@solar.abengoa.com

Master of Chemical Engineering, Abengoa, a.rodriguez@solar.abengoa.com

Abstract
From some years ago there is a very big increase of thermosolar power generation industry and with it its
associated thermal storage systems. They are crucial to ensure the success of the technology allowing
dispatchability enough to supply energy when its demanded. Two different technologies are currently
implemented commercially regarding thermal energy storage: the steam accumulator for direct steam
generation plants and the double tank of molten salts either for parabolic trough with thermal oil or the
molten salt tower technology.
Due to diversified demand profiles (with respect to type, amount, and power of needed energy) each
energy storage (electrical, thermal, mechanical or chemical storage) requires a specific, optimal solution
regarding efficiency and economics.
Keywords: Storage, thermal storage, molten salt, steam accumulator, double tank, phase change material

1.

Introduction

For thermal energy storage systems it can be derived, that there is more than one storage technology
needed to meet different applications. Consequently, a broad spectrum of storage technologies, materials
and methods are needed. The overall target in designing TES systems is the reduction of investment cost
and the enhancement of efficiency and reliability. To achieve these objectives, material, design and
system integration aspects have to be considered in equal measure.
The assessment of identification and selection of the optimal TES system only is not focus on the storage
material, further important components of the power plant also have to be included in this study: the
containment, and mainly the heat exchanger and structural parts for charging and discharging, and
furthermore devices and sub-components, which are needed for operation and integration such as pumps,
valves, control devices etc.
A key issue in the design of a thermal energy storage system is its thermal capacity. However, selection
of the appropriate system depends on many cost-benefit considerations, technical criteria and
environmental criteria.
Cost: the storage material itself, the heat exchanger for charging and discharging the system and the cost
of the space and/or enclosure for the TES.
Technical point of view: high energy density in the storage material (storage capacity); good heat
transfer between heat transfer fluid (HTF) and the storage medium (efficiency); mechanical and chemical
stability of storage material (must support several charging/discharging cycles); compatibility between
HTF, heat exchanger and/or storage medium (safety); complete reversibility of a number of
charging/discharging cycles (lifetime); low thermal losses; ease of control.

 Technology: operation strategy; maximum load; nominal temperature and specific enthalpy drop in
load; integration into the power plant.
Power of the cycle and hours number of TES requested
Efficiency of the power plant, thus this study should include the assessment of the solar field efficiency
and cycle efficiency.
Investment cost of the solar field and of the storage system and as well as an estimation of O&M cost of
the overall plant.
Assessment of the trend of the HTF cost and of the material storage cost in the market.
Abengoa is the only company whose commercial portfolio uses different storage concept. Based in this
experience, the aim of this study is to confirm the need of having different storage technologies available in the
market. The study compares both technologies explaining which are the main advantages, disadvantages, challenges
and particularities of each one. It is directed to analyse different aspects related with its plant configuration,
operational issues, performances and costs associated.

2.

Storage with steam accumulator

2.1. General overview

Direct Steam Generation (DSG) in parabolic trough power plants is a technological option with great
development potential, as it eliminates the need for intermediate heat transfer liquids while increasing
overall plant efficiency as well as reducing cost, increasing performance and becoming a more
environmentally friendly technology.
This is due, in part, to the fact that the water inside the receiver tubes absorbs the energy reflected, and
changes from liquid state into saturated steam and, subsequently, into superheated steam. The steam
produced in the solar field is fed directly to the turbine without the need for a heat exchanger. This
eliminates the oil/water heat exchanger train and incorporates water/steam separators. In addition, the
limitations on the maximum solar field temperature imposed by the degradation of the thermal oil (400
C) disappear and, therefore, the technology allows access to more efficient high temperature power
cycles.
DSG technology also improves solar field performance since the average operating temperature of the
solar field is slightly lower than that of the oil system, which reduces the thermal losses. Moreover, the
jump of temperature required in the oil/water exchanger is eliminated in the DSG plant. Furthermore,
investment costs are reduced due to the elimination of the exchange system, expansion tank, the thermal
oil itself and other systems related to the utilization of oil.
Thus, DSG technology allows an increase in production while reducing capital investment compared to
plants that utilize thermal oil as the heat transfer fluid. The savings in the cost of energy could be in the
region of ten percent.
A major design consideration for parabolic trough power plants is thermal energy storage (TES) options.
Currently the best option for DSG is the use of a steam accumulator storage system. A steam accumulator
is a direct storage system eliminating intermediate equipment. It is based on the Ruth accumulator system
where the steam is directly stored at high pressure in accumulator tanks.
The accumulator system is charged or filled with the saturated steam produced at maximum pressure from
the evaporator solar field and this steam transfers the heat from the solar field to the fluid contained inside
the tank. During the discharge process, the pressure inside the tank drop generating a flash evaporation of
steam and this steam is sent to the turbine. Two accumulators are discharged, one of them will produce

