PROGRESS IN PHOTOVOLTAICS: RESEARCH AND APPLICATIONS

Prog. Photovolt: Res. Appl. 2014; 22:427–439

Published online 14 September 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/pip.2287

RESEARCH ARTICLE

Modeling of concentrating photovoltaic and thermal

systems
Henning Helmers1*, Andreas W. Bett1, Jürgen Parisi2 and Carsten Agert3
1
Fraunhofer Institute for Solar Energy Systems ISE, Heidenhofstr. 2, 79110 Freiburg, Germany
2
University of Oldenburg, Department of Physics, Energy and Semiconductor Research Laboratory, Carl-von-Ossietzky-Str. 9-11,
26129 Oldenburg, Germany
3
NEXT ENERGY, EWE-Research Centre for Energy Technology, Carl-von-Ossietzky-Str. 15, 26129 Oldenburg, Germany

ABSTRACT
An energy balance model for concentrating photovoltaic and thermal (CPVT) systems is presented. In the model, the CPVT
system and its environment are represented using a set of input parameters. The main outputs of the model are the system’s
electrical and thermal efficiencies. The model accounts for optical losses. Thermal losses are derived from a thermal
network model of the hybrid receiver. The solar cell performance is modeled as a function of the temperature and the
irradiance. The robustness of the model is demonstrated by a sensitivity analysis of all input parameters.
The influence of the operating temperature on the electrical and thermal performances and the overall efficiency of the
CPVT system are discussed. The limiting cases of maximum electrical and thermal power outputs are presented. Further,
the influence of the concentration ratio on the electrical and thermal performance and on the partitioning of these two power
outputs is analyzed in detail. It is shown that high concentration reduces the thermal losses considerably and increases
the electrical efficiency. At concentration ratios above 300, the system operates with an overall efficiency of 75% at
temperatures up to 160
C. Copyright © 2012 John Wiley & Sons, Ltd.

KEYWORDS
photovoltaic/thermal; co-generation; combined heat and power; solar concentrator; concentrating photovoltaics; modeling

*Correspondence
Henning Helmers, Fraunhofer Institute for Solar Energy Systems ISE, Heidenhofstr. 2, 79110 Freiburg, Germany.
E-mail: [email protected]

1. INTRODUCTION receiver systems use a dish or a field of heliostats to

concentrate sun-light onto a comparably larger photovoltaic
Highly concentrating photovoltaic (HCPV) systems are dense array receiver. Therefore, in central receiver systems,
currently entering the market on industrial scale. These passive cooling is insufficient and active cooling is applied
systems include concentrating optics with high concentra- to protect the receiver from overheating.
tion ratios (C) well above 300. Consequently, the incident An active thermal loop further enables heat transfer
solar power collection is decoupled from its conversion from the central receiver to a thermal load. A schematic
into actual power output. Hence, in HCPV systems, the drawing of such a combined concentrating photovoltaic
proportional costs of the photovoltaic converter are signif- and thermal (CPVT) system is shown in Figure 1. Because
icantly reduced. This cost reduction allows the utilization the heat is acquired as usable energy, the total solar power
of expensive but highly efficient multi-junction solar cells conversion efficiency is significantly increased. In addi-
in an economical manner. Solar cells with efficiencies tion, the high concentration ratios enable to supply heat
approaching 40% are currently commercially available at temperatures in the magnitude of 100
C and above.
[1,2]. On the module level, solar power conversion effi- Heat at these temperatures is of particular interest in
ciencies above 30% have been reported [3–5]. various applications. Direct applications range from domestic
Modular HCPV systems are usually passively cooled, heating to industrial process heating. Regarding the latter, a
and thus a large part of the already collected solar power large heat demand has been identified in manifold industrial
is lost as heat to the environment. In contrast, central sectors at temperatures below 160
C [6,7]. In addition to

Copyright © 2012 John Wiley & Sons, Ltd. 427

Modeling of CPVT systems H. Helmers et al.

Furthermore, the influence of the concentration ratio on the

thermal and electrical performance and on the partitioning
of these two power outputs is investigated, and the
advantages of high concentration are elaborated.

2. ENERGY BALANCE MODEL

The theoretical description presented in this paper is based on
an energy balance approach under steady-state conditions.
This approach implies the basic assumptions of steady
ambient conditions and steady load behavior.
Figure 2 shows a schematic drawing of the energy
balance. The assembly of the tracker, concentrator, and
receiver is considered to constitute the CPVT system.
The entire installation includes all thermal and electrical
Figure 1. Schematic drawing of a concentrating photovoltaic components behind the receiver, such as the supply piping,
and thermal system. intermediate heat exchangers, storage units, and pumps, as
well as the electrical wiring and inverter, and all controls
and monitoring equipment.
these direct applications, also polygeneration approaches It should be emphasized that the CPVT system defines
exist. These approaches aim to further generate either cooling the relevant boundaries of the presented model. No further
power or drinking water from the heat using solar cooling and power losses such as parasitic electrical power consump-
air conditioning [8–10] or solar desalination [11,12] tech- 0 el

nologies. High efficient double-effect absorption chillers tion from components ( P par ) or electrical and thermal
operate at temperatures in the range of 100–160
C [13].
0 el 0 th
power losses at the installation level (P loss and P loss ) were
Thermal desalination using membrane distillation considered. These losses strongly depend on the design,
requires temperatures from 60 to 90
C [14]. dimensioning and application of the entire CPVT
The operating temperature of the CPVT system deter- installation.
mines the available applications of the produced heat and In the following paragraphs, the model is described in
their respective performances. However, the operating detail. All input parameters of the model are listed in Table I.
temperature also influences the overall system perfor-
mance. An energy balance model has been developed to
2.1. General energy balance
investigate the influences of the temperature and the
concentration ratio on the system performance in detail.
The thermal power output Pth
out of a solar thermal collector
Non-concentrating combined photovoltaic and thermal
is given by
(PVT) collectors have previously been treated with
mathematical descriptions (e.g., Reference [15–17]).
Zondag et al. [18] presented a study comparing mathemat- Pth _ p ΔT
out ¼ mc (1)
ical descriptions of a non-concentration flat-plate PVT
collector at various levels of detail. Even so, the literature
on the performance modeling of CPVT systems remains
fragmented, especially in the case of highly concentrating
systems. Rosell et al. [19] set up a model for a low-
concentration system (Si cells, parabolic trough, C = 30).
Coventry [20] presented a mathematical description of a
low-concentration CPVT system (Si cells, Fresnel trough,
C = 11). Kribus et al. [21] published an evaluation of
a high-concentration CPVT system (triple-junction
cells, paraboloidal dish, C = 500). However, none of
these studies investigated the influence of concentra-
tion on the power output of a CPVT system in
particular. Figure 2. Schematic drawing of the energy balance of a concen-
In this paper, an energy balance model of a CPVT trating photovoltaic and thermal installation. Primed quantities
system is presented in detail. The model is applied to are related to power outputs and losses at the installation level.
analyze the hybrid performance, and focuses particularly Within the boundaries of the concentrating photovoltaic and
on the partitioning of the thermal and the electrical power thermal system (consisting of the tracker, concentrator, and
outputs and how they depend on the operating temperature. receiver), these losses are not considered.

