Optical characterization of a fixed mirror solar concentrator prototype by the ray-

tracing procedure
Ramon Pujol-Nadal and Víctor Martínez-Moll

Citation: Journal of Renewable and Sustainable Energy 6, 043105 (2014); doi: 10.1063/1.4890219
View online: http://dx.doi.org/10.1063/1.4890219
View Table of Contents: http://scitation.aip.org/content/aip/journal/jrse/6/4?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in

Analysis of high concentrator photovoltaic modules in outdoor conditions: Influence of direct normal irradiance,
air temperature, and air mass
J. Renewable Sustainable Energy 6, 013102 (2014); 10.1063/1.4861065

Parabolic trough concentrators for hot water generation: Comparison of the levelized cost of production
J. Renewable Sustainable Energy 5, 023114 (2013); 10.1063/1.4795402

Severe energy crises and solar thermal energy as a viable option for Pakistan
J. Renewable Sustainable Energy 5, 013104 (2013); 10.1063/1.4772637

Hafnium and tantalum carbides for high temperature solar receivers

J. Renewable Sustainable Energy 3, 063107 (2011); 10.1063/1.3662099

Solar and wind energy resources and prediction

J. Renewable Sustainable Energy 1, 043105 (2009); 10.1063/1.3168403

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 JOURNAL OF RENEWABLE AND SUSTAINABLE ENERGY 6, 043105 (2014)

Optical characterization of a fixed mirror solar concentrator

prototype by the ray-tracing procedure
Ramon Pujol-Nadal and Vıctor Martınez-Moll
Departament de Fısica, Universitat de les Illes Balears, ctra. de Valldemossa km 7,5,
07122 Palma de Mallorca, Illes Balears, Spain
(Received 14 January 2014; accepted 26 June 2014; published online 14 July 2014)

The Fixed Mirror Solar Concentrator (FMSC) is a solar concentrator with static
reflector and moving receiver whose design emerged in the seventies as an effort to
reduce electricity production costs in solar thermal power plants. Solar
concentrators based on this geometry were constructed in the seventies and
eighties. A review of these prototypes is presented, highlighting the lack in
research probably brought about by halting these projects, and the two main flaws
of this research: poor theoretical analysis of the FMSC geometry and the
unfortunate choice of the zero-width mirror limit hypothesis. In this paper, another
methodology is presented to evaluate the FMSC geometry behavior: the construc-
tion of a FMSC prototype using lightweight materials with finite-width mirrors
(nine mirrors) and an optical characterization by 3D ray-tracing tools. The results
show good concordance between simulated and experimental data, showing that
FMSC prototypes can be characterized optically by accurate ray-tracing tools.
C 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4890219]
V

I. INTRODUCTION
It is commonly accepted that there is a great potential for the use of solar thermal collectors in
industrial applications as well as other process heat uses such as solar cooling.1–4 Although the founda-
tions of the solar thermal technology are well established, there are still a great number of open issues
that demand thorough research activity and innovative technological solutions. One of these issues is
the lack of a real market of reliable, cost effective, solar collectors designed specifically to meet the
requirements of industrial applications, which are, in some aspects, quite different from domestic hot
water or space heating applications. In addition, the International Energy Agency (IEA), in order to
expand these technologies for heating and cooling applications, proposed a roadmap as part of a process
that must evolve to take into account new technical and scientific developments, policies, and interna-
tional collaborative efforts.5 Two of the action items that are mentioned by the IEA and recommended
for research entities are the development of the integration of solar collectors in building surfaces, and
the expansion of the development of collectors that cover temperature gap between 100 and 250  C.
In this context, the company Tecnologıa Solar Concentradora SL (www.tsc-concentra.com),
in collaboration with the University of the Balearic Islands (UIB), has developed an innovative
solar concentrator with fixed mirror and tracking absorber called the CCStaR (Concentrating
Collector with Stationary Reflector). The CCStaR project started in 2006 with the aim to bring
onto the market a solar collector capable of working in the range of 100  C to 200  C, that is,
at the same time easily integrable onto light building roofs. The two principal target markets
for the development were industrial process heat applications and double stage solar cooling.
For the supply of heat in this temperature range, there are mainly two designs currently
used: parabolic trough and Fresnel reflector collectors. Small size designs can be efficiently
used on flat roofs (i.e., SopoNovaTM,6 HelioDynamics7). But another possibility was to explore
a different approach based on the concept of the Fixed Mirror Solar Concentrator (FMSC). The
original geometry of the FMSC emerged in the seventies as an effort to reduce electricity
production costs in solar thermal power plants.8,9 The FMSC concept, which is illustrated in
Fig. 1, was the starting point for the CCStaR project.

1941-7012/2014/6(4)/043105/14/$30.00 6, 043105-1 C 2014 AIP Publishing LLC

V

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-2 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

The main advantage of this kind of solution is that the largest element of the collector, the
mirror, remains fixed to the building structure, thus reducing the wind loads and simplifying the
collector integration. The FMSC geometry has long been analyzed with mathematical analytical
methods,10 but these models could not predict accurately the real optical behavior of the proto-
types constructed in the seventies (as it is showed in Sec. III), and therefore the optical efficien-
cies obtained by the real systems fell under the expected values. The large number of flat mir-
rors and other effects, such as mirror dispersion, Sun’s shape, and circumsolar radiation
distribution, have a large effect on the optical efficiency of the system. Those effects cannot be
easily taken into account by using analytical models.
The best way to predict the optical behavior of a complex system like that of a FMSC col-
lector, with variable geometry, is through the use of 3D ray-tracing algorithms. An optical anal-
ysis of the FMSC using 3D ray-tracing has been presented by the authors in a previous
publication.11
In this paper, the characterization of a real FMSC prototype with 3D ray-tracing is pre-
sented for the first time, and the model results are compared with the experimental ones. The
agreement between experimental and theoretical results is fair. Once the theoretical model has
been validated, the optical behavior of the system will be shown.