saturated steam which will be superheated in a heat exchanger using the higher pressure saturated steam
leaving the second accumulator. This system allows for a very quick response of the storage medium, and
will have a good performance during transients.
The excess steam produced in the saturated solar field will be used to charge the steam accumulator, this
allows higher mass flow of dry steam to enter the superheater during the charging phase as well as a
higher temperature at the turbine inlet.
During the start-up process in a DSG plant the evaporator solar field has to be preheated and producing a
minimum steam mass flow rate in order to focus the superheater solar field progressively, while
maintaining a part of the solar field defocused. However, this negative impact could be reduced by using
saturated steam from the storage system. Furthermore, the energy remaining in the steam accumulators
after the end of the discharge process could be used to preheat lines and auxiliary systems reducing the
start up time for the next operating day.
2.2. Advantages and disadvantages
The accumulator system will be composed of several numbers of tanks depending on the desired storage
capacity and a heat exchanger system where the saturated steam will be superheated during the
discharging process. This concept is very simple from the operational point of view.
One of the main advantages of this storage system is that the storage fluid is water, which eliminates the
negative environmental impacts and reduces the uncertainty in the price of the storage fluid. This
advantage is seen when comparing with motel comparing with molten salts that are based on nitrates and
the oil systems. Also since the storage fluid is water, all of the equipment used in these storage systems is
entirely conventional and therefore the performance of the equipment compared with a molten salt system
is higher.
Another of the advantages is the O&M cost what is significantly reduced by (1) the energy consumption
of pumping the storage fluid, which is almost negligible compared to a molten salt pump, and (2) the heat
tracing system, which is not required neither in pipes nor tanks.
The disadvantages of the accumulator design include the cost and complexity of manufacturing the tanks
and the relationship of the volume to the energy stored. The conventional steam accumulators need a huge
amount of stainless steel in the manufacturing of the tanks due to a high dependence on the steel cost.
This making the scaling up complicated due to the high pressure discharge required. Also the large
specific volume in relation to amount of the thermal energy stored is another disadvantage to consider.
On the other hand, the possibility of using new materials in the storage tanks gives these systems a great
potential to reduce current costs based on carbon steel tanks.
The next picture shows a basic diagram of the plant with storage system.

Fig. 1. Scheme of a DSG plant with steam accumulator

Another main disadvantage of the steam accumulators in comparison to molten salt double tank systems
is, that the discharge will occur at a lower pressure than the nominal pressure of the cycle. For this reason,
the energy stored by the steam accumulators has to be greater than the two-tank system for the same
amount of energy production.
In any case, the net benefit of the DSG technology and the simplicity of the steam storage system justify
and advise this solar technology, in comparison with the two-tank system for oil, make this an interesting
solar technology to pursue further.
3.

