Energy Conversion and Management 86 (2014) 11341146

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Optical and thermal evaluations of a medium temperature parabolic

trough solar collector used in a cooling installation
Moncef Balghouthi , Abdessalem Bel Hadj Ali, Seif Eddine Trabelsi, Amenallah Guizani
Centre des Recherches et des Technologies de lEnergie CRTEn Bordj Cdria, Tunisia

a r t i c l e

i n f o

Article history:
Received 23 December 2013
Accepted 29 June 2014

Keywords:
Parabolic trough solar collector
Photogrammetry
Deformation
Slope errors
Intercept factor
Optical performance
Thermal efciency

a b s t r a c t
Concentrated solar power technology constitute an interesting option to meet a part of future energy
demand, especially when considering the high levels of solar radiation and clearness index that are available particularly in Tunisia. In this work, we study a medium temperature parabolic trough solar collector
used to drive a cooling installation located at the Center of Researches and Energy Technologies (CRTEn,
Bordj-Cedria, Tunisia). Optical evaluations of the collectors using photogrammetric techniques were performed. The analysis and readjustments of the optical results were conducted using a Matlab code. Therefore, slope errors ranged from 3 to +27 milliradian and the height deviations from the ideal shapes of
the parabolic trough collector were 2.5 mm in average with a maximum of 7.5 mm. The intercept factor
was determined using both the method of the total optical errors and the camera target method leading
respectively to 0.62 and 0.7. Thus, the values of the overall optical efciency were 0.48 and 0.514.
Conversely, a thermal performance testing of the parabolic trough collector was conducted leading to
the thermal efciency and the heat losses evaluations. The instantaneous thermal efciency reached a
maximum of 0.43 but it did not exceed the value of 0.30 when the reector becomes dirty by dust deposition. This study was also an opportunity for suggesting some recommendations for the enhancement of
the PTC performances.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
In the solar thermal applications needing relatively high temperatures, the energy is optically concentrated before being converted into heat. The sunlight is concentrated in the focal plane,
with the aim of maximizing the energy ux on the absorber surface. At present the Parabolic Trough Collector (PTC) can be considered as the most advanced solar thermal technology. It represents
the most mature solar technology to generate heat at temperatures
up to 400 C for solar thermal electricity generation [1]. The other
kind of PTC is destined to provide heat to processes that need temperatures between 100 and 250 C. These applications are mainly
industrial process heat, such as cleaning, drying, evaporation, distillation, pasteurization, sterilization, cooking, among others, as
well as heat driven refrigeration and cooling. Typical aperture
widths are between 1 and 3 m, total lengths vary between 2 and
10 m by row and geometrical concentrating ratios are between
15 and 20 [2]. The PTCs of this group are called medium
temperature collectors.
Corresponding author. Tel.: +216 97570277; fax: +216 430934.
E-mail addresses: balghouthi_moncef@yahoo.fr, moncef.balghouthi@crten.rnru.
tn (M. Balghouthi).
http://dx.doi.org/10.1016/j.enconman.2014.06.095
0196-8904/ 2014 Elsevier Ltd. All rights reserved.

As far as the importance of the medium temperature parabolic

trough solar collector applications, special concerns were attributed by some organizations and researchers to this kind of solar
collectors. In fact, the International Energy Agencys (IEA) developed the Task 33/IV program to improve and optimize medium
temperature solar-thermal collectors for solar industrial processes
heat. They reported that most solar applications for industrial processes have been used on a relatively small scale and are mostly
experimental in nature [3]. In addition, Cabrera et al. [4] performed
a literature survey on worldwide applications of the medium temperature PTCs to drive air conditioning and refrigeration facilities.
They reported that, despite the relatively important solar fraction
given by the PTCs compared to other solar collector technologies,
the yearly rate of grow of this type of installations is still low.
Recently, Minder [5] presented a medium temperature CSP eld
for indirect steam generation used for milk process industry in
Switzerland. The eld area is 115 m2 and the system uses thermal
oil as heat transfer uid and works up to 190 C. Besides, Sagade
et al. [6] described the experimental results of the prototype
parabolic trough destined for process heat applications made of
berglass-reinforced plastic with its aperture area coated by aluminum foil. They tested the steel receiver coated with black proxy
material. They achieved an instantaneous efciency of 51% and 39%

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

1135

Nomenclature
C
the geometric concentrating factor
DNI
the incident normal solar radiation (W/m2)
F
focal plan position of the parabola
f
focal length of the parabola
p
Parameter of the parabola
Sx
the mirror surface local slope
Tm
average temperature of the receiver uid (C)
Ta
ambient temperature (C)
Xa, Ya, Za
absolute coordinates
X, Y, Z
local coordinates
DSx
the local slope error
DZ
the height deviations regarding the ideal collector
Da
the slope deviations
a
the surface curvature
roptical standard deviation of the total optical errors (mrad)
rslope
standard deviation of slope errors (mrad)
rspecular standard deviation of specular errors (mrad)

with and without glass cover, respectively. Others authors [79]

