Applied Thermal Engineering 31 (2011) 872e879

Contents lists available at ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Alternative cycles based on carbon dioxide for central receiver solar power plants
R. Chacartegui*, J.M. Muñoz de Escalona, D. Sánchez, B. Monje, T. Sánchez
Thermal Power Group (GMTS), Escuela Técnica Superior de Ingenieros, Camino de los descubrimientos s/n, 41092 Sevilla, Spain

a r t i c l e i n f o a b s t r a c t

Article history: Research in concentrated thermal solar power plants of all types and, in particular, those based on central
Received 24 July 2010 receiver, namely solar tower plants, has experienced great impetus in the last decade, reaching full
Accepted 6 November 2010 commercial operation with the PS10 plant in Spain. In spite of previous demonstration plants testing
Available online 13 November 2010
different receivers and power cycle layouts, this first commercial power plant adopted a cavity receiver
generating saturated steam and therefore penalising cycle efficiency in order to gain plant reliability.
Keywords:
According to the experience gained, if a competitive Levelised Cost of Electricity is to be reached, capital and
Solar tower
maintenance costs must be reduced and efficiencies must be increased. To achieve these goals, modifying
CRS
Combined cycle
the power cycle is deemed essential, whether using superheated steam or alternative fluids.
Carbon dioxide In this work, the use of supercritical and transcritical carbon dioxide cycles for this application is
Supercritical CO2 proposed. Three different cycles are considered, the first two of which are stand-alone closed cycle gas
S-CO2 turbines using carbon dioxide. The third proposal is a combined cycle that comprises a topping carbon
Organic Rankine Cycle dioxide gas turbine and a bottoming Organic Rankine Cycle. Preliminary results show that these cycles
ORC are promising technologies for solar tower plants, having the potential to compete in terms of efficiency
and costs with other conventional technologies.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction
Concentrating solar thermal power (CSP) plants have been given
great impetus in the last years, especially in countries like Spain
where the installed CSP power capacity currently exceeds 430 MW
and is expected to double by 2011, with plants in different
construction stages [1]. The development of these power plants has
been associated to the adaptation of proven steam power genera-
tion technologies combined with particular concentrating solar
power components. Among others, the latter components include
heliostats or solar receivers for central receiver solar power plants
(CRS) [2e4] and linear collectors, oil pumps and oil to water/steam
heat exchangers for parabolic trough power plants. For the first
type of plants, the high solar flux hitting the receiver (averaging
between 300 and 1000 kW/m2) enables operating at rather high
temperatures of up to 1000
C [5], even if the maximum receiver
temperature is limited to around 700
C at the current stage of
development [2]. Higher temperatures have nonetheless been
obtained in experimental or demonstration plants like the Directly
Irradiated Annular Pressurised (DIAP) facility [6] of the Weizmann
Institute (Israel), with temperature exceeding 1300
C when pres-
surised air at 10e30 bar and multistage receivers are used [7], with
air temperature in the range of 800e1000
C. In this sense, an
* Corresponding author. Tel.: þ34 954 48 72 42; fax: þ34 954 48 72 43.
E-mail address: ricardo@esi.us.es (R. Chacartegui).
analysis by Segal and Epstein [8] concluded that the optimum
power plant performance would correspond to a receiver temper-
ature close to 1600 K, what would allow using conventional gas
turbine and combined cycle technologies.
Alternative power cycles, or cycles that make use of non-
conventional fluids, are a different option to achieve higher efficien-
cies without reaching such high temperatures in the receiver. Among
them, the supercritical and transcritical closed Brayton cycles
working with carbon dioxide are deemed interesting. This cycle has
been studied for the last 40 years, since firstly proposed by Feher and
Angelino [8,9], for nuclear power production in gas reactors, though
its applicability to solar power plants has also been explored [11e13].
Thermodynamically, the main advantage of the Brayton carbon
dioxide cycle relies on its high useful to expansion work ratio (i.e.
much lower compression work than expansion work) which is in the
range 0.7e0.85 when compressor inlet is in supercritical conditions.
At cycle level, different layouts were studied by Carstens et al. [14] and
Dostal et al. [15] in order to increase cycle efficiency. From the point of
view of major equipment, the necessary features of turbomachinery
were analyzed by Vilim et al. [16] and Gong et al. [17] and heat transfer
and heat exchanger layouts were analyzed by Utamura [18].
In the same category of alternative cycles, Organic Rankine Cycles
(ORC) yield higher efficiencies than conventional steam cycles when
heat delivery is at temperatures below 370
C [19] and when a low
power output does not allow exploiting the highest efficiency of
more complex steam turbine designs (reheat or feedwater heating

