Advanced Review

Solar thermal CSP technology

Manuel Romero ∗ and José González-Aguilar

Solar thermal concentrating solar power (CSP) plants, because of their capacity
for large-scale generation of electricity and the possible integration of thermal
storage devices and hybridization with backup fossil fuels, are meant to supply a
significant part of the demand in countries of the solar belt. Nowadays, the market
penetration of solar thermal electricity is steeply increasing, with commercial
projects in Spain, USA, and other countries, being the most promising technology
to follow the pathway of wind and photovoltaics to reach the goals for renewable
energy implementation in 2020 and 2050. In the first commercial projects involving
parabolic-trough technology, some improvements are being introduced like the
use of large molten-salt heat storage systems able to provide high degrees of
dispatchability to the operation of the plant, like the plants Andasol in Guadix,
Spain, with 7.5 h of nominal storage, or the use of direct steam generation loops
to replace thermal oil at the solar field. In the near future, the research and
innovation being conducted within the field of linear Fresnel collectors may lead to
high temperature systems able to operate up to 500◦ C and produce cost-effective
superheated steam. Central receiver systems are opening the field to new thermal
fluids like molten salts (Gemasolar tower plant in Seville, Spain) with more than
14 h of nominal storage and air, and new solar receivers like volumetric absorbers,
allowing operation at temperatures above 1000◦ C. All these factors can lead to
electricity generation cost reduction of CSP plants by 30–40% for the period 2010–
2020, according to public roadmaps and cost analysis made by the International
Energy Agency in 2010.  C 2013 John Wiley & Sons, Ltd.

How to cite this article:

WIREs Energy Environ 2014, 3:42–59. doi: 10.1002/wene.79

INTRODUCTION other relevant programs like the `Solar Mission’ in

India recently approved and going for 22 GW-solar,
S olar thermal electricity or STE (also known as
CSP or concentrating solar power) is expected
to impact enormously on the world’s bulk power
with a large fraction of thermal.2 Only in Southern
Europe, the technical potential of STE is more than
1000 GW and in Northern Africa it is immense.3
supply by the middle of the century.1 Nowadays,
Worldwide, the exploitation of less than 1% of the
the high-temperature thermal conversion of concen-
total solar thermal power plant potential would be
trated solar energy is rapidly increasing, with many
enough to meet the recommendations of the United
commercial projects taken up in Spain, USA, and
Nations’ Intergovernmental Panel on Climate Change
other countries such as India, China, Israel, Australia,
for long-term climate stabilization.4
Algeria, and Italy. Spain with 2400 MW connected to
One megawatt installed concentrating solar
the grid in 2013 is taking the lead on current commer-
thermal power avoids annually 688 tons of CO2
cial developments, together with USA where a target
compared with a combined cycle conventional plant,
of 4500 MW for the same year has been fixed and
which uses natural gas and 1360 tons of CO2 com-
pared with a conventional coal/steam plant. A 1-m2
mirror in the primary solar field produces 400 kWh
The authors have declared no conflicts of interest in relation to this of electricity per year, avoids 12 tons of CO2 com-
article.
pared with a conventional coal/steam plant and con-
* Correspondence to: [email protected] tributes to a saving of 2.5 tons of fossil fuels during its
IMDEA Energı́a, Avda. Ramón de la Sagra 3, Móstoles, Spain 25-year operation lifetime. The energy payback time
DOI: 10.1002/wene.79 for the materials of CSP systems is less than 1 year,5

42  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

and most solar-field materials and structures can be

recycled and used again for further plants.
But in terms of electric grid and quality of bulk
power supply, it is the ability to provide dispatch on
demand that makes STE stand out from other renew-
able energy technologies like photovoltaics or wind.
Thermal energy storage systems store excess thermal
heat collected by the solar field. Storage systems, alone
or in combination with some fossil fuel backup, keep
the plant running under full-load conditions. This ca-
pability of storing high-temperature thermal energy
leads to economically competitive design options, as
only the solar part has to be oversized. This STE plant
feature is tremendously relevant, as penetration of so-
lar energy into the bulk electricity market is possible
only when substitution of intermediate-load power
plants of about 4000–5000 h/year is achieved. The
combination of energy on demand, grid stability, and
high share of local content that lead to creation of
local jobs provides a clear niche for STE within the
renewable portfolio of technologies. A clear indica-
tor of the globalization of such policies is that the F I G U R E 1 Flow diagram for a typical solar thermal power plant.
International Energy Agency (IEA) is sensitive to STE
within low-carbon future scenarios for the year 2050.
At the IEA’s Energy Technology Perspectives 2010,6 black body. An ideal solar receiver would thus have
STE is considered to play a significant role among negligible convection and conduction losses. In the
the necessary mix of energy technologies needed to case of a solar thermal power plant, the solar energy is
halving global energy-related CO2 emissions by 2050, transferred to a thermal fluid at an outlet temperature
and this scenario would require capacity additions high enough to feed a heat engine or a turbine that
of about 14 GW/year (55 new solar thermal power produces electricity. The solar field is usually designed
plants of 250 MW each). for a normal incident radiation of 800–900 W/m2 .
Annual normal incident radiation varies from 1600
to 2800 kWh/m2 , allowing from 2000 to 3500 an-
nual full-load operating hours with the solar element,
CONCENTRATING SOLAR SYSTEMS depending on the available radiation at the particular
Even though solar radiation is a source of high tem- site.10
perature and exergy at origin, with a high radiosity Solar reflective concentrators follow the basic
of 63 MW/m2 , sun-to-earth geometrical constraints principles of Snell’s Law of reflection.11 In a spec-
lead to a dramatic dilution of flux and to irradiance ular surface, like the mirrors used in solar thermal
available for terrestrial use, only slightly higher than power plants, the reflection angle equals the angle of
1 kW/m2 with a consequent supply of low tempera- incidence. The most practical and simplest primary
tures to the thermal fluid. It is therefore an essential geometrical concentrator typically used in STE sys-
requisite for solar thermal power plants and high- tems is the parabola. Even though there are other con-
temperature solar chemistry applications to make use centrating devices like lenses or compound parabolic
of optical concentration devices that enable the ther- concentrators,12 the reflective parabolic concentra-
mal conversion to be carried out at high solar fluxes tors and their analogues are the systems with the
and with relatively low heat losses.7, 8 A simplified greatest potential for scaling up at a reasonable cost.
model of a STE plant is depicted in Figure 1. Four CSP technologies are today represented
STE systems consist of a large reflective surface at pilot and commercial scale13, 14 : parabolic-trough
collecting the incoming solar radiation and concen- collectors (PTCs), linear Fresnel reflector (LFR)
trating it onto a solar receiver with a small aperture systems, power towers or central receiver systems
area.9 The solar receiver is a high-absorptance ra- (CRS), and dish/engine systems (DE). All the existing
diative/convective heat exchanger that emulates as plants mimic parabolic geometries with large mirror
closely as possible the performance of a radiative areas (Figure 2).

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 43

 Advanced Review wires.wiley.com/wene

F I G U R E 2 Schematic diagrams of the four STE systems currently scaled up to pilot and demonstration sizes.
PTC and LFR are 2D concentrating systems in systems, 23% for CRS, and 30% for DE. Annual
which the incoming solar radiation is concentrated capacity factors may be designed between values of
onto a focal line by one-axis tracking mirrors. They about 20% for systems without thermal energy stor-
are able to concentrate the solar radiation flux 30–80 age and more than 70% for systems making use of
times, heating the thermal fluid up to 450◦ C, with large storage units.13 With current investment costs,
power conversion unit sizes of 30–280 MW, and all STE technologies are generally thought to require
therefore, they are well suited for centralized power a public financial support strategy for market de-
generation at dispatchable markets with a Rankine ployment. At present direct capital costs of STE and
steam turbine/generator cycle. CRS optics is more power generation costs are estimated to be two to
complex, as the solar receiver is mounted on top of a three times those of fossil-fuel power plants, how-
tower and sunlight is concentrated by means of ever, industry roadmaps advance 60% cost reduction
a large paraboloid that is discretized into a field of before 2025.10 In fact, governments at some coun-
heliostats. This 3D concentrator is therefore off-axis tries like Spain are already accelerating the process of
and heliostats require two-axis tracking. Concentra- drastic tariff reduction with the goal of STE, PV, and
tion factors are between 200 and 1000 and unit sizes wind energy becoming tariff equivalent in less than
are between 10 and 200 MW, and they are therefore one decade.
well suited for dispatchable markets and integration Every square meter of STE field can produce
into advanced thermodynamic cycles. A wide variety up to 1200 kWh thermal energy per year or up to
of thermal fluids, like saturated steam, superheated 400 kWh of electricity per year. That means a cumu-
steam, molten salts, atmospheric air, or pressurized lative savings of up to 12 tons of carbon dioxide and
air, can be used, and temperatures vary between 300 2.5 tons of fossil fuel per square meter of CSP system
and 1000◦ C. Finally, DE systems are small modular over its 25-year lifetime.19
units with autonomous generation of electricity by
Stirling engines or Brayton mini-turbines located at
the focal point. Dishes are parabolic 3D concentra- PARABOLIC TROUGHS
tors with high concentration ratios (1000–3000) and
unit sizes of 5–25 kW. Their current market niche is The Parabolic-Trough Collector
in both distributed on-grid and remote/off-grid power PTCs are linear-focus concentrating solar devices
applications.15–18 suitable for working in the 150–400◦ C temperature
Typical nominal solar-to-electric conversion ef- range.20 The current research with new thermal flu-
ficiencies may move between 20% for PTC and LFR ids intends to increase the operating temperature up

