ArticlePDF Available

Concentrated Solar Power: Actual Performance and Foreseeable Future in High Penetration Scenarios of Renewable Energies

Authors:

Abstract and Figures

Analyses proposing a high share of concentrated solar power (CSP) in future 100% renewable energy scenarios rely on the ability of this technology, through storage and/or hybridization, to partially avoid the problems associated with the hourly/daily (short-term) variability of other variable renewable sources such as wind or solar photovoltaic. However, data used in the scientific literature are mainly theoretical values. In this work, the actual performance of CSP plants in operation from publicly available data from four countries (Spain, the USA, India, and United Arab Emirates) has been estimated for three dimensions: capacity factor (CF), seasonal variability, and energy return on energy invested (EROI). In fact, the results obtained show that the actual performance of CSP plants is significantly worse than that projected by constructors and considered by the scientific literature in the theoretical studies: a CF in the range of 0.15–0.3, low standard EROI (1.3:1–2.4:1), intensive use of materials—some scarce, and significant seasonal intermittence. In the light of the obtained results, the potential contribution of current CSP technologies in a future 100% renewable energy system seems very limited.
Content may be subject to copyright.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
1
Concentrated Solar Power: actual performance and foreseeable future in high penetration
scenarios of renewable energies
Carlos de Castroa,b,* Iñigo Capellán-Pérezb
September 2018
aApplied Physics Department, Escuela de Arquitectura, Av Salamanca, 18, University of
Valladolid, 47014, Valladolid, Spain. ccastro@termo.uva.es
bResearch Group on Energy, Economy and System Dynamics, University of Valladolid, Spain
Author Accepted Manuscript from the paper published in the journal “BioPhysical Economics
and Resource Quality:
Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual performance and
foreseeable future in high penetration scenarios of renewable energies”. September 2018.
BioPhysical Economics and Resource Quality (2018) 3:14. https://doi.org/10.1007/s41247-018-
0043-6
Abstract
Analyses proposing a high share of Concentrated Solar Power (CSP) in future 100% renewable
energy scenarios rely on the ability of this technology, through storage and/or hybridization, to
partially avoid the problems associated with the hourly / daily (short-term) variability of other
variable renewable sources such as wind or solar photovoltaic. However, data used in the
scientific literature are mainly theoretical values. In this work, the actual performance of CSP
plants in operation from publicly available data from 4 countries (Spain, the USA, India and
UAE) has been estimated for 3 dimensions: capacity factor, seasonal variability and Energy
Return on Energy Invested (EROI). In fact, the results obtained show that the actual
performance of CSP plants is significantly worse than that projected by constructors and
considered by the scientific literature in the theoretical studies: a capacity factor in the range
of 0.15-0.3, low standard EROI (1.3:1-2.4:1), intensive use of materials some scarce-, and
significant seasonal intermittence. In the light of the obtained results, the potential
contribution of current CSP technologies in a future 100% renewable energy system seems
very limited.
© 2017. This manuscript version is made available under the CC-BY-NC-ND 4.0 license
http://creativecommons.org/licenses/by-nc-nd/4.0/
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
2
Table of contents
1. Introduction .......................................................................................................................... 2
2. Concentrated solar power plant technologies ...................................................................... 5
3. Capacity factor (CF) ............................................................................................................... 6
4. Monthly/seasonal variability ................................................................................................. 8
5. EROI ..................................................................................................................................... 13
5.1. EROI expression ................................................................................................................ 13
5.2. Energy Used by CSP system (EnUtot) ............................................................................... 14
5.2.1. Scenarios ................................................................................................................... 17
5.2.2. Current EnU of CSP plant .......................................................................................... 18
5.2.3. EuN of CSP plants in high RES penetration scenarios ............................................... 19
5.3. The g factor (average efficiency in the transformation of primary energy to electricity
and the quality of the energy) ................................................................................................. 20
5.4. Estimation of EROI ............................................................................................................ 21
6. Discussion and conclusions ..................................................................................................... 24
Acknowledgements ..................................................................................................................... 26
References ................................................................................................................................... 26
Appendix A .................................................................................................................................. 32
Appendix B .................................................................................................................................. 33
1. Introduction
The transition to Renewable Energy Sources (RES) is an indispensable condition to achieve
sustainable socio-economic systems. Most governments are developing policy frameworks to
promote the penetration of renewable energy sources to improve energy security (increasingly
threatened by the depletion of fossil fuels), while mitigating emissions to limit anthropogenic
climate change and other negative externalities of conventional energy sources (Capellán-
Pérez et al., 2014; IPCC, 2014; Johansson, 2013; REN21, 2015; WEO, 2014). Among renewables,
wind and solar are estimated to have the greatest potential (de Castro et al., 2013; IPCC, 2011;
Smil, 2010), with projections often assuming that the resource base provides no practical
limitation if adequate investments are forthcoming (e.g., (IEA and IRENA, 2017; IPCC, 2011)).
At the same time, wind and solar are the RES most critically affected by the intermittency of
the source in the short (e.g. hours, day/night), medium (days/weeks) and long-term (e.g.
winter/summer, annual) (Capellán-Pérez et al., 2017b; MacKay, 2013; Trainer, 2017a, 2013,
2012, 2010; Wagner, 2014). In this context, concentrating solar power (CSP) with thermal
energy storage (TES) can partially compensate for the short-term variability of other RES due
to its ability to store energy and dispatch energy following the demand. Due to its ability to
provide an hourly/daily flexible capacity, CSP is expected to complement PV and wind,
substantially increasing their penetration potential, especially in locations with adequate solar
resources. The performance of CSP is also enhanced when coupling the plant to a back-up
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
3
system, typically natural gas (Denholm and Mehos, 2015; García-Olivares, 2016; García-
Olivares et al., 2012; Jacobson and Delucchi, 2011; NREL, 2012). On the other hand, CSP plants
are more expensive (IEA and IRENA, 2013; IRENA, 2018; Trainer, 2017b; Turchi, 2010) and, at
present, represent a less universal solution than other RES in general and than PV in particular.
This is because: (1) they only use direct irradiance (DNI) (PV also uses diffuse irradiance); (2)
they require higher levels of irradiance with low cloudiness to be economically optimal, most
suitable locations corresponding to arid zones; and (3) they adapt less well to terrain
unevenness (Deng et al., 2015; Hernandez et al., 2015).1 Thus, CSP investments are expected
to be profitable just when a relatively high renewable penetration is targeted in the electricity
mix (Brand et al., 2012).
Globally, the installation of new capacity has grown at a pace of over 30% per year between
2005-2015, but in 2015 the growth was 9.7%, and in 2016 just 2.3% (REN21, 2017, 2016). Also,
for the first time, all of the facilities added in 2015 and 2016 incorporated TES capacity, a
feature now seen as central to maintaining the competitiveness of CSP through the flexibility
of hourly/daily dispatchability (REN21, 2016, 2017). At the end of 2015 there were over 4.8
GW of CSP in operation globally, which produced almost 10 TWh in that year (as opposed to
over 250 TWh produced by all solar technologies) (IRENA db, 2017); <0.04% of the global
electricity produced in that year (BP, 2017). Spain and the USA at present account for most of
the CSP installed power (~80%); however, facilities are under construction in several countries
such as Australia, Chile, China, India, Israel, Mexico, Saudi Arabia and South Africa. Thus, it is
commonly expected that this technology will spread over the next few years in those countries
with high irradiance levels (REN21, 2016, 2017). However, the global deployment level of CSP
is still currently low and uncertain.
Due to the aforementioned factors, CSP with TES is thus usually seen as a key technology to
design or approach 100% RES power systems. Table 1 shows the estimated contribution of CSP
by different studies in the literature proposing global 100% RES scenarios (Delucchi and
Jacobson, 2011; García-Olivares, 2016; Greenpeace et al., 2015; Jacobson et al, 2016; Jacobson
and Delucchi, 2011; WWF, 2011). These studies typically assume that large quantities of
electricity could be technically transported on a continental scale between areas of high
renewable resources (e.g. solar from deserts and wind from marine platforms) to the regions
of consumption.2 In terms of energy generation, these studies project generation from CSP to
range from 1 to 5 TWe (9,000 44,000 TWh/yr or 30-160 EJ/yr), which is between 40% and
almost 2 times the current global electricity generation by all sources. The projected share of
CSP ranges between 12% (WWF, 2011) to 42% (García-Olivares, 2016) of the total energy
generation. The capacity factor (CF, i.e., the ratio of an actual electrical energy output over a
given period of time to the maximum possible electrical energy output over the same amount
of time) considered by these studies ranges from 0.31 to 0.75 and is thus supposed to be
around 2 to 3 times bigger than the CF of other RES, such as wind and solar PV respectively.
1 Additionally, restrictions on water use in arid regions that often have the most appropriate solar
resources for CSP would reduce plant efficiency due to the implementation of dry-cooling technologies.
2 However, these large scale intercontinental infrastructures are challenged by geopolitical and
economic barriers, as well as concerns over energy and food security (for a detailed discussion, see
(Capellán-Pérez et al., 2017b)).
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
4
CSP energy
generation
Total energy
generation
Share CSP
CF
Study
TWe
TWe
%
(García-Olivares, 2016)
5
12
42
0.4-0.75
(Delucchi and Jacobson,
2011; Jacobson and
Delucchi, 2011)
2.3
11.5
20
0.31
(Greenpeace et al.,
2015)
1.6
9.7
16
0.63
(WWF, 2011)
1
8.3
12
0.46
(Jacobson et al, 2016)
1.8
11.8
15
0.53
Table 1: Contribution of CSP in global 100% RES scenarios. CF: power plant capacity factor.
Other studies focusing on scenarios of strong penetration of RES at country-level also assume a
big participation of CSP when there are good solar irradiance resources. For instance, Lenzen
et al., (2016) gives a 49% share of electricity production for a 100% RES electricity transition
model for Australia (with plant CF of 0.3 without TES, and a CF of 0.6 with 15 hours of storage
by TES). Elliston et al., (2012) also gives a 40% penetration share of CSP with CF of 0.6 for
Australia; whereas NREL (2012) gives a share of 12% of electricity production (CF=0.51) for a
90% RES electricity penetration scenario for the USA.
Most studies in the literature usually apply theoretical assumptions for modeling RES systems
that have been shown to overestimate the performance of real systems (Clack et al., 2017;
Moriarty and Honnery, 2016; Trainer, 2017a, 2013, 2012, 2010). For example, for wind,
Arvesen and Hertwich (2012) concluded that “there appears to be a general tendency of wind
power LCAs to assume higher capacity factors than current averages from real-world
experiences”. Boccard (2009) found that, despite the capacity factor of wind power usually
being assumed to be in the 3035% range of the name plate capacity, the mean realized value
for Europe between 2003 and 2007 was below 21% (findings consistent through the period
2000-2014 (IRENA db, 2017)).
The energy return on energy invested (EROI), estimated for theoretical or particular plants, in
particular for PV, has been contested when compared with the EROI of national RES systems,
with a tendency to lower the expectations (Ferroni and Hopkirk, 2016; Palmer, 2013; Prieto
and Hall, 2013; Weißbach et al., 2013). An ongoing discussion over this important issue is
taking place at present (Ferroni et al., 2017; Raugei et al., 2017, 2015; Weißbach et al., 2014).
However, to our knowledge, no study has to date focused on the real performance of CSP. In
this work, we fill this gap in the literature by estimating the capacity factor, seasonal variations
and EROI values of CSP plants in operation. Ultimately, the aim of the paper is to provide
ground for discussion on the potential contribution of CSP to a 100% RES system.
The capacity factor is a parameter that critically affects the life-cycle analyses that estimate the
energy and material requirements (such as the energy payback time (EPT) and EROI) as well as
the environmental impacts (e.g. global warming potential, acidification, eutrophication, loss of
biodiversity, noise, human and ecosystem toxicity, land requirements, etc.) and economic
costs. For example, a comparison of the real capacity factor at global level of CSP could be
quickly estimated using the aforementioned data for 2015 (IRENA db, 2017): an installation
base of 4.4GW at the end of the year with an annual production of 9TWh; assuming that the
new capacity was added uniformly throughout 2015, this would give a CF = 0.24, which
contrasts with the usually considered values in the literature (range 0.25-0.75, depending on
the technology and geographical location of the plant (Burkhardt et al., 2011; Corona et al.,
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
5
2016, 2014; García-Olivares, 2016; IEA and IRENA, 2013; Klein and Rubin, 2013; Lechón et al.,
2008; Pihl et al., 2012; Turchi, 2010; Viebahn et al., 2011; Weinrebe et al., 1998), see Table 2 in
section 3). This comparison suggests that similar discrepancies between theoretical and real
performance to those existing for other RES might also be present for CSP.
The real performance of CSP is analyzed through the collection of publicly available data from
34 individual CSP power plants in operation in 4 countries (Spain, the USA, India and United
Arab Emirates (UAE)) (IRENA db, 2017), which amounts to >40% of the total CSP power
capacity installed in the world, as of the end of 2016, and the national-aggregated production
of Spain. The obtained results are compared with the values used in the peer-review literature,
the online global list of CSP projects from NREL (2017) and the data provided by the
constructors of the power plants.
In a second stage, annual and monthly electricity production are analyzed for diverse CSP
plants and a new indicator of performance for variable RES is proposed based on Capellán-
Pérez et al., (Capellán-Pérez et al., 2017b) called Sv3, defined as the ratio of the electricity
generation from the worst month in a year versus the average monthly electricity generation
in that same year, therefore, the lower the Sv, the higher the seasonal intermittence. This
performance indicator will be compared with other variable RES for the Spanish and USA
power systems.
In a third stage, we re-estimate the EROI of different CSP power plants studied in the
literature, taking into account the real capacity factors previously found and recalculating the
total Energy Used (EnUtot, the total energy used in the construction, operation and disposal of
the CSP system), taking on board three key factors usually not considered in the literature in
enough detail: (1) including most materials involved in the construction and operation of CSP
plants; (2) the use of particular values of embodied energy in materials (MJ/kg) used by CSP
and not common to other RES technologies; and (3) considering CSP technologies using
abundant materials. Finally, in the light of the obtained results, the potential contribution of
CSP in 100% RES systems is discussed.
The paper is organized as follows. Section 2 overviews the CSP technologies, while sections 3
to 5 review the performance factors of real CSP power plants in operation: capacity factor
(section 3), monthly and seasonal variability and intermittence (section 4) and EROI (section 5).
Finally, section 6 discusses the implications of the results.