1359-4311/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2010.11.008
 R. Chacartegui et al. / Applied Thermal Engineering 31 (2011) 872e879
Fig. 1. TeS diagram of a simple recuperative carbon dioxide topping cycle (layout 1).
among others). Combined cycles with topping recuperative gas
turbines and bottoming ORCs have been reported as an alternative to
conventional combined cycles by Chacartegui et al. [20] and to low
temperature solar thermal electric generation by Gang et al. [21].
This paper is focused on the analysis of alternative cycles for solar
power plants with intermediate temperature central receiver (TIT
below 1100 K) in order to improve the performance of the power
plant. In all cases, carbon dioxide cycles have been considered, ether
in stand-alone or combined cycle layouts with ORC bottoming
cycles. The results show the interest of these cycles, which are
envisaged as promising technologies for solar tower facilities.
2. Carbon dioxide cycles
2.1. Cycle description and modelling
Two carbon dioxide cycles for intermediate temperature are
analyzed in this paper. First, a stand-alone closed recuperative
Brayton cycle is considered, namely layout 1 eL1-, Fig. 1. Then,
a second system, namely layout 2 eL2-, incorporates a two stage
compression. After an initial compressor (05-06), a fraction of the
flow is bypassed, cooled down and compressed (06-01-02), while the
remaining flow is compressed with the same pressure ratio without
being cooled (06-07). This recompression cycle is aimed at enhancing
Fig. 2. TeS diagram of the supercritical carbon dioxide cycle (layout 2).
873
Table 1
Main assumptions of carbon dioxide and ORC cycles.
Maximum pressure 225 bar
Minimum temperature 303e308 K
Pinch Point (or minimum economiser temperature difference) 10 K
Heat recovery exchangers efficiency 85%
Heat exchangers pressure drop 2%
Mechanical losses 2%
CO2 turbine efficiency 90%
CO2 compressor efficiency 80%
ORC condenser temperature 303 K
ORC turbine efficiency 87%
ORC pump efficiency 80%
heat transfer and attenuating internal pinch point problems in the
low temperature heat exchangers [15,22](Fig. 2).
A lumped volume approach has been used to model cycle
performance, applying mass and energy conservation to each
individual component of the system. For the fluids, real dense gas
behaviour has been considered when working in the vicinity of the
critical point (low compressibility factor). All these features have
been incorporated into a computer code using Engineering Equa-
tion Solver (EESÒ) in order to facilitate post-processing of results
and genetic algorithms optimisation. Results from the model have
been compared against available data in literature [10,15] showing
satisfactory agreement. The main assumptions of the model are
shown in Table 1:
The CO2 cycles were studied under compressor transcritical and
supercritical inlet conditions. The maximum value of the com-
pressor intake pressure (CIP) was limited to 75 bar.
2.2. Stand-alone carbon dioxide gas turbine analysis
This section shows results of the genetic algorithm optimisation
carried out to maximise cycle efficiency [23]. First, Fig. 3 illustrates
the impact of compressor inlet pressure on specific work and hot
delivery temperature from the recuperator for different turbine inlet
temperatures (TIT). These results apply to transcritical,
P01 < 73.5 bar, and supercritical, P01 > 73.5 bar, cycles following
layout 1 and show that there is little effect on specific work whereas
the recuperative potential is very sensitive to compressor inlet
pressure. Actually, increasing pressure at state 01 brings about
a dramatic drop in hot delivery temperature at the recuperator, T04p.
Fig. 4 shows specific work vs efficiency plots for layouts 1 and 2,
L1 and L2 respectively. For the same turbine inlet temperatures
considered in Fig. 3, an optimisation process has been carried out
Fig. 3. CO2 cycle (layout 1). Effect of compressor inlet pressure (P01) on specific work
and recuperator hot delivery temperature (T04p). Carbon dioxide cycle (layout 1).
874 R. Chacartegui et al. / Applied Thermal Engineering 31 (2011) 872e879

Fig. 6. TeS diagram of the transcritical carbon dioxide cycle and ORC combined cycle
(layout 3).
Fig. 4. CO2 cycle (layouts 1 and 2). Efficiency vs specific work for different turbine inlet
temperature (TIT).
cycle shown in Fig. 6, which seems to be slightly less complex than
the carbon dioxide only cycle of layout 2 studied before.
where compressor inlet conditions remain constant at 308 K and
75 bar and pressure ratio is left to vary from 1.5 to 3. A clear
3.1. Organic working fluids
improvement in cycle performance is observed when layout 2 is
adopted, yielding higher efficiencies and specific works than layout
A preliminary analysis of candidate working fluids for the Organic
1 for the same TIT.
Rankine Cycle is presented in this Section, where only dry fluids are
TeQ diagrams of the recuperative heat exchangers are shown in
taken into consideration. First, a summary of some important
Fig. 5, corresponding to layouts 1 and 2 optimised for 1100 K turbine
parameters, including fluid properties and cycle specifications, is
inlet temperature and transcritical operation. For each stream, mean
shown in Table 2. This relevant information includes maximum
heat capacity rates are given, defined as mass flow rate multiplied by
temperature beyond which the fluid starts to degrade (Tmax) [24],
_ p ). It is worth noting that hot
specific heat at constant pressure (m$C
critical temperature and pressure (Tcrit, Pcrit), condenser pressure (P1)
and cold stream temperatures are closer for the optimised L2,
and maximum temperature and pressure that avoid condensation
yielding higher heat recovery efficiency. Nevertheless, in spite of L2
within the turbine (Tmax, Pmax). In all cases, saturated live vapour at
having two serial recuperators, more heat is recuperated in layout 1
turbine inlet is considered, thus avoiding the superheated region that
due to the different optimum pressure ratios that bring about
does not provide further advantages for the dry fluids analyzed as
a lower turbine exhaust temperature in case L2. The low delivery
was shown in [20]. Additionally, it is worth noting that Isopentane,
temperature of the recuperator hot stream in layout 2 (T04pp), con-
R245fa and Isobutane yield cycles where pressure is always above
strained by compressor delivery temperature, limits heat recovery
ambient (P1 > 1), thus preventing non-condensable gases (air) from
from the exhaust of this layout.
leaking into the system.
Table 3 provides additional properties of the fluids, mainly
3. Combined cycle CO2eORC related with hazard: lower and upper flammability limits (LFL and
UFL, which are the leaner and richer fluid/air mixtures that will
In this section, the analysis of combined cycles composed by burn), autoignition temperature, global warming potential (GWP),
a topping carbon dioxide cycle and a bottoming ORC cycle is pre- ozone depletion potential (ODP) and flammability class according
sented. For the former, layout 1 was selected yielding the combined to the National Fire Protection Association of the US (NFPA). From
this information, it is concluded that R245fa and Isobutane are the
least and most hazardous fluids among those included in the
analysis.