II. PREVIOUS FMSC PROTOTYPES AND RELATED WORK

The FMSC was conceived by Russell at General Atomic in the early seventies.8,9 In
Bansal’s thesis, an analytic mathematical modelling was used to predict the daily and annual
performance characteristics of the system in different geographical locations,10 and conse-
quently all prototypes built in the seventies were supported by the theoretical models presented
in this thesis. In 1975, the Arizona State University received a small scale model of the FMSC
constructed by General Atomic.12 The prototype was made with reduced dimensions with the
main purpose of analysing the flux profile variations in the focal zone. The prototype had an
aperture of 1.1 m  3.05 m with 97 small mirror strips of 1.14 cm wide. It was shown that 70%
of the flux incident fell within a 2 cm wide strip (almost twice that of the mirror width), in
view of the results (that were not predicted in the theoretical analysis10); it was suggested that
a secondary reflector could significantly improve the performance of this concentrator and could

FIG. 1. Optical principle of the FMSC. The receiver moves in a circular path on the generating circle where flat mirrors are
positioned in order to reflect sunlight to a focus point. (a) The generating circle has a radius R, the focal length is F ¼ 2R,
and the reflector width is W. (b) The receiver is positioned by the hf angle. The position angle of the receiver is the double
than that of the transverse incidence angle (hf ¼ 2ht). Note that the receiver itself rotates on an angle hr ¼ ht. Reproduced by
permission from R. Pujol-Nadal and V. Martınez-Moll, ASME J. Sol. Energy Eng. 134(3), 031009 (2012). Copyright 2012
by the American Society of Mechanical Engineers and Reproduced by permission from R. Pujol-Nadal and V. Martınez-
Moll, Appl. Opt. 52(30), 7389–7398 (2013). Copyright 2013 by the Optical Society.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-3 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

achieve concentration ratios of around 20. In 1975, another FMSC prototype of 50.2 m2 with
metal frame-mounted was constructed at the Georgia Institute of Technology; the prototype
used air as heat transfer-medium to reach 400  C, delivered heat to a tank of stones for heat
storage at about 260  C, and served as a part of their solar demonstration program.13
Alternatively, General Atomic constructed a cast concrete mirror for FMSC prototypes with the
purpose of reducing manufacturing costs.14,15 The main purpose was to ensemble a low cost
reflector with cast concrete; however, casting concrete over glass mirrors did not appear to
result in a sufficiently durable bond, and an additional coat of paint or other film was necessary
to provide long-term protection to the silver backing of the mirrors. The heat receiver used was
a compound parabolic secondary concentrator to increase the concentrator factor and to
improve the uniformity of the heat distribution on the receiver tube. The measurements showed
that about 95% of the light was concentrated into the aperture width of the secondary concen-
trator, reaching a concentration of about 23 (as it was predicted in Ref. 12). However, as to be
expected from a prototype model, other problems were encountered. The polished aluminium
used was unsatisfactory, the reflectivity was not better than 0.88 and dropped rapidly due to
tarnishing that occurred under weather exposure. The pyrex glass covering the heat pipe cavity
cracked when the receiver was into focus, but a simple design repair eliminated this glass prob-
lem in subsequent tests.15 Development work was discussed for an implementation of 260 m2
of FMSC for a demonstration collector field,16 while an evaluation of the prototype12 after 3 yr
in operation was presented,17 showing a reduction of 16% in the captured energy by the effect
of dust after two weeks without cleaning the mirrors. On the other hand, it was shown that the
angular alignment of the mirror slats was the most critical of all the design dimensions of the
FMSC geometry. The authors stated that the only logical orientation for a FMSC was with its
axis in east-west direction.17 Experimental measurements of heat loss from several configura-
tions of the heat receiver for FMSC design where presented.18 In 1977, a computer program
was developed to simulate the performance of a power plant for electricity production.19
General Atomic manufactured in 1978 FMSC modules of precast concrete and glass mirrors to
supply a 260 m2 collector field to Sandia Laboratories for their Solar Total Energy
Demonstration Facility, and Scientific Atlanta, Inc. developed another construction approach
based on supporting the glass mirrors with inexpensive metal stampings for the same demon-
stration facility.20–22 From experiments conducted in the Sandia Laboratories, the precast con-
crete FMSC had a peak noon efficiency of 47%, and the metal stamping had an efficiency of
41%.23 In this report, the FMSC collectors had the lowest performance of the eight different
concentrating collectors tested at Sandia Laboratories. The main problem of the two FMSC pro-
totypes was the difficulty encountered in having to position the multiple small mirror segments
accurately enough to get reflected light within the receiver aperture; on the other hand, the ther-
mal losses of the General Atomic prototype were lower, indicating that it would also be a good
high-temperature performer if the basic focus problems of multiple small mirrors could be
resolved. In 1979, a theoretical study for 100 MW electric generation plant was performed using
FMSC to supply energy for steam generation.24 For mirror specularity, the analysis concluded
that 10 mrad was acceptable, and low-cost precast concrete remained as the reference panel
fabrication method. Various heat transport fluids were considered, and draw salt was selected
because it resulted in the lowest system cost and it had the greatest end use applicability.
After great efforts in studying the behavior of FMSC, the Line-Focus concentrating
Collector Program from the seventies from Sandia Laboratories concluded in 1980 that the par-
abolic trough was the best line concentrator suitable for application scenarios which require me-
dium temperature.25 The main argument to justify discarding the FMSC design was that a fixed
mirror concentrator collects less energy over a year than system with a tracking reflector like
parabolic trough.
Parallel to the work done during the seventies in the U.S. on FMSC, other research pro-
grams were conducted in France and Argentina.26,27 During 1980, a full-scale thermal-electric
power plant was constructed in Corsica (France) with FMSC, with a total aperture area of
1176 m2. The concentrator consisted of 71 flat reflecting facets 43 mm width  1 m long and
glued onto two concrete supports. The collector aperture plane was oriented to east-west