44  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

F I G U R E 3 Typical parabolic-trough collector.

to 500◦ C.21 The concentrated radiation heats the fluid main problem with this oil is its high solidification
that circulates through the receiver tube, thus trans- temperature (12◦ C), which requires an auxiliary heat-
forming the solar radiation into thermal energy in the ing system when oil lines run the risk of cooling be-
form of the sensible heat of the fluid. Figure 3 shows low this temperature. Because the boiling tempera-
a typical PTC and its components. ture at 1013 mbar is 257◦ C, the oil circuit must be
Collector rotation around its axis requires a pressurized with nitrogen, argon, or some other in-
drive unit. One drive unit is usually sufficient for sev- ert gas when oil is heated above this temperature.
eral parabolic-trough modules connected in series and Although there are other suitable thermal oils for
driven together as a single collector. Drive units com- slightly higher working temperatures with lower so-
posed of an electric motor and a gearbox combination lidification temperatures, they are too expensive for
are used for small collectors (aperture area <100 m2 ), large solar plants.
whereas powerful hydraulic drive units are required The typical PTC receiver tube is composed of an
to rotate large collectors. A drive unit placed on the inner steel pipe surrounded by a glass tube to reduce
central pylon is commanded by a local control unit to convective heat losses from the hot steel pipe.9 The
track the sun. At present, all commercial PTC designs steel pipe has a selective high-absorption (>90%),
use a single-axis sun-tracking system.22 low-emission (<30% in the infrared) coating that re-
Thermal oils are commonly used as the work- duces radiation thermal losses. Receiver tubes with
ing fluid in these collectors for temperatures above glass vacuum tubes and glass pipes with an antire-
200◦ C because at these operating temperatures, nor- flective coating achieve higher PTC thermal efficiency
mal water would produce high pressures inside the and better annual performance, especially at higher
receiver tubes and piping. This high pressure would operating temperatures. Receiver tubes with no vac-
require stronger joints and piping, and thus raise uum are usually used for working temperatures be-
the price of the collectors and the entire solar field. low 250◦ C because thermal losses are not so critical
However, the use of demineralized water for high at these temperatures. Because of the manufacturing
temperatures/pressures is currently under investiga- constraints, the maximum length of a single receiver
tion and the feasibility of direct steam generation at pipe is less than 6 m, so that the complete receiver
100 bar/400◦ C in the receiver tubes of PTCs has al- tube of a PTC is composed of a number of single re-
ready been proven in an experimental stage.23 For ceiver pipes welded in series up to the total length of
temperatures below 200◦ C, either a mixture of wa- the PTC. The total length of a PTC is usually within
ter/ethylene glycol or pressurized liquid water can be 100–150 m.
used as the working fluids because the pressure re- Two PTC designs specially conceived for large
quired in the liquid phase is moderate. solar thermal power plants are the LS-3 (owned
The oil most widely used in PTCs for temper- by the Israeli company SOLEL Solar Systems) and
atures up to 395◦ C is VP-1, which is a eutectic mix- EuroTrough (owned by the EuroTrough Consor-
ture of 73.5% diphenyl oxide/26.5% biphenyl. The tium), both of which have a total length of 100 m and

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 45

 Advanced Review wires.wiley.com/wene

F I G U R E 4 Parabolic-trough plant introducing a molten-salt circuit with two storage tanks to increment capacity factor.

a width of 5.76 m, with back-silvered thick-glass mir- The maturity of PTC systems is confirmed by
rors and vacuum absorber pipes. American Solargenix the solar electricity generating systems (SEGS) plants.
design has an aluminum structure. However, other The plants SEGS II–IX, which use thermal oil as the
collector designs are recently becoming commercially working fluid (HTF technology), were designed and
available in the short-to-medium term like the ones implemented by the LUZ International Limited com-
developed by the companies Solargenix, Albiasa, or pany from 1985 to 1990. All the SEGS plants are lo-
Sener.9, 21 The main constraint when developing the cated in the Mojave Desert, Northwest of Los Angeles
mechanical design of a PTC is the maximum tor- (California). With their daily operation and over 2.2
sion at the collector ends because high torsion would million square meters of PTCs, SEGS plants are this
lead to a smaller intercept factor and lower optical technology’s best example of commercial maturity
efficiency. and reliability. Their plant availability is over 98%
and their solar-to-electric annual efficiency is in the
range of 14–18%, with a peak efficiency of 22%.20
Electricity Generation with PTCs Thanks to the continuous improvements in the SEGS
The suitable PTC temperature range and their good plants, the total SEGS I cost of $0.22/kWhe for elec-
solar-to-thermal efficiency up to 400◦ C make it pos- tricity produced was reduced to $0.16/kWhe in the
sible to integrate a parabolic-trough solar field in a SEGS II and down to $0.09/kWhe in SEGS IX24 by
Rankine water/steam power cycle to produce elec- the year 1996.
tricity. The simplified scheme of a typical solar ther- With the revival of commercial STE projects
mal power plant using parabolic troughs integrated since 2006 in US and Spain, a new generation of
in a Rankine cycle is shown in Figure 4. The technol- SEGS-type plants has come to the arena. This is
ogy commercially available at present for parabolic- the case of the Nevada Solar One (NSO) project of
trough power plants is the heat transfer fluid (HTF) 75 MWe in the US, the Ibersol project in Puertol-
technology, which uses oil as the heat carrier between lano, Spain, or the Shams One 100 MWe plant in
the solar field and the power block. Abu Dhabi.25 NSO started grid-connected operation
Although parabolic-trough power plants usu- in June 2007 and it is considered a milestone in the
ally have an auxiliary gas-fired heater to produce opening of the second market deployment of PTC
electricity when direct solar radiation is not avail- technology in the world after SEGS experience. Since
able, the amount of electricity produced with natural then, more than 40 PTC plants (about 50 MWe each)
gas is always limited to a reasonable level. This limit are being constructed and starting operation in Spain
changes from one country to another: 25% in Cali- between the period 2007–2013 and more than 2 GW
fornia (USA), 15% in Spain, and no limit in Algeria. on track in the US.2 Main characteristics of NSO plant
Typical solar-to-electric efficiencies of a large solar are shown in Table 1.
thermal power plant (>30 MWe ) with PTCs are be- In spite of their environmental benefits, there are
tween 15% and 22%, with an average value of about still some obstacles to the commercial use of this tech-
17%. The yearly average efficiency of the solar field nology. The main barriers at present are the high in-
is about 50%. vestment cost (2500–5000 $/kW, depending on plant

46  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

T A B L E 1 Main Characteristics of Nevada Solar One Plant

Solar Field Power Block

Solar collector assemblies 760 Turbine generator gross output 75 MWe

Aperture area (m) 5 Net output to utility 72 MWe
Aperture area (m2 ) 470 Solar steam inlet pressure 86.1 bar
Length (m) 100 Solar steam reheat pressure 19.5 bar
Concentration ratio 71 Solar steam inlet temperature 371◦ C
Optical efficiency 0.77
Number of mirror segments 182,400
Number of receiver tubes 18,240
Field aperture (m2 ) 357,200
Site area (km2 ) 1.62
Field inlet temperature (◦ C) 300
Field outlet temperature (◦ C) 390