3.2. CO2eORC combined cycle analysis

The performances of CO2eORC combined cycles working with

different organic fluids are analyzed in this section, where the main
modelling assumptions of Section 2, Table 1, are considered for the

Table 2
Organic Rankine cycles data.

Organic fluid Tmax (K) Tcrit (K) T3max (K) Pcrit (bar) P3max (bar) P1 (bar)
Toluene 671.9 591.7 569 41.26 31.16 0.05
Cyclohexane 560.7 553.6 536 40.75 32.67 0.16
n-Hexane e 507.9 491 30.58 26.63 0.25
Isopentane e 460.4 448 33.7 27.87 1.09
R245fa e 427.2 403 36.51 23.33 1.77
Fig. 5. CO2 cycle (layouts 1 and 2). TeQ diagram within the recuperator (maximum Isobutane 413.3 407.8 380 36.4 22.5 4.05
cycle temperature 1100 K).
 R. Chacartegui et al. / Applied Thermal Engineering 31 (2011) 872e879 875

topping carbon dioxide system. Nevertheless, it is important to

note that, with respect to Section 2, the main parameters of this
latter cycle vary from stand-alone to combined cycle operation as
a result of the optimisation process. For instance, optimum pressure
ratio is higher for combined cycle applications, yielding tran-
scritical carbon dioxide cycles.

3.2.1. Bottoming cycle analysis

The analysis of a bottoming ORC system incorporated into the
combined cycle in Fig. 6 is now presented. The operating conditions
of the topping cycle are assumed to be those resulting from opti-
mising its performance for 1100 K turbine inlet temperature and
45 bar compressor inlet pressure. The percentage of waste heat
from topping cycle exhaust that is used to generate vapour at the
bottoming cycle is quantified by the Heat Recovery Vapour
Generator (HRVG) efficiency (hHRVG). The fraction of this recuper-
ated energy that is then converted into useful shaft work by the
Organic Rankine Cycle is quantified by an ORC efficiency (hORC).
Hence, a global bottoming system efficiency (hBC) is defined as the
fraction of available waste heat that is converted into shaft power in Fig. 7. ORC cycle efficiency vs live vapour temperature (T3) for the selected fluids.
the HRVG þ ORC. The effects of hHRVG and hORC on hBC are analyzed
in the following figures.
a parabolic shaped plot of the global bottoming cycle efficiency (hBC)
ðh  h4 Þ  ðh2  h1 Þ where an optimum value of T3 is found. Two aspects of this hBC vs T3
hORC ¼ 3
(1) plot are worthy of note. First, the effect of live vapour temperature on
h3  h2p
HRVG efficiency is dominant so the fluid with highest hHRVG, n-
Hexane, also yields highest hBC. Second, when bottoming cycle effi-
_ ORC ðh3  h2 Þ
m
hHRVG ¼   (2) ciency increases so does the optimum temperature T3, which shifts
_
mCO 2 h04p  href rightwards in Fig. 8.
Combined cycle efficiency is plotted in Fig. 9 as a function of
maximum cycle efficiency (T03); i.e. maximum turbine inlet
hBC ¼ hHRVG $hORC (3)
temperature at the topping cycle. An interesting feature of this plot is
The effect of live vapour temperature (T3) on cycle efficiency (hORC) that there is not a single fluid that outperforms the others over the
is first illustrated in Fig. 7 for the fluids selected, showing that each entire temperature range. On the contrary, the organic fluid of choice
fluid fits a different temperature range. Hence, it is those fluids with depends on the operating conditions of the topping cycle even
higher live vapour temperature that yield higher efficiencies (for though Isopentane and R245fa seem to optimise performance for
instance, toluene) due to their higher expansion work in the turbine, high and low maximum temperature combined cycles (T03 < 900 K
in spite of also having a higher vaporisation enthalpy. Globally, the and T03 > 900 K respectively). This is in agreement with the afore-
former effect is dominant and therefore hORC increases with T3. mentioned dominant effect of HRVG performance on bottoming
A deeper insight into bottoming cycle performance when cycle efficiency.
Toluene, Cyclohexane and n-Hexane are used is now given, since
these three fluids provide higher cycle efficiencies in Fig. 7. Thus, 3.2.2. Combined cycle optimisation
Fig. 8 illustrates the effect of live vapour temperature on global Bottoming system performance has been reviewed for a partic-
bottoming cycle efficiency (hBC), right vertical axis, and Heat ular topping system set up in the previous section. Now, both
Recovery Vapour Generator efficiency (hHRVG), left vertical axis. In systems are subjected to simultaneous optimisation, a summary of
all cases, the reference operating conditions of the topping cycle
remain constant. First thing to observe is that hHRVG falls rapidly as
live vapour temperature increases due to the higher “stack”
temperature (T04pp) brought about by a constrained hot side inlet
temperature (T04p) and higher cold side saturation temperature
(T3). In other words, fluids achieving highest ORC efficiency, Fig. 7,
show lowest HRVG efficiency, Fig. 8. This behaviour is typical of
heat recovery boilers in conventional combined cycle power plants.
The aggregated effect of these two opposite trends of HRVG and
ORC efficiencies with respect to live vapour temperature yields

Table 3
Relevant properties of organic working fluids.