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-4 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

orientation with latitude-tilt inclination (42 ). The heat receiver assembly consisted of a long
extruded aluminum tube inside where the absorber tube and thermal insulation were located.
Geometrical concentration was about 30, a peak flux intensity of 50 suns was measured and
spot with 10 cm, while optical efficiency was about 65%. The fluid outlet temperature was regu-
lated at 250  C. The authors26 reported that the main advantage of using FMSC was that solar
collectors could be made on-site. In Argentina, only theoretical analyses were performed,28–30
and no experimental results have been found.
The FMSC prototypes constructed up to this time had small mirrors in order to achieve
high concentration ratios; because all the prototypes supported by the theoretical analysis in
Bansal’s thesis10 assumed a zero-width mirror limit. In his thesis, the energy losses by self-
shadowing (called edge losses) were calculated by analytical methods for estimating the geo-
metric interception of solar rays by the reflector steps. Experiments with the first constructed
reflector led to the conclusion that a secondary reflector was needed for the receiver, although
the installation of this secondary reflector only created more shadows on the primary reflector
surface. These results suggested that instead of a zero-width mirrors design, a finite-width mir-
rors reflector could be constructed that would not need a large number of mirrors that would
complicate its manufacture and positioning precision. Moreover, there is no point having very
narrow mirrors if the positioning error causes a double focus amplitude of the mirror’s width,
as has been seen in the previous studies. Bansal concluded that a more detailed analysis was
needed, and three dimensional accuracy using ray-tracing techniques were recommended for the
non-continuous FSMC geometry. Nevertheless, the author concluded that the FMSC was a
promising concept.
It was not until 2012 when the authors of this paper presented an optical analysis of FMSC
using 3D ray-tracing tools.11 In the optical study, the FMSC geometry was defined by using
three design parameters: the number of mirrors N (the limit of zero-width mirror was avoided),
the ratio of focal length to the reflector width F/W, and the intercept factor c (in order to repre-
sent different receiver widths, as defined as the fraction of the reflected radiation incident upon
the absorbing surface of the receiver). A standard evacuated tube with a flat fin was considered
as a receiver; it was shown that optical efficiencies of up to 73% were possible with this geom-
etry. In a second paper,31 thermal efficiencies of the FMSC geometry were exposed at a work-
ing temperature of 200  C, and dependent upon the location considered, (Munich, Palma de
Mallorca, and El Cairo), an annual thermal efficiency of 39%, 44%, and 48%, respectively, can
be obtained. One of the main results in Ref. 31 was that designs with 13 mirrors gave almost
the same energy than designs with more mirrors like 51. Therefore, it is not necessary to con-
struct designs with huge number of mirrors according to this research if finite-width mirrors hy-
pothesis is considered.

III. THE CCStaR V0 PROTOTYPE

A FMSC prototype was built with the following design parameters: N ¼ 9 and F/W ¼ 1.5.
In Fig. 2, a schematic of the mirrors layout is shown. The centre of each mirror and the tilt is
reported (only right side mirrors are shown because the reflector is symmetric).
The prototype was called CCStaR V0 (Concentrating Collector with Stationary Reflector
Version 0). The prototype was orientated on North-South and 15 tilted inclination. The most
important aspects of CCStaR V0 are exposed below. The receiver had a typical configuration
of a flat plate collector, with a net absorber surface of 0.12 m width and 2.7 m length, see Fig.
3(a) for the receiver constructed. The reflector had a length of L ¼ 3 m and a width of W ¼ 1 m.
A low cost concentrator was constructed with the reflector made from a piece of extruded poly-
styrene cut by numerical control which was glued the nine flat mirrors 11.1 cm width. See Fig.
3(b) for the CCStaR V0 prototype, where it shows how the solar concentrator is located on a
sandwich panel, a typical structure for a cover in an industrial ship. The physical properties are
exposed in Table I, where the normal incidence absorptance and angular absorptance depend-
ence have been estimated from the references exposed in Table I, and the refraction coefficient,

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-5 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

FIG. 2. Schematic of the mirrors in the right side. The distances are presented in mm.

FIG. 3. (a) Flat receiver constructed for the CCStaR V0. (b) CCStaR V0 prototype installed for experimental testing,
North-South axis orientation, and 15 tilted.

TABLE I. Material properties of CCStaR V0.