428 Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd.
DOI: 10.1002/pip
H. Helmers et al. Modeling of CPVT systems

Table I. Overview of the input parameters of the model. factor S = Aap/Atotal that accounts for shading of the optics
Category Subject Symbols
by the receiver and necessary mounting structures leading
to an effective aperture area Aap, the mirror reflectance Ropt,
Hydraulics Thermal loop Tin, p, V an intercept factor g that accounts for spillage losses as
Cooling fluid properties cp, r well as possible losses due to homogenizing secondary
Optics Concentrator Atotal, Ropt, S optics, reflection from the cover of the receiver Rcov, reflec-
Intercept factor g tion at the absorber Rabs, and an alignment factor # that
Alignment and tracking # accounts for further optical losses in a real system caused
Receiver Geometry Arec, aPV by possible misalignment of the optics and receiver or
Encapsulation Bencap, hencap, ecov non-ideal tracking accuracy:
Reflectance Rcov, RPV, Rnon-PV
Solar cells
PV(TRef), tk,rel
Thermal design hhe
opt ¼ SRopt gð1  Rcov Þð1  Rabs Þ# (8)
Ambient conditions Gb, Tamb, vWind
Because the absorber is only partially packed with solar
cells, as described by the packing factor aPV, Rabs is given
where m_ ¼ rV_ is the mass flow rate through the collector, by
which is determined from the density of the cooling fluid
r and its volume flow rate V._ cp is the specific heat capacity Rabs ¼ RPV aPV þ Rnon-PV ð1  aPV Þ (9)
of the cooling fluid. ΔT = Tout  Tin is the temperature
difference between the outlet temperature Tout and the inlet where RPV and Rnon-PV are the reflectances of the solar
temperature Tin at the receiver. The mean fluid temperature cells and the area that is not covered by solar cells,
inside the receiver Tm is given by respectively.
It should be noted that the fraction of the incident light
Tin þ Tout
Tm ¼ (2) that is absorbed inside the encapsulation is not considered
2 among the optical losses. This power is converted into heat
The basic energy balance equation of the investigated and thus, from the viewpoint of a co-generating system,
system with incoming radiation power Pin, total power not lost.
output Pout, and power losses Ploss can be written as follows:
2.3. Thermal losses
Pin ¼ Pout þ Ploss (3)
The thermal losses at the receiver can be divided into
The incoming radiation power Pin is given by convective and radiative losses:

Pin ¼ Atotal Gb (4) th; conv

loss ¼ Ploss
Pth þ Pth;
loss
rad
(10)

where Atotal is the total collector area of the system and Gb is Both of these loss mechanisms are described by a heat
the incident beam radiation, that is, the direct normal transfer coefficient h, which accounts for heat transfer from
irradiance. The total power output of the co-generating system the cover of the receiver (at temperature Tcov) to the envi-
consists of a thermal part Pth el
out and an electrical part Pout : ronment. The reference temperature for the convective
losses is the ambient temperature Tamb. Regarding radiative
Pout ¼ Pth
out þ Pout
el
(5) losses, the receiver undergoes radiative exchange with the
mirror and the surrounding terrain, which are approxi-
The power losses can be divided into losses of either mated as black bodies at ambient temperature Tamb as well.
optical or thermal nature (Popt th
loss and Ploss ):
Pth;
loss
conv
¼ hconv Arec ðTcov  Tamb Þ (11)
Ploss ¼ Popt
loss þ Pth
loss (6)

Pth;
loss
rad
¼ hrad Arec ðTcov  Tamb Þ (12)
2.2. Optical losses

To derive the optical losses of the system, the total optical The characteristic area used in Equations (11) and (12)
efficiency
opt of the system is defined as follows: is the total receiver area Arec, which implies the assumption

 that the edges and the rear side of the receiver are ideally
Popt
loss ¼ Pin  1 
opt (7) insulated with surface temperatures equal to that of the
environment.
The optical efficiency considers the following factors, From Equations (10) to (12) and (4), the following
listed in the order of the incident light path: a shading proportionality is derived:

Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd. 429
DOI: 10.1002/pip
Modeling of CPVT systems H. Helmers et al.

 
Pth Arec 1
PV ðTabs Þ ¼
PV ðTRef Þ 1 þ tk;rel ðTabs  TRef Þ (18)
loss
/ ¼ (13)
Pin Atotal Cgeo
where
PV(TRef) is the efficiency at a reference temperature
where Cgeo is the geometric concentration ratio. This result TRef. TRef is chosen to be 25
C to provide a direct link to
clearly demonstrates that the thermal losses decrease in measurements taken under standard test conditions. tk,rel
inverse proportion to the concentration ratio. is the relative temperature coefficient of the photovoltaic
The convective heat transfer coefficient is modeled as a efficiency.
function of the wind speed vWind using the following The linear relation in Equation (18) requires special
correlation (in SI units) [22]: attention when treating multi-junction solar cells composed
of low-band-gap subcells, such as the germanium subcell
hconv ¼ 2:8 þ 3vWind (14) in the industrial standard triple-junction solar cell: Because
the photo-voltage decreases as the temperature increases,
The heat transfer coefficient of radiative losses to the the photo-voltage of a low-band-gap material becomes
environment is modeled as follows [23]: almost 0 V at high temperatures. Consequently,
PV(Tabs)
must exhibit nonlinear behavior at a certain limit temp-
 