Organic fluid M [g/mol] Autoignition [K] LFL [%v] UFL [%v] NFPA GWP ODP
Toluene 92.14 695 1.1 7.1 3 NA NA
Cyclohexane 84.16 518 1.3 8.4 3 NA NA
n-Hexane 86.18 496 1.1 7.5 3
Isopentane 72.15 693 1.4 7.6 3
R245fa 134 e e e 0 950 0
Isobutane 58 693 1.8 8.4 4 3 0 Fig. 8. Effect of live vapour (T3) temperature on HRVG and bottoming cycle efficiencies
(n-Hexane, Cyclohexane and Toluene).
876 R. Chacartegui et al. / Applied Thermal Engineering 31 (2011) 872e879
Fig. 9. Combined cycle efficiency vs topping cycle maximum temperature (T03). Results
apply to independent optimisation of topping and bottoming cycle.
which is presented in Fig. 10. This parallel optimisation is per-
formed for pressure ratio and live vapour temperature in the bot-
toming and topping cycles respectively.
A simple comparison of Figs. 9 and 10 shows that results for
single and simultaneous optimisations are different. Now, R245fa,
Isopentane and Cyclohexane yield highest combined cycle effi-
ciency for T03 < 840 K, 840K < T03 < 1040 K and T03 > 1040 K
respectively. Additionally, for higher turbine inlet temperatures,
Toluene is envisaged as the best candidate fluid.
From the previous results, it is concluded that, as a general rule
of thumb, organic fluids whose critical temperature is closer to the
operating live vapour temperature provide higher cycle efficiency.
This is owed to the expansion ratio in the turbine increasing more
than proportionally than the corresponding vaporisation enthalpy.
This is not universal though and, for instance, n-hexane does not
follow this rule for the temperature range under consideration
here. Such singularity of n-Hexane is related to the performance of
the Heat Recovery Vapour Generator, as it is the only fluid analyzed
for which the minimum temperature difference between hot and
cold fluids, namely pinch point, is found at the hot end of the
economiser. On the contrary, the pinch point is found either at the
cold end or somewhere in the middle of the economiser for the
remaining fluids. This particular feature (of n-Hexane) limits the
maximum live vapour temperature and, therefore, a further
Fig. 10. Combined cycle efficiency vs topping cycle maximum temperature (T03).
Results apply to simultaneous (parallel) optimisation of topping and bottoming cycle.
reduction of the corresponding enthalpy of vaporisation. Accord-
ingly, bottoming efficiency falls.
3.2.3. Interest of internal heat recovery in the bottoming cycle
The potential for internal heat recovery within an Organic
Rankine Cycle is rather low, around 50 K in terms of temperature
difference between turbine exhaust and condenser temperatures
(T4eT2), so the effect of incorporating a recuperator is expected to
have a low positive impact on combined cycle efficiency. Neverthe-
less, in order to assess this presumed behaviour, the performances of
recuperative and non-recuperative bottoming cycles using Iso-
pentane are reported in Table 4.
Table 4 shows that using a recuperative ORC cycle increases
bottoming cycle efficiency (hORC) by 3 percentage points with
respect to using a non-recuperative cycle. However, at the same
time, a higher inlet temperature to the HRVG (T2p) penalises hHRVG.
Globally, both effects are balanced and the corresponding bot-
toming cycle global efficiencies do not differ from one another
significantly. Accordingly, insight of the additional cost/complexity
without relevant efficiency gain, it is concluded that the recuper-
ative layout is not advisable for the combined cycle under analysis.
The same behaviour is observed in all the analyzed fluids.
4. Comparison of results
The results from all systems studied in previous sections are
brought together into a single plot of system efficiency against
maximum (combined) cycle temperature. Thus, Fig. 11 shows h vs
T03 plots for stand-alone CO2 with L1 and L2 layouts and combined
CO2eORC systems with different organic working fluids. In all
cases, results are shown for the optimised systems.
These results show an efficiency gain of 7e13 percentage points of
the combined cycle layout compared with the simpler stand-alone
gas turbine with layout 1 and a 2e3 percentage points with respect
to the more complex layout 2. Moreover, within the temperature
range of interest for solar applications (T03 < 1100 K), R245fa is
considered the best candidate due to its thermodynamic perfor-
mance and additional advantages: non-flammability and higher
than atmospheric condenser pressure amongst others.
5. Application to central receiver solar power plants
5.1. Evaluation of central receiver solar plant
Carbon dioxide gas turbines, either simple or combined cycle,
and steam power plants are now compared when applied to oper-
ating/boundary conditions typical of central receiver solar power
plants. For carbon dioxide based systems, stand-alone units and
CO2eORC(R245fa) combined cycles are taken as reference, with the
features shown in previous sections. For steam, two different
systems are considered. First, a superheated live steam Rankine
cycle with five feedwater heaters and reheat. Second, a saturated
live steam non-reheat turbine with four feedwater heaters as found
in the commercially available PS10 power plant near Sevilla, Spain
[5]. In all cases, the global efficiency of the resulting solar power
plant is expressed in a simple and direct way as the product of three
partial efficiencies, namely optical for the solar field, thermal for the
solar receiver and cycle for the power block [8,25].
hglobal ¼ hopt $hrec $hcycle (4)
Results of the cycles under analysis are summarised in Table 5 for
a reference ground-level solar radiation of 1000 W/m2 and a repre-
sentative value of the solar to thermal energy conversion efficiency.
This value, which combines optical and receiver efficiencies
 R. Chacartegui et al. / Applied Thermal Engineering 31 (2011) 872e879 877

Table 4
Combined CO2eIsopentane cycle performance (recuperative and non-recuperative).