2. Concentrated solar power plant technologies
CSP is an electricity generation technology that uses heat provided by solar irradiation
concentrated on a small area. Using mirrors, sunlight is reflected to a receiver where heat is
collected by a thermal energy carrier (primary circuit), and subsequently used directly (in the
case of water/steam), or via a secondary circuit to power a turbine and generate electricity. At
present, there are four available CSP technologies: parabolic trough collector, solar power
tower, linear Fresnel reflector and parabolic dish systems (Zhang et al., 2013). In this study, we
refer to them as Parabolic, Tower, Fresnel and Dish technologies, respectively.
The CSP performance can be enhanced by the incorporation of two complementary
technologies: Thermal energy storage (TES) and backup systems. Storage avoids losing the
daytime surplus energy while extending the production after sunset. TES collects the excess
3 In Capellán-Pérez et al. (2017), Sv is called Seasonal Variation, that could be a rather confusing term;
here we use this performance factor in this sense: the more seasonal variation the lower Sv, or
inversely, the more approaching Sv=1 the lesser seasonal variation of electricity production.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
6
heat in the solar field and sends it to a heat exchanger, which warms the heat transfer fluid
going from the cold tank to the hot tank. When needed, the heat from the hot tank can be
returned to the heat transfer fluid and sent to the steam generator. A fuel backup system
(typically based on natural gas) helps to regulate production and/or guarantee a desired
generation capacity, especially in demand peak periods. CSP plants equipped with backup
systems that produce electricity are called hybrid plants. For more information on CSP
technologies see Zhang et al., (2013).
CSP plants produce electricity from a thermal process that can be supported by non-solar
sources. Those plants that use natural gas as back-up can use it to preheat the thermal fluid or
to maintain the heat of the molten salts or the material that can be used as storage; it can also
use the natural gas to produce electricity. Some CSP projects propose to replace the use of
natural gas with biomass so that they can be coherently classified as renewable sources (e.g.,
Borges Termosolar (Lleida, Spain) uses hybridization with biomass of forest residues and
natural gas). The hybridization with natural gas increases efficiency and therefore the capacity
factor (CF), defined here as the ratio of electric power supplied on average in a year by the
power plant and the nominal capacity of the plant that we will take as the gross power of the
turbine. The back-up with natural gas, and especially the use of storage, allows electricity to be
generated relatively independently of the instantaneous solar radiation.
Since we intend to characterize CSP plants as renewable in a context of future scenarios of
100% RES, the output energy of the plant that we consider will be the net electric power
produced from the solar field. In the case of hybrid plants with natural gas, if this produces
electricity, the share of the production coming from the natural gas is deducted. In the event
that natural gas is used to support the storage or to maintain the heat of the thermal fluid,
that natural gas will be accounted for as self-consumption of the plant. The fact that many CSP
plants use natural gas for the preheating of thermal fluid instead of electricity is somewhat
paradoxical, if we take into account the fact that the plant produces electricity and is therefore
connected to an electrical grid from which it could take that energy. However, these plants
generally prefer to build a parallel gas network (sometimes kilometers away from a natural gas
source) with all the energy and material costs involved.
As the solar efficiency of the hybrid system increases, it becomes more difficult to quantify the
contribution of the support system, so the net electricity produced that we consider will be
greater than what a pure renewable system would generate; therefore, we will probably be
conservative/optimists in some of the estimates.
3. Capacity factor (CF)
Table 2 shows the values of the capacity factor estimated from real production data (“Real CF
column) for different technologies and individual CSP plants from 4 countries (Spain, the USA,
India and UEA) as well as for the complete CSP system of the USA and Spain. Subsequently, the
obtained values are compared with: (1) the range values published in the scientific literature
(column “Literature CF”), and (2) the foreseen CF by constructors (“Expected CF” column) as
found in the NREL’s list of CSP plants (NREL, 2017) from the announced electricity production
and the gross power of the plant or the construction companieswebsites or projects.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
7
Tecnology
Storage
Expected CF
Literature CF
Real CF
Parabolic,
Tower
0.20
Parabolic
0.5h
0.2
0.42-0.51
0.18
Parabolic
6h
0.38
0.27
Parabolic
no
0.26
0.28
Parabolic
no
0.24
0.16
Parabolic
no
0.24
0.25-0.5
0.21
Parabolic
no
0.17
Parabolic
no
0.17
0.18
Tower
10h
0.52
0.55-0.71
0.14
Tower
no
0.31
0.25-0.28
0.19
Tower
no
0.02
Dish Stirling
no
0.25-0.28
0,19
Parabolic
no
0.24
0.25-0.9
0.20
Parabolic,
Tower,
Fresnel
0.25
parabolic
7.5
0.374
0.30
parabolic
7.5
0.4
0.42-0.51
0.32
parabolic
no
0.228
0.25-0.5
0.16
fresnel
0.5h
0.185
0.22-0.24
0.15
parabolic
no
0.27
0.25-0.9
0.19
-
0-10h
0.2-0.5
0.25-0.75
0.15-0.30
Table 2: Estimates of the CF of several individual CSP plants, sets of plants and global USA and
Spanish CSP systems: expected values from the industry, values used in the scientific literature
and the results obtained in the work for real plants. The type of technology has been indicated
as well as if plants have storage (molten salts).
The “Real CFcolumn is the one calculated for the net solar production: for the USA Total, we
take the period Oct 2016-Sep 2017 (the last year of data from EIA (US EIA db, 2018)) (plants
that at present are not in operation are excluded, also the Still Water plant due to lack of data
for the year 2017 as of January 2018). For Maricopa, we take 11 months of 2010; this
demonstration plant was decommissioned in 2011. For SHAMS, we take the average of 2014,
2015 and 2016 (Alobaidli et al., 2017; Sanz, 2017). For the Sierra plant (closed at present), we
take the average over its (short) life time. For the Godawari plant in India, we take the monthly
data from April 2015 to March 2016 (Solanki, 2016). For the Spanish CSP system, the
calculation refers to the year 2017, except December (REE, 2018). For the individual Spanish
plants, data are for 2016 from the Ministry of Energy (Ministerio de Energía, 2018). All data are
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
8
net solar production, which for hybrid plants is estimated by deducting 4.4%4 of the electricity
production attributable to natural gas, in accordance with Spanish national legislation
(Ministerio de Industria, Energía y Turismo, 2014), as well as assuming 10% self-consumption.
The column of Literature CFrefers to the range that has been found in the scientific
literature. For the parabolic technology with storage (Burkhardt et al., 2011; Corona et al.,
2014; Lechón et al., 2008; Pihl et al., 2012; Turchi, 2010; Viebahn et al., 2011), for the parabolic
technology without storage (IEA and IRENA, 2013; Klein and Rubin, 2013; Weinrebe et al.,
1998), for Tower technology with storage (Corona et al., 2016; IEA and IRENA, 2013; Lechón et
al., 2008; Pihl et al., 2012; Viebahn et al., 2011), for Tower technology without storage (IEA and
IRENA, 2013), for Fresnel technology (IEA and IRENA, 2013), and for Dish Stirling technology
(García-Olivares, 2016; IEA and IRENA, 2013). For the UEA SHAMS plant and the Godawari
India plant, García-Olivares (García-Olivares, 2016) considers 0.75 in subtropical deserts and
quotes Trieb (2006) who gives 0.9 as possible for these latitudes.
Table 2 shows that the CF of real plants currently in operation is in the range of 0.15-0.3 and
this is a lower value than those expected by the industry (0.2-0.5) or those usually used in the
academic literature (0.25-0.75). In general, the CF of the Spanish plants is better than that of
the USA ones (despite a lower average solar irradiance in Spain). This may be due to the fact
that Spanish plants are usually hybridized with natural gas and the method applied to estimate
the net renewable electricity produced is probably underestimating the natural gas
contribution.
In the light of the obtained results, the current average CF level of CSP plants in the present
electricity system (with low penetration of RES variables) is around 0.2 for plants without
storage and 0.25 for plants with storage.
4. Monthly/seasonal variability
CSP with storage and/or hybridization can partially avoid the problems associated with the
hourly/daily (short-term) variability of other variable renewable sources, such as wind or PV. In
order to investigate the scale of medium and long-term variability, the monthly output of
some real CSP plants currently in operation is reported and compared with the variability of
other intermittent RES.
Figure 1 shows the monthly electricity generation of the Genesis solar plant (USA) from
November 2013 to September 2017, with large fluctuations between summer and winter. The
same pattern is identified for the 7 plants SEGSIII-IX (Figure 2). In the latter, the fact that the
back-up power from natural gas is mainly used in high productivity months, i.e., exacerbating
the seasonal variability, is also visible.
4 4.4% = 15%·0.2907, 15% being the legal maximum of primary energy to be supplied by gas, and 0.2907
the efficiency factor of gas combustion. Assuming that the maximum has been reached, this is
reasonable, given that most CSP plants in Spain have been penalized in the past for surpassing the 15%
level (CNMC, 2016).
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
9
Figure 1: Monthly production (data from the EIA) of the Genesis plant. It is a plant without
storage and without hybridization, although it consumes approximately 2% of what it produces
with natural gas that serves to keep the thermal fluid hot. This natural gas not only prevents
the possibility of that fluid solidifying, but also helps to activate the electrical production in the
first hours of the morning. Note that, between monthly minimums and maximums, there may
be a difference factor of more than 5.
We apply the performance indicator Sv to evaluate the variations of output along the year.
This indicator was defined in Capellán-Pérez et al., (Capellán-Pérez et al., 2017b) at a
theoretical country-level, and here it is applied at plant-level. Sv is a sensitive indicator to be
taken into account when assessing a hypothetical mix of high penetration of renewable
energies, as it gives an idea of the fluctuations that must be dealt by the whole system when it
is lower than 1. In particular, the Sv can be used to estimate the overcapacity required to deal
with seasonal variability.
0
20,000
40,000
60,000
80,000
100,000
nov-13
Jan 2014
mar-14
may-14
jul-14
sep-14
nov-14
Jan 2015
mar-15
may-15
jul-15
sep-15
nov-15
Jan 2016
mar-16
may-16
jul-16
sep-16
nov-16
Jan 2017
mar-17
may-17
jul-17
sep-17
MWh
Monthly Genesis solar plant production
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
10
Figure 3: Seasonal variability of PV, CSP, Hydro and Wind for electricity production in Spain
(2014-2017): (a) Monthly electricity production per installed power and (b) Sv. Own work from
(REE, 2018).
Figure 3 shows that the monthly/seasonal (long-term) variability of CSP in Spain is much higher
than for other RES (PV, wind, hydro). In the whole of Spain -which has 50 CSP plants and since
2013 is without new facilities- the Sv is <0.2, which is much lower than for photovoltaic (Sv ~
0.55), wind (Sv = 0.5-0.72) and hydro (Sv=0.5-0.61), also with small capacity additions in the
0
0.1
0.2
0.3
0.4
MWh/MW
PV CSP Hydro Wind
Spanish monthly electricity production per installed power
0.00
0.20
0.40
0.60
0.80
Dmnl
PV CSP Hydro Wind
Sv for different RES technologies in Spain
a
b
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
11
last few years (Figure 3). For the USA, from December 2016 to November 2017, Sv was 0.33 for
CSP as against 0.69, 0.62 and 0.51 for Hydro, Wind and PV, respectively (data elaborated from
EIA: https://www.eia.gov/electricity/monthly/). Sv data for the USA must be taken with
precaution for PV and wind, because installed power continues to grow at an important pace.
Table 3 collates the Sv for some individual CSP plants from the USA, Spain, UAE and India. We
can roughly estimate Sv = 0.27 (the average of Spain and the USA, see Table 3) for plants in
latitudes such as those of Spain and the USA, and Sv = 0.66 for plants in low latitudes (average
of UEA and India). Solar Power Tower type plants seem to have a better Sv5 (although the CF is
not improved), this and the better average radiation of the USA plants may be the reason that
the USA average is better than the Spanish one. Given that, currently, more than 80% of the
installed power belongs to Spain and the USA, the world average Sv at present is <0.3.
Assigning 2/3 of the future potential of CSP in scenarios of high penetration of RES to areas of
low latitude (high irradiance) and 1/3 to the rest, we would have an estimated Sv of around 0.5
for these scenarios.
Plant or system
Sv
The USA
0.33
SEGS III-IX
0.03
Nevada Solar One
0.20
Mohave
0.12
Ivanpah 1
0.55
Ivanpah 2
0.49
Ivanpah 3
0.57
Genesis
0.20
Solana
0.37
Spain
0.20
Enerstar
0.21
Puerto Herrado 1,2
0.25
SHAMS (UAE)
0.77
Godawari (India)
0.55
Table 3. Calculated values of Sv. Sources: see references in text to build Table 2.
The electricity production variations in Spain comparing daily productions, instead of monthly,
can be enormous. On the best days of the year (generally near the summer solstice, e.g.
22/06/2016, see Figure 4a), the average production can exceed 1,400MWe of instant power,
reach the maximum between 10:30 and 20:30 and, with the help of storage and natural gas
back-up, this is maintained for the rest of the day until 5:30 a.m. at 700MWe (1,000MW
installed of the total of 2,300MW have storage of 7.5 hours or more); from 5:30 to 8:00, it
decreases to 266MWe (here the gas necessarily intervenes, otherwise it would fall to zero).
This pattern may occur for several days (no clouds over Spain).
On the other hand, the production of some several days on a row of December and January is
practically zero (e.g. 25, 26, 27, 28 and 29 December 2017) (see Figure 4b). The average
5 Towers use a double tracking solar technology against the one tracking used by Parabolic plants.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
12
production for those days (coinciding with the deep storm "Bruno" that entered Spain at that
time) was about 18MWe (80 times lower than in the best days), with an equivalent CF =
0.0078. The storage of > 10 hours usually proposed to deal with hourly/daily variability, would
increase the variability on the monthly/seasonal level, since on the cloudy days of winter the
production would be practically null, while on the best days of summer, hourly storage would
increase the average production (average CF increases with TES). In other words: hourly/daily
storage exacerbates seasonal variability.
Figure 4: Instantaneous power generated by the 50 solar thermal plants in Spain on
12/06/2016 (a) and 25-29/12/2017 (b). On days 28 and 29 the production does not fall to zero
during the night due to the hybridization of some plants with natural gas. Elaborated from the
figures generated on (REE, 2018).
Section 6 includes a discussion of the implications of this seasonal variability for the CF of CSP
plants in the context of scenarios of high RES penetration.