size and thermal storage capacity) and the minimum Two pioneer projects introducing thermal stor-
size of the power block required for high thermody- age are the plants Andasol-I and Andasol-II.28 This
namic efficiency. However, these barriers are shared PTC plants installed in Guadix, Spain have a nom-
by all the solar thermal power technologies currently inal power of 50 MWe each and an oversized solar
available.26 field (510,120 m2 mirrors surface area) with an inte-
One of the strategies to mitigate risk perception grated 1010 MWth molten-salt thermal storage sys-
and increment the capacity factor of the power block tem to extend the plant’s full-load operation 7.5 h
is to integrate a parabolic-trough solar field in the beyond daylight hours leading to a capacity factor of
bottoming cycle of a combined-cycle gas-fired power 41%. Figure 4 shows Andasol plant flow diagram and
plant. This configuration is called the integrated solar Figure 5 shows an aerial view of the tanks. As it can
combined cycle system (ISCCS). Although the contri- be observed, the introduction of a third circuit with
bution of the solar system to the overall plant power molten salts adds more complexity with a number
output is small (approximately 10–15%) in the ISCCS of new heat exchangers and heat tracing to avoid the
configuration, it seems to be a good approach to mar- salts thawing. It is still early to know the final result in
ket penetration in some developing countries, which terms of energy management and operational robust-
is why the government of Algeria has promoted an ness. Early operational data demonstrate the capacity
ISCCS plant and the World Bank, through its Global of the plant to supply electricity several hours after
Environment Facility (GEF), funded ISCCS plants in sunset. Other countries such as Italy are planning fur-
Morocco and Egypt.27 ther steps with the use of molten salts both for thermal
storage and solar field, like the 5-MW demonstration
project Archimede.29
Thermal Energy Storage and New Thermal Even though molten salts are nowadays the pre-
Fluids ferred option for demonstration and first commercial
In many countries, the market penetration of STE projects with storage in the US and Spain, there are
systems is based on feed-in-tariffs or green certificates other options under development and assessment like
linked to significant restrictions or regulations regard- the use of concrete or other solid bed materials,30 and
ing the use of hybrid concepts like ISCCS schemes. the use of phase change media (PCM).31 PCM pro-
Because of that, the use of thermal energy storage vide a number of desirable features, for example, high
systems with an oversized solar field is pursued to op- volumetric storage capacities and heat availability at
timize economics and dispatchability of PTC plants. constant temperatures. Energy storage systems using
For temperatures of up to 300◦ C, thermal min- the latent heat released on melting eutectic salts or
eral oil can be stored at ambient pressure, and is the metals have often been proposed, but never carried
most economical and practical solution. Synthetic and out on a large scale due to difficult and expensive
silicone oils, available for up to 410◦ C, have to be internal heat exchange and cycling problems. Heat
pressurized and are expensive. Then, molten salts can exchange between the HTF and the storage medium
be used between 220 and 560◦ C at ambient pressure, is seriously affected when the storage medium solid-
but require parasitic energy to keep them liquid. ifies. Encapsulation of PCMs has been proposed to

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 47

Advanced Review wires.wiley.com/wene

F I G U R E 5 Aerial view of power island and detail of molten-salt storage tanks and heat exchangers of Andasol plant in Guadix, Spain. This
50 MW plant stores thermal energy excess in 28,500 tons of nitrate molten salts able to provide up to 7.5 equivalent hours of operation at nominal
capacity. Reproduced by permission of ACS/Cobra Energı́a (Spain).

improve this. Combining the advantages of direct- generation with horizontal PTCs at 100 bar/400◦ C.33
contact heat exchange and latent heat, hybrid salt- Two precommercial projects are under development
ceramic phase change storage media have recently in Spain to demonstrate the technical feasibility of di-
been proposed. The salt is retained within the sub- rect steam generation combined with a power block.21
micron pores of a solid ceramic matrix such as mag-
nesium oxide by surface tension and capillary force.
Heat storage is then accomplished in two modes: by
LINEAR FRESNEL REFLECTORS
the latent heat of the salt and also the sensible heat of
the salt and the ceramic matrix.32 LFRs are composed by an array of linear (or slightly
PCM are extremely suitable for direct steam bent) mirror strips that independently move and col-
generation PTC plants. If the superheated steam re- lectively focus on absorber lines suspended from el-
quired to feed the steam turbine in the power block evated towers. Reflective segments are close to the
were produced directly in the receiver tubes of the ground and can be assembled in a compact way up
PTCs, the oil would be no longer necessary, and tem- to 1 ha/MW. This technology aims at achieving the
perature limitation and environmental risks associ- performance of parabolic troughs with lower costs.
ated with the oil would be avoided. Simplification They are characterized by having a fix linear focus
and cost reduction of overall plant configuration is where the absorber is static.14 However, optical ef-
then evident as only one fluid is used. This effect ficiency is lower than that of parabolic troughs due
combined with the increment of efficiency after re- to a higher impact of the incidence angle and the co-
moving intermediate heat exchanger might lead to a sine factor. Consequently, operating temperature at
reduction of 15% of the cost of the electricity pro- the working fluid is usually lower, typically between
duced. The disadvantages of this concept originate 150 and 350◦ C. By this reason, linear Fresnel technol-
from the thermo-hydraulic problems associated with ogy has been historically applied to generate saturated
the two-phase flow existing in the evaporating section steam via direct in-tube steam generation, and use
of the solar field. An additional disadvantage is the into ISCCS or in regenerative Rankine cycles, though
use of much heavier absorber tubes and interconnect- current R&D is aiming at higher temperatures above
ing lines required for the high pressure steam. Never- 400◦ C.34
theless, experiments performed at in a 2-MWth loop LFRs typically make use of lower cost nonvac-
at the Plataforma Solar de Almerı́a (PSA) in Spain uum thermal absorbers where the stagnant air cavity
have proven the technical feasibility of direct steam provides significant thermal insulation, light reflector

48  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

F I G U R E 6 Scheme of compact LFR system with a multitower array and dynamic aiming strategy of mirror strips.

support structures close to the ground, low cost flat in a 1350 m2 segment not connected to the coal fired
float glass reflector, and low cost manual cleaning be- plant, and was used to trial initial performance, and
cause the reflectors are at human height.35 The LFRs it first produced steam at 290◦ C in July 2004. The ex-
also have much better ground utilization, typically us- pansion to 9 MWth was completed by 2008. Activities
ing 60–70% of the ground area compared with about of the company moved to the US and were contin-
33% for a trough system, and lower O&M costs ued by Ausra, Palo Alto. Ausra established a factory
because of more accessible reflectors. of components, tubular absorbers and mirrors in Las
Vegas and built the Kimberlina 5-MW demonstration
plant in Bakersfield at the end of 2008. In 2010, Ausra
Historical evolution of LFR systems was purchased by AREVA that is presently committed
After some pioneering experiences by Francia36 and to the commercial deployment of this technology.
Di Canio et al.,37 the first serious development un- The third technology player after SPG/MAN
dertaken on the compact LFR system was proposed and AREVA is the German company NOVATEC So-
at the University of Sydney in 1993. The concept is lar, formerly NOVATEC Biosol. The technology of
composed by a single field of reflectors together with NOVATEC is based on its collector Nova-1 aimed
multiple linear receivers (see Figure 6). Each reflector to produce saturated steam at 270◦ C. They have de-
is able to change their focal point from one receiver veloped a serial production factory for prefabricated
to another during the day to minimize shading and components, a 1.4 MW small commercial plant, PE-1,
shadow losses in the dense reflector field. This system in Puerto Errado, Murcia, Spain, which has been grid
covers about 71% of the ground compared with 33% connected since March 2009, and a second 30 MW
for parabolic-trough systems.38 In 2000, the company commercial plant, PE-2, (with a mirror surface of
Solarmundo built a 2400 m2 LFR prototype collector 302,000 m2 ) has been completed and is in operation
field with such a technology at Liege, Belgium, but test since August 2012. NOVATEC is promoting 50 MW
results were not reported. Later the company moved plants mixing PTC and LFR fields, where LFR pro-
to Germany and was renamed Solar Power Group vides preheating and evaporation and PTC field takes
(SPG). SPG signed an exclusivity cooperation agree- over superheating. The company claims that this hy-
ment with DSD Industrieanlagen GmbH (renamed to bridization results in 22% less land use and higher
MAN Ferrostaaal Power Industry in 2005). A 800 kW profitability. In March 2011, ABB acquired a 35%
linear Fresnel pilot operating at 450◦ C has already shareholding in Novatec Solar. In September 2011,
been tested in PSA, Spain.39, 40 Novatec Solar claimed that its technology has suc-
Back in Australia, in early 2002, a new com- cessfully generated superheated steam at temperatures
pany, Solar Heat and Power Pty Ltd. (SHP), made above 500◦ C at its 1.4 MW demonstration plant in
extensive changes to the engineering design of the re- Murcia, Spain, by implementing a receiver containing
flectors to lower cost and has become the first to com- vacuum absorber.41
mercialize LFR technology. SHP initiated in 2003 for In last few years, new companies have explored
Macquarie Generation, Australia’s largest electricity the application of Fresnel technologies for electricity
generator, a demonstration project of 103 MWth (ap- generation. SkyFuel Inc., Albuquerque, New Mexico
proximately 39 MWe ) plant with the aim of supply- is developing the Linear Power TowerTM , a Fresnel
ing preheat to the coal fired Liddell power station. based-on concept designed to use a high-temperature
Phase 1 of the project, completed in 2004, resulted molten-salt HTF and it incorporates thermal energy