TIT(K) Non-recuperative Recuperative

hCC hTC hBC hORC hHRVG hCC hTC hBC hORC hHRVG
1100 47.49 38.34 16.08 18.07 88.96 47.51 38.34 16.12 21.24 75.87
1050 46.23 37.03 15.78 18.07 87.34 46.22 37.03 15.78 21.24 74.27
1000 44.71 35.55 15.32 18.07 84.77 44.7 35.55 15.31 21.23 72.09
950 42.85 33.89 14.58 18.07 80.68 42.88 33.89 14.63 21.24 68.9
900 40.69 31.99 13.73 17.16 80.04 40.69 31.99 13.73 19.77 69.47
850 38.28 29.81 12.92 16.17 79.91 38.29 29.81 12.93 19 68.05
800 35.57 27.28 12.15 15.84 76.68 35.57 27.28 12.15 17.94 67.72

(hsolth ¼ hopt $hrec ), is based on results reported in literature for
state of the art heliostats and central receivers. Nevertheless, even
though the reported combined solar-to-thermal efficiency varies in
the range 60e69% [25,26], a more conservative value of 62% is
adopted here. Further discussion on this selection is given in the
next section.
For the superheated steam cycles, 600 and 850 K live steam
temperatures are considered while the remaining temperatures and
pressures are taken from typical subcritical steam power plants with
similar layout. For the PS10-like case, live steam temperature is
imposed by the need to have saturated steam at the rated operating
pressure. For all the carbon dioxide based systems, two different
turbine inlet temperatures have been considered: 850 and 1000 K.
These different reference temperatures derive from the dissim-
ilar characteristics of solar receivers and working fluids to be used in
each case. Thus, for steam power plants, saturated steam receivers
formed by bundles of pipes with horizontal or vertical arrangement
and their evolution to reheated steam generation have been
considered. For the alternative CO2 cycles, a volumetric receiver for
gas heating [27e29] is used instead owed to the inherent difficulties
of the aforementioned steam receivers to work with gas due to the
very low heat transfer coefficient and thermal inertia of this latter
fluid [5]. These volumetric receivers present several advantages like
simplicity and control of the solar shape, which is easier with
a single phase receiver than with a biphasic one, especially when
reheat is used, as demonstrated in Solar One [5].
From the results shown in Table 5, the stand-alone L2 CO2
system and the combined cycle formed by a topping L1 CO2 turbine
and a bottoming R245fa cycle give highest efficiencies and, there-
fore, they have the potential to be competitive options against the
next generation of steam-based solar tower plants. Efficiencies of
Fig. 11. Comparison of cycle efficiencies vs maximum temperature (T03) plots for
stand-alone carbon dioxide systems (layouts 1 and 2) and CO2eORC combined cycles.
the same order or even slightly higher are expected for the former
systems but with a simpler and more compact layout than those
currently commercialised and based on steam technology.
5.2. Some technoeconomical considerations
As it has been stated in the previous sections, the proposed cycle
has not been developed yet and the estimated performance and cost
of technology shown here are based on theoretical studies and
projections departing from the existing experience with similar
technology. This approach has been successfully applied by the
authors to the analysis of turbomachinery and heat exchangers in
carbon dioxide based cycles in a recent publication [30].
In the commercial CRS thermal solar power plant, tower and
receiver account for 14% of the total cost of the plant while the
fraction corresponding to the power block is estimated at around
32% [5]. In sight of these values, additional considerations about
costs and performance of both blocks with the proposed tech-
nology will be later discussed.
For the power block, and even though dedicated turboma-
chinery for supercritical carbon dioxide applications is not available
yet, it can be anticipated that radial turbomachinery is likely to be
more suitable than axial devices for the power range of current
Central Receiver Solar power plants. These radial turbines and
compressors are favoured by an extremely compact design for the
range of specific works presented in Figs. 3 and 4 in which they
achieve higher internal efficiencies than other working fluids like
air or helium, thus increasing the potential of supercritical CO2
cycles [15,30,31].
The most challenging part of system development is compressor
design though, since operating in the vicinity of the critical point
yields abrupt variations of fluid properties within the machine [17].
In this regard, it must be highlighted that the proposed combined
cycle avoids this problem thanks to the optimised transcritical
solution, which clearly simplifies the design of new turbomachinery.
At the other side of the proposed combined cycle, ORC systems
are already commercialised by manufacturers like ORMAT or Tur-
boden, despite continuous product development and potential
scale up/down benefits that can derive from adapting either
working fluid or component design to the singularities and power
output of a power plant in particular. Still, equipment costs,
900e2000 $/kW depending on power output, and performances
can be estimated rather accurately based on existing cycles.
With regard to costs in a carbon dioxide cycle, heat exchangers
play a fundamental role: solar receiver, recuperator and heat
recovery vapour generator. The recuperator is likely to be of the
Printed Circuit type (PCHE), which is currently being considered for
nuclear applications and, in some cases, withstands extreme
pressures and temperatures as high as 500 bar and 1073 K [32].
These heat exchangers are very compact in size and achieve
a remarkably high effectiveness, in some cases exceeding 97% [33].
Thus, in a water/water application, the cost of a PCHE can be up to
878 R. Chacartegui et al. / Applied Thermal Engineering 31 (2011) 872e879

Table 5
Estimated performance of a central receiver solar power plant using different power cycles.