0
500
1000
1500
2000
2500
0 5 10 15 20
Electricity production of CSP in Spain (MWe)
Hour of the day (06/12/2016)
a
0
20
40
60
80
100
120
140
160
180
200
0,00 1,00 2,00 3,00 4,00 5,00
Electricity production of CSP in Spain (MWe)
Days after 12/25/2017
b
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
13
5. EROI
This section focuses on the estimation of EROI for CSP plants. Section 5.1 reviews the different
EROI expressions used in the literature and justifies the expression applied in the analysis.
Section 5.2 focuses on the estimation of Energy Used (EnU), taking as starting point previous
works, which are complemented with literature review. Section 5.3 discusses which g factor
(quality of electricity) to be applied, and finally section 5.4 presents the obtained results.
To represent the implications for EROI of different levels of deployment of CSP, three scenarios
are taken into account: scenario 1 considers current situation (reduced deployment near
points of consumption and overcapacities and/or outside storage required to deal with
intermittencies not considered), while Scenarios 2 and 3 refer to future scenarios with large
scale deployment of CSP plants in hot deserts characterized by high irradiance, high winds,
etc., (i.e., conditions similar to those of SHAMS 1 of UEA) (García-Olivares, 2016) . Both
scenario 2 and 3 consider overcapacity requirements to deal with seasonal intermittency, the
use of common materials and higher distribution losses than for scenario 1. While scenario 2
refers to regional distribution scenario 3 refers to international distribution (hence higher
distribution losses for scenario 3 than for scenario 2) and less conservative embodied energies
of some materials.
5.1. EROI expression
Until the present in the literature, there have been few studies specifically calculating the EROI
of the CSP (Weißbach et al., 2013). However, there are several works that work with Life Cycle
Assessment (LCA), in which the Cumulated Energy Demand” (CED) or Cumulated Exergy
Demand (CExD) is calculated and, from them, the Energy Payback Time (EPT or EPBT), which is
the time measured in months or years in which the plant generates as much electrical energy
as the electrical equivalent of the primary energy consumed.
CED is a term with origin in the LCA community. But there, CED is defined including all the
primary energy harvested in the operation phase (in our case, the solar radiation over the CSP
mirrors), that has no sense to calculate EPBT or EROI from CED and, therefore, is excluded of
CED estimations from EPBT and EROI literature. To avoid confusion of the different “CEDs”
being used in the literature, and given priority to the historical precedence to the CED defined
by LCA community, we change the term to EnU (Energy Used) instead of CED when the
purpose is to estimate EPBT or EROI.
According to different authors, Energy Payback Time is defined differently for CSP plants:
 =


(eq. 1)
(e.g. (Corona et al., 2014; Krishnamurthy and Banerjee, 2012; Lechón et al., 2008; Viebahn,
2013; Weißbach et al., 2013), where Enet is the yearly net electricity output (MJ/year), EnUc is
the energy usedin its mineral extraction, manufacturing, construction and dismantling of the
CSP plant (MJ), EnUo is the energy used associated with the operation and maintenance
(MJ/yr), and g is a quality factor that compares the electricity generated with the primary
energy consumed in the EnU. Note that the EnUs are given here in units of energy (MJ) and not
in the ratios (MJ/KWh) that we will use in the next section that would be the EnUtot/Enet.
Although exergy do not capture all the irreversibilities, if energy quality is taken into
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
14
consideration it will be better to use exergy and not primary energy (Weißbach et al., 2013),
but there are few LCAs that use exergy (Ehtiwesh et al., 2016).
The other usual definition of the Energy Payback Time (e.g. (Burkhardt et al., 2011; Heath et
al., 2011; Raugei et al., 2015) is:
 =
/
(eq. 2)
where EnUtot is the sum of all primary energy (or exergy) supplied by sources across the LCA
of the CSP plant and EnUtot = EnUc+EnUo·Life time.
Note that both definitions are different and give rise to different values of the Energy Payback
Time.
Weisbach et al ., (Weißbach et al., 2013) proposes using the LCA methodology calculated by
the EnU to estimate the EROI of different energy technologies. From the EPBT, the relationship
would be established:
 =  
 =  ·
 ·
(eq. 3)
In this paper, we propose not to use the EPT of (eq.1), as it can lead to physically impossible
results. EPT or EPBT must be defined as positive (as well as the EROI). However, a system using
more energy in operation and maintenance than the energy it provides (e.g. EnUo > Enet/g) is
physically possible (although without much economic sense) but would give negative results
according to eq. 1. Therefore, equation 1 should be discarded.
5.2. Energy Used by CSP system (EnUtot)
The literature usually estimates the energy requirements to build, maintain and dismantle a
CSP plant and uses the so-called Cumulative Energy Demand (CED) (here EnU) and, more
recently, the Cumulative Exergy Demand (CExD) (Ehtiwesh et al., 2016) (here ExU). For this
calculation, a list of the minerals or materials necessary during the lifetime assigned to the
plant is normally made and each of them is assigned the embodied energy intensity (MJ/kg),
according to the Life Cycle Assessment (LCA) methodology. This energy is usually given in units
related to the electrical production of the plant: MJ/KWh, which generates some confusion,
because different authors assign different values to the electrical production of similar plants,
which makes inter-comparability difficult. In addition, as we have analyzed (see section 3), all
the reviewed studies use an overestimated CF, which therefore causes the overestimation of
the electrical production that the plant will produce during its life-time. Table 4 shows the EnU
values found in the literature (“EnUpublished”) as well as their correction, considering instead
a CF = 0.25 (“EnUCED corrected”). When we use the term EnU without quotation marks we
refer to cumulative energy demand in MJ or other energy dimensional unit. When we use the
term “EnU, we refer to MJ/KWh as is often used in the literature (although has no energy
units).
Asdrubali et al., (2015) found a wide span in the values of EnUpublished in the literature,
over 1 magnitude order of difference between the lower (0.2 MJ/kWh) and the higher (2.8
MJ/kWh) ranges; recent studies fit with this wide range (see Table 4).
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
15
Autor/technology
“EnU”
published
“EnU” corrected
(CF=0.25)
(Corona et al., 2016) hybrid Tower
1.337
2.19
(Burkhardt et al., 2011; Heath et al., 2011)
Parabolic
0.40-0.43
0.75-0.82
(Lechón et al., 2008) Parabolic-Tower
2.45-2.79
4.27-7.92
(Ehtiwesh et al., 2016) Parabolic
0.198
0.792
(Corona et al., 2014) Parabolic hybrid
1.15-3.20
1.47-4.81
(Asdrubali et al., 2015) (review range)
0.16-2.78
Table 4: “Energy Used” (“EnU”) of CSP plants in MJ/kWh found in the literature and its value if
the CF were 0.25 instead of the one assumed in their theoretical works. The data of Asdrubali
et al. (2015) are the extremes found in his review. The range for Corona et al. (2014) refers to
different grades of hybridization with natural gas. The range of Burkhardt et al. 2011 refers to
wet and dry technologies, respectively. The Lechón et al. (2008) data are for Parabolic and
Tower, respectively. Note that, when corrected, the Tower “EnU” is much greater than
Parabolic. Ehtiwesh et al. (2016) data refers to Exergy (destroyed useful energy), which results
in 20% more than the same calculation performed with Energy. Thus, in the case of using
exergy which in our opinion is more consistent with the EROI calculations that weigh the
quality of the energy source- and if the share were maintained, the rest of the values would be
multiplied by approximately 1.2.
The re-estimation of EnU”, taking into account real values for CF, increases the range to 0.8-
7.9 MJ/kWh, i.e., an increase of 4x for the lower bound and almost 3 times for the upper
bound of the range. Thus, the consideration of real values for CF is likely to affect the EPBT and
EROI values previously published in the literature.
In this section, the EnU of a standard type of CSP plant is estimated for two cases (1) current
plants (section 5.2.2.) and (2) plants in the context of high RES penetration scenarios (section
5.2.3). The applied methodology includes the review of previously published works, industry
data and LCA databases. Previous works do not always take into account the same materials
and energy costs associated with the LCA of the plant, but they can be combined, especially
when some materials that others do take into account are missing in their calculations. Thus,
Pihl et al., (2012) disaggregates mainly at the mineral level in detail and other authors
disaggregate at the level of more elaborate materials (plastics, steels, etc.). In addition, as we
will argue, some energy intensities of the materials (MJ/kg) that have been taken are different
from the reality, or the reality that we would expect in a future of high penetration of
renewable energy sources. That is why we have estimated the value of the EnU from the set of
authors reflected in Table 4, for the material requirements per installed MW of a parabolic-
type plant of 50 MW with TES, analogous to Andasol (Granada, Spain) or La Africana (also in
Spain). Although Tower technology has a better Sv than Parabolic, we chose Parabolic with TES
because it is the most proven technology at a commercial level. With a similar or better CF
than other technologies, it is the most used for EnU and EROI estimations in the literature and
probably has better EROI than other technologies (Lechón et al., (Lechón et al., 2008) using the
same methodology, as both technologies show a slightly better EnU for Parabolic than for
Tower. Since it considers a theoretical CF greater for the Tower, if the real CF is similar, this
gives a much worse EnU for Tower (see Table 4)). The estimates are performed for three
different scenarios, depending on the potential deployment level and geographical location of
CSP over the next few decades.
Since aluminum is a much more abundant material than silver, thinking of scenarios of high
solar penetration and trying to replace the more than foreseeable shortage of silver, there are
proposals for scenarios of high penetration of RES with common materials (García-Olivares,
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
16
2016; García-Olivares et al., 2012). In this sense, ReflecTech, (2012) compares two types of
parabolic mirrors, one classic (flat glass coated), based on silver (embodied energy of
35.24MJ/kg), with one based on recycled aluminium (embodied energy 24.29MJ/kg). However,
in reality virgin ultrapure aluminium is used instead of Al recycled in the fabrication of mirrors
given its higher purity which implies a better reflectivity (Vargel, 2004). Thus, in this work the
embodied energy of the mirrors is re-calculated taking (ReflecTech, 2012) as a starting point
but using instead data for virgin aluminium, obtaining 85.5MJ/kg if considering the value used
by (Keough, 2011) data. The reflectivity is very dependent of purity: ultra-pure 99.99%
aluminium has an 85% reflectivity versus 75% of aluminium with 99.6% purity (Vargel, 2004).
The difference between the reflectivity of silver versus ultrapure aluminum is almost 12%,
which would be directly reflected in the overall efficiency of the plants, while the embodied
energies associated with the mirrors would be higher.
On the other hand, Ehtiwesh et al., (2016) consider 21.05MJ/kg for the energy intensity of the
"molten salt" of the TES technology. However, if these salts are taken from the synthesis of
ammonia and urea, which in turn come from natural gas, de Castro et al., ( 2013) reasoned
that a strong scaling of this technology would exceed the reserves of the mines and that to
synthesize it, more than 50MJ/kg would be required, since only the synthesis of urea and
ammonia, from which these salts would be made, requires 40-50MJ/kg. Heath et al., (2011)
considers that the associated emissions of CO2 (in one LCA) from the synthesis of salts versus
those from mines would multiply these by a factor of almost 5; therefore, if we use the
associated emissions as a proxy for embodied energy, the synthesis could require about
100MJ/kg.
Also, some authors who gave low EnUs” in Table 4 (e.g. (Ehtiwesh et al., 2016)) do not take
into account site preparation (removed lands, access roads, fences, waste ponds, retaining
walls, etc.). For example, Turchi (2010), in an economic project, does take into account all
these preparations, but takes a foot (0.3048m) of earth removed in the entire occupation of
the plant. If we took 2Mm3 of land removed, as was done in La Africana of 50MW (La Africana,
2018) and 0.45MJ/Kg (from Hammond and Jones (2008)), it gives us a very conservative
measure of energy for site preparation.
Heath et al., (2009) also consider materials that have little "weight", such as glass wool,
refractory glass, calcium silicate and "small" machines, such as pumps, that would increase
energy consumption by 2.5% over the total without taking into account this type of material.
Reviewed studies do not correctly take into account the energy costs of the over-sizing of
current lines, roads, fiberglass cabling, fences, natural gas conduction lines, etc., that leave the
plant and connect them with the rest of the energy system of the country or region. These
plants are located in deserts or semi-deserts, generally relatively close to towns and cities.
However, as these areas are filled, the distances will be greater and far from the centers of
consumption. So Kuenlin et al., (2013) calculate that the impact of the trans-regional lines,
necessary for scenarios of high global penetration of renewables, can exceed 15% of the costs
of the plant, which would surely be reflected in the embodied energies of these high and
medium voltage lines. The Mohave project (Douglas et al., 2010) needs more than 150
kilometers of fiber optic cables to stabilize the electricity network, 32 monopoles of about 32
meters high, new paved roads, a fence of more than 2 meters along its 750 Ha of occupation,
2500m2 of waste treatment ponds with a base of 50cm of compacted silt surrounded by 60 cm
high cement, 20 buildings for workers outside the solar field systems and the thermal power
plant (control units, assembly factory of the modules, etc.). The 66 permanent workers of this
250MW plant have to travel several miles a day to go to the nearest towns where the
necessary social infrastructures exist. Colonizing deserts requires greater energy effort than
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
17
other ecosystems (hence the density of population is always small). Turchi (2010) makes an
economic study of the civil works, in addition to the industrial one, of a 103MW park (similar
to that of Mohave), finding that operation and maintenance (O&M) have an economic cost of
30% of the construction, a number which could be used as a proxy to obtain the energy
requirements of the O&M. If we consider the use of materials in the construction and
maintenance phases, Pihl et al., (2012) gives some requirements in the maintenance phase of
approximately 20% of associated embodied energy taking 25 years of plant life.
The data we have used for the materials embodied energies (MJ/Kg) comes, in general, from
Hammond and Jones (2008), which uses the LCA criterion of Cradle to Gate (some of Cradle to
site) for semi-fabricated components: sections, sheets, rods, etc., which go to the construction.
To quote that article: "Highly fabricated and intricate items are beyond this report". Therefore,
our methodology is probably conservative, compensating for the possible future improvement
in the efficiency increase of the embodied energies.
5.2.1. Scenarios
We distinguish 3 scenarios for an Andasol or La Africana standard plant of 50MW: Scenario 1
represents the current situation (reduced deployment near points of consumption), while
Scenarios 2 and 3 refer to future scenarios with large scale deployment of CSP plants in hot
deserts characterized by high irradiance, high winds, etc., (i.e., conditions similar to those of
SHAMS 1 of UEA) (García-Olivares, 2016). Scenarios 2 and 3 take into account the use of
aluminum mirrors instead of silver (as in Scenario 1), given the potential scarcity of the latter in
large-scale deployment scenarios of CSP (de Castro et al., 2013; García-Olivares, 2016; García-
Olivares et al., 2012), as well as the loss of reflectivity of the aluminum mirrors relative to the
silver mirrors and the damage suffered by mirrors due to severe winds in deserts. The three
scenarios assume a life time of the plant of 25 years and a CF of 0.25. In particular:
- Scenario 1 considers molten salts from mines, silver mirrors and a regional distribution
of electricity (embodied energy and losses of 5%) as the current plants with better CF.