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 49

Advanced Review wires.wiley.com/wene

storage. In Europe, the French company CNIM is de- cies. To achieve break even costs for electricity with
veloping its own technology for the boiler part of current LFR technology, the cost target for the Fres-
the plant focusing on direct superheated steam gen- nel solar power plants needs to be about 55% of the
eration and linear Fresnel principle. A first prototype specific costs of parabolic-trough systems.43
(800 m2 ) has been built, commissioned, and pretested
in 2010.42
CENTRAL RECEIVER SYSTEMS
In power towers or CRSs, incident sunrays are tracked
Future Technology Development by large mirrored collectors (heliostats), which con-
and Performance Trends centrate the energy flux onto radiative/convective heat
exchangers called solar receivers, where energy is
Even though some solid commercial programmes are transferred to a thermal fluid, mounted on top of
underway on LFR, still it is early to have consol- a tower (Figure 2). Plant sizes of 10–200 MW are
idated performance data with respect to electricity chosen because of economy of scale, even though
production. The final optimization would integrate advanced integration schemes are claiming the eco-
components development to increment temperature nomics of smaller units as well.44 The high solar
of operation and possible hybridization with other flux incident on the receiver (averaging between 300
STE systems like parabolic troughs. and 1000 kW/m2 ) enables operation at relatively high
There are many possible types of receiver, in- temperatures of up to 1000◦ C and integration of ther-
cluding evacuated tube and PV modules, but the most mal energy into more efficient cycles in a step-by-step
cost-effective system seems to be an inverted cavity approach. CRS can be easily integrated in fossil plants
receiver. In the case of SHP technology, the absorber for hybrid operation in a wide variety of options and
is a simple parallel array of steam pipes at the top has the potential for generating electricity with high
of a linear cavity, with no additional redirection of annual capacity factors through the use of thermal
the incoming light from the heliostats to minimize storage. With storage, CRS plants are able to operate
optical losses and the use of hot reflectors. In the over 4500 h/year at nominal power.45
case of SPG technology, the absorber is a single tube
surmounted by a hot nonimaging reflector made of
glass, which must be carefully manufactured to avoid Technology of Heliostats and Solar
thermal stress under heating, and exhibit some opti- Receivers
cal loss. Both systems can produce saturated steam The collector field consists of a large number of mir-
or pressurized water. At present AREVA and NO- rors, called heliostats, with two-axis tracking and a
VATEC are looking for new absorbers able to work local control system to continuously focus direct solar
at temperatures above 450◦ C. By 2015, according to radiation onto the receiver aperture area. Heliostats
developers, linear Fresnel can be expected to be op- fields are characterized by their off-axis optics. Be-
erating with superheated steam at 500◦ C yielding an cause the solar receiver is located in a fixed position,
efficiency improvement of up to 18.1% relative to the entire collector field must track the sun in such a
current saturated steam operation at 270◦ C.10 way that each and every heliostat individually places
For reflectors, automation is a key issue that its surface normal to the bisection of the angle sub-
has been demonstrated by NOVATEC. Additional tended by sun and the solar receiver.
effort should be given to the optimized demonstra- Heliostat field optical efficiency includes the co-
tion of multitower arrays to maximize ground cov- sine effect, shadowing, blocking, mirror reflectivity,
erage ratios. However, it is the lack of reliable in- atmospheric attenuation, and receiver spillage.46 Be-
formation regarding annual performance and daily cause of the large area of land required, complex op-
evolution of steam production that should be tar- timization algorithms are used to optimize the annual
geted as a first priority. Still some concerns remain energy produced by unit of land, and heliostats must
regarding the ability to control steam production be- be packed as close as possible so the receiver can be
cause of the pronounced effect of cosine factor in small and concentration high. Because the reflective
this kind of plants. This dynamic performance would surface of the heliostat is not normal to the incident
also affect the potential integration with other STE rays, its effective area is reduced by the cosine of the
systems such as PTC or CRS. Until now, most com- angle of incidence; the annual average cosine varies
parative assessments vis-à-vis parabolic troughs are from about 0.9 at two tower heights distance north of
not economically conclusive, revealing the need to the tower to about 0.7 at two tower heights south of
use much larger fields to compensate lower efficien- the tower. Of course, annual average cosine is highly

50  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

dependent on site latitude. Consequently, in sites close eSolar, Aora, or Cloncurry are introducing heliostat
to the Equator a surround field would be the best op- units of only a few m2 . The small heliostats have bet-
tion to make best use of the land and reduce the tower ter optical efficiency and the ones by eSolar can be
height. North fields improve performance as latitude flat mirrors compared with curved and canted facets
increases (South fields in the Southern hemisphere), in the large heliostats. This advantage and the easier
in which case, all the heliostats are arranged on the transportation to the site with minimal installation
North side of the tower. works can lead to a further decrease in the heliostats
The combination of all the above-mentioned costs. Brightsource with an ambitious program of
factors influencing the performance of the heliostat large projects is making use of single-facet 7.3-m2
field should be optimized to determine an efficient lay- heliostats54 and the company eSolar with a multi-
out. There are many optimization approaches to es- tower plant configuration presents a highly innova-
tablish the radial and azimuthal spacing of heliostats tive field with ganged heliostats of extremely small
and rows.47 One of the most classic, effective, and size (1.14 m2 each) that implies the large number
widespread procedures is the ‘radial staggered’ pat- of 12,180 units for a single 2.5 MW tower.55 If
tern, originally proposed by the University of Hous- such small heliostats may reach installed costs below
ton in the seventies.48 Integral optimization of the 200 $/m2 , it can only be understood under aggres-
heliostat field is decided by a tradeoff between cost sive mass production plans and preassembly during
and performance parameters. Heliostats, land and ca- manufacturing process by reducing on-site mounting
bling network must be correlated with costs. Cost and works. Annual performance and availability of those
performance also often have reverse trends, so that highly populated fields are still under testing.
when heliostats are packed closer together, blocking In a solar power tower plant, the receiver is the
and shadowing penalties increase, but related costs heat exchanger where the concentrated sunlight is in-
for land and wiring decrease. A classical code in use tercepted and transformed into thermal energy use-
since the eighties for optimization of central receiver ful in thermodynamic cycles. Radiant flux and tem-
subsystems is DELSOL3.49 perature are substantially higher than in parabolic
Mature low-cost heliostats consist of a reflect- troughs, and therefore, high technology is involved
ing surface, a support structure, a two-axis tracking in the design, and high-performance materials should
mechanism, pedestal, foundation, and control system. be chosen. The solar receiver should mimic a black
The development of heliostats shows a clear trend body by minimizing radiation losses. To do so cav-
from the early first generation prototypes, with a ities, black-painted tube panels or porous absorbers
heavy, rigid structure, second-surface mirrors, and able to trap incident photons are used. In most de-
reflecting surfaces of around 40 m2 ,50 to designs signs, the solar receiver is a single unit that central-
with large 100–120 m2 reflecting surfaces, lighter izes all the energy collected by the large mirror field,
structures, and lower-cost materials.51, 52 Heliostats and therefore high availabilities and durability are a
of 120 m2 were finally adopted for the first commer- must. Just as cost reduction is the priority for further
cial tower power plants PS10 and PS20 promoted development in the collector field, in solar receivers,
by the company Abengoa Solar.53 Since the first- the priorities are thermal efficiency and durability.
generation units, heliostats have demonstrated beam Typical receiver absorber operating temperatures are
qualities below 2.5 mrad (standard deviation of re- between 500 and 1200◦ C and incident flux covers a
flected rays including all heliostat errors but not in- wide range between 300 and over 1000 kW/m2 .9, 56
cluding intrinsic errors due to the solar disk) that There are different solar receiver classifications
are good enough for practical applications in so- criteria depending on the construction solution, the
lar towers, so the main focus of development is di- use of intermediate absorber materials, the kind of
rected at cost reduction. Estimated production costs thermal fluid used or heat transfer mechanisms. Ac-
of large area glass/metal heliostats for sustainable cording to the geometrical configuration, there are
market scenarios are around $130–200/m2 . Large basically two design options, external and cavity-type
area glass/metal units make use of glass mirrors sup- receivers. In a cavity receiver, the radiation reflected
ported by metallic frame facets. from the heliostats passes through an aperture into
Recently, some developers are introducing sub- a box-like structure before impinging on the heat
stantial changes in the conception of heliostat design. transfer surface. Cavities are constrained angularly
A number of projects based on the paradigm of max- and subsequently used in North-field (or South-field)
imum modularity and mass production of compo- layouts. External receivers can be designed with a
nents are claiming small-size heliostats as a compet- flat plate tubular panel or are cylindrically shaped.
itive low-cost option. Companies like Brightsource, Cylindrical external receivers are the typical solution