Element Property

Receiver cover glass thickness 0.002 m

Receiver cover glass refraction coefficient 1.526
Receiver cover glass extinction coefficient 8 m1
Normal incidence absorptance (from Ref. 33) a0 ¼ 0.90
 1:2
Angular absorptance dependence (adopted from Ref. 34) a 1
¼ 1  0:057
a0 cos hi
Receiver insulation Polyurethane (0.04 m thick)
Mirror thickness 0.002 m
Mirror glass refraction coefficient 1.526
Mirror glass extinction coefficient 8 m1
Mirror reflectance layer 0.95 (silver)

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-6 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

the extinction coefficient, and the mirror reflectance layer values have been assumed from typi-
cal values exposed in the book32 and from the manufacturer’s material documentation.
The receiver rotation angle hr, see Fig. 1(a), was obtained through a fixed mechanical link
with the tracking arm shaft, using a constant transmission ratio of 1=2 with respect to hf. This
mechanical link reduces the number of interactions with the controller and simplifies the elec-
tronics needed to adjust the reflector arm.
The main purpose for the construction of this prototype was to test with a real system in
order to compare and validate the optical performance with a developed 3D ray-tracing pro-
gram, while making proof of the easily integration onto typical industrial roofs by using light-
weight materials for the construction of the solar concentrator, instead of the heavier ones used
in the seventies prototypes exposed above. Furthermore, this solution permits an installation ori-
entation along the North-South axis.

IV. TESTING METHODOLOGY

A thermal testing of the prototype was conducted in order to measure the optical behavior
while operating in real conditions. The thermal testing methodology used was based on the quasi-
dynamic model exposed on the part 6.3 of the European Standard EN 12975–2,35 where the aver-
age fluid temperature was closed to the ambient temperature in order to measure the optical behav-
ior of the system. Then the power gained by unit net area is given by the following expression:
•
Q
¼ F0 ðqsaÞen Kb GbT þ F0 ðqsaÞen Kd GdT ; (1)
Aa
where F0 is the collector heat removal factor, (qsa)en is the effective product of the reflectance
of the mirror q, the solar transmittance of the glass cover s, and the solar absorptance of the
absorber at normal incidence a (referred with the en sub index), GbT and GdT are the beam and
diffuse solar irradiance, respectively (both on the collector plane referred with the T sub index).
The product F0 ðqsaÞen Kb can be referred to as the optical efficiency relative to beam solar irra-
diation gb, and the second component product F0 ðqsaÞen Kd can be referred to as the optical effi-
ciency relative to the diffuse solar irradiation gd. The Kb and Kd factors are the incidence angle
modifier (IAM) for beam and diffuse solar irradiation, respectively.
The IAM provides a performance factor when the sun is not perpendicular to the collector.
The IAM terms are calculated by the ratio between the optical efficiency in any sun position
and the optical efficiency in a normal incidence. The commonly accepted model for determin-
ing the IAM for axial collectors is a biaxial model where the angle modifier is given by the
product of transverse and longitudinal components to the planes of the receiver according to
the angles defined in Fig. 4: K(ht,hl) ¼ K(ht,0)K(0,hl). However, for the FMSC geometry, it was
shown that the IAM can be factorized as K(ht,hl) ¼ K(ht,0)K(0,hl)f(ht,hl) where f(ht,hl) is a func-
tion that takes into account the longitudinal end losses.11
In order to test the prototype at ambient temperature, a bench-test was constructed based
on the closed loop circuit proposed in EN 12975–2.35 Sensors were used to monitor the input
and output of the collector (flow rate meter, temperature sensors, radiometers) and were con-
nected to a data logger with a recording frequency of 1 s. The inner and outer temperatures
were measured with Negative Temperature Coefficient sensors, the global solar irradiance GT
was measured with a secondary standard pyranometer on the plane of the collector, and the
direct normal incidence solar irradiance GDNI was measured with a first class pyrheliometer
mounted on a sun-tracker close to the collector, see Table II for uncertainty.
The measured power gained from the prototype can be calculated from the following expression:
• •
Q exp ¼ Cd V ðTo  Tin Þ ; (2)
•
where C is the fluid heat capacity, d is the fluid density, V is the volumetric flow, and Tin and
To are the inlet and outlet temperature, respectively.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-7 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

FIG. 4. ht and hl are the transverse and longitudinal angles that are the projected incidence angles on the two reference
planes perpendicular and along the axis of the collector. hi is the incident angle. Reproduced by permission from R. Pujol-
Nadal and V. Martınez-Moll, ASME J. Sol. Energy Eng. 134(3), 031009 (2012). Copyright 2012 by the American Society
of Mechanical Engineers and Reproduced by permission from R. Pujol-Nadal and V. Martınez-Moll, Appl. Opt. 52(30),
7389–7398 (2013). Copyright 2013 by the Optical Society.

V. EXPERIMENTAL RESULTS
In order to optically characterize the prototype, the first step was to determine the receiver
optical efficiency to obtain the collector heat removal factor F0 in Eq. (1). Ray-tracing simula-
tions have been conducted in this section and the subsequent one.
A forward ray-tracing code was implemented and was presented in Ref. 11. In the ray-
tracing, the geometry of the solar concentrator is described by discrete elements with triangular
surfaces. Four kinds of surfaces can be introduced: specular surfaces, opaque surfaces, interface
surfaces (to model pieces of glass), and absorber surfaces (the receiver). The program calculates
ray trajectories from one source (called the sun window) that emits to all the surfaces of the
system, and only ray-optics propagation is taken into account. The angular size of the sun is
modeled according to the Buie equations,36 and Fresnel effects are handled using a Monte
Carlo approach. The program can calculate the beam optical efficiency and the radiation distri-
bution on the absorber. The ray-tracing code was updated in order to take the diffuse solar radi-
ation into account, assuming radiation from an isotropic sky. In this case, the rays traced are
those sent from all over the sky’s half sphere (2p sr) instead of one fixed direction as when
direct solar radiation is simulated. The four types of material surfaces, the angular size of the
sun, and the intersection detection algorithm were described in Ref. 11.
A convergence analysis was conducted before each numerical experiment in order to deter-
mine the minimum number of rays that needed to be computed in the ray-tracing to obtain a

TABLE II. Measurement uncertainty estimation.