As mentioned above, the calculations will be conservative in terms of the total EnU
requirements.
- Scenario 2 considers molten salts from the synthesis of urea with an embodied energy
of 50MJ/kg, ultrapure virgin aluminum mirrors (85.5MJ/Kg) and the same regional
distribution of losses in the electrical network as in Scenario 1.
- Scenario 3 considers molten salts from the synthesis of urea with an embodied energy
of 100MJ/kg, an international distribution of electricity from deserts to points of high
consumption, with embodied energy and losses of 15% of the total EnU found (García-
Olivares, 2016; Trieb, 2006).
Section 5.2.2. reports the results obtained for the current CSP plants, while section 5.2.3.
reports the results obtained considering the need for overcapacities in high RES penetration
scenarios.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
18
5.2.2. Current EnU of CSP plant
Table 5 reports the contribution by phase to the EnUfor each scenario. Results are given in
(MJ/kWh) per MW installed in order to compare our results with the literature (Table 4).
“EnU” (MJ/KWh)
Scenario 1
Scenario 2
Scenario 3
Source
steels
0.37
0.37
0.37
(Ehtiwesh et al.,
2016)
concrete
0.02
0.02
0.02
(Ehtiwesh et al.,
2016)
plastics
0.10
0.10
0.10
(Montgomery, 2009)
Syntetic oil
0.17
0.17
0.17
(Hammond and
Jones, 2008; Pihl et
al., 2012)
molten salts
0.20
0.47
0.94
(Ehtiwesh et al.,
2016) corrected
Ag based mirror
0.08
0.00
0.00
(Ehtiwesh et al.,
2016; ReflecTech,
2012)
Al based mirror
0.00
0.26
0.26
(Keough, 2011;
ReflecTech, 2012)
site preparation
0.49
0.49
0.49
(La Africana, 2018)
other material and
machineries
0.03
0.03
0.03
(Heath et al., 2009)
broken mirrors
0.00
0.03
0.07
(Radan, 2016)
water (distilled 0.2MJ/kg)
0.13
0.13
0.13
(Hammond and
Jones, 2008, p. 20;
Kuenlin et al., 2013;
Turchi, 2010)
Cu, Mg and other metals
0.02
0.02
0.02
(Hammond and
Jones, 2008; Pihl et
al., 2012)
rock
0.02
0.02
0.02
(Hammond and
Jones, 2008; Pihl et
al., 2012)
Operation phase
0.19
0.19
0.19
(Hammond and
Jones, 2008; Pihl et
al., 2012)
Dismantling and disposal
0.05
0.06
0.07
(Burkhardt et al.,
2011; Heath et al.,
2011)
Dry cooled performance
0.14
0.18
0.22
(Heath et al., 2011)
Al mirror reflectivity loss
0.00
0.36
0.43
(García-Olivares
2016)
Grid needs and losses
0.18
0.18
0.54
(García-Olivares,
2016; Trieb, 2006)
“EnU”tot
2.18
3.07
4.06
Table 5: Energy Used (EnU) for materials
The EnUlevels obtained for scenarios 1 and 2 are in the high bound of the literature review
(Table 4) and over the high bound for scenario 3. The latter is consistent with the fact that
scenario 3 is assessing CSP in conditions which have not been studied in other studies.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
19
In relation to the contribution to the EnU for each phase/material processing varies
depending on the scenario (see table 5 and the detailed results in Appendix A):
For both scenarios 1 and 2, site preparation, steels and molten salts account for ~45-50% of
the EnU, followed by the energy requirements associated to the operation phase, grid needs
and losses, and in the case of scenario 2, the Al based mirrors and Al mirror reflectivity loss.
For scenario 3, molten salts, grid necessities and site preparation represent ~50% of the EnU,
followed by Al mirror reflectivity loss, steels, Al based mirror and Dry cooled performance
(together accounting for ~80% of the total EnU).
The issues associated with intermittency, need of back up etc., and the associated embodied
energies, are not considered in Table 5. These additional requirements could be currently low
in most countries, given the relatively low penetration of intermittent RES, although relatively
to the own intermittent RES penetration this costs could be important (e.g. intermittent RES
impose some adaptation of the grid system, therefore, some investment and some energy
cost). In any case, in high RES scenarios, it is necessary to take them into consideration. The
next section is dedicated to estimating the EnU under these conditions.
5.2.3. EuN of CSP plants in high RES penetration scenarios
Apart from hydro pumping storage (PHS), storage systems to compensate for the seasonal
variations are not yet available and alternative technologies of large-scale storage are still in
the R&D phase (Wagner, 2014). Thus, apart from flexible demand management, overcapacity
is probably the best strategy to deal with seasonal and annual variation. As mentioned in
section 4, the Sv can be used to estimate the required overcapacity to deal with seasonal
variability in systems with high penetration of solar technologies. The criterion of Capellán-
Pérez et al., (Capellán-Pérez et al., 2017b) to ensure that the month of lowest irradiance meets
the average annual demand for electricity required, allows the difference between the
irradiance of winter against that of summer to be taken into account. In accordance with this
criteria, the inverse of Sv is the overcapacity (f) to be added to a mix with strong variable RES
penetration (see eq. 3 in Capellán-Pérez et al., (Capellán-Pérez et al., 2017b)), then the CF at
the plant level is:
 =12·
 (eq. 4)
Ave being the monthly average power production and Pow the gross nominal power of the
plant or present system. If overcapacity is required for seasonal variation, the effective CF
(CF,eff) at the system level with strong RES penetration will be:
, =12·
·= · =12·
 (eq. 5)
Where Lowest is the power production in the lowest monthly production of an average annual
electricity production.
In a future with a high penetration mix of RES with strong penetration of seasonal variables,
RES, CF,eff will be an underestimation of the effective CF of the CSP system, if we consider that
no other intermittencies are taken into account (see below). This could be an overestimation
only if the ratio of strong seasonal variables RES over non seasonal variables is low (Capellán-
Pérez et al., 2017b). However, all analyzed high penetration RES scenarios from Table 1, and
other assessed high RES penetration scenarios, have a very high use of wind and solar relative
to biomass and geothermal power (with an Sv that could be close to one, if so desired, because
monthly production does not depend on geographical latitude or climatology).
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
20
Thus, Lenzen et al., (2016) goes from a CF = 0.6 for a CSP plant to a CF = 0.35 for the CSP
system when the electrical system is 100% RES. Delucchi and Jacobson ( 2011) give an initial CF
of 0.31 for plants, but when they build their 100% RES model in Jacobson and Delucchi (2011),
they reach a final CF = 0.18 for the CSP system. In both studies, the CF is reduced to 60% of its
initial value. If we take equation 5 to compare, these values would be equivalent to Sv = 0.6.
To estimate them in high RES penetration scenarios, other methodologies have been
considered for the calculations of the embodied energy of the materials used for construction,
operation and dismantling (see Table 6).
“EnU” (MJ/KWh)
Scenario 1
Scenario 2
Scenario 3
Materials in construction, operation,
dismantling and disposal, with grids
2.18
3.07
4.06
Overcapacity for intermittence, back
up, etc. (Lenzen and Deluchi models)
1.45
2.05
2.71
Overcapacity for seasonal
intermittence (this work)
2.18
3.07
4.06
EnUtot (Lenzen/Deluchi models)
3.63
5.12
6.77
“EnU”tot (this work)
4.36
6.14
8.12
Table 6: Total “EnU” (Energy Used) estimated in tree scenarios for materials and overcapacity
for two methodologies.
According to our calculations of CF, eff = Sv · CF ~ 0.5 · CF (see eq. 6 and section 4), which
would require twice the infrastructure to give the same CF, eff without the need for
overcapacity. Then the EnUtot would be multiplied by 2. In the same way, taking the CF
reduction factor of the Lenzen et al., (2016) and Jacobson-Delucchi models, the EnUtot would
increase by 67%.
5.3. The g factor (average efficiency in the transformation of primary energy to
electricity and the quality of the energy)
Different authors use different criteria for the value of g. Thus, most authors take g as the
average efficiency in the transformation of primary energy to electricity and this is usually
taken as that of the specific country where the plant is studied. This efficiency depends on the
electricity mix of each country, as well as the evolution over time of the efficiency of different
electricity transforming plants (taking a past or present value tends to give lower values of the
EPBT than if values of the future are taken, where a relative increase in electricity as a final
energy use, or an increase in the efficiency of thermal plants, would give values greater than g
at present). We call this criteria, "primary energy replacement".
Given that there is no term of quality in the classic definition of the EROI and given that, to
assure more consistency, the LCA should use exergy and not primary energy, Weißbach et al.,
(Weißbach et al., 2013) uses the value 1 for the g factor. We call this criteria, "direct electricity
output". Thus, we define an EROIg=1 following the ”direct electricity output” criteria (Weißbach
et al., 2013), taking the destroyed exergy of equal quality as electricity (g = 1). On the other
hand, we call it EROIg=0.456 when following Raugei et al., (2015) reasoning when they criticize
Weisbach's methodology and consider that the EnU must be modified by a factor: “the
average ‘life cycle efficiency’ of the grid(“primary energy replacement” criteria). With data
from 2013 worldwide, we find that the electricity production is 2,232 Mtoe for primary
electricity of 5,111Mtoe; then the factor on a global scale is 0.456 (IEA, 2018) (also available at
https://www.iea.org/Sankey/). Therefore, g = 0.456 and EROIg=1 = EROIg=0.456/0.456
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
21
However, Prieto and Hall (2013) argue that, although electricity actually has more quality as an
energy source than others, on a global scale, only ¼ of the final energy needs are covered by
electricity and the rest by other forms (heat, mechanical ...), and it is not obvious that they
have that factor g but, in the case of heat, it could be greater than 1. For the specific case of
CSP plants, we can suspect that this is effectively the case when they usually hybridize with
natural gas, probably for reasons of economic efficiency, but in this case indicates that, for this
part of the consumption, the factor of "quality" g would be greater than 1, since real plants are
using natural gas to preheat the thermal fluid instead of electricity, despite the fact that the
plant produces electricity and could only consume electricity (not strictly requiring to be
connected to a natural gas pipeline).
Here, we propose and finally use a new criterion following the arguments of Prieto and Hall
(2013), that we will call “final to primary energy” criteria. We call it EROIg=0.687, and that is to
take g by directly comparing the final energy consumed (discounting the non-energy uses that
are mixed in the statistics) with the primary energy (also discounting the proportion that does
not go to energy) provided by the IEA in its Sankey diagram for the World. The result is a g =
0.687.
This also agrees with general global studies which aim to evaluate the substitution of the
current energy system towards a renewable one with a total electrification of the system:
Thus, García-Olivares et al., (García-Olivares et al., 2012) conclude that we would need almost
70% of the primary energy that we consume today in an electric form to provide the same
services.
Therefore, in this work, we consider EROI = EROIg=0,687 = EROIg=1/0.687.
An alternative, beyond the scope of this work but consistent with our criteria here exposed, is
to consider g dynamically, given that the average efficiency in the transformation of primary
energy to electricity will change (increase) during the transition to renewables. However, this
requires a more complex modeling, for example, the approach taken to build the MEDEAS
models (Capellán-Pérez et al., 2017a).
For sensitivity purposes and because it is an ongoing debate, in Appendix B some results are
re-elaborated assuming different values of g (g=1 and g=0.456).
5.4. Estimation of EROI
Hall et al., (2014) propose different calculations for the EROI, distinguishing between EROIst
(standard), EROIpou (point of use) and EROIext (extended). The latter extends the boundaries
of the calculations and is more coherent if the aim is to compare complete systems and not
particular plants. Since complete systems usually lack data, or we must hypothesize what the
complete system would look like in future energy mixes, the EROIext usually starts from
indirect estimates based on associated economic costs and then, through some energy
intensity function (on a national or global scale), calculates the complete energy costs that do
not usually appear in the LCA (e.g. any necessary economic transaction, such as the paid work
of the project engineer, requires an indirect energy consumption associated with the energy
intensity of said economic transaction). In the extended calculations of the EROI, estimates
associated with support infrastructures that a future energy mix based mainly on renewables
would require (e.g. external storage infrastructures or overcapacities) and that today are not
necessary may also be included. Here, we use this last criterion, which in reality would not
generate an EROIext, but a conservative estimation of the EROIpou. Therefore, from the
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
22
EROIext approach, the presented calculations would be very conservative, since EROIst>
EROIpou> EROIext.
A total of 11 studies have been reviewed that report analyses for 15 CSP power plants of
different typologies and functioning under different conditions (Burkhardt et al., 2011; Corona
et al., 2016, 2014; Ehtiwesh et al., 2016; García-Olivares, 2016; Heath et al., 2011; IEA and
IRENA, 2013; Krishnamurthy and Banerjee, 2012; Lechón et al., 2008; Montgomery, 2009;
Viebahn, 2013; Weißbach et al., 2013). Table 7 shows the reported EROIst in the papers
(“Reported EROI” column), if the papers do not report the EROIst, the column shows the
estimates of EROIst that we can deduct based on the reported data. The “Standarized EROIst”
column shows our re-estimation, if possible considering the real values for CF found in section
3 and considering g=0.687.