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 51

 Advanced Review wires.wiley.com/wene

Cavity convective heat exchangers absorb the concentrated

Cylindrical
solar radiation.58 The solar radiation is not absorbed
on an outer surface, but inside the structure ‘vol-
ume’. The heat transfer medium (mostly air) is forced
through the porous structure and is heated by convec-
tive heat transfer.59 Volumetric absorbers are usually
made of thin heat-resistant wires (in knitted or layered
grids) or either metal or ceramic (reticulated foams,
Billboard Volumetric etc.) open-cell matrix structures. Good volumetric ab-
sorbers are very porous, allowing the radiation to pen-
etrate deeply into the structure. Thin substructures
(wires, walls, or struts) ensure good convective heat
transfer. A good volumetric absorber produces the
F I G U R E 7 Different configurations of solar receivers. From left to so-called ‘volumetric effect’, which means that the ir-
right and top to bottom: (a) external tubular cylindrical, (b) cavity radiated side of the absorber is at a lower temperature
tubular, (c) billboard tubular, and (d) volumetric. than the medium leaving the absorber. Under specific
operating conditions, volumetric absorbers tend to
have an unstable mass flow distribution.60 Receiver
adopted for surround heliostat fields. Figure 7 shows arrangements with mass flow adaptation elements
examples of cylindrical external, billboard external (e.g., perforated plates) located behind the absorber
and cavity receivers. can reduce this tendency, as well as appropriate se-
Receivers can be directly or indirectly irradiated lection of the operating conditions and the absorber
depending on the absorber materials used to transfer material.61 A number of initiatives have formulated
the energy to the working fluid.57 Directly irradiated air-cooled volumetric schemes for STE plants, both
receivers make use of fluids or particle streams able for atmospheric pressure,62, 63 and for pressurized
to efficiently absorb the concentrated flux. Particle re- systems,56 though still commercial plants with these
ceiver designs make use of falling curtains or fluidized technologies are missing. Although air-cooled open
beds. In many applications, and to avoid leaks to the volumetric receivers are a promising way of produc-
atmosphere, direct receivers should have a transpar- ing superheated steam, the modest thermal efficiency
ent window. Windowed receivers are excellent solu- at the receiver (74% nominal and 61.4% annual av-
tions for chemical applications as well, but they are erage) must still be improved.64 At present, all the
strongly limited by the size of a single window, and benefits from using higher outlet temperatures are sac-
therefore clusters of receivers are necessary. rificed by radiation losses at the receiver, leading to
The key design element of indirectly heated low annual electricity production, so it is clear that
receivers is the radiative/convective heat exchange volumetric receiver improvements must reduce losses.
surface or mechanism. Basically, two heat transfer Selection of a particular receiver technology is
options are used, tubular panels and volumetric sur- a complex task, as operating temperature, heat stor-
faces. In tubular panels, the cooling thermal fluid age system, and thermodynamic cycle influence the
flows inside the tube and removes the heat collected design. In general, tubular technologies allow either
by the external black panel surface by convection. It is high temperatures (up to 1000◦ C) or high pressures
therefore operating as a recuperative heat exchanger. (up to 120 bar), but not both.65 Directly irradiated or
Depending on the HTF properties and incident so- volumetric receivers allow even higher temperatures
lar flux, the tube might undergo thermo-mechanical but limit pressures to below 15 bar. In addition, for
stress. As heat transfer is through the tube surface, high solar concentrations and high temperatures, the
it is difficult to operate at an incident flux above use of secondary concentrators, that allow reducing
600 kW/m2 (peak). Table 2 shows how only with thermal losses and improving the flux profile onto
high thermal conductivity liquids like sodium it is the aperture, is necessary in some cases. However,
possible to reach operating fluxes above 1 MW/m2 . secondary concentrators lead to a limited view cone.
Air-cooled receivers have difficulties working with This means that of all the possible heliostat positions
tubular receivers because of the lower heat transfer around the tower, only those within the ellipse, re-
coefficients. To improve the contact surface, a differ- sulting from the section boundary of the view cone
ent approach based on wire, foam or appropriately with the ground plane, are usable.66 Consequently,
shaped materials within a volume are used. In volu- the use of receivers with secondary concentrators on
metric receivers, highly porous structures operating as the aperture is impacting on heliostat layout.

52  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

T A B L E 2 Operating Temperature and Flux Ranges of Solar Tower Receivers

Fluid Water/Steam Liquid Sodium Molten Salt (nitrates) Volumetric Air
2
Flux (MW/m )
Average 0.1–0.3 0.4–0.5 0.4–0.5 0.5–0.6
Peak 0.4–0.6 1.4–2.5 0.7–0.8 0.8–1.0
Fluid outlet temperature (³C) 490–525 540 540–565 (700–1,000)

Experience in CRSs at 50% load.9 The system makes use of 624 heliostats
Although there have been a large number of STE of 121 m2 each, distributed in a North-field configu-
tower projects, only a few have culminated in the ration, a 100-m high tower, a 15 MWh heat storage
construction of an entire experimental system.9, 46, 67 system and a cavity receiver with four 4.8 × 12 m
The thermal fluids used in the receiver are saturated tubular panels. Although the system makes use of a
or superheated steam, nitrate-based molten salts and saturated steam turbine working at low temperature,
air. The most extensive precommercial experience has the nominal efficiency of the power block (30.7%) is
been collected by several European projects located in relatively good. This efficiency is the result of opti-
Spain at the premises of the PSA67 and the 10 MW mized management of waste heat in the thermody-
Solar One68 and Solar Two plants69 in the US. At namic cycle. The combination of optical, receiver,
present water/steam and molten salts are the HTFs and power block efficiencies leads to a total nomi-
being selected for the first generation of commercial nal efficiency at design point of 21.7%. Total annual
plants. efficiency decreases to 15.4%, including operational
losses and outages. PS10 is a milestone in the CRS de-
Water/Steam Plants: from PS10 and PS20 ployment process, as it is the first solar power tower
Projects to Superheated Steam plant developed for commercial exploitation. Com-
Production of superheated steam in the solar receiver mercial operation started on June 21, 2007. Since
has been demonstrated in several plants, such as So- then, the plant is performing as designed. The con-
lar One, Eurelios, and CESA-1, but operating expe- struction of PS20, a 20 MWe plant with the same
rience showed critical problems related to the con- technology as PS10, followed. PS20 started operation
trol of zones with dissimilar heat transfer coefficients in May 2009. With 1255 heliostats (120 m2 each)
like boilers and superheaters.67 Better results regard- spread over 90 ha and with a tower of 165 m high,
ing absorber panel lifetime and controllability have the plant is designed to produce 48.9 GWh/year.
been reported for saturated steam receivers. The good Saturated steam plants are considered a tempo-
performance of saturated steam receivers was quali- rary step to the more efficient superheated steam sys-
fied at the 2–MWth Weizmann receiver that produced tems. Considering the problems found in the eighties
steam at 15 bar for 500 h in 1989.70 Even though with superheated steam receivers, the current trend is
technical risks are reduced by saturated steam re- to develop dual receivers with independent absorbers,
ceivers, the outlet temperatures are significantly lower one of them for the preheating and evaporation and
than those of superheated steam, making applications another one for the superheating step. The experience
where heat storage is replaced by fossil fuel backup accumulated with heuristic algorithms in central con-
necessary. trol systems applied to aiming point strategies at he-
PS10, the first commercial CRS plant in the liostat fields allows achieving a flexible operation with
world, adopted the conservative scheme of produc- multiaperture receivers. The company Abengoa Solar,
ing saturated steam to limit risk perception and avoid developer of PS10 and PS20, is at present designing
technology uncertainties. The 11 MW plant, located a superheated steam receiver for a new generation of
near Seville in South Spain, was designed to achieve water/steam plants.71 But the most advanced strategy
an annual electricity production of 23 GWh at an is the program announced by the BrightSource In-
investment cost of less than 3500€/kW. The project dustries (Israel) Ltd. (BSII) that it has already built a
made use of available, well-proven technologies like demonstration plant of 6 MWth located at the Negev
the glass-metal heliostats and the saturated steam cav- desert in June 2008.54 The final objective of BSII is
ity receiver to produce steam at 40 bar and 250◦ C.53 to promote plants producing superheated steam at
The plant is a solar-only system with saturated steam 160 bar and temperature of 565◦ C (named DPT550).
heat storage able to supply 50 min of plant operation With those characteristics, they expect up to 40%