PS10 (saturated steam) Steam CO2 (type 1) CO2 (type 2) CO2 (type 1) e ORC
(R245fa)
TIT [K] 524 600 850 850 1000 850 1000 850 1000
Cycle efficiency [%] 30.7 36.63 41.93 30.59 33.8 36.39 42.48 38.82 43.96
WT/mass flow [kW kg1 s] 608.9 770.8 1274 88.11 114.2 107.5 155.8 117.2 161.6
Q [kJ$kg1] 1986.3 2105 3038 293.9 343.9 295.3 366.8 295.6 360.2
Global Efficiency [%] 19 22.62 25.89 18.89 20.87 22.47 26.23 23.71 27.14
Net Power [W m2] 136.46 226.11 258.95 185.12 205.06 224.79 262.29 244.95 276.34

thirty times smaller than the equivalent shell and tube heat exchangers in the system except for the heat addition process,
exchanger [34], and in other applications can reach a 85% smaller whether with a CO2 volumetric receiver or by means of an inter-
footprint [35]. mediate heat transfer fluid, for which Inconel 617 or similar is
There also exists previous experience in designing receiver to recommended. Economically, reference [41] estimates the cost of
work with incondensable gases. For instance the SOLGATE project a supercritical carbon dioxide recuperator at $16/KWe while for the
[36], where a pressurised volumetric air receiver was used with intermediate heat exchanger this value falls to $6.4/KWe, provided
a secondary concentrator receiver technology called REFOS. SS 347 is used. If Incoloy-800 had to be used for temperatures
Following this concept, the configuration would be a group of around 1273 K, an increase in cost by a factor of seven is projected.
parallel interconnected single receiver modules subdivided into low
temperature (up to 873 K), medium-temperature (up to 1073 K) and
6. Conclusions
high-temperature units (up to 1273 K). According to their temper-
ature level, these unit receivers would be located in the low, medium
In this paper, the application of carbon dioxide Brayton to
and high-flux region of the focal spot [27]. The receiver finally
concentrated solar power plants with central receiver has been
adopted in the proposed supercritical carbon dioxide system should
studied yielding the following conclusions:
combine the basis of REFOS technology with the PCHE design due to
the very high pressures found in the heating process.
 Amongst the two closed Brayton cycles considered, the more
The size of a receiver of any type relies strongly upon optical
complex layout L2 working with supercritical conditions at
considerations even if for supercritical carbon dioxide it is expected
compressor inlet and with improved heat recovery achieves
to have the small size typical of the aforementioned PCHE heat
higher cycle efficiency. The improvement with respect to
exchangers. In reference [27], estimated costs are 16 k€/m2, 33 k
simpler schemes can be as high as 7e12 percentage points,
€/m2 and 37.5 k€/m2 for the low, medium and high-temperature air
depending on turbine inlet temperature.
modules respectively. The costs for CO2 receivers at a future
 For the CO2 and ORC combined cycle proposed in this work,
commercial stage are expected to be in the same order. Other
a simpler topping cycle layout (L1) is selected. For this layout,
interesting option is making use of an intermediate fluid like, for
designs with lower compressor inlet pressures and higher
instance, molten salts. In this case, the expected receiver efficiency
pressure ratios yield similar specific work, as shown in Section
is in the range from 75 to 90% [5,29,37,38], similar to other volu-
2, but higher turbine exhaust temperature, thus favouring
metric receivers and slightly lower than pipe receivers used with
bottoming cycle performance. This combined cycle layout has
water/steam, 80e93% [5,39].
lower complexity than layout L2 in the topping cycle and takes
Finally, some considerations are given with respect to the Heat
advantage of already commercial ORC technology in the bot-
Recovery Vapour Generator (HRVG) which is specific of the
toming system.
combined cycle proposed. Table 6 shows the dependence of HRVG
 Combined cycle analysis shows that recuperative ORC systems
design and performance with respect to the cold side working fluid
do not provide a significant advantage due to rather low Heat
and for the optimised solution. Interestingly, Isobutene requires
Recovery Vapour Generator efficiency.
minimum heat transfer area and, therefore, its cost is lowest.
 There is not a single organic fluid able to optimise combined
However, R245fa yields higher efficiency at a minimum extra cost,
cycle efficiency independently of turbine inlet temperature of
still far from the required surface for other fluids. This interesting
the topping cycle. As a general rule of thumb, there exist
behaviour adds up to the previous thermodynamic advantages
a parallel trend between topping cycle (T03) and critical
leading to the conclusion that R245fa is the most interesting fluid of
temperature of the organic fluid. Thus, highest combined cycle
choice for the bottoming Organic Rankine Cycle.
efficiency when high turbine inlet temperatures are used in
Some final considerations are given with regard to materials
combination with organic fluids with high critical temperature,
that meet the high pressure and temperature requirements typical
the opposite being also true. Nevertheless, some exceptions
of supercritical carbon dioxide cycles. Thus, stainless steels SS 316
exist, like n-Hexane, that are usually related with the location
(873 K) or SS 347 (948 K) [40,41] can be used for the different heat
of the minimum temperature difference within the economiser
Table 6
of the Heat Recovery Vapour Generator.
Summary of HRVG characteristics for a number of organic working fluids in the  Amongst the organic fluids considered, Cyclohexane, Iso-
bottoming cycle (results expressed per unit mass flow of CO2 in the topping cycle). pentane and R245fa in decreasing temperature (T03) order
Fluid Global HRVG optimise combined cycle efficiency.