Reported EROI
Standardized
EROIst
(CF=0.25 &
g=0.687)
Burkhardt et al., 2011
27.7-30
-a
Corona et al., 2014 (hybrid range)
14.1-17.5
1.47-3.41
Ehtiwesh et al., 2016
20.2
7.74
Heath et al., 2011
30
6.97
IEA and IRENA, 2013
60
Krishnamurthy and Banerjee, 2012
7.44-12.2
Lechón et al., 2008 (Tower)
24.6
1.88
Lechón et al., 2008 (Parabolic)
24
2.14
Weißbach et al., 2013(Parabolic)
21
≈1
(if buffered) (EROIpou)
9.6
Viebahn, 2013 (range)
10.9-67.6
5.02
Montgomery, 2009
22
García-Olivares, 2016
18
Table 7: This table gives the calculation of the EROI that the authors give directly or can be
deduced from their estimates of other parameters. Some authors give several technologies or
hypotheses (first column). The second column repeats the calculations using equation 4 where
appropriate, a CF = 0.25 if the technology has storage and CF = 0.2 without storage. The factor
g here is taken as 0.687. a Empty cells refer to cases where it was not possible to recalculate
the EROI.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
23
From the table EnUtot calculated in this study, and taking the value g = 0.687, we would obtain
the following EROI levels for each scenario (Table 8):
EROI
Scenario 1
present
system
scenario 2
future
system
Scenario 3
future
system
EROIst (Materials in construction,
operation, dismantling and disposal
including grids)
2.4
1.7
1.3
EROIpou (present CSP)
<2.4
-
-
EROIpou (system, high penetration
Lenzen and Deluchi models)
1.4
a
1.0
0.8
EROIpou (system, high penetration
this work)
1.2
a
0.85
0.65
Table 8: Conservative (reality probably lower) estimated values of the EROIst (first row) and
the EROIpou (second, third and four rows) for the CSP according to three scenarios and
different methodologies, taking the quality factor of electricity g = 0.687. The EROIpou for the
present system is not estimated (see text). a Supposing the scenario 1 (“present system”: Ag
based mirrors, mined salts, CSP installations near consumer centers) but with the over cost of
high penetration of CSP to deal with intermittences, see text for details of scenarios.
In relation to EROIst, the obtained values are 2.4:1 for scenario 1, 1.7:1 for scenario 2 and 1.3:1
for scenario 3. We only estimate the EROIpou for high penetration escenarios; The EROIpou for
the present system is not estimated, it will add relatively to the EROIst the embodied energy
that the energy system demand to deal with the intermittencies that CSP impose over the
system grids (see figure 4). RES penetration in Spain and USA is concomitant with the lowering
of CF of the quick response of new natural gas power installations, therefore, the electric grid
system is adapting with energetic cost to the present penetration. But, because the RES
variables are not the main sources of electricity, it is very difficult to estimate what it is the
over cost attributable to CSP (for instance the energy cost of PHS) against the rest of energy
sources (most fossil and nuclear fuels). From NREL 2012 model (for USA) one can deduct that
the cost tend to increase with the penetration of RES variables in relative terms (the % of over
cost increase with the % of RES penetration) and that the present cost (with less than 20% of
RES variables in the electricity mix) of storage and overcapacities (e.g. natural gas power
plants) atributable to RES variables are not null. For both methodologies considered
(Lenzen/Delucchi and this work), only Scenario 16 would (barely) have an EROIpou>1:1, while
both Scenarios 2 and 3 would fall into EROIpou<1:1. EROIpou take into account storage needs
at hourly intermittence level (by TES technology) and the overcapacity needed to deal with
seasonal intermittence, but not the storage needed for day/week intermittence (e.g. pumped
hydro storage for several days as reflected in Figure 4b), therefore it is probably conservative.
Tables B1 and B2 in Appendix B re-estimate the obtained values in Tables 7 and 8 respectively
considering alternative values for the g factor in the literature (g=1 and g=0.456). The obtained
EROI levels following g=1 criteria are lower. The recalculation of EROI from other studies of the
literature applying g=1 criteria provide EROI values below the ratio 3:1 (Table B1); while the
EROIst obtained in Scenario 1 is already as low as 1.65:1 (Table B2). On the other hand,
following g=0.456 criteria, the EROI levels are slightly improved. The recalculation of EROI from
other studies of the literature with g=0.456 provide EROI values which provide a wide range of
2.2-11.66:1 (Table B1). In relation to the EROIst obtained in Scenarios 1, 2 and 3, the value in
6 Although scenario 1 refers to present conditions, we suposse here that this present conditions could
be extrapolated to high penetration of CSP in the electricity mix, this result in very conservative
estimations of EROI (optimistic)
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
24
Scenario 1 reaches 3.62:1, while for EROIpou of the system in Scenario 1 decreases until 2:1
(Table B2). Hence, the consideration of different values for the g factor does not modify the
main conclusions of the analysis presented in this paper.
6. Discussion and conclusions
As mentioned in the introduction, analyses proposing a high share of CSP in future 100% RES
scenarios rely on the ability of this technology, through storage and/or hybridization, to
partially avoid the problems associated with the hourly/daily (short-term) variability of other
renewable variable sources, such as wind or PV (e.g. (Delucchi and Jacobson, 2011; García-
Olivares, 2016; Greenpeace et al., 2015; Jacobson et al, 2016; Jacobson and Delucchi, 2011;
WWF, 2011)). However, this advantage seems to be more than offset by the overall
performance of real CSP plants. In fact, the obtained results from CSP plants in operation,
using publicly available data from 4 countries (Spain, the USA, India and UAE) show that the
actual performance of CSP plants is shown to be significantly worse than projected by
constructors and considered by the scientific literature in the theoretical studies: capacity
factors in the same order as wind and PV, low EROI, intensive use of materials some scarce-
and significant seasonal intermittence. The consideration of these factors would likely modify
the conclusions of the analyses reviewed in Table 1.
In particular, real data shows that the capacity factor attributable to the solar energy of the
CSP plants is currently in the range of 0.15-0.3, representing a strong reduction in relation to
the range of usually expected values by the industry (0.2-0.5) and the common values used in
the academic literature (0.25-0.75) (for the currently most studied technology, Parabolic with
TES, the theoretical CF in the literature is around twice the real performance). This bias may
seem especially paradoxical in the case of the scientific literature; given that there has been
publicly available data for many power plants for years (e.g., SEGS plants in the USA have been
operating since the end of the 1980s). CF is a key parameter that critically affects the life-cycle
analyses that estimate the energy and material requirements (such as the energy payback time
(EPBT) and EROI) as well as the environmental impacts and economic costs.
Depending on the technology, the seasonal variability can be even worse than for wind or PV,
as has been shown for the case of Spain and the USA, where the output can also be zero for
many days in winter. Given that storage systems on the required scale to compensate for the
seasonal variations are not yet available and alternative technologies of large-scale storage are
still in the R&D phase (Wagner, 2014), the solution would require a combination of
overcapacities and flexible demand.
On the other hand, low latitude locations with high irradiances, such as hot deserts, are more
difficult to colonize (wind, dust/sand, extreme temperatures, water scarcity, etc.). Although
the seasonal variability in the studied plants in India and UEA improves in relation to those in
the USA and Spain, the fluctuations are still of the same order of magnitude as for other RES,
such as wind and PV. In desert areas, dust storms can also cover large regions during several
days (e.g. over the Sahara and the Arabian Peninsula in March 2010 (NASA EO, 2010)). These
dust storms explain why the analyzed CSP plant in UEA SHAMS 1 has a lower annual average
DNI than the CSP plants from the USA (located at higher latitudes). In regions affected by the
monsoon, such as India, the observed Sv in the Godawari CSP plant is 0.55, which is lower than
the expected from the variation in total irradiance at country level (Sv for India of 0.76
(Capellán-Pérez et al., 2017b)).
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
25
The current operation of CSP plants with back-up tends to exacerbate the magnitude of the
seasonal variability. In order to maximize revenues over the plant’s lifetime, back-up is more
important in high productivity months (i.e., summer) than low ones (i.e., winter). However, the
management of the energy system with a high penetration of intermittent RES will have to be
strongly regulated in order to prevent these behaviors, which will be against the efficiency of
the whole system.
The results obtained in this work allow us to compare the theoretical Sv values estimated in
Capellán-Pérez et al., (Capellán-Pérez et al., 2017b) for solar PV for different countries with the
actual Sv of real systems. The conservatively estimated Sv calculated in that study, based on
solar radiation for PV, is Sv=0.70 for Spain, which is higher than the actual Sv value found for
solar PV ~0.55 (see Table 3). In this sense, Capellán-Pérez et al., (Capellán-Pérez et al., 2017b)
seems to underestimate the land requirements for solar power plants; thus worsening the
potential environmental impacts related to their high-scale deployment.
In relation to EROIst (standard), the values obtained from applying a conservative
methodology are significantly lower than for other RES systems, in the order of 2.4:1 for
current systems, decreasing to 1.7:1 using common materials and reaching 1.3:1 when
considering common materials and transcontinental exports from deserts to high consumption
areas (such as those proposed by the DESERTEC project). In the case of CSP with back-up
systems from biomass; given the low EROI of biomass (de Castro et al., 2014) in relation to
natural gas, the EROI of this renewable hybrid system would be even lower, probably lower
than 1:1. If we consider the EROIpou (point of use) system, considering the needs of
overcapacities to deal with seasonal variation in high RES penetration scenarios, we estimate
the EROIpou to be around 1:1, which converts this technology into a carrier and not a source
of energy.
The reviewed scientific literature has tended to lean on one another to set CSP performance,
with a lack of critical revision of the initial theoretical assumptions, which could be justified
when data from real plants was not available, but this is no longer the case. The literature
analyzing real power plants and national systems of renewable power technologies show that
a bias towards overestimation of the performance of these technologies generally exists in the
scientific literature. However, it is crucial to correctly inform society about the decisions to be
taken in order to make a rapid transition to a renewable global energy system to avoid climate
change and other environmental impacts, as well as the physical limitation of fossil and
nuclear resources, remaining critical of the self-advocated solutions and avoiding wishful
thinking.
We would like to conclude this assessment with a comment on the relationship between
economic (monetary) and energetic costs. Learning rates, i.e. the percentage reduction in
monetary cost that occurs when cumulated output doubles, are usually interpreted as a proxy
of technological improvement of a technology. CSP plants are today much more expensive
than other renewable technologies (IRENA, 2018). However, a recent assessment has
concluded that there has been a sharp improvement in CSP of around 20% globally in the last
years (Lilliestam et al., 2017). Indeed, further (monetary) cost decreases of the CSP
technologies are usually expected for the next years (IRENA, 2018; Lilliestam et al., 2017).
However, this trend has still to be confirmed given that the aforementioned study uses a small
sample of plants and countries, and targets a short period of time (5 years), including
speculative data such as the expected costs from projects still to be completed. Given that
there is a relationship between monetary and energetic costs, the high cost of CSP plants is
consistent with the low EROI obtained in this study (Hall and Klitgaard, 2012; Heun and de Wit,
2012). However, the improvement of the (monetary) learning rates does not necessarily
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
26
correspond with a reduction in the material and energetic intensity of a technology. Other key
factors are the geographical location of the power plant (different levels of DNI), the cost of
commodities (raw materials) as well as the local context of the manufacturer's and operator’s
country. For example, for the case of solar PV, Pillai (2015) found that the reduction in average
production cost and price of solar panels of around 21% per year in the period 20052012 has
been driven by more factors than technological improvement (reduction in the use of
polysilicon and improvement of panel efficiencies mainly), such as the reduction in the price of
polysilicon, increasing market penetration of lower cost firms from China (where production
cost of firms were 22.4% lower than that of firms from other countries during the period of the
study), and increases in industry investment. In this sense, Indian and Chinese CSP planned
power plants are major downwards outliers in the data reported by Lilliestam et al., (2017).
Despite several decades of deployment, CSP is often depicted as a novel technology which
much room for technological improvement. In this sense, it is recommended a stronger focus
on the factors behind the variation of (monetary) learning rates of alternative technologies in
future works. From the point of view of the viability of the full energy system (and society) the
relevant magnitude is the EROI of the system, including the additional costs related with
overcapacities, grids and storage to deal with the intermittency of RES generation (Day et al.,
2018; Lambert et al., 2014; Palmer, 2017), which is not possible to capture with monetary
indicators such as the levelised cost of electricity of an individual technology (LCOE).
Hence, in the light of the obtained results (low capacity factor and EROI, intensive use of
materials some scarce- and the significant seasonal intermittence), the potential contribution
of current CSP technologies in a future 100% RES system seems very limited.
Acknowledgements
This work has been partially developed under the MEDEAS project, funded by the European
Union's Horizon 2020 Research and Innovation Programme under Grant agreement no.
691287. Iñigo Capellán-Pérez also acknowledges financial support from the Juan de la Cierva
Research Fellowship of the Ministry of Economy and Competitiveness of Spain (no. FJCI-2016-
28833).
Conflict of interest statement
On behalf of all authors, the corresponding author states that there is not conflict of interest.
References
Alobaidli, A., Sanz, B., Behnke, K., Witt, T., Viereck, D., Schwarz, M.A., 2017. Shams 1 - Design
and operational experiences of the 100MW - 540°C CSP plant in Abu Dhabi. AIP Conf.
Proc. 1850, 020001. https://doi.org/10.1063/1.4984325
Arvesen, A., Hertwich, E.G., 2012. Assessing the life cycle environmental impacts of wind
power: A review of present knowledge and research needs. Renew. Sustain. Energy
Rev. 16, 59946006. https://doi.org/10.1016/j.rser.2012.06.023
Asdrubali, F., Baldinelli, G., D’Alessandro, F., Scrucca, F., 2015. Life cycle assessment of
electricity production from renewable energies: Review and results harmonization.
Renew. Sustain. Energy Rev. 42, 11131122.
https://doi.org/10.1016/j.rser.2014.10.082
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
27
Boccard, N., 2009. Capacity factor of wind power realized values vs. estimates. Energy Policy
37, 26792688. https://doi.org/10.1016/j.enpol.2009.02.046
BP, 2017. BP Statistical Review of World Energy June 2017, Statistical Review of World Energy.
British Petroleum.
Brand, B., Stambouli, A.B., Zejli, D., 2012. The value of dispatchability of CSP plants in the
electricity systems of Morocco and Algeria. Energy Policy 47, 321331.
Burkhardt, J.J., Heath, G.A., Turchi, C.S., 2011. Life Cycle Assessment of a Parabolic Trough
Concentrating Solar Power Plant and the Impacts of Key Design Alternatives. Environ.
Sci. Technol. 45, 24572464. https://doi.org/10.1021/es1033266
Capellán-Pérez, I., de Blas, I., Nieto, J., De Castro, C., Miguel, L.J., Mediavilla, M., Carpintero, Ó.,
Rodrigo, P., Frechoso, F., Cáceres, S., 2017a. MEDEAS Model and IOA implementation
at global geographical level (Deliverable MEDEAS project,
http://www.medeas.eu/deliverables No. D4.1). GEEDS, University of Valladolid.