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 53

Advanced Review wires.wiley.com/wene

conversion efficiency at the power block for unit sizes Molten-Salt Systems: Solar Two
between 100 and 200 MW. The receiver is cylindrical, and Gemasolar
dual, and with a drum. The first commercial project For high annual capacity factors, solar-only power
under development is Ivanpah Solar in California plants must have an integrated cost-effective ther-
of 392 MW. The complex is comprised of three sep- mal storage system. One such thermal storage sys-
arate plants built in phases between 2010 and 2013. tem employs molten nitrate salt as the receiver HTF
Planned conversion efficiency from solar to electricity and thermal storage media. To be usable, the op-
is 20%. erating range of the molten nitrate salt, a mixture
The combination of recent initiatives on small of 60% sodium nitrate and 40% potassium nitrate,
heliostats, compact modular multitower fields, and must match the operating temperatures of modern
production of superheated steam may be clearly vi- Rankine cycle turbines. In a molten-salt power tower
sualized in the development program of the company plant, cold salt at 290◦ C is pumped from a tank at
eSolar. This company proposes a high degree of mod- ground level to the receiver mounted atop a tower
ularity with power units of 46 MW covering 64 ha, where it is heated by concentrated sunlight to 565◦ C
consisting of 16 towers and their corresponding he- (Figure 8). The salt flows back to ground level into an-
liostat fields sharing a single central power block. other tank. To generate electricity, hot salt is pumped
With replication, modularity sizes up to 500 MW and from the hot tank through a steam generator to make
upward may be obtained.72 Within the development superheated steam. The superheated steam powers a
program of STE in the US, eSolar has four plants to- Rankine-cycle turbine. The collector field can be sized
talizing 334 MW.2 Two modules of 2.5 MW each to collect more power than is demanded by the steam
were installed in 2009 by eSolar in Lancaster, Cali- generator system, and the excess salt is accumulated
fornia, and are already in operation. Each receiver has in the hot storage tank. With this type of storage
two independent cavities and the heliostat layout con- system, solar power tower plants can be built with
sists of identical arrays to maximize replication and annual capacity factors up to 70%. As molten salt
modularity. Each tower is associated with 12,180 flat has a high energy storage capacity per volume (up
heliostats of 1.14 m2 each.55 The receivers are dual- to 500–700 kWh/m3 ), they are excellent candidates
cavity, natural-circulation boilers. Inside the cavity, for solar thermal power plants with large capacity
the feedwater is preheated with economizer panels factors. Even though nitrate salt has a lower specific
before entering the steam drum. A downcomer sup- heat capacity per volume than carbonates, they still
plies water to evaporator panels where it is boiled. store 250 kWh/m3 . The average heat conductivity of
The saturated water/vapor mixture returns to the nitrates is 0.52 W/mK and their heat capacity is about
drum where the steam is separated, enters superheater 1.6 kJ/kg K. Nitrates are a cheap solution for large
panels, and reaches 440◦ C at 6.0 MPa. Each receiver storage systems.
absorbs a full-load power of 8.8 MWth . Overall plant The largest demonstration of a molten-salt
efficiency expected by eSolar would be 23% solar to power tower was the Solar Two project, a 10 MW
electricity.72 power tower located near Barstow, California. The

F I G U R E 8 Schematic of a molten-salt central receiver system with cylindrical tubular receiver (EPGS: electric power generating system).

54  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

purpose of the Solar Two project was to validate the T A B L E 3 | Technical Specifications and Design Performance of
technical characteristics of the molten-salt receiver, the Gemasolar Project
thermal storage, and steam generator technologies, Technical Specifications
improve the accuracy of economic projections for
commercial projects by increasing the capital, oper- Heliostat field reflectant surface 304,750 m2
ating, and maintenance cost database, and distribute Number heliostats 2650
information to utilities and the solar industry to fos- Land area of solar field 142 ha
ter a wider interest in the first commercial plants. A Receiver thermal power 120 MWth
110 MWht two-tank molten-salt thermal storage sys- Tower height 145 m
Heat storage capacity 15 h
tem was installed, a 42 MWth receiver, a 35 MWth
Power at turbine 17 MWe
steam generator system (535◦ C, 100 bar).73 Power NG burner 16 MWth
The plant began operating in June 1996. The
project successfully demonstrated the potential of Operation
nitrate-salt technology. Some of the key results were: Annual electricity production 112 MWhe
receiver efficiency was measured at 88%, the ther- Production from fossil (annual) 15%
mal storage system had a measured round-trip effi- Capacity factor 74%
ciency of over 97%, and the gross Rankine-turbine
cycle efficiency was 34%, all of which matched per-
formance projections. On one occasion, the plant (115 m2 each heliostat) is oversized to supply 15-h
operated around-the-clock for 154 h straight.74 On equivalent heat storage capacity. The plant is designed
April 8, 1999, testing and evaluation of this demon- to operate around-the-clock in summertime, leading
stration project was completed and subsequently was to an annual capacity factor of 74%. Fossil backup
shut down. One attempt to prove scaled-up molten- corresponding to 15% of annual production will be
salt technology is the Gemasolar project being pro- added. The levelized energy costs are estimated to be
moted by Torresol Energy, a joint venture between the approximately $0.16/kWh. Gemasolar represents a
Spanish SENER and MASDAR initiative from Abbu breakthrough for solar technology in terms of time-
Dhabi.75 Table 3 summarizes the main technical spec- dispatch management.76
ifications of Gemasolar project and Figure 9 shows an
aerial view of the plant.
With only 17 MWe , the plant that connected
CONCLUSION AND OUTLOOK
to the grid in summer 2011 is designed to produce From the seventies to the nineties, the development
112 GWhe /year. A large heliostat field of 304,750 m2 of STE technologies remained restricted to a few

F I G U R E 9 Aerial view of Gemasolar plant, property of Torresol EnergyC Torresol Energy, located in South Spain. At present, it is the largest
commercial solar central receiver system with a circular-shape heliostat field. Reproduced by permission of Torresol Energy.

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 55

Advanced Review wires.wiley.com/wene

countries and only a few, though important, re- ond generation technologies, like direct steam gener-
search institutions and industries were involved. The ation or molten-salt systems, even though in some
situation has dramatically changed since 2006 with cases still some innovations are under assessment
the approval of specific feed-in-tariffs or power pur- in early commercialization plants or demonstration
chase agreements in Spain and the US. Both coun- projects.
tries with more than 6 GW of projects under devel- Parabolic trough is the technology widely used
opment and more than 2 GW in operation at the nowadays in commercial projects, though other tech-
end of 2012 are undoubtedly leading the commer- nologies like linear Fresnel collectors and CRSs are
cialization of STE.77 Other countries such as India, developing the first grid-connected projects and re-
China, Australia, and Italy adopted the STE process. veal promising impacts on cost reduction.13 The
Subsequently, a number and variety of engineering reduction in electricity production costs should be
and construction companies, consultants, technolo- a consequence, not only of mass production but
gists, and developers committed to STE are rapidly also of scaling-up and R&D. A technology roadmap
growing. A clear indicator of the globalization of promoted by the European Industry Association
STE commercial deployment for the future energy ESTELA10 states that by 2015, when most of the
scenario has been elaborated by the IEA. This con- improvements currently under development are ex-
siders STE to play a significant role among the nec- pected to be implemented in new plants, energy pro-
essary mix of energy technologies for halving global duction boosts greater than 10% and cost decreases
energy-related CO2 emissions by 2050.6 This scenario up to 20% are expected to be achieved. Furthermore,
would require capacity addition of about 14 GW/year economies of scale resulting from plant size increase
(55 new solar thermal power plants of 250 MW each). will also contribute to reduce plants’ CAPEX per MW
However, this new opportunity is introducing an im- installed up to 30%. STE deployment in locations
portant stress to the developers of STE. In a period of with very high solar radiation further contributes to
less than 5 years, in different parts of the world, these the achievement of cost competitiveness of this tech-
developers of STE are forced to move from strategies nology by reducing costs of electricity up to 25%.
oriented to early commercialization markets based on All these factors can lead to electricity generation
special tariffs, to strategies oriented to a massive pro- cost savings up to 30% by 2015 and up to 50% by
duction of components and the development of large 2025, reaching competitive levels with conventional
amounts of projects with less profitable tariffs. This sources (e.g., coal/gas with stabilized electricity costs
situation is speeding up the implementation of sec- <10€c/kWh).

ACKNOWLEDGMENTS
This review has been carried out within the framework of the SOLGEMAC Project (S2009
ENE-1617), partially funded by ‘Comunidad de Madrid’ and ‘European Social Fund’.

REFERENCES
1. Lovegrove K, Stein W. Concentrating Solar Power 4. Philibert C. International Energy Technology Col-
Technology. Principles, Developments and Applica- laboration and Climate Change Mitigation, Case
tions. No. 21. Woodhead Publishing Series in Energy. Study 1: Concentrating Solar Power Technolo-
Cambridge, UK: Woodhead Publishing Limited; 2012, gies. Paris, France: OECD/IEA Information Paper.
ISBN: 9781845697693. COM/ENV/EPOC/IEA/ SLT; 2004, 8.
2. Herring G. Concentrating Solar Thermal Power Gains 5. Burkhardt JJ, Heath GA, Turchi CS. Life cycle assess-
Steam in Spain, as Momentum Builds for Major ment of a parabolic trough concentrating solar power
Projects in the US, North Africa, the Middle East, Asia plant and the impacts of key design alternatives. Envi-
and Australia. Photon International; 2009, 46–52. ron Sci Technol 2011, 45:2457–2464.
3. Nitsch J, Krewitt W, Langniss O. Renewable energies 6. IEA. Energy Technology Perspectives 2010—Scenarios
in Europe. Encyclopedia of Energy. San Diego: Else- and Strategies to 2050. Paris, France: International En-
vier; 2004, 5, 313–331. ergy Agency; 2010. ISBN 978-92-64-08597-8.