Physical value measured Uncertainty

Direct normal irradiance GDNI 63%

Sloped global irradiance GT 61.5%
Inner temperature tin 60.1  C
Outer temperature te 60.1  C
Ambient temperature ta 60.1  C
Volumetric flow 61%

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-8 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

variability of the results lower than 0.004. But the biggest contribution to the results is due to
the errors associated to the input optical parameters. A sensitivity study has been conducted to
assess the influence of the optical parameters on the ray-tracing results. We only considered
errors for the normal incidence absorptance (a0 ¼ 0.90 6 0.02), extinction coefficient
(k ¼ 8 6 2 m1), and mirror reflectance layer (q ¼ 0.95 6 0.02). Considering the upper and the
lower limit errors exposed above, the ray-tracing results varies in 60.05 for both the optical ef-
ficiency when the CCStaR V0 is simulated on normal beam irradiation, and the receiver on dif-
fuse irradiation. Therefore, we considered an error of 60.05 for the ray-tracing results as a con-
sequence of the input parameters.

A. Receiver optical efficiency

The receiver was tested at ambient working temperature in order to find its optical effi-
ciency and more specifically the value of F0 .
The power by unit net area gained by the receiver is given by the following expression:
•
Qrec
¼ F0 ðsaÞen Kbrec GbT þ F0 ðsaÞen Kdrec GdT ; (3)
Arec

where the Kbrec and Kdrec factors are the IAM for beam and diffuse solar irradiation of the re-
ceiver, respectively.
Testing results provided four series of data that satisfied the condition of stable inlet tem-
perature according to Standard EN 12975–2.35 In Table III, the conditions of these measure-
•
ments are exposed, and the power gained by the receiver is shown as Qrec; exp using Eq. (2).
Four stable points were obtained.
In addition, the receiver was analysed by the ray-tracing program exposed in Ref. 11. The
diffuse optical efficiency obtained by ray-tracing, with 107 rays, was ðsaÞen Kdrec ¼ 0:635, and
the beam optical efficiency at normal incidence, with 106 rays, was ðsaÞen ¼ 0:831. The value
of ðsaÞen Kbrec for the four stable points is exposed in Table III because it depends on the solar
position angles ht and hl. Finally, the power gained by the receiver calculated by ray-tracing is
•
shown in Table III, and has been determined by Qrec;rt ¼ Arec ½ðsaÞen Kbrec GbT þ ðsaÞen Kdrec GdT .
Since this program did not take into account the collector heat removal factor F0 , it was deter-
•
0 Qrec; exp
mined for each stable point as: F ¼ • , see Table III for the four F0 values calculated.
Qrec;rt
Since four values of F0 were obtained, we considered the collector heat removal factor
equal to the slope in a linear regression with zero intercept value. In Fig. 5, the four stable
0
points and the uncertainty are plotted. The linear regression gave a value of F ¼ 0:968 and a
2
determination coefficient r ¼ 0.988. This value falls within the range considered by studies of
flat-plate collectors, for examples consult Ref. 32.

B. Testing results and ray-tracing validation

Having determined the optical efficiency of the receiver and the collector heat removal fac-
tor F0 ¼ 0.968, we proceeded to measure the efficiency of the full optics system. The solar con-
centrator was installed in a building of the University of Balearic Islands, oriented in North-
South axis direction and with an inclination of 15 . Prior to the start of an experimental run,
TABLE III. Conditions of the measurements for the receiver and results.

• •
Data series tm ta GbT (W) GdT (W) hT (deg) hL (deg) ðsaÞen Kbrec Qrec; exp (W) Qrec;rt (W) F0

1 36.0 31.9 866 128 1.3 13.6 0.832 251 6 8 259 6 11 0.969
2 34.8 31.3 840 141 6.1 13.5 0.835 248 6 8 256 6 11 0.967
3 33.8 30.5 830 139 9.8 13.4 0.827 242 6 8 251 6 11 0.965
4 30.5 30.6 814 144 12.7 13.3 0.827 240 6 8 247 6 11 0.970

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-9 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

FIG. 5. Comparison between experimental power and ray-tracing power for the receiver and linear regression to obtain F0 .

both the mirrors and the receiver were thoroughly cleaned, hence it was assumed that there
were no losses attested to the accumulation of dust.
CCStaR V0 was tested at ambient working temperature so the results could be compared
with the theoretical ray-tracing model. Thus, the theoretical power gained determined with ray-
tracing is given by the following equation (diffuse irradiation in Eq. (1) can be neglected for
the prototype configuration):
•
Qrt ¼ g0b Kb GbT Aa ; (4)

where g0b ¼ F0 ðqcsaÞen , the IAM term Kb was calculated by ray-tracing program for each sun
position during the testing campaign, and the beam irradiation was obtained from the measure-
ments. Subsequently, the power gained obtained by ray-tracing Eq. (4) could be compared with
the power measured from the experimental measurements Eq. (2).
The real geometry and the physical properties in Table I were introduced as inputs into the
ray-tracing program. A total optical error was included in the mirror imperfections considering
a Gaussian distribution with a variance value of r ¼ 10 mrad; this includes the whole dispersion
caused by the next sources errors: slope, scattering, tracking, and alignment.37 The size of the
sun was modelled by the Buie equations36 with a circumsolar ratio (CSR) of 0.05. In this simu-
lation, 106 rays were emitted for each sun position.
From the testing, at working temperature near to the ambient, nine series data satisfied the
condition of stable inlet temperature according to Standard EN 12975–2.35 Fig. 6 shows the ex-
perimental thermal power measured and the thermal power predicted by ray-tracing, Eqs. (2)
and (4), respectively, and the close agreement between the simulated and the experimental val-
ues can be seen. In Table IV, the conditions of these measurements are exposed.
In a collector testing, the inlet temperature stability is very difficult when irradiation condi-
tions vary; in our case, the irradiation was very stable, but the IAM varied due to the sun posi-
tion changes, and the effect on the instability was similar to the changes in irradiation as can
be seen in Sec. VI when the IAM values are shown. So for this reason, maintaining stability of
the inlet temperature is very difficult for systems with variable geometry. Even so, it can be
seen that there is a good concordance between simulation and experimental results with a corre-
lation coefficient r ¼ 0.957 for these nine experimental points.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-10 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