Capellán-Pérez, I., de Castro, C., Arto, I., 2017b. Assessing vulnerabilities and limits in the
transition to renewable energies: Land requirements under 100% solar energy
scenarios. Renew. Sustain. Energy Rev. 77, 760782.
https://doi.org/10.1016/j.rser.2017.03.137
Capellán-Pérez, I., Mediavilla, M., de Castro, C., Carpintero, Ó., Miguel, L.J., 2014. Fossil fuel
depletion and socio-economic scenarios: An integrated approach. Energy 77, 641666.
https://doi.org/10.1016/j.energy.2014.09.063
Clack, C.T.M., Qvist, S.A., Apt, J., Bazilian, M., Brandt, A.R., Caldeira, K., Davis, S.J., Diakov, V.,
Handschy, M.A., Hines, P.D.H., Jaramillo, P., Kammen, D.M., Long, J.C.S., Morgan, M.G.,
Reed, A., Sivaram, V., Sweeney, J., Tynan, G.R., Victor, D.G., Weyant, J.P., Whitacre,
J.F., 2017. Evaluation of a proposal for reliable low-cost grid power with 100% wind,
water, and solar. Proc. Natl. Acad. Sci. 114, 67226727.
https://doi.org/10.1073/pnas.1610381114
CNMC, 2016. RESOLUCIÓN POR LA QUE SE APRUEBA LA LIQUIDACIÓN DEFINITIVA DE LAS
PRIMAS EQUIVALENTES, LAS PRIMAS, INCENTIVOS Y COMPLEMENTOS A LAS
INSTALACIONES DE PRODUCCIÓN DE ENERGÍA ELÉCTRICA DE TECNOLOGÍA SOLAR
TERMOELÉCTRICA CORRESPONDIENTE A LOS EJERCICIOS 2009, 2010 Y 2011 (No.
LIQ/DE/ 143 /1 5). Comisión Nacional de los Mercados y la Competencia.
Corona, B., Miguel, G.S., Cerrajero, E., 2014. Life cycle assessment of concentrated solar power
(CSP) and the influence of hybridising with natural gas. Int. J. Life Cycle Assess. 19,
12641275. https://doi.org/10.1007/s11367-014-0728-z
Corona, B., Ruiz, D., San Miguel, G., 2016. Life Cycle Assessment of a HYSOL Concentrated Solar
Power Plant: Analyzing the Effect of Geographic Location. Energies 9, 413.
https://doi.org/10.3390/en9060413
Day, J.W., D’Elia, C.F., Wiegman, A.R.H., Rutherford, J.S., Hall, C.A.S., Lane, R.R., Dismukes, D.E.,
2018. The Energy Pillars of Society: Perverse Interactions of Human Resource Use, the
Economy, and Environmental Degradation. Biophys. Econ. Resour. Qual. 3, 2.
https://doi.org/10.1007/s41247-018-0035-6
de Castro, C., Carpintero, Ó., Frechoso, F., Mediavilla, M., de Miguel, L.J., 2014. A top-down
approach to assess physical and ecological limits of biofuels. Energy 64, 506512.
https://doi.org/10.1016/j.energy.2013.10.049
de Castro, C., Mediavilla, M., Miguel, L.J., Frechoso, F., 2013. Global solar electric potential: A
review of their technical and sustainable limits. Renew. Sustain. Energy Rev. 28, 824
835. https://doi.org/10.1016/j.rser.2013.08.040
Delucchi, M.A., Jacobson, M.Z., 2011. Providing all global energy with wind, water, and solar
power, Part II: Reliability, system and transmission costs, and policies. Energy Policy
39, 11701190. https://doi.org/10.1016/j.enpol.2010.11.045
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
28
Deng, Y.Y., Haigh, M., Pouwels, W., Ramaekers, L., Brandsma, R., Schimschar, S., Grözinger, J.,
de Jager, D., 2015. Quantifying a realistic, worldwide wind and solar electricity supply.
Glob. Environ. Change 31, 239252. https://doi.org/10.1016/j.gloenvcha.2015.01.005
Denholm, P., Mehos, M., 2015. The Role of Concentrating Solar Power in Integrating Solar and
Wind Energy.
Douglas, K., Boyd, J., Byron, J., Eggert, A., Weisenmiller, R., Vaccaro, K., 2010. Abengoa Mojave
Solar Project. Commission Decision. (No. CEC-800-2010-008-CMF. DOCKET NUMBER
09-AFC-5). California Energy Commission.
Ehtiwesh, I.A.S., Coelho, M.C., Sousa, A.C.M., 2016. Exergetic and environmental life cycle
assessment analysis of concentrated solar power plants. Renew. Sustain. Energy Rev.
56, 145155. https://doi.org/10.1016/j.rser.2015.11.066
Elliston, B., Diesendorf, M., MacGill, I., 2012. Simulations of scenarios with 100% renewable
electricity in the Australian National Electricity Market. Energy Policy 45, 606613.
https://doi.org/10.1016/j.enpol.2012.03.011
Ferroni, F., Guekos, A., Hopkirk, R.J., 2017. Further considerations to: Energy Return on Energy
Invested (ERoEI) for photovoltaic solar systems in regions of moderate insolation.
Energy Policy 107, 498505. https://doi.org/10.1016/j.enpol.2017.05.007
Ferroni, F., Hopkirk, R.J., 2016. Energy Return on Energy Invested (ERoEI) for photovoltaic solar
systems in regions of moderate insolation. Energy Policy 94, 336344.
https://doi.org/10.1016/j.enpol.2016.03.034
García-Olivares, A., 2016. Energy for a sustainable post-carbon society. Sci. Mar. 80, 257268.
https://doi.org/10.3989/scimar.04295.12A
García-Olivares, A., Ballabrera-Poy, J., García-Ladona, E., Turiel, A., 2012. A global renewable
mix with proven technologies and common materials. Energy Policy 41, 561574.
https://doi.org/10.1016/j.enpol.2011.11.018
Greenpeace, GWEC, SolarPowerEurope, 2015. Energy [R] evolution-A sustainable world energy
outlook 2015. Greenpeace, GWEC, SolarPowerEurope.
Hall, C.A.S., Klitgaard, K.A., 2012. Energy and the Wealth of Nations: Understanding the
Biophysical Economy. Springer New York, New York, NY.
Hall, C.A.S., Lambert, J.G., Balogh, S.B., 2014. EROI of different fuels and the implications for
society. Energy Policy 64, 141152. https://doi.org/10.1016/j.enpol.2013.05.049
Hammond, G.P., Jones, C.I., 2008. Embodied energy and carbon in construction materials.
Proc. Inst. Civ. Eng. - Energy 161, 8798.
Heath, G., Turchi, C., Decker, T., Burkhardt, J., Kutscher, C., 2009. Life cycle assessment of
thermal energy storage: two-tank indirect and thermocline, in: ASME 2009 3rd
International Conference on Energy Sustainability Collocated with the Heat Transfer
and InterPACK09 Conferences. American Society of Mechanical Engineers, pp. 689
690.
Heath, G.A., Turchi, C.S., Burkhardt III, J.J., 2011. Life cycle assessment of a parabolic trough
concentrating solar power plant and impacts of key design alternatives. SolarPACES
conference, Granada, Spain. Golden: NREL. Retrieved from http://www. nrel.
gov/docs/fy11osti/52186. pdf.
Hernandez, R.R., Hoffacker, M.K., Murphy-Mariscal, M.L., Wu, G.C., Allen, M.F., 2015. Solar
energy development impacts on land cover change and protected areas. Proc. Natl.
Acad. Sci. 201517656. https://doi.org/10.1073/pnas.1517656112
Heun, M.K., de Wit, M., 2012. Energy return on (energy) invested (EROI), oil prices, and energy
transitions. Energy Policy 40, 147158. https://doi.org/10.1016/j.enpol.2011.09.008
IEA, 2018. IEA World Energy Statistics and Balances, World Energy Statistics and Balances
(database). IEA/OECD, Paris (France).
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
29
IEA, IRENA, 2017. Perspectives for the Energy Transition. Investment Needs for a Low-Carbon
Energy System. International Energy Agency and International Renewable Energy
Agency.
IEA, IRENA, 2013. Concentrating Solar Power. IEA-ETSAP and IRENA Technology Brief E10
(Technology Brief No. 10). IEA ETSAP and IRENA.
IPCC, 2014. Climate Change 2014: Mitigation of Climate Change. Fifth Assess. Rep. Intergov.
Panel Clim. Change.
IPCC, 2011. Special Report on Renewable Energy Sources and Climate Change Mitigation.
Cambridge University Press, United Kingdom and New York (USA).
IRENA, 2018. Renewable Power Generation Costs in 2017. International Renewable Energy
Agency, Abu Dhabi.
IRENA db, 2017. IRENA Resource (Database). International Renewable Energy Agency,
http://resourceirena.irena.org.
Jacobson et al, 2016. 100% Clean and Renewable Wind, Water, and Sunlight (WWS) All-Sector
Energy Roadmaps for 139 Countries of the World.
Jacobson, M.Z., Delucchi, M.A., 2011. Providing all global energy with wind, water, and solar
power, Part I: Technologies, energy resources, quantities and areas of infrastructure,
and materials. Energy Policy 39, 11541169.
https://doi.org/10.1016/j.enpol.2010.11.040
Johansson, B., 2013. Security aspects of future renewable energy systems–A short overview.
Energy 61, 598605. https://doi.org/10.1016/j.energy.2013.09.023
Keough, 2011. Austempered Ductile Iron (ADI) A Green Alternative. American Foundry
Society www.afsinc.org, Schaumburg, Illinois, USA.
Klein, S.J.W., Rubin, E.S., 2013. Life cycle assessment of greenhouse gas emissions, water and
land use for concentrated solar power plants with different energy backup systems.
Energy Policy 63, 935950. https://doi.org/10.1016/j.enpol.2013.08.057
Krishnamurthy, P., Banerjee, R., 2012. Energy analysis of solar thermal concentrating systems
for power plants, in: International Conference on Future Electrical Power and Energy
Systems, Lecture Notes in Information Technology. pp. 509514.
Kuenlin, A., Augsburger, G., Gerber, L., Maréchal, F., 2013. Life Cycle Assessment and
Environomic Optimization of Concentrating Solar Thermal Power Plants. Presented at
the 26th International Conference on Efficiency, Cost, Optimization, Simulation and
Environmental Impact of Energy Systems (ECOS2013).
La Africana, 2018. Africana Energia. http://africanaenergia.es/index.php/es/africana-
energia.html (retrieved 15-1-2018).
Lambert, J.G., Hall, C.A.S., Balogh, S., Gupta, A., Arnold, M., 2014. Energy, EROI and quality of
life. Energy Policy 64, 153167. https://doi.org/10.1016/j.enpol.2013.07.001
Lechón, Y., de la Rúa, C., Sáez, R., 2008. Life Cycle Environmental Impacts of Electricity
Production by Solarthermal Power Plants in Spain. J. Sol. Energy Eng. 130, 021012-
0210127. https://doi.org/10.1115/1.2888754
Lenzen, M., McBain, B., Trainer, T., Jütte, S., Rey-Lescure, O., Huang, J., 2016. Simulating low-
carbon electricity supply for Australia. Appl. Energy 179, 553564.
https://doi.org/10.1016/j.apenergy.2016.06.151
Lilliestam, J., Labordena, M., Patt, A., Pfenninger, S., 2017. Empirically observed learning rates
for concentrating solar power and their responses to regime change. Nat. Energy 2,
17094. https://doi.org/10.1038/nenergy.2017.94
MacKay, D.J.C., 2013. Solar energy in the context of energy use, energy transportation and
energy storage. Philos. Trans. R. Soc. Lond. Math. Phys. Eng. Sci. 371, 20110431.
https://doi.org/10.1098/rsta.2011.0431
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
30
Ministerio de Energía, 2018. Estadísticas Eléctricas Anuales: Eléctricas 2016-2018. Generación
eléctrica 2016 [ODS].
Ministerio de Industria, Energía y Turismo, 2014. Orden IET/1882/2014, de 14 de octubre, por
la que se establece la metodología para el cálculo de la energía eléctrica imputable a la
utilización de combustibles en las instalaciones solares termoeléctricas (No. BOE-A-
2014-10475), BOE. Ministerio de Industria, Energía y Turismo.
Montgomery, Z., 2009. Environmental Impact Study: CSP vs. CdTe thin film photovoltaics.
Masters Proj. Submitt. Partial Fulfillment Requir. Master Environ. Manag. Degree
Nicholas Sch. Environ. Duke Univ.
Moriarty, P., Honnery, D., 2016. Can renewable energy power the future? Energy Policy 93, 3
7. https://doi.org/10.1016/j.enpol.2016.02.051
NASA EO, 2010. Massive Dust Storm Sweeps Across Africa. NASA Earth observatory:
https://earthobservatory.nasa.gov/NaturalHazards/view.php?id=43200&eocn=image
&eoci=related_image.
NREL, 2017. Concentrating Solar Power Projects. National Renewable Energy Laboratory,
http://www.nrel.gov/csp/solarpaces/ (Retrieved 6-1-2017).
NREL, 2012. Renewable Electricity Futures Study (Entire Report) (4 vols. No. NREL/TP-6A20-
52409). National Renewable Energy Laboratory, Golden, CO, USA.
Palmer, G., 2017. A Framework for Incorporating EROI into Electrical Storage. Biophys. Econ.
Resour. Qual. 2, 6. https://doi.org/10.1007/s41247-017-0022-3
Palmer, G., 2013. Household Solar Photovoltaics: Supplier of Marginal Abatement, or Primary
Source of Low-Emission Power? Sustainability 5, 14061442.
https://doi.org/10.3390/su5041406
Pihl, E., Kushnir, D., Sandén, B., Johnsson, F., 2012. Material constraints for concentrating solar
thermal power. Energy, Integration and Energy System Engineering, European
Symposium on Computer-Aided Process Engineering 2011 44, 944954.
https://doi.org/10.1016/j.energy.2012.04.057
Pillai, U., 2015. Drivers of cost reduction in solar photovoltaics. Energy Econ. 50, 286293.
https://doi.org/10.1016/j.eneco.2015.05.015
Prieto, P.A., Hall, C.A.S., 2013. Spain’s Photovoltaic Revolution: The Energy Return on
Investment, 2013th ed. Springer.