Tcov 4  Tamb 4 erature. However, extrapolating the measured data from
hrad ¼ secov  (15) Nishioka et al. [25], it can be assumed that this nonlinear-
ðTcov  Tamb Þ
ity occurs at temperatures above 200
C for concentration
where s is the Stefan–Boltzmann constant and ecov is the ratios above 150. Furthermore, it has been demonstrated
emissivity of the cover. experimentally for temperatures up to 120
C and concen-
tration ratios up to 200 that the linear relationship of
2.4. Electrical performance Equation (18) is a good approximation to the efficiency
of a triple-junction solar cell [26].
Regarding the total electrical power output Pelout , the total
2.4.2. Photovoltaic efficiency as a function of
electrical efficiency of the system
el is defined by the fol-
irradiance
lowing equation:
To account for the efficiency’s dependence on irradiance,
Pelout ¼ Pin 
el (16) both the photovoltaic efficiency at the reference temperature

PV(TRef) and its temperature coefficient tk,rel in Equation
To model
el, the portion of the radiation that actually (18) are modeled as functions of the irradiance incident on
hits the solar cells (defined as
opt,PV) is multiplied by the the solar cells GPV. GPV is given by
actual photovoltaic efficiency of the receiver. This product
is expressed as follows: Pin 
opt;PV Atotal Gb 
opt;PV
GPV ¼ ¼ / Cgeo (19)
  APV Arec aPV

el ¼ S  Ropt  g  ð1  Rcov Þ  1  Bencap  #  aPV  x 
PV
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} where APV is the total solar cell area of the receiver.

opt;PV
A general expression for the solar cell efficiency
PV is
(17)
Pmp
where Bencap describes the absorptance of the encapsula-
PV ¼ (20)
Pin;PV
tion. x is a constant loss factor that expresses the average
effect of inhomogeneous illumination on the photovoltaic
performance of the receiver, especially with respect to where Pmp is the maximum power of the solar cells. The
specific module interconnection.
PV is the photovoltaic incident radiation power on the solar cells Pin,PV depends
efficiency of the solar cells. linearly on the irradiance GPV:
To realistically simulate the photovoltaic efficiency of
the solar cells, the two major known dependencies are Pin;PV ¼ APV GPV (21)
implemented in the model: (i) the photovoltaic efficiency is
a function of the cell temperature Tabs; and (ii) it is a function The following expression describes the dependence of
of the irradiance GPV that the solar cells are exposed to, Pmp on the irradiance:
which is correlated with the concentration ratio.
Pmp ¼ Isc Voc FF (22)
2.4.1. Photovoltaic efficiency as a function of
temperature
Regarding cell temperature, a common practice is to with short-circuit current Isc, open-circuit voltage Voc, and
model the solar cell efficiency as a linear function of the fill factor FF. Approximating the multi-junction solar cell
cell temperature [24]. This correlation is expressed as using the one-diode model leads to the following relations
follows: for Isc and Voc:

430 Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd.
DOI: 10.1002/pip
H. Helmers et al. Modeling of CPVT systems

Isc ffi Iph ¼ cAPV GPV (23) tk;rel ðGPV Þ ¼ b4 þ b5 lnGPV (29)

cAPV
Voc ffi nVT ln GPV (24)
I0 where b4 < 0 and b5 > 0 are fitting parameters.

2.5. Temperatures
with the photo-generated current Iph, the dark saturation
current I0 ≪ Iph, the ideality factor n, the thermal voltage
From the previous sections, three relevant receiver
VT = kTabs/q, and a collection constant c.
temperatures are identified: The cover temperature Tcov
The ideal FF is only a function of the Voc. The FF can
determines the thermal losses (compare Equations (11)
be described to a good approximation using the normalized
and (12)). The absorber temperature Tabs determines the
open-circuit voltage voc = Voc/(nVT), as follows [27]:
photovoltaic efficiency (compare Equation (18)) and thus
the electrical power output. Finally, for a given inlet
voc  lnðvoc þ 0:72Þ temperature Tin, the mean fluid temperature Tm defines the
FF ffi (25)
voc þ 1 thermal power output (compare Equations (1) and (2)).
The receiver is modeled as a one-dimensional thermal
network (Figure 3), similar as suggested, for example, in
However, as the concentration increases, the power loss
References [19,20,23,30]. The one-dimensional approach
ΔP in the parasitic series resistance Rs becomes significant.
implies that the lateral temperature distribution across the
With Joule’s law and r = Imp/Isc (i.e., the ratio of the current
receiver is expressed by the respective mean temperature.
at maximum power to the short-circuit current) the power
The rear side of the receiver and the edges are assumed
loss can be expressed as follows:
to be ideally insulated. Regarding the vertical dimension,
the nodes of the thermal network are connected with
ΔP ¼ Rs Imp 2 ffi Rs ðcrAPV GPV Þ2 (26) thermal resistances expressed as heat transfer coefficients
hi. The heat transfer coefficients for convection hconv and
radiation hrad to the environment are given from Equations
When the series resistance losses are considered, Equation (14) and (15). hencap defines the conductive heat transfer
(20) becomes coefficients through the encapsulation. hhe is determined
by the sum of the thermal resistances of the interface
Pmp  ΔP

PV ¼ (27)
Pin;PV

Finally, inserting Equations (21)–(26) into Equation (27)

provides the following modeling function for the photovoltaic
efficiency:

PV ðGPV Þ ¼ b1  lnðb2 GPV Þ

lnðb2 GPV Þ  lnf lnðb2 GPV Þ þ 0:72g
 b3 GPV
lnðb2 GPV Þ þ 1
(28)

The constants b1 = cnVT, b2 = cAPV/I0, and b3 = Rsc2r2APV

are used as independent fitting parameters.
The literature dealing with the dependence of the
temperature coefficient on the irradiance is scarce. A
review of the available measured data for triple-junction solar
cells [26,28,29] shows that the cells become significantly less
sensitive to temperature as the concentration increases.
The temperature coefficient of the efficiency is primarily Figure 3. Schematic drawing of the thermal network model of
determined by the Voc of which the temperature coefficient the receiver that is used to determine the relevant temperatures
is proportional to the logarithm of the irradiance [29]. Moti- of the cover Tcov, absorber Tabs, and fluid Tm. The thermal resis-
vated by this finding, tk,rel(GPV) is modeled by fitting the fol- tances are modeled using the respective heat transfer coeffi-
lowing equation to the empirical data: cients hi. The arrows indicate the heat sources and sinks.

Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd. 431
DOI: 10.1002/pip
Modeling of CPVT systems H. Helmers et al.

water/heat exchanger, the layer structure of the heat ex- 2.6. Application of the model
changer, and the thermal interface material between the so-
lar cells and the heat exchanger. The model has been applied to investigate the system
The heat sources and sinks in the network are imple- performance of an example system. To do so, the input
mented at the respective nodes. The absorbed power inside parameters are set to the ones listed in Table II. Where
the encapsulation is calculated as available, the values are taken from literature or data
sheets, and reasonable estimations have been made when
Pencap
abs ¼ Pin 
opt Bencap (30) data was unavailable. The concentration ratio is varied by
variation of the total collector area for a fixed size of
The major heat source is provided by the power that is the receiver.
absorbed at the absorber, which is given by The solar cells are assumed to be lattice-matched triple-
junction cells. The corresponding parameters b1 through b5
  were obtained by fitting Equations (28) and (29) to
abs ¼ Pin 
opt  1  Bencap
Pabs (31)
available data from [36] and [29], respectively. The result-
ing photovoltaic efficiency
PV(GPV, Tabs) as a function of
The electrical power output Pelout from the solar cells is
the irradiance and temperature is presented in Figure 6.
treated as a heat sink at the absorber.
The non-photovoltaic area of the receiver is assumed to
Finally, the fluid cycle transports heat into and out of
be painted with a highly absorbent solar paint. The
the receiver, as given by
assumed encapsulation consists of 0.5 mm high-transparency
silicone covered by 1 mm borosilicate glass. Regarding the
_ p Tin=out
Pfluid;in=out ¼ rVc (32) active cooling, a high-performance micro-channel heat
exchanger is assumed. The assumed thermal interface be-
where the density r = r(Tm, p) and the specific heat cp = cp tween the solar cells and the heat exchanger is SnAg solder.
(Tm, p) of the cooling fluid are modeled as functions of the The cooling fluid is water, and its thermodynamic properties
mean fluid temperature Tm and pressure p. are modeled according to IAPWS-IF97 [33]. Furthermore, a
This one-dimensional network approach provides an clean solar mirror with ideal alignment and tracking is as-
elegant way to derive the relevant temperatures because it sumed. Shading is neglected. The ambient conditions are
leads to a set of equations that can be solved by linear modeled using typical values for concentrator sites.
algebra:

! !
2.7. Sensitivity analysis
HT ¼ P (33)
A sensitivity analysis has been performed to investigate the
where the heat transfer matrix H is given by sensitivity of the model to its input parameters. At fixed
!
0 1 point X in the parameter space, the magnitude of the
hconv þ hrad þ hencap hencap 0 influence of a specific parameter xi on the (electrical,
H¼@ hencap hencap þ hhe hhe A thermal or overall) system efficiency
j can be assessed
0 hhe hhe by the elasticity of
j with respect to xi, given by
(34)

 @
j xi
!
The temperature vector T is defined as E
j ; xi ¼ (37)
@xi
j !
X
0 1
! @ Tcov A
T ¼ Tabs (35) The value of the elasticity corresponds to the relative
Tm change of the efficiency per unit of relative change of the
input parameter. This one-factor-at-a-time approach does
! not consider interdependence between parameters. For
The power vector P, that represents heat sources and
sinks, is given by !
the sensitivity analysis, X has been set to the values in
0 encap 1 Table II, and a concentration ratio of Cgeo = 500 and an
! @ Pabs þ ðhabs conv þ hrad ÞTamb
A
inlet temperature of Tin = 80
C have been used.
P¼ Pabs  Poutel (36) In Figure 4, the percentage change of the overall system
Pfluid;in  Pfluid;out efficiency
Σ =
el +
th is plotted against the percentage
change of the input parameter value. Each parameter’s
The quantities Pel out ¼ Pout ðTabs Þ; h rad = h rad (T cov ), assumed value is altered by 15%. In the representation
el

r = r(Tm), and cp = cp(Tm) are functions of the respective presented in Figure 4, the slope at the y-intercept equals
temperatures. Consequently, Equation (33) is implicit and the respective elasticity of each parameter. The elasticities
an iterative calculation procedure is conducted to deter- of the electrical, thermal, and overall efficiencies with
mine the exact values of the temperatures. respect to all parameters are given in Table III.

432 Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd.
DOI: 10.1002/pip
H. Helmers et al. Modeling of CPVT systems

Table II. Parameters of the example system used for the application of the model. Where available, the values were taken from
literature or data sheets, and reasonable estimations have been made when data were unavailable (compare text).

Symbol Unit Value Explanation

aPV 1 0.98 From Reference [31]

Arec mm² 150  150 Comparable to Reference [32]
Bencap 1 0.013 Own estimation
cp J/kg/K cp(Tm, p) After IAPWS-IF97 [33]
Gb W/m² 850 After Reference [34]
hencap W/m²/K 2.40 10^2 Own estimation
hhe W/m²/K 2.69 10^4 Own estimation
p bar 8 No evaporation below 170
C [33]
Rcov 1 0.07 Borosilicate glass
Rnon-PV 1 0.04 Solar paint
Ropt 1 0.944 From Reference [35]
RPV 1 0.08 Concentrator solar cell
S 1 1 Shading neglected
tk,rel %rel/K See Figure 6 After Reference [29]


Tamb C 20 After Reference [34]
vwind m/s 3 After IEC 62108 [37]
V_ l/min 8 Own estimation
g 1 0.95 Own estimation
ecov 1 0.88 From Reference [38]

PV(25
C) % See Figure 6 After Reference [36]

opt 1 0.76 From Equation (8)
# 1 1 Ideal alignment and tracking
x 1 0.911 After Reference [32]
r kg/m³ r(Tm, p) After IAPWS-IF97 [33]

An increase in these parameters directly increases the op-

tical efficiency of the system, and as a result, the system
performance is increased. The opposite, but considerably
less pronounced behavior holds for the reflectances
Rcov and Rabs, the latter of which is mainly determined
by RPV and aPV. The elasticities of the overall efficiency
with respect to these reflectances range in magnitude from
-0.04 to -0.09%rel/%rel. The model is robust against
changes to the input values of all other parameters. The
elasticities of the overall efficiency with respect to these
parameters are below 0.01%rel/%rel and thus the overall
efficiency changes less than 0.15%rel for a change of
15%rel in the value of the input parameter.
Also, the electrical and thermal efficiencies show the
Figure 4. Plot of the percentage change of the overall system highest elasiticities with respect to the optical parameters
efficiency
Σ (i.e., the sum of electrical and thermal efficiencies) described earlier. Further, the solar cell parameters b1 and
against the percentage change of the parameter value of xi for b4, and the inhomogeneity factor x have a major influence
the most significant parameters. Both percentages are relative on the model. However, these parameters only affect the
!
to their baseline values at the fixed point X in the parameter partitioning of the electrical and thermal power outputs,
space. Each parameter’s value is altered by 15%. In this repre-
and therefore have no impact on the overall efficiency.
sentation, the slope of each parameter at the y-intercept equals
the respective elasticity.