56  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

7. Winter CJ, Sizmann RL, Vant-Hull LL, eds. So- 21. Fernandez-Garcia A, Zarza E, Valenzuela L, Perez
lar Power Plants. Berlin, Germany: Springer-Verlag; M. Parabolic-trough solar collectors and their appli-
1991, ISBN 3-540-18897-5. cations. Renew Sustain Energy Rev 2010, 14:1695–
8. Kodama T. High-temperature solar chemistry for con- 1721.
verting solar heat to chemical fuels. Prog Energy Com- 22. Rabl A. Active Solar Collectors and Their Applica-
bust Sci 2003, 29:567–597. tions. New York: Oxford University Press; 1985, 59–
9. Romero M, Zarza E. Concentrating solar thermal 66. ISBN: 0-19-503546-1.
power. In: Kreith F, Goswami Y, eds. Handboook of 23. Eck M, Zarza E, Eickhoff M, Rheinländer J, Valen-
Energy Efficiency and Renewable Energy. Chapter 21. zuela L. Applied research concerning the direct steam
Boca Raton, Florida: CRC Press, Taylor & Francis generation in parabolic troughs. Solar Energy 2003,
Group; 2007, 1–98. 74:341–351.
10. Kearney AT. Solar Thermal Electricity 2025. STE In- 24. Kearney DW, Cohen GE. Current experiences with the
dustry Roadmap for the European Solar Thermal Elec- SEGS parabolic trough plants. In: Becker M, Böhmer
tricity Association (ESTELA). Brussels, Belgium: Eu- M, eds. Proceedings of the Eighth International Sym-
ropean Solar Thermal Electricity Association; 2010. posium on Solar Thermal Concentrating Technologies.
Available at: www.estelasolar.eu. (Accessed July 15, Vol. 1. Cologne, Germany, 1996. Heidelberg, Ger-
2013). many: C.F. Müller; 1997, 217–224.
11. Rabl A. Active Solar Collectors and Their Applica- 25. Goebel O. Shams one 100 MW CSP plant in Abu
tions. New York: Oxford University Press; 1985, 59– Dhabi. Proceedings SolarPACES 2009 (CD). Ref.
66. ISBN: 0-19-503546-1. manuscript: 15523. Berlı́n, Germany; September 15–
12. Welford WT, Winston R. High Collection Non- 18, 2009. Stuttgart, Germany: DLR. ISBN 978-3-00-
Imaging Optics. New York: Academic Press; 1989. 028755-8.
13. Romero M, González Aguilar J. Solar thermal power 26. Romero M. Solar thermal power plants. In: IUPAP
plants: from endangered species to bulk power pro- Working Group on Energy, ed. Report on Research
duction in sun-belt regions. Chapter 3. In: Rao KR, ed. and Development of Energy Technologies, October 6;
Energy & Power Generation Handbook. New York: 2004, 96–108.
ASME Three Park Avenue; 2011, 10016–5990. 27. Horn M, Führing H, Rheinländer J. Economic anal-
14. Mills D. Advances in solar thermal electricity technol- ysis of integrated solar combined cycle power plants.
ogy. Solar Energy 2004, 76:19–31. A sample case: the economic feasibility of an ISCCS
15. Mancini T, Heller P, Butler B, Osborn B, Schiel W, power plant in Egypt. Energy 2004, 29:935–1011.
Goldberg V, Buck R, Diver R, Andraka C, Moreno J. 28. Relloso S, Delgado E. Experience with molten salt ther-
Dish-stirling systems: an overview of development and mal storage in a commercial parabolic trough plant.
status. Int J Solar Energy Eng 2003, 125:135–151. Andasol-1 commisioning and operation. Proceedings
16. Lopez C, Stone K. Design and performance of the SolarPACES 2009 (CD). Ref. manuscript: 11396.
Southern California Edison Stirling Dish. Solar Engi- Berlı́n, Germany; September 15–18, 2009. Stuttgart,
neering, Proceedings of ASME International Solar En- Germany: DLR. ISBN 978-3-00-028755-8.
ergy Conference. Maui, Hawaii; 1992, 945–952. New 29. Falchetta M, Mazzei D, Crescenzi T, Merlo L. Design
York: ASME. ISBN 0-7918-762-2. of the Archimede 5 MW molten salt parabolic trough
17. Keck T, Heller P, Weinrebe G. Envirodish and solar plant. Proceedings SolarPACES 2009 (CD). Ref.
Eurodish—system and status. Proceedings of the ISES manuscript: 11608. Berlı́n, Germany; September 15–
Solar World Congress. Göteborg, Sweden; 2003. 18, 2009. ISBN 978-3-00-028755-8.
Freiburg, Germany: International Solar Energy Soci- 30. Laing D, Steinmann W-D, Tamme R., Richter C. Solid
ety. ISBN: 91-631-4740-8. media thermal storage for parabolic trough power
18. Stine W, Diver RB. A Compendium of Solar plants. Solar Energy 2006, 80:1283–1289.
Dish/Stirling Technology. Report SAND93–7026. Al- 31. Laing D, Bahl C, Bauer T, Lehmann D, Steinmann
buquerque, New Mexico: Sandia National Laborato- WD. Thermal energy storage for direct steam gen-
ries; 1994. eration. Proceedings SolarPACES 2009 (CD). Ref.
19. Geyer M. Panel 1 briefing material on status of ma- manuscript: 12055. Berlı́n, Germany; September 15–
jor project opportunities. The current situation, issues, 18, 2009. Stuttgart, Germany: DLR. ISBN 978-3-00-
barriers and planned solutions. International Execu- 028755-8.
tive Conference on Expanding the Market for Con- 32. Liu M, Saman W, Bruno F. Review on storage ma-
centrating Solar Power (CSP) – Moving Opportunities terials and thermal performance enhancement tech-
into Projects; Berlin, Germany. June 19–20; 2002. niques for high temperature phase change thermal
20. Price H, Luepfert E, Kearney D, Zarza E, Cohen G, storage systems. Renew Sustain Energy Rev 2012, 16:
Gee R, Mahoney R. Advances in parabolic trough so- 2118–2132.
lar power technology. Int J Solar Energy Eng 2002, 33. Zarza E, Valenzuela L, Leon J, Hennecke K, Eck M,
124:109–125. Weyers HD, Eickhoff M. Direct steam generation in