FIG. 6. Simulated and measured thermal power for the nine stable series.

VI. RAY-TRACING RESULTS

Once the theoretical model of the CCStaR V0 was validated with the ray-tracing program,
the optical behavior of the prototype was analysed. First, the maximum profitable energy by ge-
ometrical construction (self-shadowing and geometrical interception), was determined by ray-
tracing (issuing 106 rays) in order to quantify the losses generated by geometrical construction,
which gave a value of f ¼ 0.7293. This low value is caused mainly by the self-shadowing of
the receiver (the gross width is 0.185 m), and that the length of the receiver is lower than the
reflector (2.7 m front 3.0 m).
On the other hand, the intercept factor c can be determined assuming the mirror dispersion
(r ¼ 10 mrad) and the size of the sun (CSR ¼ 0.05) considering an ideal behavior for the reflec-
tance (q ¼ 1) and for the absorptance (a ¼ 1), a glass refraction coefficient n ¼ 1, and an extinc-
tion coefficient k ¼ 0 m1. In this situation, the optical efficiency for beam irradiation at normal
incidence simulated by ray-tracing gave a value of 0.6854, which is the value of c. Although c
takes a very low value, it must be noted that the largest contribution comes from the self-
shadowing and geometrical interception (f ¼ 0.7293).
Following this step, the IAM curves are shown, along with the goodness of the IAM facto-
rization and the distribution of radiation onto the receiver.

TABLE IV. Conditions of the measurements for the CCStaR V0.

Data series tm ta GbT (W) hT (deg) hL (deg)

1 28.0 31.6 775 25.3 18.6

2 27.7 31.3 764 20.1 18.8
3 27.8 29.3 774 18.9 18.9
4 30.2 31.0 774 14.9 19.0
5 31.5 31.3 770 6.0 19.1
6 24.8 30.8 783 1.6 19.2
7 27.1 31.0 775 4.5 19.1
8 30.5 31.5 780 5.8 19.1
9 30.2 30.2 774 7.7 19.1

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-11 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

FIG. 7. Transverse and longitudinal IAM values of the CCStaR V0 prototype calculated by ray-tracing.

A. The IAM curves

Figure 7 shows the results for the transverse and longitudinal IAM obtained by ray-tracing
issuing 106 rays. It can be observed that the transverse IAM is close to one for incidence angles
up to 65 , which is the transverse incidence angle when the intersection between the receiver
and the reflector occurs.11 The longitudinal IAM drops rapidly when the incidence angle
increases; this is because the edge losses are very high for the configuration constructed. (The
relationship between length L and focus F is L/F ¼ 2.)
With respect to the optical efficiency at normal incidence, a low value g0b ¼ 46.1% was
obtained by ray-tracing (this value includes the F0 value obtained experimentally). This value
was caused mainly by the factor of f ¼ 0.7293. But improving this factor, i.e., f ¼ 1 (which is
the value for a receiver composed by an evacuated tube with absorber material in the rear sur-
face and the same length of the reflector11), a peak efficiency of almost 63% could be reached.

B. Factorization approximation for the IAM

The IAM for all combinations of the transverse and longitudinal angles has been calculated
and the results have been plotted in Fig. 8, the factorized approximation IAM from
K(ht,hl) ¼ K(ht,0)K(0,hl)f(ht,hl) has also been calculated to compare this approximation with the
exact value obtained by ray-tracing (green points in Fig. 8). The correlation between the IAM
factorization and the ray-tracing results was r ¼ 0.975. Thus, the IAM factorized approximation
is valid for this prototype, as has been proven for theoretical FMSC designs.11
It should be noted that the optical behavior of the prototype is very irregular and not easily
predictable by analytic methods, which can be calculated accurately with computer ray-tracing
tools. Moreover, with the IAM factorized approximation, the optical behavior of the FMSC ge-
ometry can be obtained similarly to the bi-axial collectors.

C. Radiation distribution on the absorber

Finally, the local flux concentration on the receiver was calculated issuing 107 rays in the
ray-tracing program. Figure 9 shows the radiation distribution for three transversal angles of
incidence. It can be observed that irradiation reaches a maximum concentration of six suns.
Unlike the results reported for theoretical FMSC designs,11 the maximum radiation for the
CCStaR V0 did not occur at the normal incident angle. This is because the CCStaR V0 has

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-12 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

FIG. 8. IAM values calculated by ray-tracing program and by factorization for the overall angular range.

self-shadowing when the receiver is in the high position (normal incidence), in spite of the the-
oretical design considered in Ref. 11, which has no self-shadowing effect (the receiver was an
evacuated tube with absorber material in the rear surface). It should be noted that energy losses
are produced in the tails of the bell, where a local flux concentration of two suns are achieved
at normal incidence (hi ¼ 0 ).