3. DISCUSSION
The analysis demonstrates that the model is most sensitive
to the optical parameters of the system. The model exhibits The model is used to analyze the influences of the operating
elasticities of the overall efficiency of approximately temperature and the concentration ratio on the system’s per-
+1%rel/%rel to the reflectance of the mirror Ropt, the shading formance. To illustrate the influences of these two
factor S, the intercept factor g, and the alignment factor #. parameters on the thermal behavior, Figure 5 presents the

Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd. 433
DOI: 10.1002/pip
Modeling of CPVT systems H. Helmers et al.

Table III. The elasticities of the electrical, thermal, and overall represents different geometric concentration ratios Cgeo.
efficiencies with respect to all input parameters of the model. Several observations can be made from this figure: The
The table is sorted in descending order of the elasticity of the thermal efficiency decreases as the temperature increases
overall efficiency. because of the increasing thermal losses. However, the
xi E(
el, xi) E(
th, xi) E(
Σ, xi) magnitude of these thermal losses decreases drastically
as the concentration ratio increases (compare Equation
[%rel/%rel] [%rel/%rel] [%rel/%rel] (13)). In the temperature range shown (up to a mean fluid
Ropt +1.01 +1.00 +1.00
temperatures of 160
C), radiation losses, which exhibit
S +1.01 +1.00 +1.00 nonlinear behavior with temperature (compare Equation
g +1.01 +1.00 +1.00 (15)), can be neglected in good approximation. This
# +1.01 +1.00 +1.00 approximation is especially valid at high concentrations.
Cgeo +0.01 > 0.01 < +0.01 At Tm = Tamb, at which point thermal losses do not exist,
Gb +0.01 > 0.01 < +0.01 the thermal efficiency is equal to the optical efficiency
Tamb > 0.01 < +0.01 < +0.01 of 76%.
hencap > 0.01 < +0.01 < +0.01 Finally, at high concentration ratios above 300,
hhe < +0.01 > 0.01 < +0.01 thermal losses become nearly negligible at mean fluid
b1 +1.06 0.57 < +0.01 temperatures up to 160
C. As a result, the maximum
x +1.01 0.54 < +0.01 thermal efficiency in this temperature range is nearly
V_ < +0.01 > 0.01 < +0.01 independent of temperature and equal to the optical
b5 +0.14 0.07 < +0.01 efficiency.
b2 +0.05 0.03 < +0.01 To analyze the total electrical efficiency of the system,
p > 0.01 < +0.01 > 0.01 its relation to the photovoltaic conversion efficiency of
b3 0.05 +0.03 > 0.01 the solar cells is examined. The functions utilized for
b4 0.21 +0.11 > 0.01

PV(GPV) and tk,rel(GPV) are plotted in Figure 6. As can
Arec > 0.01 < +0.01 > 0.01
be seen, the photovoltaic efficiency increases with the
Rnon-PV < +0.01 > 0.01 > 0.01
concentration ratio up to a maximum value. The concen-
Bencap 0.01 < +0.01 > 0.01
tration ratio related to this maximum efficiency is
ecov < +0.01 > 0.01 > 0.01
vwind < +0.01 > 0.01 > 0.01
determined by the specific cell design, especially the front
Tin 0.09 +0.04 > 0.01
grid metallization. At higher concentrations, the effi-
aPV +1.01 0.61 0.04 ciency decreases because the increasing series resistance
Rcov 0.08 0.08 0.08 losses become significant.
RPV < +0.01 0.13 0.09 Furthermore, the photovoltaic efficiency decreases with
increasing temperature. This loss in efficiency originates
primarily from the decrease in the open-circuit voltage as
the temperature increases. However, the efficiency loss
maximum thermal efficiency
th,max of the system in pure
thermal operation with Pelout ¼ 0, that is, either short-circuit
or open-circuit operation. The abscissa of this plot is the
reduced temperature Tm  Tamb. The family of curves

Figure 6. Temperature coefficient tk,rel (dash-dotted line) and

the photovoltaic efficiency
PV(25
C) (solid line) used for the
simulation as functions of the irradiance GPV. The symbols
Figure 5. The maximum thermal efficiency curve in pure thermal represent the empirical data from Reference [29] (stars) and
operation (i.e., Pout
el
¼ 0Þ is plotted as a function of the reduced Reference [36] (squares), respectively. Additionally, the resulting
temperature Tm  Tamb. The dashed arrow illustrates the effect photovoltaic efficiency
PV(GPV, Tabs) as a function of the
of increasing the concentration ratio. irradiance and temperature is plotted as a family of curves.

434 Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd.
DOI: 10.1002/pip
H. Helmers et al. Modeling of CPVT systems