Storage with Molten salts

3.1. General overview

Molten salts are the most widespread fluid for thermal storage in CSP commercial applications due to
good thermal properties and reasonable cost. Nowadays, molten salts provides a thermal storage solution
for most of the technologies available on the market due to this fluid could be used as direct and indirect
storage depending of the selected plant philosophy.
Both, trough and tower technologies, use the double tank system as thermal storage configuration. Molten
salts are used as indirect storage in parabolic trough facilities which works with oil as heat transfer fluids
and as direct storage for tower concepts which molten salts are also used as circulating heat transfer fluid.
Other concepts under development like the parabolic trough with molten salts as heat transfer fluid which
it could be comparable to the tower with molten salts regarding thermal storage point of view.
In general, molten salts storage system offers the possibility to provide electrical production at constant
conditions thanks to maintain the storage material in different tanks when it is charged or discharged. In
addition it becomes an interesting solution due it has very high energy density per specific volume and
very high thermal inertia due to its characteristic thermal properties of high heat capacity and low thermal
conductivity. Due those thermal properties the system can be designed with minimum thermal losses
which represent higher global effectiveness.
The double tank of molten salts requires less specific volume for the same energy stored thanks to the
higher thermal capacity of the salts, specifically when it is used as direct storage medium where inventory
is minimized due to temperature gradients between hot and cold focus are bigger. On the other hand,
double tank storage concept involves intermediate equipments in the system configuration as heat
exchangers. In this way, two extra heat exchangers (thermal oil to salts and thermal oil to steam) in the
case of indirect storage and one (salts-steam) in the case of direct storage are needed for these
configurations.
The most common fluids for double tank storage system are sodium and potassium nitrates mixtures with
a weight composition which optimizes cost and thermal properties. These mixtures, which have prices
significantly stable in the market, are well known from decades ago with wide bibliographic information
and proven feasibility at pilot scale. Regarding materials compatibility, corrosion phenomena should be
taken into account due to impurity content of these mixtures but it can be assured the good performance
with the most common materials used in the industry.
Molten salts as storage material has inherent risks due to high freezing point of these fluids. Electric heat
tracing systems and tank heaters are installed to minimize freezing risks. These equipments involve
high parasitic consumption in terms of maintain the salts hot enough to avoid freezing or plugs even when
the system is completely discharged.
Going in deep to each different technology there are some different particularities depending on which it
is related. Below there are described the most significant ones for each solution.
3.2 Parabolic trough with thermal oil as heat transfer fluid technology

The double tank of molten salts represents an optimum system for this technology due it matches
perfectly the thermal sensible behaviour of the thermal oil used currently (see figures). Thermal oil
operation temperatures are from 300C to 400C approximately and molten salts are efficient and
operable enough at those temperatures.

410
HTF
390

Molten Salts

370

Temperature (C)

350
330
310
290
270
250
Length of the HX

Fig. 2. Profile of temperature in the heat exchanger

The power cycle to be used with this system could be with preheater, evaporator, superheater and
reheater. Depending on the cycle design common efficiencies reached with this technology are around
37%.
Thanks to the utilisation of efficient heat exchangers the hysteresis between charge and discharge it could
be reduce to a few degrees (around 10C) thus the system is able to generate higher than 90% of the
nominal conditions and as commented before it is also able to maintain constant conditions during the
whole discharge.
From different experiences with this system several assumptions can be confirmed: the system is able to
supply energy at constant conditions; there are no big concerns about corrosion, always related with
chlorides content on salts; degradation of the salts or other components related with the total impurities in
the salts; it is a system with high thermal inertia with the benefit it could represent regarding thermal
losses and there are no major toxicity problems than the NOx control within the tanks (strongly related
with the magnesium content on the salts).
On another hand it has been proven that the system needs of significant time to change from charge to
discharge conditions and in relation to that and with the heat exchanger design, it is difficult to produce or
to design a system to produce at partial loads in order to produce jointly with the regular solar field
production.

Finally it has to be mentioned that it is one of the most cost-effective systems for the different
technologies apart of the well-known particularities it has thus it becomes the desired thermal storage
system for parabolic trough technology with thermal oil as heat transfer fluid.