VII. CONCLUSIONS
A review of the FMSC studies and prototypes has been given, and it can be seen that one
of the most relevant parameters of this geometry was the alignment of the narrow mirrors;

FIG. 9. Local flux concentration on the receiver calculated by ray-tracing.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-13 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

nevertheless, this factor could not be taken into account by the analytical models used in the
first studies. The Line-focus program of Sandia Laboratories discards this geometry for solar-
electric generation, but in France a pilot demonstration was executed and demonstrated that
FMSC is a good candidate for solar power generation. Theoretical studies conducted in the sev-
enties overestimated the captured energy of the FMSC. Gaps in knowledge about the FMSC
have been identified, and it was shown that the limit of zero-width for the mirrors complicated
the construction of the prototypes, and obtained poor optical efficiencies.
A new methodology to analyze FMSC prototypes has been presented, with the construction
of a prototype using lightweight materials in order to validate a numerical model based on 3D
ray-tracing procedure. The prototype is simpler than those constructed in the seventies because
zero-width limit mirrors were avoided. The peak efficiency achieved was similar to the proto-
types tested at Sandia Laboratories, but by improving the geometrical dimensions 63% peak ef-
ficiency could be achieved. A good performance on transverse IAM was achieved (nearly to
one in the principal incidence angles).
A numerical model by ray-tracing was validated with experimental results. The use of this
method allows us to find the optical behavior of the system for sun positions where the collec-
tor has not been tested. An IAM approximation factorization similar than the bi-axial collectors
has been proved too.
In this paper, it has been proven that an installation of the FMSC on North-South orienta-
tion is also possible, and with the use of its lightweight materials, makes it a good candidate to
provide energy in medium range temperature applications with easily integration onto building
roofs.

ACKNOWLEDGMENTS
The authors plan future experiments to analyse a prototype based on the Curved Slats Fixed
Mirror Solar Concentrator geometry (CSFMSC), which is an evolution of the FMSC where flat
mirrors are replaced by curved mirrors.38–40

1
N. V. Ogueke, E. E. Anyanwu, and O. V. Ekechukwu, “A review of solar water heating systems,” Renewable Sustainable
Energy Rev. 1(4), 043106 (2009).
2
S. Mekhilef, R. Saidur, and A. Safari, “A review on solar energy use in industries,” Renewable Sustainable Energy Rev.
15(4), 1777–1790 (2011).
3
R. Gabbrielli and F. Zammori, “Potential for cogeneration through solar energy in the tissue industry: Technical and eco-
nomic aspects,” ASME J. Sol. Energy Eng. 134(1), 011015 (2012).
4
O. A. Jaramillo et al., “Parabolic trough concentrators for hot water generation: Comparison of the levelized cost of
production,” Renewable Sustainable Energy Rev. 5(2), 023114 (2013).
5
See http://www.iea.org/publications/freepublications/publication/2012_SolarHeatingCooling_Roadmap_FINAL_WEB.pdf
for IEA, “Technology Roadmap: Solar Heating and Cooling,” 2012 (last accessed December 12, 2013).
6
See http://simonsgreenenergy.com.au/wp-content/themes/energy/pdf/Data_Sheet_SopoNova_Web.pdf for Sopogy,
"SopoNovaTM MicroCSPTM: Industrial and Utility Solar Collector," 2014.
7
See http://www.ecobuilding-club.net/downloads/RTD/HelioDynamics.pdf for HelioDynamics Ltd, “Mid-temperature
concentrating solar thermal collectors,” 2014.
8
J. L. Russell, E. P. DePlomb, and R. K. Bansal, “Principles of the fixed mirror solar concentrator,” 2nd ed., General
Atomic Co., San Diego, CA, 1974, Report No. GA-A12903.
9
J. L. Russell, Jr., “Investigation of a central station solar power plant,” General Atomic Company Report No. GA-
A12759, August 31, 1973.
10
R. K. Bansal, “Theoretical analysis of fixed mirror solar concentrator,” Ph.D. thesis (Arizona State University, 1974).
11
R. Pujol-Nadal and V. Martınez-Moll, “Optical analysis of the fixed mirror solar concentrator by forward ray-tracing
procedure,” ASME J. Sol. Energy Eng. 134(3), 031009 (2012).
12
S. Y. Harmon et al., “Characteristics of the concentrated solar flux produced by the FMSC prototype,” in Sharing the
Sun: Solar Technology in the Seventies; Proceedings of the Joint Conference, Winnipeg, Canada, August 15–20 (1976),
Vol. 2, pp. 291–303.
13
J. R. Williams, “Experimental solar heat supply system with fixed mirror concentrator for the heating and cooling of
buildings,” Issued as Progress Report No. 1 and Final report, Project No. E-25-644, Georgia Institute of Technology,
School of Mechanical Engineering, Atlanta, Georgia, 1975, see http://hdl.handle.net/1853/37324 (last accessed
December 12, 2013).
14
J. L. Russell, Jr., J. R. Schuster, and G. H. Eggers, “Development status of the fixed mirror solar concentrator,” in
Proceedings of the 12th Intersociety Energy Conversion Engineering Conference (1977), Vol. 2, pp. 1141–1146.
15
G. H. Eggers et al., “Solar collector field subsystem program on the Fixed Mirror Solar Concentrator,” Final Report for
the period March 28, 1976 through September 30, 1976, General Atomic Project No. 3251, 1976.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33
 043105-14 R. Pujol-Nadal and V. Martınez-Moll J. Renewable Sustainable Energy 6, 043105 (2014)