with temperature is reduced as the concentration ratio by the electrical power output. Thus, as the temperature
increases. Consequently, in high concentration systems, increases, the increase in thermal losses is accompanied
the effect of the temperature on the electrical performance by a reduction in the electrical efficiency. At high
is significantly reduced. concentration ratios above 100, this effect leads to an
This effect is illustrated in Figure 7, in which the increase in the thermal efficiency with temperature,
maximum electrical efficiency for operation at maximum because the thermal losses become insignificant whereas
electrical power output is presented. The family of curves additional thermal energy is gained from the decrease in
illustrates the combined influence of the temperature and electrical efficiency. Finally, the thermal efficiency is
the irradiance on the photovoltaic conversion efficiency. nearly independent of the concentration ratio in the
The electrical efficiency decreases as the operating temper- range of Cgeo = 100 . . . 1000.
ature increases. However, the sensitivity to temperature When the overall efficiency, that is, the sum of the
(given by the slope in Figure 7) is significantly reduced electrical (Figure 7) and thermal (Figure 8) efficiencies, is
at high concentration ratios. considered, the same family of efficiency curves shown
The thermal efficiency for operation at maximum in Figure 5 is obtained. Thus, for operation at maximum
electrical power output is shown in Figure 8. Because electrical power output, the ordinate of Figure 5 should
the electrical power output of the photovoltaic cells be renamed the overall efficiency
Σ.
defines a heat sink in the receiver, this graph differs from Indeed, hybrid operation changes the partitioning of the
the graph shown in Figure 5. Compared with pure absorbed incident power into electrical and thermal power
thermal operation, the thermal power output is reduced outputs; it does not introduce further power losses. The
operations at maximum electrical power output (compare
Figures 7 and 8) and maximum thermal power output (i.
e., short-circuit or open-circuit operation, compare
Figure 5) define the limiting cases in which either the
electrical or thermal power output is maximized. The parti-
tioning of the total power output of a CPVT system into the
electrical and thermal power outputs is freely adjustable
between these two limits, depending on the specific
application.
It is emphasized that the generated heat Pth out and the
generated electricity Pelout are not equivalent. It is clear that
electricity has a higher value than heat, and heat at a high
temperature has a higher value than heat at a low tempera-
ture. Thermodynamically, this fact is expressed in the
definition of exergy, leading to weighting factors of unity
for electricity and the Carnot factor for heat. However,
Figure 7. Electrical efficiency for operation at maximum electrical the assessment criteria for an appropriate optimization of
power output as a function of the reduced temperature Tm  Tamb. a CPVT system’s power output depend strongly on the
The dashed arrows illustrate the influence of increasing the chosen point of view, which can be economically, ecolog-
concentration ratio. ically or a combination of both. Moreover, the system
design, namely the specific thermal application, influences
the definition of assessment criteria because the possible
substitution of other conventional technologies has to be
considered, for example, when the heat is used to drive
air conditioning or desalination systems. For a deeper
analysis and different approaches of valuating heat and
electricity, the reader is referred to the literature [39–43].

4. SUMMARY AND CONCLUSION

A theoretical model of CPVT systems has been presented.
This model is based on an energy balance treatment of the
tracker, concentrator, and receiver assembly. A sensitivity
analysis of all model parameters has been performed, and
Figure 8. Thermal efficiency for operation at maximum electrical the model is demonstrated to be robust to changes in the
power output as a function of the reduced temperature Tm  Tamb. input parameters. The strongest influence on the output of
The dashed arrow illustrates the influence of increasing the the model arises from the parameters that determine the
concentration ratio. optical efficiency of the system.

Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd. 435
DOI: 10.1002/pip
Modeling of CPVT systems H. Helmers et al.

The influences of the operating temperature and the hencap W/m²/K Conductive heat transfer coefficient through
concentration ratio on the system performance have been the encapsulation
discussed in detail. The partitioning of the absorbed inci- hhe W/m²/K Conductive heat transfer coefficient through
dent power into electrical and thermal power outputs the heat exchanger
depends on the mode of operation. The limiting cases of hrad W/m²/K Radiative heat transfer coefficient to the
maximum electrical and maximum thermal power output environment
have been presented. Depending on the application, the I0 A Dark saturation current
Imp A Current at maximum power output
total power output can be divided between these two limits.
Iph A Photo-generated current
It has been shown that high concentration ratios offer
Isc A Short-circuit current
several advantages to a hybrid PVT system. At high
m_ l/min Mass flow of the cooling fluid
concentration ratios above 300, the thermal losses become
n 1 Ideality factor
nearly negligible for mean fluid temperatures up to 160
C.
p bar Pressure in the thermal loop
Consequently, operation at these temperatures, which is of Pabs
abs
W Power absorbed by the absorber
special interest in various polygeneration approaches, encap
Pabs W Power absorbed in the encapsulation
becomes feasible. Furthermore, the electrical efficiency Pfluid,in/out W Heat flow into/out of the receiver due to flow
increases as the concentration ratio increases. This behav- of cooling fluid
ior is especially valid at higher temperatures because the Pin W Radiation power input
temperature coefficient decreases as the concentration ratio Pin,PV W Incident radiation power on the solar cells
increases. Finally, in high-concentration CPVT systems, Ploss W Total power loss
opt
overall conversion efficiencies of incident solar power to Ploss W Optical power loss
usable power output of 75% can be achieved at operating Ploss
th
W Total thermal power loss to environment
temperatures up to 160
C. th; conv
Ploss W Thermal power loss due to convection
th; rad
The presented model allows the optimization of CPVT Ploss W Thermal power loss due to radiation
systems. The effect of changes to any design parameter Pmp W Maximum power of the solar cells
on the total system performance can be derived. Further- Pout W Total power output of the CPVT system
more, for a given design and set of operating parameters, Pout
el
W Electrical power output of the CPVT system
the model enables the computation of the annual electrical Pout
th
W Thermal power output of the CPVT system
and thermal energy yields. ΔP W Power loss due to series resistance
0 el
P loss W Electrical power losses at the installation level
0 th
P loss W Thermal power losses at the installation
level
NOMENCLATURE 0 el
P par W Parasitic electrical power consumption at the
installation level
Physical constants r 1 Ratio of current at maximum power to short-
circuit current
k J/K Boltzmann constant Rabs 1 Mean reflectance of the absorber
q C Electron charge Rcov 1 Reflectance of the cover of the receiver
s W/m²/K4 Stefan–Boltzmann constant Rnon-PV 1 Reflectance of non-photovoltaic receiver area
Ropt 1 Reflectance of the concentrating mirror
Latin letters RPV 1 Reflectance of the solar cells
Rs Ω Series resistance
aPV 1 Packing factor S 1 Shading factor
Aap m² Effective aperture area tk,rel %rel/K Relative temperature coefficient of the solar
Arec m² Total receiver area cell efficiency
APV m² Total solar cell area Tabs K Mean temperature of the absorber/solar cells
Atotal m² Total collector area Tamb K Ambient temperature
Bencap 1 Absorptance inside encapsulation Tcov K Temperature of the cover of the receiver
c A/W Collection constant Tin K Inlet temperature of the fluid at the receiver
cp J/kg/K Specific heat capacity of the cooling fluid Tm K Mean temperature of the fluid inside the receiver
C 1 Concentration ratio Tout K Outlet temperature of the fluid at the receiver
Cgeo 1 Geometric concentration ratio TRef K Reference temperature
FF 1 Fill factor ΔT K Temperature rise of the cooling fluid across
G W/m² Irradiance the receiver
Gb W/m² Incident beam irradiance (direct voc 1 Normalized open-circuit voltage
normal irradiance) vwind m/s Wind speed
GPV W/m² Irradiance incident on the solar cells Voc V Open-circuit voltage
hconv W/m²/K Convective heat transfer coefficient to the VT V Thermal voltage
environment

(Continues) (Continues)

436 Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd.
DOI: 10.1002/pip
H. Helmers et al. Modeling of CPVT systems

Greek letters IEEE Journal of Photovoltaics 2011; 1(2): 213–218.