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 57

Advanced Review wires.wiley.com/wene

parabolic troughs: final results and conclusions of the National Laboratories; 1986. Available at: http://prod.
DISS project. Energy 2004, 29:635–644. sandia.gov/techlib/access-control.cgi/1986/868009.
34. Platzer WJ, Linear Frensel collector as an emerging op- pdf. (Accessed July 15, 2013).
tion for concentrating solar thermal power. Proceed- 47. Sánchez M, Romero M. Methodology for generation
ings of the ISES Solar World Congress; 2009. Freiburg, of heliostat field layout in central receiver systems
Germany: International Solar Energy Society. based on yearly normalized energy surfaces. Solar En-
35. Kalogirou SA. Solar thermal collectors and applica- ergy 2006, 80:861–874.
tions. Prog Energy Combust Sci 2004, 30:231–295. 48. Lipps FW, Vant-Hull LL. A cellwise method for the
36. Francia G, Pilot plants for solar steam generation sta- optimization of large central receiver systems. Solar
tions. Solar Energy 1968, 12:51–64. Energy 1978, 20:505–516.
37. Di Canio DG, Treytl WJ, Jur FA, Watson CD. Line Fo- 49. Kistler BL. A User’s Manual for DELSOL3: A
cus Solar Thermal Central Receiver Research Study— Computer Code for Calculating the Optical Perfor-
Final Report, DOE/ET/20426–1. US Department of mance and Optimal System Design for Solar Thermal
Energy; 1979. Santa Clara, California: FMC Corpora- Central Receiver Plants. Sandia Report, SAND-86–
tion. 8018. Livermore, California: Sandia National Lab-
38. Mills DR, Morrison GL. Compact linear Fresnel re- oratories; 1986. Available at: http://prod.sandia.gov/
flector solar thermal powerplants. Solar Energy 1999, techlib/access-control.cgi/1986/868018.pdf. (Accessed
68:263–283. July 15, 2013).
39. Bernhard R, Hein S, de Lalaing J, Eck M, Eick- 50. Mavis CL. A Description and Assessment of Helio-
hoff M, Pfänder M, Morin G, Häberle A. Linear stat Technology. SAND87–8025, Livermore, Califor-
Fresnel collector demonstration on the PSA, Part nia: Sandia National Laboratories; 1989.
II—commissioning and first performance tests. Pro- 51. Romero M, Conejero E, Sánchez M. Recent expe-
ceedings of the 14th Solar Paces Symposium. Las riences on reflectant module components for inno-
Vegas; 2008. Golden, Colorado: NREL. Available vative heliostats. Solar Energy Mater 1991, 24:320–
at: http://www.nrel.gov/docs/gen/fy08/42709CD.zip. 332.
(Accessed July 15, 2013). 52. Monterreal R, Romero M, Garcı́a G, Barrera G. De-
40. Hautmann G, Selig M, Mertins M. First European velopment and testing of a 100 m2 glass-metal helio-
linear Fresnel power plant in operation—operational stat with a new local control system. In: Claridge DE,
experience and outlook. Proceedings SolarPACES Pacheco JE, eds. Solar Engineering. New York: ASME;
2009 (CD). Ref. manuscript: 16541. Berlin, Germany; 1997, 251–259. ISBN: 0-7918-1556-0.
September 15–18, 2009. ISBN 978-3-00-028755-8. 53. Osuna R, Fernández V, Romero M, Sanchez M. PS10:
41. Selig M. Commercial CSP plants based on Fres- A 11-MW solar tower power plant with saturated
nel collector technology. Proceedings of SolarPACES steam receiver. Proceedings 12th SolarPACES Inter-
2011. Ref. manuscript: 24660. Granada Spain; 2011. national Symposium. CD-Rom. 6–8 October. Oaxaca,
Madrid, Spain: CIEMAT. Mexico; 2004, S3–102. Cuernavaca, Mexico: Instituto
42. Lehaut C. Construction, start-up and performance Investigaciones Electricas. ISBN: 968-6114-18-1.
tests of a direct steam generation CSP Fresnel mod- 54. Silberstein E, Magen Y, Kroyzer G, Hayut R, Huss H.
ule. Proceeding of SolarPACES 2010. Ref. manuscript. Brightsource solar tower pilot in Israel’s Negev oper-
0008. Perpignan, France; 2010. Odeillo, France: ation at 130 bar @ 530◦ C superheated steam. Pro-
PROMES-CNRS. ceedings SolarPACES 2009 (CD). Berlı́n, Germany;
43. Dersch J, Morin G, Eck M, Häberle A. Compar- September 15–18, 2009. Stuttgart, Germany: DLR.
ison of linear Fresnel and parabolic trough collec- ISBN 978-3-00-028755-8.
tor systems—system analysis to determine break even 55. Schell S. Design and evaluation of eSolar’s he-
costs of linear Fresnel collectors. Proceedings So- liostat fields. Proceedings SolarPACES 2009 (CD).
larPACES 2009 (CD). Ref. manuscript: 15162. Berlin, Berlı́n, Germany; September 15–18, 2009. Stuttgart,
Germany; September 15–18, 2009. Stuttgart, Ger- Germany: DLR. ISBN 978-3-00-028755-8.
many: DLR. ISBN 978-3-00-028755-8. 56. Romero M, Buck R, Pacheco JE. an update on solar
44. Romero M, Marcos MJ, Téllez FM, Blanco M, central receiver systems, projects, and technologies. Int
Fernández V, Baonza F, Berger S. Distributed power J Solar Energy Eng 2002, 124:98–108.
from solar tower systems: a MIUS approach. Solar En- 57. Becker M, Vant-Hull LL. Thermal receivers. In: Win-
ergy 2000, 67:249–264. ter CJ, Sizmann RL, Vant-Hull LL, eds. Solar Power
45. Kolb GJ. Economic evaluation of solar-only and hy- Plants. Berlin, Germany: Springer-Verlag; 1991, 163–
brid power towers using molten-salt technology. Solar 197. ISBN 3-540-18897-5.
Energy 1998, 62:51–61. 58. Avila-Marin AL. Volumetric receivers in solar thermal
46. Falcone PK. A Handbook for Solar Central Receiver power plants with central receiver system technology:
Design. SAND86–8009. Livermore, California: Sandia a review. Solar Energy 2011, 85:891–910.

58  2013 John Wiley & Sons, Ltd. Volume 3, January/February 2014

 WIREs Energy and Environment Solar thermal CSP technology

59. Agrafiotis CC, Mavroidis I, Konstandopoulos AG, 68. Radosevich LG, Skinrood AC.The power production
Hoffschmidt B, Stobbe P, Romero M, Fernández- operation of Solar One, the 10 MWe solar thermal
Quero V. Evaluation of porous silicon carbide mono- central receiver pilot plant. J Solar Energy Eng 1989,
lithic honeycombs as volumetric receivers/collectors of 111:144–151.
concentrated solar radiation. Solar Energy Mater Solar 69. Pacheco JE, Gilbert R. Overview of recent results of the
Cells 2007, 91:474–488. Solar Two test and evaluations program. Renewable
60. Palero S, Romero M, Castillo JL. Comparison of ex- and Advanced Energy Systems for the 21st Century
perimental and numerical air temperature distributions RAES’99. Maui, Hawaii; April 11–15. RAES99–7731.
behind a cylindrical volumetric solar absorber module. New York: ASME; 1999. ISBN: 0-7918-1963-9.
J Solar Energy Eng 2008, 130:011011-1-8. 70. Epstein M, Liebermann D, Rosh M, Shor AJ. Solar
61. Marcos MJ, Romero M, Palero S. Analysis of air return testing of 2 MW (th) water/steam receiver at the Weiz-
alternatives for CRS-type open volumetric receiver. mann Institute solar tower. Solar Energy Mater 1991,
Energy 2004, 29:677–686. 24:265–278.
62. Hoffschmidt B, Fernandez V, Pitz-Paal R, Romero 71. Fernandez-Quero V, Osuna R, Romero M, Sanchez
M, Stobbe P, Téllez F. The development strategy of M, Ruiz V, Silva M. EURECA: advanced receiver for
the hitrec volumetric receiver technology—up-scaling direct superheated steam generation in solar towers, as
from 200 kWth via 3 MWth up to 10MWel. 11th an option for increasing efficiency in large low cost di-
SolarPACES International Symposium on Concen- rect steam generation plants. Proceedings of the 2005
trated Solar Power and Chemical Energy Technolo- Solar World Congress ISES-2005; August 6–12, Or-
gies. Zurich, Switzerland; September 4–6, 2002, 117– lando, Florida. Boulder, Colorado: Pub. American So-
126. Zurich, Switzerland: ETH. ISBN: 3-9521409- lar Energy Society; 2005. ISBN 0-89553-177-1.
3-7. 72. Tyner CE, Pacheco JE. eSolar’s power plant architec-
63. Romero M, Marcos MJ, Osuna R, Fernández V. De- ture. Proceedings SolarPACES 2009 (CD). Berlı́n, Ger-
sign and implementation plan of a 10 MW solar many; September 15–18, 2009. Stuttgart, Germany:
tower power plant based on volumetric-air technol- DLR. ISBN 978-3-00-028755-8.
ogy in Seville (Spain). SOLAR ENGINEERING 2000- 73. Kelly B, Singh M. Summary of the final design for
Proceedings of the ASME International Solar Energy the 10 MWe solar two central receiver project. Solar
Conference. Madison, Wisconsin; June 16–21. New Engineering. Vol. 1. New York: ASME; 1995, 575.
York: ASME; 2000. ISBN: 0791818799. 74. Pacheco JE, Reilly HE, Kolb GJ, Tyner CE. (2000),
64. Tellez F, Romero M, Heller P, Valverde A, Reche JF, Summary of the solar two test and evaluation program.
Ulmer S, Dibowski G. Thermal performance of SolAir Proceeding of the Renewable Energy for the New Mil-
3000 kWth ceramic volumetric solar receiver. Proceed- lennium. Sydney, Australia; March 8–10, 2000, 1–11.
ings 12th SolarPACES International Symposium. CD- 75. Ortega JI, Burgaleta JI, Tellez F. Central receiver sys-
Rom. Oaxaca, Mexico; October 6–8, 2004, S9–206. tem (CRS) solar power plant using molten salt as heat
ISBN: 968-6114-18-1. transfer fluid. In: Romero M, Martı́nez D, Ruiz V,
65. Kribus A. Future directions in solar thermal elec- Silva M, Brown M, eds. Proceedings of the 13th Inter-
tricity generation. Solar Thermal Electricity Genera- national Symposium on Concentrated Solar Power and
tion. Madrid, Spain: Colección documentos CIEMAT. Chemical Energy Technologies. Seville, Spain; June 20.
CIEMAT; 1999, 251–285. ISBN: 84-7834-353-9. Pub. Madrid, Spain: CIEMAT; 2006. ISBN: 84-7834-
66. Schmitz M, Schwarzbözl P, Buck R, Pitz-Paal Assess- 519-1.
ment of the potential improvement due to multiple 76. Burgaleta JI, Arias S, Salbidegoitia IB. Operative ad-
apertures in central receiver systems with secondary vantages of a central tower solar plant with thermal
concentrators. Solar Energy 2006, 80:111–120. storage system. Proceedings SolarPACES 2009 (CD).
67. Grasse W, Hertlein HP, Winter CJ. Thermal solar Ref. manuscript: 11720. Berlin, Germany; September
power plants experience. In: Winter CJ, Sizmann RL, 15–18, 2009. Stuttgart, Germany: DLR. ISBN 978-3-
Vant-Hull LL, eds. Solar Power Plants. Berlin, Ger- 00-028755-8.
many: Springer-Verlag; 1991, 215–282. ISBN 3-540- 77. REN21. Renewables 2013 Global Status Report. Paris:
18897-5. REN21 Secretariat; 2013. ISBN 978-3-9815934-0-2.

Volume 3, January/February 2014  2013 John Wiley & Sons, Ltd. 59