3.3 Tower technology with molten salts as heat transfer fluid

For tower technology with molten salts as heat transfer fluid it seems compulsory to install a double tank
of molten salts thermal storage system. Due the temperature gradient on the salts side it is higher than
with thermal oil (300C - 565C instead of 300C - 400C) the system is able to store more thermal energy
with the same volume making the system more cost-effective. It has to be mentioned that the upper limit
of 565C it is given by the molten salts which are not stable at higher temperatures.
One distinguishing characteristic of the system is that it is able to work with Rankine cycles with
preheater, evaporator, superheater but not with reheater with pressures lower that 65bar thus the salts
would freeze in the process of reheating. All those equipment are common from both the power block
steam cycle and the thermal storage system. The common efficiencies that can be achieved with those
cycles are around 40%.
Another particularity of that system is that it works only in one direction: always exchanging heat with
the water-steam in discharge mode due the charge it is done directly in the receiver, because of that this
system does not suffer any hysteresis between the charge and the discharge thus the discharge is able to
generate power at nominal conditions.
As commented before the system usually needs of a preheater, an evaporator and a superheater, and
depending on the unit the thermal behavior between the steam and the salts is different (see figures). For
the preheater and the superheater the thermal exchange occurs in good conditions in a counter flow heat
exchanger but for the evaporator the heat exchange occurs from a lowering temperature heat source to a
constant temperature sink (the evaporating steam) with the loss of affective heat transfer area that this
represents.

Fig. 3. Profile of temperature in a steam generator with molten salt

The possible problems associated are mainly the same as with the parabolic trough technology with
thermal oil but taking into account bigger temperatures which it is more restrictive from the point of view
of corrosion and degradation temperatures.
Regarding the time response of the system and the possibility to work in parallel with receiver production
it has to be mentioned that depending on the design of the preheater, evaporator and superheater it will be
more or less feasible to work at part loads but when those equipment allow to do it the system is able to
supply extra flow to the receiver in order to maintain the nominal conditions in a relatively short time.
Finally about the economics it is true that the system is the one with lowest cost even it is needed one
stainless steel tank. This is because it is avoided one heat exchanger and because the smaller size of the
system needed for the same capacities due to the bigger temperature gradient.

4.

Other promising options for the near future1

4.1. General overview

At first of this paper, direct steam generation technology was presented as one of the most promising
options for renewable power plants improving the convention plants based on oil system. This is due to
the fact that with water as the heat transfer fluid first of all the thermal oil it is avoided with the cost and
toxicity that it represents and it is possible to reach higher temperatures and pressures at the BOP with the
increment on the efficiency associated.
In order to achieve those high temperatures, pressures and efficiencies from storage system at the BOP it
is used to design those plants with a superheated steam Rankine cycle having two differentiated zones or
modules depending on the conditions of the steam: the saturated module and the superheated module.
The thermal storage system currently commercially available is the steam accumulator described before
and only for the production of saturated steam but there are other options under study and development
which are focused the production of the steam in the two different modules: the saturated section and the
superheated section.
As commented before one of the big problems of the direct steam generation storage is the electricity
generation at constant and nominal conditions, although it can be achieved with the solutions presented
above it represents a big volume utilization due to the low energy density of the systems.
With the objective of improving this performance, one interesting technology to solve it is the one based
on phase change materials. A system with one phase change material would be able to produce saturated
steam at constant conditions while the storage material is crystallising.
From more than 10 years ago few pilots have been constructed, operated and tested attempting to solve
most of the problems associated to this technology as the low thermal conductivity of the phase change
material, the volume expansion management, the thermal degradation, corrosion issues and so on but
most of them have resulted in material failures or excessive costs.
Related to the problem of producing steam at constant and nominal conditions it has to be considered that
it is also needed the superheated module in order to produce electricity at the same nominal conditions
during the charge.
4.2. Next generation of storage technology
In order to be able to produce electricity at nominal conditions during the whole discharge of the direct
steam storage systems some alternatives have been considered:

Phase change material system for the saturated block and solid sensible heat accumulator to
superheat the steam
Phase change material system for the saturated block and phase change material cascade to
superheat the steam
Phase change material system for the saturated block and double tank of molten salts to
superheat steam
Steam accumulators for the saturated block and solid sensible heat accumulator to superheat the
steam
Steam accumulators for the saturated block and double tank of molten salts to superheat steam

5. Conclusion
Thermal energy storage is the tool that the concentrated solar plant has to be dispatched. The thermal
energy storage technology has to be defined depending on some factors as the heat transfer fluid, the
condition of the power block and other parameter described in the paper.
It is very important to have knowledge of the different technologies existing in order to select the
optimum storage in each plant. Steam accumulator and molten salt storage are the two commercial option
that the market offer and each one has a market share,
New developments are been carried out in order to optimize the existing design but a good portfolio of
storage technology allow to build the optimum plant per project.

References
[1] IEA ECES Annex 19, final report 2010