DOI: 10.1109/jphotov.2011.2172775
b1  5 Parameters of the photovoltaic efficiency fit 5. Araki K, Uozumi H, Egami T, Hiramatsu M, Miyazaki Y,
g 1 Intercept factor
Kemmoku Y, Akisawa A, Ekins-Daukes NJ, Lee HS,
ecov 1 Emissivity of the cover

el 1 Electrical efficiency of the system
Yamaguchi M. Development of concentrator modules

opt 1 Optical efficiency of the system with dome-shaped Fresnel lenses and triple-junction

opt,PV 1 Effective optical efficiency regarding the solar concentrator cells. Progress in Photovoltaics:
cells Research and Applications 2005; 13(6): 513–527.

PV 1 Photovoltaic conversion efficiency of the solar DOI: 10.1002/pip.643
cells 6. Schweiger H. POSHIP—The Potential of Solar Heat

th 1 Thermal efficiency of the system
for Industrial Processes. AIGUASOL Enginyeria:

Σ 1 Overall efficiency of the system
# 1 Alignment factor
Barcelona, 2001.
x 1 Inhomogeneity loss factor 7. Vannoni C, Battisti R, Drigo S. Potential for Solar
r kg/m³ Density of the cooling fluid Heat in Industrial Processes. University of Rome
“La Sapienza”: Madrid, 2008.
8. Mittelman G, Kribus A, Dayan A, Solar cooling with
concentrating photovoltaic/thermal (CPVT) systems.
ACKNOWLEDGEMENTS Energy Conversion and Management 2007; 48(9):
2481–2490. DOI: 10.1016/j.enconman.2007.04.004
The authors would like to thank Alexander Boos and 9. Balaras CA, Grossman G, Henning H-M, Infante
Simon P. Philipps for fruitful discussions. Henning
Ferreira CA, Podesser E, Wang L, Wiemken E. Solar
Helmers gratefully acknowledges the scholarship support
air conditioning in Europe—an overview. Renewable
of the German Federal Environmental Foundation (DBU).
This work was partly funded by the Federal Ministry for and Sustainable Energy Reviews 2007; 11(2): 299–314.
the Environment, Nature Conservation and Nuclear Safety DOI: 10.1016/j.rser.2005.02.003
(BMU) under the “KoMGen” project (#0327567A). The 10. Henning H-M. Solar assisted air conditioning of
authors are responsible for the content of this paper. buildings—an overview. Applied Thermal Engi-
neering 2007; 27(10): 1734–1749. DOI: 10.1016/j.
applthermaleng.2006.07.021
11. Mittelman G, Kribus A, Mouchtar O, Dayan A. Water
REFERENCES desalination with concentrating photovoltaic/thermal
1. Guter W, Kern R, Köstler W, Kubera T, Löckenhoff (CPVT) systems. Solar Energy 2009; 83(8): 1322–1334.
R, Meusel M, Shirnow M, Strobl G. III-V multijunc- DOI: 10.1016/j.solener.2009.04.003
tion solar cells—new lattice-matched products 12. Cipollina A, Micale G, Rizzuti L. Seawater Desalina-
and development of upright metamorphic 3 J cells. tion. Springer: Berlin Heidelberg, 2009.
Proceedings of the 7th International Conference on 13. Grossman G. Solar-powered systems for cooling,
Concentrating Photovoltaic Systems CPV-7, Las dehumidification and air-conditioning. Solar Energy
Vegas, USA, AIP (1407), 2011; 5–8. DOI: 10.1063/ 2002; 72(1): 53–62. DOI: 10.1016/s0038-092x(01)
1.3658282 00090-1
2. Ermer JH, Jones RK, Hebert P, Pien P, King RR, 14. Koschikowski J, Wieghaus M, Rommel M. Membrane
Bhusari D, Brandt R, Al-Taher O, Fetzer C, Kinsey GS, Distillation for Solar Desalination, Vol. 7. Springer:
Karam N. Status of C3MJ + and C4MJ production Berlin Heidelberg, 2009; 165–188.
concentrator solar cells at spectrolab. IEEE Journal of 15. Florschuetz LW. Extension of the Hottel–Whillier
Photovoltaics 2012; 2(2): 209–213. DOI: 10.1109/ model to the analysis of combined photovoltaic/ther-
jphotov.2011.2180893 mal flat plate collectors. Solar Energy 1979; 22(4):
3. van Riesen S, Gombert A, Gerster E, Gerstmaier T, 361–366. DOI: 10.1016/0038-092x(79)90190-7
Jaus J, Eltermann F, Bett AW. Concentrix solar’s 16. Cox III CH, Raghuraman P. Design considerations for
progress in developing highly efficient modules. flat-plate-photovoltaic/thermal collectors. Solar En-
Proceedings of the 7th International Conference of ergy 1985; 35(3): 227–241. DOI: 10.1016/0038-092x
Concentrating Photovoltaic Systems CPV-7, Las (85)90102-1
Vegas, Nevada, USA, AIP (1407), 2011; 235–238. 17. Bergene T, Løvvik OM. Model calculations on a flat-
DOI: 10.1063/1.3658334 plate solar heat collector with integrated solar cells,
4. Kinsey GS, Nayak A, Liu M, Garboushian V. Increasing Solar Energy 1995; 55(6): 453–462. DOI: 10.1016/
power and energy in Amonix CPV solar power plants. 0038-092x(95)00072-y

Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd. 437
DOI: 10.1002/pip
Modeling of CPVT systems H. Helmers et al.