Effectiveness Specific area Heat transfer

With respect to the application of the proposed combined cycle
[%] m2/(kg/s CO2) kW/(kg/s CO2) to central receiver solar power plants, the following conclusions are
Isobutane 93.7 4 247.8 drawn:
R245fa 95.8 5.7 253.6
Isopentane 94.5 5.93 240.8
Cyclohexane 90.7 22.09 262.1
 The use of CO2eORC combined cycles in central receiver solar
power plants operating with maximum temperatures close to
 R. Chacartegui et al. / Applied Thermal Engineering 31 (2011) 872e879
1000 K provides a power increase of almost 7 percentage points
with respect to superheated (850 K) steam cycles with reheat
and five feedwater heaters for the same incident radiation. With
respect to the commercially available PS10 technology, this
means a twofold increase in power output and efficiency.
 The combined cycle proposed is simpler and more compact
than the steam cycle with reheat and feedwater heating used in
the comparison.
 With this configuration, make-up water and the corresponding
demineralised water facility are not required. However, CO2
and ORC storage tanks and pumping devices are still necessary.
 The proposed carbon dioxide cycle works at a high pressure of
225 bar, which is lower than state of the art supercritical steam
boilers. This pressure level is thus technologically affordable
and, even if capital costs are expected to increase, this effect is
attenuated by the reduced size of the unit in comparison to the
aforementioned steam boilers.
 The proposed carbon dioxide combined cycle works at a higher
than atmospheric low pressure, hence avoiding air leakages
into the system.
 In locations with limited water availability, and in order to
achieve the required condenser temperature in the ORC cycle,
the most feasible solution to reduce the demand of cooling
water is to use an air cooled condenser to drop temperature
down to 323 K and then, in serial, a conventional water
condenser to further reduce temperature to 305e308 K.
References
[1] http://www.protermosolar.com
[2] R. Buck, C. Barth, M. Eck, W.D. Steinmann, Dual-receiver concept for solar
towers, Solar Energy 80 (2006) 1249e1254.
[3] S. Alexopoulos, B. Hoffschmidt, Solar tower power plant in Germany and
future perspectives of the development of the technology in Greece and
Cyprus, Renewable Energy 35 (2010) 1352e1356.
[4] F. Colzi, S. Petrucci, G. Manzolini, R. Chacartegui, P. Silva, S. Campanari, D.
Sánchez, Modeling on/off-design performance of solar tower plants using
saturated steam, ES2010e90399, Proceedings of the ASME 2010 4th Inter-
national Conference on Energy Sustainability ES2010 May 17e22, 2010,
Phoenix, Arizona, USA.
[5] M. Romero, E. Zarza, Concentrating Solar Thermal Power, Handbook of Energy
Efficiency and Renewable Energy. Taylor & Francis Group, 2007, Chapter 21.
[6] A. Kribus, R. Zaibel, D. Carey, A. Segal, J. Karni, A solar-driven combined cycle
power plant, Solar Energy 62 (2) (1998) 121e129.
[7] A. Kribus, P. Doron, R. Rubin, J. Karni, R. Reuven, S. Duchan, E. Taragan,
A multistage solar receiver: the route to high temperature, Solar Energy 67
(1e3) (1999) 3e11.
[8] A. Segal, M. Epstein, Optimized working temperatures of a solar central
receiver, Solar Energy 75 (2003) 503e510.
[9] E.G. Feher, The supercritical thermodynamic power cycle, Energy Conversion
8 (1968) 85e90.
[10] G. Angelino, Carbon Dioxide Condensation Cycles for Power Production, ASME
Paper 68-GT-23, 1968.
[11] X.R. Zhang, H. Yamaguchi, An experimental study on evacuated tube solar
collector using supercritical CO2, Applied Thermal Engineering 28 (2008)
1225e1233.
[12] H. Yamaguchi, X.R. Zhang, K. Fujima, M. Enomoto, N. Sawada, Solar energy
powered rankine cycle using supercritical CO2, Applied Thermal Engineering
26 (2006) 2345e2354.
[13] J. Wang, Z. Sun, Y. Dai, S. Ma, Parametric optimization design for supercritical
CO2 power cycle using genetic algorithm and artificial neural network,
Applied Energy 87 (2010) 1317e1324.
[14] N.A. Carstens, P. Hejzlar, M.J. Driscoll, Control System Strategies and Dynamic
Response for Supercritical CO2 Power Conversion Cycles, MIT-GFR-038, 2006.
[15] V. Dostal, M.J. Driscoll, P. Hejzlar, A Supercritical Carbon Dioxide Cycle for
Next Generation Nuclear Reactors, MIT-ANP-TR-100, 2004.
[16] R.B. Vilim et al., Dynamic System Analysis of a Supercritical CO2 Compression
Loop, Proceedings of ICAPP ‘08 Anaheim, CA USA, June 8e12, 2008, Paper 8343.
[17] Y. Gong, N.A. Carstens, M.J. Driscoll, I.A. Matthews, Analysis of Radial Compressor
Options for Supercritical CO2 Power Conversion Cycles, MIT-GFR-034, 2006.
[18] M. Utamura, Thermal-hydraulic characteristics of microchannel heat
exchanger and its application to solar gas turbines, Proceedings of ASME Turbo