Raugei, M., Carbajales-Dale, M., Barnhart, C.J., Fthenakis, V., 2015. Rebuttal: “Comments on
‘Energy intensities, EROIs (energy returned on invested), and energy payback times of
electricity generating power plants’ Making clear of quite some confusion.” Energy
82, 10881091. https://doi.org/10.1016/j.energy.2014.12.060
Raugei, M., Sgouridis, S., Murphy, D., Fthenakis, V., Frischknecht, R., Breyer, C., Bardi, U.,
Barnhart, C., Buckley, A., Carbajales-Dale, M., Csala, D., de Wild-Scholten, M., Heath,
G., Jæger-Waldau, A., Jones, C., Keller, A., Leccisi, E., Mancarella, P., Pearsall, N., Siegel,
A., Sinke, W., Stolz, P., 2017. Energy Return on Energy Invested (ERoEI) for
photovoltaic solar systems in regions of moderate insolation: A comprehensive
response. Energy Policy 102, 377384. https://doi.org/10.1016/j.enpol.2016.12.042
REE, 2018. REE webpage. Red Eléctrica de España, www.ree.es/ (Retrieved 15-1-2018).
ReflecTech, 2012. Embodied energy (RefkecTech White Paper).
REN21, 2017. Renewables Global Futures Report. Great debates towards 100% renewable
energy. Paris: REN21 Secretariat.
REN21, 2016. Renewables 2016. Global Status Report. REN 21, Paris.
REN21, 2015. Renewables 2015. Global Status Report. REN 21, Paris.
Sanz, B., 2017. How Shams 1 CSP project Produce more electricity beyond expectation? Here
the benificial O&M Experience.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
31
Smil, V., 2010. Energy Transitions: History, Requirements, Prospects. Praeger, Santa Barbara,
California, USA.
Solanki, J., 2016. Godawari CSP plant: an overview of performance. Sun Focus 4, 68.
Trainer, T., 2017a. Some problems in storing renewable energy. Energy Policy 110, 386393.
https://doi.org/10.1016/j.enpol.2017.07.061
Trainer, T., 2017b. ’Can renewables meet total Australian energy demand: A “disaggregated”
approach. Energy Policy 109, 539544. https://doi.org/10.1016/j.enpol.2017.07.040
Trainer, T., 2013. Can Europe run on renewable energy? A negative case. Energy Policy 63,
845850. https://doi.org/10.1016/j.enpol.2013.09.027
Trainer, T., 2012. A critique of Jacobson and Delucchi’s proposals for a world renewable energy
supply. Energy Policy 44, 476481. https://doi.org/10.1016/j.enpol.2011.09.037
Trainer, T., 2010. Can renewables etc. solve the greenhouse problem? The negative case.
Energy Policy 38, 41074114. https://doi.org/10.1016/j.enpol.2010.03.037
Trieb, F., 2006. Trans-Mediterranean interconnection for concentrating solar power. German
Aerospace Center (DLR), Institute of Technical Thermodynamics and Section Systems
Analysis and Technology Assessment, Stuttgart.
Turchi, C., 2010. Parabolic Trough Reference Plant for Cost Modeling with the Solar Advisor
Model (SAM) (Technical Report No. NREL/ TP-550-47605). NREL.
US EIA db, 2018. USA Energy Statistics (Database). US Energy Information Administration,
http://www.eia.gov.
Vargel, C., 2004. Corrosion of aluminium. Elsevier.
Viebahn, P., 2013. Solarthermische Kraftwerkstechnologie für den Schutz des Erdklimas,
SOKRATES-Projekt. Stuttgart.
Viebahn, P., Lechón, Y., Trieb, F., 2011. The potential role of concentrated solar power (CSP) in
Africa and EuropeA dynamic assessment of technology development, cost
development and life cycle inventories until 2050. Energy Policy, At the Crossroads:
Pathways of Renewable and Nuclear Energy Policy in North Africa 39, 44204430.
https://doi.org/10.1016/j.enpol.2010.09.026
Wagner, F., 2014. Considerations for an EU-wide use of renewable energies for electricity
generation. Eur. Phys. J. Plus 129, 114. https://doi.org/10.1140/epjp/i2014-14219-7
Weinrebe, G., Boehnke, M., Trieb, F., 1998. Life cycle assessment of an 80 MW SEGS plant and
a 30 MW Phoebus power tower. Sol. Eng. 417424.
Weißbach, D., Ruprecht, G., Huke, A., Czerski, K., Gottlieb, S., Hussein, A., 2014. Reply on
“Comments on ‘Energy intensities, EROIs (energy returned on invested), and energy
payback times of electricity generating power plants’ Making clear of quite some
confusion.” Energy Complete, 10041006.
https://doi.org/10.1016/j.energy.2014.02.026
Weißbach, D., Ruprecht, G., Huke, A., Czerski, K., Gottlieb, S., Hussein, A., 2013. Energy
intensities, EROIs (energy returned on invested), and energy payback times of
electricity generating power plants. Energy 52, 210221.
https://doi.org/10.1016/j.energy.2013.01.029
WEO, 2014. World Energy Outlook 2014. OECD / IEA, Paris.
WWF, 2011. The energy report: 100% renewable energy by 2050. WWF, Ecofys, OMA.
Zhang, H.L., Baeyens, J., Degrève, J., Cacères, G., 2013. Concentrated solar power plants:
Review and design methodology. Renew. Sustain. Energy Rev. 22, 466481.
https://doi.org/10.1016/j.rser.2013.01.032
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
32
Appendix A
Energy Used (EnU)
Scenario 1 Scenario 2 Scenario 3
MJ/KWh
%
MJ/KWh
%
MJ/KWh
%
steels
0.37
17.0%
0.37
12.1%
0.37
9.1%
concrete 0.02
0.9%
0.02
0.7%
0.02
0.5%
plastics
0.1
4.6%
0.1
3.3%
0.1
2.5%
Synthetic oil 0.17
7.8%
0.17
5.5%
0.17
4.2%
molten salts
0.2
9.2%
0.47
15.3%
0.94
23.2%
Ag based mirror 0.08
3.7%
-
-
-
-
Al based mirror
-
-
0.26
8.5%
0.26
6.4%
site preparation 0.49
22.5%
0.49
16.0%
0.49
12.1%
other material and machineries
0.03
1.4%
0.03
1.0%
0.03
0.7%
broken mirrors 0
0.0%
0.03
1.0%
0.07
1.7%
water (distilled 0.2MJ/kg)
0.13
6.0%
0.13
4.2%
0.13
3.2%
Cu, Mg and other metals 0.02
0.9%
0.02
0.7%
0.02
0.5%
rock
0.02
0.9%
0.02
0.7%
0.02
0.5%
Operation phase
0.19
8.7%
0.19
6.2%
0.19
4.7%
Dismantling and disposal
0.05
2.3%
0.06
2.0%
0.07
1.7%
Dry cooled perfomance 0.14
6.4%
0.18
5.9%
0.22
5.4%
Al mirror reflectivity loss
0
0.0%
0.36
11.7%
0.43
10.6%
Grid necessities and losses 0.18
8.3%
0.18
5.9%
0.54
13.3%
EnUtot
2.18
100.0%
3.07
100.0%
4.06
100.0%
Table A1: Contribution to the EnU for each phase/material processing of a CSP power
plant standard type Andasol for scenarios 1, 2 and 3.
Author Accepted Manuscript: Carlos de Castro, Iñigo Capellán-Pérez: “Concentrated Solar Power: actual
performance and foreseeable future in high penetration scenarios of renewable energies”. BioPhysical
Economics and Resource Quality (2018) 3:14
https://doi.org/10.1007/s41247-018-0043-6
33
Appendix B
Reported
EROI
Recalculated
EROI
(CF=0.25,
g=0.687, Sv=1)
Recalculated
EROI
(CF = 0,25,
g=1, Sv=0.5)
Recalculated
EROI
(CF=0.25,
g=0,456, Sv=1)
Corona et al., 2014 (hybrid
range) 14.1-17.5 1.47-3.41 0.50-1.17 2.21-5.12
Ehtiwesh et al., 2016
20.2
7.74
2.66
11.66
Heath et al., 2011
30
6.97
2.40
10.50
Lechón et al., 2008 (Tower)
24.6
1.88
0.65
2.83
Lechón et al., 2008
(Parabolic) 24 2.14 0.74 3.22
Viebahn, 2013 (range)
10.9-67.6
5.02
1.73
7.56
Table B1: Sensitivity analysis for some other parameters for other two extreme cases (see
table 7 and text for the chosen parameters). Sv= 0.5 consider overcapacity for seasonal
intermittence and Sv= 1 do not consider overcapacity, back up, etc.
Scenario 1
scenario 2
Scenario 3
EROIg=1
EROIg=0.456
EROIg=1
EROIg=0.456
EROIg=1
EROIg=0.456
EROIst (Materials in
construction, operation,
dismantling and disposal
including grids)
1.65
3.62
1.17
2.56
0.89
1.96
EROIpou (System, high
penetration, Lenzen and
Deluchi models)
0.96
2.11
0.687
1.51
0.55
1.21
EROIpou (System, high
penetration, this work)
0.82
1.81
0.58
1.28
0.45
0.98
Table B2: Sensitivity analysis for the parameter “g”. The first row, is the EROIst without needs
for additional storage or overcapacities or back-ups due to intermittence of CSP, it is
equivalent to choose Sv=1. The second row takes into account the overcapacities and storage
following both, Lenzen and Deluchi models. The third row follow our criteria for overcapacities
needed due to seasonal intermittence only (Sv=0.5). See text for explanation.
... In the recent past, solar energy has had huge deployment prospects worldwide and specifically in developing countries, since most of them situated in global solar belt. Among solar harvesting technologies, Concentrated Solar Power (CSP) has proven to be very promising due to its concentrating nature and the ability to store energy at low costs allows it to be installed in different climatic zones and provide peak load electricity (De Castro and Capellán-Pérez, 2018). Various types of CSP plants exist, depending upon the focal mechanism. ...
... Pakistan, till date, does not have any installed CSP powerplant; besides, to the best of the authors' knowledge, policies, cost structures or tariff regimes for future installation do not exist. For cost inputs in the model, capacity-weighted averages of PTC and ST power plants commissioned globally during years 2018-2020 (IRENA, 2021) were explored and verified from literature for different regions of the world including USA (Kurup and Craig, 2015), Iran (Hirbodi et al., 2020), Spain, India, UAE (De Castro and Capellán-Pérez, 2018) and others (Camera, 2019). Afterwards, projected reductions in capital generation costs for the year 2025, considering field-specific research and development activities and economies of scale were applied to cost estimates and were used as input for financial analysis (Dieckmann et al., 2017). ...
Article
In the current article, which is first to a two-part study, techno-economic evaluation, viability and optimization of concentrated solar power plants considering design parameters of solar multiple and integration of thermal energy storage is performed. High frequency (10 min), validated ground datasets are used for numerical simulations using System Advisor Model. Techno-economic performance analysis of dry and wet cooling system equipped solar tower and parabolic trough collectors is assessed for three different climatic zones i.e. temperate, semi-arid and sub-tropical. The optimization problem intends to maximize technical potential alongside minimizing net cost of electricity production. It is found that on average, 60 and 75% increase in annual electricity generation, and 5 and 7% decrease in levelized cost of electricity is observed with addition of 6 h of storage for solar tower and trough plants, respectively. For cost-effective solution, optimal solar multiple and thermal energy storage combination is in range of 3–3.2 and 15.4–15.9 for solar tower, and 2.3–3 and 10.6–11.6 h for trough plants. Wet-cooled solar towers perform the best in each climatic zone offering 5–6% lower cost of electricity production and 3–7% higher electricity generation potentials. Moreover, for mutually exclusive projects, levelized cost of electricity is a better metric to gauge economic viability than net present value. The follow-up study (Part B) focuses on environ-economic optimization of concentrated solar power plants. Furthermore, outcomes from both optimization objectives are utilized for multi-objective optimization based ‘balanced’ solution assessment in the next part.
... A number of sources estimate that world resources of the required metals are insu cient for batteries to play a major storage role. [13], [14], [15], [16], [17] [18], [19], [20] It takes 160 grams of lithium metal to store 1 kWh, so if Australia needed to store 575 GWh then 3.7 kg per person would be required. At this rate, for a world of 8 billion the total would be 29 million tonnes, for electricity storage alone. ...
Preprint
Full-text available
Wind generation data enables periods of low wind input to the Australian energy supply system to be analysed in terms of the amount of supply that would have to be drawn from storage to meet demand over those periods. The findings challenge the common assumption that high penetration renewable supply would be achievable at a relatively low need for storage and at an easily affordable cost.
... 2. Los potenciales tecnológicos asequibles y eficacia de los sistemas de captación y transformación se han exagerado y se exageran de forma sistemática en la literatura (ver las denuncias por ejemplo en Boccard 2009; Giampietro y Mayumi, 2009;Miller et al., 2010;de Castro et al., 2013de Castro et al., y 2014de Castro y Capellán-Pérez, 2018; Seibert y Rees 2021), hasta el punto en que se emplean metodologías para aquellas evaluaciones que violan incluso los principios más básicos de la física (Miller et al., 2010;de Castro et al., 2011;de Castro, 2015;Patzek, 2004Patzek, y 2006, sin que haya una rectificación o reconocimiento de los hechos. Incluso, en algunos casos, se llega a tergiversar, a mentir y a atacar personalmente a los autores de estas críticas por investigadores y los propios editores de revistas (Seibert y Rees, 2022;. ...
Article
Full-text available
Las fuentes de energía no renovables (fósiles y nucleares) están doblemente limitadas, tanto por su finitud como por los perjuicios ecológicos y sociales que causan. Las fuentes de energía renovable tienen flujos en la biosfera muy grandes, sin embargo, los sistemas tecnológicos que las captan no son renovables y por tanto tienen limitaciones tecnológicas, ecológicas y sociales también. Aunque una buena parte de la literatura científica ha venido estimando que estos factores limitantes son pequeños frente a los políticos y económicos, aquí se muestra que esta literatura ha venido sobreestimando la capacidad tecno-sostenible obtenible a lo largo del presente siglo, en buena medida por un tecno-optimismo implícito y una falta de pensamiento sistémico. Este potencial podría ser del orden de entre la mitad y la cuarta parte del uso actual de energía, lo que apunta, dada la necesaria transición hacia fuentes renovables por problemas ambientales, a un fuerte decrecimiento de la matriz energética que sostiene nuestras sociedades a escala global.
... It should be noted that CSP farms usually attain a capacity factor between 15% and 30% 46 , even though expected industry values can reach as high as 50%. One standard deviation of the mean of the population depicted in Fig. 8a covers values ranging from 14.1% to 35.6%, which agrees with earlier reporting 46 . Even though the land use of CSP plants could be linearly fitted with respect to the rated capacity, a high deviation exists, as shown in Fig. 8b. ...