18. Zondag HA, de Vries DW, van Helden WGJ, van 29. Siefer G, Bett AW. Analysis of temperature coeffi-
Zolingen RJC, van Steenhoven AA. The thermal and cients for III–V multi-junction concentrator cells.
electrical yield of a PV-thermal collector. Solar Energy Progress in Photovoltaics: Research and Applications.
2002; 72(2): 113–128. DOI: 10.1016/s0038-092x(01) DOI: 10.1002/PIP.2285
00094-9 30. Duffie JA, Beckman WA. Solar engineering of
19. Rosell JI, Vallverdú X, Lechón MA, Ibáñez M. Design thermal processes. Wiley: New Jersey, 2006.
and simulation of a low concentrating photovoltaic/ 31. Verlinden PJ, Lewandowski A, Bingham C, Kinsey
thermal system. Energy Conversion and Manage- GS, Sherif RA, Lasich JB. Performance and reliability
ment 2005; 46(18–19): 3034–3046. DOI: 10.1016/ of multijunction III–V modules for concentrator dish
j.enconman.2005.01.012 and central receiver applications. Proceedings of the
20. Coventry JS. A solar concentrating photovoltaic/thermal 4th World Conference on Photovoltaic Energy Con-
collector. PhD Thesis, Australian National University, version, Waikoloa, Hawaii, USA, 2006; 592–597.
Canberra, Australia, 2004. DOI: 10.1109/WCPEC.2006.279526
21. Kribus A, Kaftori D, Mittelman G, Hirshfeld A, 32. Löckenhoff R, Kubera T, Rasch KD. Water cooled TJ
Flitsanov Y, Dayan A. A miniature concentrating dense array modules for parabolic dishes. Proceedings
photovoltaic and thermal system, Energy Conversion of the 6th International Conference of Concentrating
and Management 2006; 47(20): 3582–3590. DOI: Photovoltaic Systems CPV-6, Freiburg, Germany,
10.1016/j.enconman.2006.01.013 AIP (1277), 2010; 43–46. DOI: 10.1063/1.3509229
22. Watmuff JH, Charters WWS, Proctor D. Solar and 33. Wagner W, Cooper JR, Dittmann A, Kijima J,
wind induced external coefficients—solar collectors. Kretzschmar HJ, Kruse A, Mares R, Oguchi K, Sato
Cooperation Mediterraneenne pour l’Energie Solaire, H, Stocker I, Sifner O, Takaishi Y, Tanishita I,
Revue Internationale d’Heliotechnique, 2nd Quarter, Trubenbach J, Willkommen T. The IAPWS industrial
1977: 56. formulation 1997 for the thermodynamic properties
23. Rabl A. Active Solar Collectors and Their Applications. of water and steam. Journal of Engineering for Gas
Oxford Universtiy Press: New York, USA, 1985. Turbines and Power 2000; 122(1): 150–184. DOI:
24. Emery K, Burdick J, Caiyem Y, Dunlavy D, Field H, 10.1115/1.483186
Kroposki B, Moriarty T, Ottoson L, Rummel S, Strand 34. Kurtz S, Muller M, Marion B, Emery K, McConnell R,
T, Wanlass MW. Temperature dependence of photo- Surendran S, Kimber A. Considerations for how to rate
voltaic cells, modules and systems. Proceedings of CPV. Proceedings of the 7th International Conference
the 25th IEEE Photovoltaic Specialists Conference, of Concentrating Photovoltaic Systems CPV-7, Las
Washington DC, USA, 1996; 1275–1278. DOI: Vegas, Nevada, USA, AIP (1407), 2011; 25–29.
10.1109/PVSC.1996.564365 DOI: 10.1063/1.3658287
25. Nishioka K, Takamoto T, Agui T, Kaneiwa M, Uraoka 35. Flabeg. Solar mirrors for CSP and CPV. Available
Y, Fuyuki T. Evaluation of temperature characteristics form: http://www.flabeg.com/uploads/media/FLA-
of high-efficiency InGaP/InGaAs/Ge triple-junction BEG_Solar_04.pdf (accessed on April 10, 2012).
solar cells under concentration. Solar Energy Materi- 36. AZUR SPACE Solar Power. Concentrator triple junction
als and Solar Cells 2005; 85(3): 429–436. DOI: solar cell 3C40C. Available from: http://www.azurspace.
10.1016/j.solmat.2004.05.008 com/index.php?mm=129 (accessed on April 10, 2012).
26. Nishioka K, Takamoto T, Agui T, Kaneiwa M, Uraoka Y, 37. IEC 62108 ed1.0. Concentrator photovoltaic (CPV)
Fuyuki T. Annual output estimation of concentrator modules and assemblies—design qualification and
photovoltaic systems using high-efficiency InGaP/ type approval, 2007.
InGaAs/Ge tripple-junction solar cells based on experi- 38. Tabor H. Radiation, convection and conduction coeffi-
mental solar cell´s characteristics and field-test meteoro- cients in solar collectors. The Bulletin of the Research
logical data. Solar Energy Materials and Solar Cells Council of Israel 1958; 6C: 155–176.
2006; 90: 57–67. DOI: 10.1016/j.solmat.2005.01.011 39. Coventry JS, Lovegrove K. Development of an
27. Green MA. Accuracy of analytical expressions for approach to compare the ’value’ of electrical and
solar cell fill factors. Solar Cells 1982; 7(3): 337–340. thermal output from a domestic PV/thermal system.
DOI: 10.1016/0379-6787(82)90057-6 Solar Energy 2003; 75(1): 63–72. DOI: 10.1016/
28. Braun A, Hirsch B, Vossier A, Katz EA, Gordon JM. S0038-092X(03)00231-7
Temperature dynamics of multijunction concentrator 40. Meunier F. Co- and tri-generation contribution to
solar cells up to ultra-high irradiance. Progress in climate change control. Applied Thermal Engineering
Photovoltaics: Research and Applications 2011; 2002; 22(6): 703–718. DOI: 10.1016/S1359-4311(02)
(published online). DOI: 10.1002/pip.1179 00032-7

438 Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd.
DOI: 10.1002/pip
H. Helmers et al. Modeling of CPVT systems

41. Schicktanz MD, Wapler J, Henning HM. Primary of Solar Energy Engineering 2008; 130(1): 011001–
energy and economic analysis of combined heating, 011005. DOI: 10.1115/1.2804618
cooling and power systems. Energy 2011; 36(1): 43. Huang BJ, Lin TH, Hung WC, Sun FS. Performance
575–585. DOI: 10.1016/j.energy.2010.10.002 evaluation of solar photovoltaic/thermal systems. Solar
42. Kribus A, Mittelman G. Potential of polygeneration Energy 2001; 70(5): 443–448. DOI: 10.1016/S0038-
with solar thermal and photovoltaic systems. Journal 092X(00)00153-5

Prog. Photovolt: Res. Appl. 2014; 22:427–439 © 2012 John Wiley & Sons, Ltd. 439
DOI: 10.1002/pip