16
G. H. Eggers and J. L. Russell, Jr., “The FMSC collector subsystem for the Sandia Solar Total Energy Facility,” in Solar
Cooling and Heating: Architectural, Engineering, and Legal Aspects; Proceedings of the Forum, Miami Beach, FL,
December 13–15, 1976 (A78-43968 19-44) (1978), Vol. 1, pp. 255–271.
17
R. Pinon and B. D. Wood, “Evaluation of a fixed mirror solar concentrator,” in Proceedings of the 1978 Annual Meeting
of the American Section of the International Solar Energy Society, Inc., August 28–31 (1978), Vol. 1, pp. 398–407.
18
W. E. Walker, J. J. Housman, and J. L. Russell, Jr., “Measurement of heat loss from a heat receiver assembly of a Fixed
Mirror Solar Concentrator,” in 13th Intersociety Energy Conversion Engineering Conference (1978), pp. 1554–1562.
19
J. L. Russell, Jr. et al., “Preliminary system analysis of a fixed mirror solar power central station: Final report,” EPRI ER-
434 (Research Project 739-1), General Atomic Company, 1977.
20
J. R. Schuster, J. L. Russell, Jr., and G. H. Eggers, “Fixed mirror solar concentrator for power generation,” in
International Symposium on Solar Thermal Power Stations, Cologne, Germany, April 11–13 (1978), pp. 1–15.
21
J. R. Schuster, G. H. Eggers, and J. L. Russell, Jr., “Operating experience with the general atomic fixed mirror solar con-
centrator,” in Proceedings of the 1978 annual meeting of the American Section of the International Solar Energy Society,
Inc., August 28–31 (1978), Vol. 1, pp. 863–870.
22
J. R. Schuster, J. L. Russell, Jr., and G. H. Eggers, “Design, construction, and testing of a fixed mirror solar concentrator
field,” in 13th Intersociety Energy Conversion (1978), pp. 1541–1547.
23
V. E. Dudley and R. M. Workhoven, “Summary report: concentrating solar collector test results, Collector Module Test
Results Collector Module Test Facility (CMTF), January-December 1978,” Sandia Laboratories Report No. SAND78-
0977, 1979.
24
J. R. Schuster, J. M. Neill, and J. Bass, “Fixed mirror solar concentrator for application to a 100 MW(e) electric generat-
ing plant,” in 14th Intersociety Energy Conversion Conference (1979), pp. 15–19.
25
V. L. Dugan, “Proceedings of the semi-annual distributed receiver systems program review: Line-focus concentrating
collector program,” Sandia Laboratories Report No. NASA-CR-163135 (1980), pp. 27–33.
26
J. L. Boy-Marcotte et al., “Construction of a 100 kW Solar Thermal-Electric Experimental Plant,” ASME J. Sol. Energy
Eng. 107, 196–201 (1985).
27
R. O. Nicolas et al., “Concentrador cilındrico fijo a espejo faceteado para aprovechamiento de la energıa solar. Estudio
teorico y construccion de un prototipo,” Comision Nacional de Energıa At omica, Rep
uclica Argentina, 1978.
28
R. O. Nicolas and J. C. Duran, “Generalization of the two-dimensional optical analysis of cylindrical concentrators,” Sol.
Energy 25(1), 21–31 (1980).
29
J. C. Duran and R. O. Nicolas, “Comparative optical analysis of cylindrical solar concentrators,” Appl. Opt. 26(3),
578–582 (1987).
30
J. C. Duran and R. O. Nicolas, “Development and applications of a two-dimensional optical analysis of non-perfect cylin-
drical concentrators,” Sol. Energy 34(3), 257–269 (1985).
31
R. Pujol-Nadal, V. Martınez-Moll, and A. Moia-Pol, “Parametric analysis of the fixed mirror solar concentrator for me-
dium temperature applications,” ASME J. Sol. Energy Eng. 136(1), 011019 (2013).
32
J. A. Duffie and W. Beckman, Solar Engineering of Thermal Processes, 2nd ed., edited by Anonymous (Wiley, New
York, 1991).
33
See http://www.almecosolar.com/Brochure/tinox_energy_new_en.pdf for Almeco-TiNOX GmbH, 2014.
34
T. Tesfamichael and E. W€ackelgård, “Angular solar absorptance and incident angle modifier of selective absorbers for so-
lar thermal collectors,” Sol. Energy 68(4), 335–341 (2000).
35
CEN, “Thermal Solar Systems and Components-Solar Collectors-Part 2: Test Methods,” European Standard EN 12975-
2:2006, European Committee for Standardisation, 2006.
36
D. Buie, C. J. Dey, and S. Bosi, “The effective size of the solar cone for solar concentrating systems,” Solar Energy
74(5), 417–427 (2003).
37
R. B. Pettit, C. N. Vittitoe, and F. Biggs, “Simplified calculational procedure for determining the amount of intercepted
sunlight in an imaging solar concentrator,” ASME J. Sol. Energy Eng. 105(1), 101–107 (1983).
38
R. Pujol-Nadal and V. Martınez-Moll, “Optical analysis of a curved-slats fixed-mirror solar concentrator by a forward
ray-tracing procedure,” Appl. Opt. 52(30), 7389–7398 (2013).
39
R. Pujol-Nadal and V. Martınez-Moll, “Parametric analysis of the curved slats fixed mirror solar concentrator for medium
temperature applications,” Energy Convers. Manage. 78, 676–683 (2014).
40
F. Sallaberry et al., “Optical and thermal characterization procedure for a variable geometry concentrator: A standard
approach,” Renewable Energy 68, 842–852 (2014).

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.93.5.131 On: Tue, 13 Jan 2015 12:36:33