Article
Full-text available
This paper introduces the annual energy density concept for electric power generation, which is proposed as an informative metric to capture the impacts on the environmental footprint. Our investigation covers a wide range of sources classified by rated power and compares different regions to establish typical spatial flows of energy and evaluate the corresponding scalability to meet future net-zero emission (NZE) goals. Our analysis is conducted based on publicly available information pertaining to different regions and remote satellite image data. The results of our systematic analysis indicate that the spatial extent of electric power generation toward 2050 will increase approximately sixfold, from approximately 0.5% to nearly 3.0% of the world’s land area, based on International Energy Agency (IEA) NZE 2050 targets. We investigate the worldwide energy density for ten types of power generation facilities, two involving nonrenewable sources (i.e., nuclear power and natural gas) and eight involving renewable sources (i.e., hydropower, concentrated solar power (CSP), solar photovoltaic (PV) power, onshore wind power, geothermal power, offshore wind power, tidal power, and wave power). In total, our study covers 870 electric power plants worldwide, where not only the energy density but also the resulting land or sea area requirements to power the world are estimated. Based on the provided meta-analysis results, this paper challenges the common notion that solar power is the most energy-dense renewable fuel source by demonstrating that hydropower supersedes solar power in terms of land use in certain regions of the world, depending on the topography.
Article
Full-text available
Wind generation data enable periods of low wind input to the Australian energy supply system to be analysed in terms of the amount of supply that would have to be drawn from storage to meet demand over those periods. The findings challenge the common assumption that high-penetration renewable supply would be achievable at a relatively low need for storage and at an easily affordable cost.
Article
The objective of this study is to include environmental impact in optimization of concentrated solar power plants previously limited to techno-economic analysis only. Performance of solar towers and parabolic trough collectors equipped with dry and wet cooling is numerically investigated at 10 min interval for three climatic zones. Based on electricity generation profiles pertinent to variation of design variables of solar multiple and thermal energy storage, the water usage is quantified techno-economically, considering water scarcity levels. Besides, five different potential impacts over lifecycle of concentrated solar and natural gas-based power plants are compared after unifying them into metric of net environmental impact avoided. Additionally, utilizing techno-economic solutions and eco-friendly solutions, multi-objective ‘balanced’ solution is proposed. It is found that levelized cost of water increases five-fold for areas with extreme water scarcity and is higher for trough than solar tower. Net environmental impact of troughs is higher and lower electricity generation causes the avoided impact to be higher than solar towers, when compared with natural gas-fired power plants. For eco-friendly solutions, solar multiple varies from 3.5 to 4 with thermal energy storage of 24 h. Dry-cooled solar towers with optimal solar multiple in range of 3–3.2 and thermal energy storage of 18.8–19.9 h are evaluated for pareto-optimality based ‘balanced’ solutions for all zones. These solutions offer minimum net levelized cost of electricity, alongside maximized electricity generation, huge return on invested capital and energy, and reducing environmental impact by over 94%. Lastly, semi-arid climates are found suitable for concentrated solar power plant installation.
Article
Given the dire consequences of climate change and the war in Ukraine, decarbonization of electrical power systems around the world must be accomplished, while avoiding recurring blackouts. A good understanding of performance and reliability of different power sources underpins this endeavor. As an energy transition involves different societal sectors, we must adopt a simple and efficient way of communicating the transition's key indicators. Capacity factor (CF) is a direct measure of the efficacy of a power generation system and of the costs of power produced. Since the year 2000, the explosive expansion of solar PV and wind power made their CFs more reliable. Knowing the long-time average CFs of different electricity sources allows one to calculate directly the nominal capacity required to replace the current fossil fuel mix for electricity generation or expansion to meet future demand. CFs are straightforwardly calculated, but they are rooted in real performance, not in modeling or wishful thinking. Based on the current average CFs, replacing 1 W of fossil electricity generation capacity requires installation of 4 W solar PV or 2 W of wind power. An expansion of the current energy mix requires installing 8.8 W of solar PV or 4.3 W of wind power.
Article
Full-text available
Although the possibility of renewable energy supply meeting 100% of demand is widely assumed the issue remains unsettled. This paper discusses some of the central questions that need clarification or possible resolution before we can decide whether the goal is achievable. These include the need for simulations, storage options, the EROI values for whole renewable supply systems, and the probable dollar costs of whole systems capable of meeting all energy demand whilst providing for intermittency. Examination of several Australian simulations indicates that significant difficulties and uncertainties remain to be resolved, and the discussion of these supports the case that 100% supply systems will be at the least very costly and might be unaffordable.
Article
Full-text available
To meet the COP21 2 °C climate target, humanity would need to complete a transition to renewable energy within the next several decades. But for decades, fossil fuels will continue to underpin many fundamental activities that allow modern society to function. Unfortunately, net energy yield from fossil fuels is now falling, and despite substantial growth in renewable energy, total global energy demand and fossil fuel consumption are still increasing. Recent studies document promising trends in net energy yield from new renewable energy, particularly wind and solar. However, most studies do not fully consider the complexities of multiple factors including production intermittency, storage, the need to replace a massive infrastructure network, and lack of fungibility of different energy sources. Also, oft-overlooked, is that human impacts have caused widespread degradation of natural ecosystems and the provisioning of ecosystem goods and services, especially affecting vulnerable areas like coastal zones and arid regions. An accelerated renewable energy transition to meet climate targets and replace declining stocks of high net yielding fossil fuels will compete with resources needed for crucial investments to mitigate already locked in climate change and environmental degradation impacts. Integrative approaches that include all costs can help balance interdependent factors such as net energy dynamics, resource allocation, and ecosystem degradation. Energy-climate investment pathways produce economic output and quality of life tradeoffs that must be considered. Accordingly, developing future energy policy requires a systems approach with global boundaries and new levels of appreciation of the complex mix of interrelated factors involved.
Article
Full-text available
Difficulties involved in some commonly advocated options for the storage of renewable electricity are discussed. As is generally recognised the most promising strategies involve biomass and pumped hydro storage, but these involve drawbacks that appear to be major limitations on the achievement of 100% renewable supply systems. Neglected aspects of the solar thermal storage solution are detailed, indicating that it is not likely to be able to make a significant contribution. Batteries, vehicle-to-grid, biomass and hydrogen based solutions also appear to have major drawbacks. Although other options not examined here might alter the outlook, the general impression arrived at is that the probability of achieving satisfactory storage provision enabling 100% renewable power supply are not promising. Provision of total energy supply from renewable sources would probably multiply the task by an order of magnitude.
Article
Full-text available
We develop roadmaps to transform the all-purpose energy infrastructures (electricity, transportation, heating/cooling, industry, agriculture/forestry/fishing) of 139 countries to ones powered by wind, water, and sunlight (WWS). The roadmaps envision 80% conversion by 2030 and 100% by 2050. WWS not only replaces business-as-usual (BAU) power, but also reduces it ∼42.5% because the work: energy ratio of WWS electricity exceeds that of combustion (23.0%), WWS requires no mining, transporting, or processing of fuels (12.6%), and WWS end-use efficiency is assumed to exceed that of BAU (6.9%). Converting may create ∼24.3 million more permanent, full-time jobs than jobs lost. It may avoid ∼4.6 million/year premature air-pollution deaths today and ∼3.5 million/year in 2050; ∼22.8trillion/year(12.7¢/kWhBAUallenergy)in2050airpollutioncosts;and22.8 trillion/year (12.7 ¢/kWh-BAU-all-energy) in 2050 air-pollution costs; and ∼28.5 trillion/year (15.8 ¢/kWh-BAU-all-energy) in 2050 climate costs. Transitioning should also stabilize energy prices because fuel costs are zero, reduce power disruption and increase access to energy by decentralizing power, and avoid 1.5°C global warming.
Article
Full-text available
A paper by Ferroni and Hopkirk (2016) provided evidence that presently available PV systems in regions of moderate insolation like Switzerland and countries north of the Swiss Alps act as net energy sink. These findings were disputed in a paper (Raugei et al., 2017). Additional clarifications in support of our conclusions are explained, including mention of weak points in the argumentation by Raugei et al. Our study is based on the concept of the extended ERoEI (ERoEIEXT) for PV systems, knowing that this is not the mainstream concept in the Life Cycle Assessment (LCA), applying the Process-Based Life Cycle Assessment. The concept of the ERoEIEXT considers many possible energy contributions needed for assessing the envisioned transition from fossil fuel to other types of energy sources and here in particular to photovoltaics in regions of moderate insolation. The conclusions of our original study remain unchanged. Any attempt to adopt an Energy Transition strategy by substitution of intermittent for base load power generation in countries like Switzerland or further north will result in unavoidable net energy loss. This applies both to the technologies considered, to the available data from the original study and to newer data from recent studies.
Article
Full-text available
The transition to renewable energies will intensify the global competition for land. Nevertheless, most analyses to date have concluded that land will not pose significant constraints on this transition. Here, we estimate the land-use requirements to supply all currently consumed electricity and final energy with domestic solar energy for 40 countries considering two key issues that are usually not taken into account: (1) the need to cope with the variability of the solar resource, and (2) the real land occupation of solar technologies. We focus on solar since it has the highest power density and biophysical potential among renewables. The exercise performed shows that for many advanced capitalist economies the land requirements to cover their current electricity consumption would be substantial, the situation being especially challenging for those located in northern latitudes with high population densities and high electricity consumption per capita. Assessing the implications in terms of land availability (i.e., land not already used for human activities), the list of vulnerable countries enlarges substantially (the EU-27 requiring around 50% of its available land), few advanced capitalist economies requiring low shares of the estimated available land. Replication of the exercise to explore the land-use requirements associated with a transition to a 100% solar powered economy indicates this transition may be physically unfeasible for countries such as Japan and most of the EU-27 member states. Their vulnerability is aggravated when accounting for the electricity and final energy footprint, i.e., the net embodied energy in international trade. If current dynamics continue, emerging countries such as India might reach a similar situation in the future. Overall, our results indicate that the transition to renewable energies maintaining the current levels of energy consumption has the potential to create new vulnerabilities and/or reinforce existing ones in terms of energy and food security and biodiversity conservation.
Book
This bold and controversial argument shows why energy transitions are inherently complex and prolonged affairs, and how ignoring this fact raises unrealistic expectations that the United States and other global economies can be weaned quickly from a primary dependency on fossil fuels. Energy transitions are fundamental processes behind the evolution of human societies: they both drive and are driven by technical, economic, and social changes. In a bold and provocative argument, Energy Transitions: History, Requirements, Prospects describes the history of modern society's dependence on fossil fuels and the prospects for the transition to a nonfossil world. Vaclav Smil, who has published more on various aspects of energy than any working scientist, makes it clear that this transition will not be accomplished easily, and that it cannot be accomplished within the timetables established by the Obama administration. The book begins with a survey of the basic properties of modern energy systems. It then offers detailed explanations of universal patterns of energy transitions, the peculiarities of changing energy use in the world's leading economies, and the coming shifts from fossil fuels to renewable conversions. Specific cases of these transitions are analyzed for eight of the world's leading energy consumers. The author closes with perspectives on the nature and pace of the coming energy transition to renewable conversions.
Article
Attempts to assess the possibility of deriving all energy from renewable sources typically deal only with the aggregate amount of energy required, and do not consider the implications and difficulties arising from the need for differing forms of energy. In a 100% renewable system some of these forms will have to be provided by conversion from others, mostly from electricity. Conversion involves inefficiencies, losses, and embedded energy costs of infrastructures, and thus energy and dollar costs. In this study an attempt has been made to determine the magnitude and effect of this general problem, by beginning with estimates of the quantities required by different sectors and of the different forms they use. How these needs might best be met in a renewable system is then considered. Although there is insufficient data to enable confident conclusions, this “disaggregated” approach indicates that a 100% renewable system to meet Australian energy demand would involve costs that would probably constitute an unacceptably large fraction of GDP.
Article
Concentrating solar power (CSP) capacity has expanded slower than other renewable technologies and its costs are still high. Until now, there have been too few CSP projects to derive robust conclusions about its cost development. Here we present an empirical study of the cost development of all operating CSP stations and those under construction, examining the roles of capacity growth, industry continuity, and policy support design. We identify distinct CSP expansion phases, each characterized by different cost pressure in the policy regime and different industry continuity. In 2008–2011, with low cost pressure and following industry discontinuity, costs increased. In the current phase, with high cost pressure and continuous industry development, costs decreased rapidly, with learning rates exceeding 20%. Data for projects under construction suggest that this trend is continuing and accelerating. If support policies and industrial structure are sustained, we see no near-term factors that would hinder further cost decreases.
Conference Paper
SHAMS 1 (“Shams” means “Sun” in Arabic) Concentrated Solar Power plant is a very successful example of a modern plant, which combines the known configuration of a parabolic trough technology with the well-established power generation technologies operated at 540°C live steam temperature while respecting the specific requirement of the daily starts and shutdowns. In addition to the high live steam temperature challenge and being located in the middle of the desert approx. 120 km south west of the city of Abu Dhabi, the plant has to face, the plant has to fact several atmospheric challenges like the high dust concentration, wind storms, and high ambient temperature. This paper, written jointly by Shams Power Company – the project and operating company and MAN Diesel & Turbo – the steam turbine original manufacturer, describes the challenges in optimizing the design of the steam turbine to fulfill the requirement of fast start up while operating the plant on daily transient pattern for minimum 30 years. It also addresses the several atmospheric challenges and how the project and operating company has overcame them. Finally, the paper gives a snap shot on the operational experience and record of the plant showing that despite the very challenging environment, the budgeted target has been exceeded in the first two years of operation.
Article
A number of analyses, meta-analyses, and assessments, including those performed by the Intergovernmental Panel on Climate Change, the National Oceanic and Atmospheric Administration, the National Renewable Energy Laboratory, and the International Energy Agency, have concluded that deployment of a diverse portfolio of clean energy technologies makes a transition to a low-carbon-emission energy system both more feasible and less costly than other pathways. In contrast, Jacobson et al. [Jacobson MZ, Delucchi MA, Cameron MA, Frew BA (2015) Proc Natl Acad Sci USA 112(49):15060-15065] argue that it is feasible to provide "low-cost solutions to the grid reliability problem with 100% penetration of WWS [wind, water and solar power] across all energy sectors in the continental United States between 2050 and 2055", with only electricity and hydrogen as energy carriers. In this paper, we evaluate that study and find significant shortcomings in the analysis. In particular, we point out that this work used invalid modeling tools, contained modeling errors, and made implausible and inadequately supported assumptions. Policy makers should treat with caution any visions of a rapid, reliable, and low-cost transition to entire energy systems that relies almost exclusively on wind, solar, and hydroelectric power.