Expo 2007, Power for Land, Sea, and Air, May 14e17, 2007, Montreal, Canada.
[19] T.C. Hung, T.Y. Shai, S.K. Wang, A review of organic Rankine cycles (ORCs) for
the recovery of low/grade waste heat, Energy 22 (7) (1997) 661e667.
879
[20] R. Chacartegui, D. Sánchez, J.M. Muñoz de Escalona, T. Sánchez, Alternative
ORC bottoming cycles for combined cycle power plants, Applied Energy 86
(2009) 2162e2170.
[21] P. Gang, L. Jing, J. Jie, Analysis of low temperature solar thermal electric
generation using regenerative organic Rankine Cycle, Applied Thermal Engi-
neering 30 (2010) 998e1004.
[22] M. Utamura, Thermodynamic Analysis of Part-Flow Cycle Supercritical CO2
Gas Turbines, ASME Paper GT2008e50151, 2008.
[23] D.E. Goldberg, Genetic Algorithms in Search, Optimization, and Machine
Learning. Addision-Wesley Publishing Company, 1989.
[24] E. Prabhu, Solar Trough Organic Rankine Electricity System (STORES) Stage 1.
Power Plant Optimization and Economics. NREL/SR-550e39433, March 2006.
[25] A. Kribus, A high-efficiency triple cycle for solar power generation, Solar
Energy 72 (1) (2002) 1e11.
[26] A. Segal, M. Epstein, Comparative performances of “tower-top” and “tower-
refrector” central solar receivers, Solar Energy 65 (4) (1999) 207e226.
[27] P. Schwarzbözl, R. Buck, C. Sugarmen, A. Ring, M. Marcos, P. Altwegg, J. Enrile,
Solar gas turbine systems: design, cost and perspectives, Solar Energy 80
(2006) 1231e1240.
[28] A. Kribus, Thermal integral micro-generation systems for solar and conven-
tional use, Journal of Solar Energy Engineering 124 (2002) 189e197.
[29] M. Romero, R. Buck, J. Pacheco, An Update on solar central receiver systems,
projects, and technologies, Journal of Solar Energy Engineering 124 (2002)
98e108.
[30] D. Sánchez, J.M. Muñoz de Escalona, R. Chacartegui, A. Muñoz, T. Sánchez,
A comparison between MCFC based hybrid systems using air and supercritical
carbon dioxide Brayton cycles with state of the art technology, Journal of
Power Sources (2010). doi:10.1016/j.jpowsour.2010.09.091.
[31] J. Cha, et al., Development of a supercritical CO2 Brayton energy conversion
system coupled with a sodium cooled fast reactor, Nuclear Engineering and
Technology 41 (2009) 1025e1044.
[32] http://www.heatric.com/
[33] J.K. Min, J.H. Jeong, M.Y. Ha, K.S. Kim, High temperature heat exchanger
studies for applications to gas turbines, Heat Mass Transfer 46 (2009)
175e186.
[34] D.A. Reay, Compact heat exchangers, enhancement and heat pumps, Inter-
national Journal of Refrigeration 25 (2002) 460e470.
[35] A.M. Johnston, W. Levy, S.O. Rumbold, Application of Printed Circuit Heat
Exchanger Technology within Heterogeneous Catalytic Reactors, Paper pre-
sented at the AIChE Annual Meeting 2001, November 2001.
[36] European commission, Solgate, Solar hybrid gas turbine electric power
system, Contract ENK5-CT-2000e00333, EUR 21615.
[37] K. Kitzmiller, F. Miller, Thermodynamic cycles for a small particle heat
exchange receiver used in concentrating solar power plants, ES2010e90229,
Proceedings of the ASME 2010 4th International Conference on Energy
Sustainability ES2010, May 17e22, 2010, Phoenix Arizona, USA.
 ski, M. Modest, A. Steinfeld, Heat transfer analysis of
[38] I. Hischier, D. Hess, W. Lipin
a novel pressurized air receiver for concentrated solar power via combined cycles,
Journal of Thermal Science and Engineering Applications 1 (2009) 1e6 041002.
[39] R. Buck, T. Bräuning, T. Denk, M. Pfänder, P. Schwarzbözl, F. Tellez, Solar-
Hybrid gas turbine-based power tower systems (REFOS), Journal of Solar
Energy Engineering 124 (2002) 2e9.
[40] C.F. McDonald, Recuperator considerations for future higher efficiency
microturbines, Applied Thermal Engineering 23 (2003) 1463e1487.
[41] M.J. Driscoll, Supercritical CO2 Plant Cost Assessment, MIT-GFR-019, 2004.
Nomenclature
CIP: Compressor inlet pressure
Cp: Specific heat at constant pressure
CRS: Central Receiver Solar plant
DIAP: Directly Irradiated Annular Pressurised
EES: Equation Engineering Solver
HRVG: Heat Recovery Vapour Generator
L1: Layout 1
L2: Layout 2
M: Mass flow
ORC: Organic Rankine Cycle
P01: Compressor inlet conditions layout 1/Supercritical compressor inlet layout 2.
P05: Minimum pressure for layout 2
PCHE: Printed Circuit Heat Exchangers
S-CO2: Supercritical CO2 cycle
T3: Bottoming cycle maximum temperature
T03: Topping cycle maximum temperature
TIT: Turbine Inlet Temperature
T04pp: Recuperator hot delivery temperature
Greek symbols
aextrac: Mass flow extraction ratio. Layout 2
hBC: Bottoming cycle efficciency
hCRS: Central Receiver Solar plant efficiency
hHRVG: Heat Recovery Vapur Generator efficiency
hOpt: Optical efficiency
hORC: Organic Rankine Cycle efficiency