studied the thermal performances of medium temperature parabolic trough particularly the thermal characterization of the receiver such as overall heat loss, end loss and thermal emittance of the
coating. In order to improve the performances of these kinds of PTC
and their usage for industrial process heat, much more investigations in the design, simulation, experimental and evaluating
technique ways are still required [10].
As the optical quality in particular the geometric precision of
the solar concentrators has a signicant impact on the efciency
and thus on the performance of the PTCs plants, many studies were
performed on the surface measurement methods of solar concentrators. Thomas et al. [11] and Xiaoa et al. [12] presented a review
of available methods for surface shape measurement of solar concentrator. They gave a detailed description of the very used techniques: the photogrammetry, the deectometry and the Video
Scanning Hartmann Optical Test (VSHOT). The most studied measurement technique was the photogrammetry which is a method
based on photographic processes and widely used for the 3-dimensional measurement of objects. The use of photogrammetry for the
parabolic trough collector shapes evaluations was performed by
Garca-Corts et al. [13], Shortis et al. [14], Fernndez-Reche
et al. [15] and Pottler et al. [16]. Digital close range photogrammetry has proven to be a precise and efcient measurement technique
for the assessment of shape accuracies of solar concentrators and
their components. The combination of high quality mega-pixel digital still cameras, appropriate software, and calibrated reference
scales in general is sufcient to provide coordinate measurements
with precisions of 1:50,000 or better. The extreme exibility of
photogrammetry to provide high accuracy 3D coordinate measurements over almost any scale makes it particularly appropriate for
the measurement of solar concentrator systems. In the last years,
close range photogrammetry has become a helpful tool to perform
this optical evaluation, mainly due to the commercial availability
of high resolution digital cameras and photogrammetry software
packages [17].
In this study, we present a medium temperature parabolic
trough solar collector used to drive a cooling installation located
at the Center of Researches and Energy Technologies (CRTEn,
Bordj-Cedria, Tunisia). In a previous study [18], dealing with the
description of the cooling installation, the results of the running
and the global performances COP were presented. Nevertheless,

rdisplacement standard deviation of receiver displacement errors

(mrad)

rtracking standard deviation of tracking errors (mrad)

rsun
Gaussian distribution for the errors caused by the sun
rtot
cn
W

gO
qm
s
ac

shape
the standard deviation of the total errors
the intercept factor at normal incidence angle
rim angles
the optical efciency
average specular reectance of the mirror
transmittance of the glass envelope
absorptance of the absorber surface coating

Subscripts and abbreviations

d
design data
m
measured data
PTSC
parabolic trough solar collector
HTF
heat transfer uid

in this work, we focused on the solar loop of the installation and

particularly the used parabolic trough solar concentrator. Optical
evaluations of the collectors using photogrammetric techniques
were performed. To establish that a parabolic trough concentrator
has a good optical quality, it was documented that tolerances must
be lower than 35 mm, and close-range photogrammetry is an
accurate enough technique to measure these surfaces as accuracies
lower than 1 mm can be easily achieved [16]. In addition, the
reector of the considered parabolic concentrator is polished aluminum without glass cover which allowed the positioning of the
targets exactly on the desired surface contrarily to the silvered
glass covered mirror when the targets are placed about 4 mm
above the reector due to the glass thickness [14]. Therefore, the
uncertainties of the results are reduced. Theses raisons put
together with the low-cost of the technique allowed us to adopt
the photogrammetry.
In the following paragraphs of the text, the procedures of
images capturing and 3D processing were presented. These procedures use combination of high quality megapixel digital still cameras, appropriate software, suitable targeting and calibrated
reference scales to provide coordinate measurements with high
precisions. The analysis and readjustments of the optical results
were conducted using a Matlab code leading to the slope errors
and the height deviations from the ideal shapes of the parabolic
trough collector. The intercept factor was determined using the
method of the total optical errors and the camera target method.
The overall optical efciency was then performed. Moreover, a
thermal performance testing of the parabolic trough collector
was presented leading to the thermal efciency and heat losses
evaluations. In addition some recommendations for the enhancement of the PTC performances were suggested.
2. General description of the solar cooling installation
The solar cooling installation is used to supply chilled water to a
research laboratory building located in the Research and Technology Center of Energy in Borj Cedria, Tunisia. It consists of 39 m2 linear parabolic trough solar collectors (PTSC) coupled to a 16 kW
double effect absorption chiller, a cooling tower, a backup heater,
two tanks for storage and drain-back storage and a set of fan-coils
installed in the building to be conditioned. A general scheme of the
installation is presented in Fig. 1.

1136

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

Fig. 1. Scheme of the solar cooling installation.

The PTSCs were installed in series and oriented in the direction

of the eastwest axis. The absorption chiller then uses thermal
energy converted from sun coming from the collector to generate
chilled water. The installation contains also a gas-red boiler used
as an auxiliary heating to boost the collector array outlet temperature when solar energy is insufcient.

3. The parabolic trough solar collector

The parabolic trough solar collectors plant constitutes the main
heat driving source of the air conditioning installation. It is composed essentially by a parabolic reective mirror, receiver tubes,
a steel support structure and a single-axis drive mechanism as
showed in Fig. 2. Solar radiations are reected by the mirrors and

focused on the absorber tubes where they are converted into thermal energy and transferred to the circulating heat transfer uid.
The converted solar thermal energy is used to drive the absorption chiller. The driving devices can track the sun, making sunlight
focusing on the receiver all the time. The reector is parabolic aluminum polished mirror, with dust-proof coating on the surface.
The installed PTSCs comprise three module assembled in series.
Each module has a 5.8 m by 2.3 m aperture opening. This
13.34 m2 aperture area and 0.68 m2 receiver area correspond to a
19.6 geometric concentration ratio. The reector mirror has a
0.89 typical reectance. The mirror is mounted on a steel support
structure articulated to steel supports tightly xed to the ground
through concrete foundations. The receiver is an absorber tube
located along the focal line of the parabolic mirror, whose surface
is treated and covered by an aluminumnitrogen/aluminum selective absorptive layer to enhance the heat collecting efciency. The
absorber tube is contained within an evacuated glass envelope to
minimize conduction, convection and radiation losses. The collector is orientated with its axis in the EastWest direction. The
advantages of this tracking mode are that very little collector
adjustment is required during the day and the full aperture always
faces the sun at noon.
4. Optical performance evaluations

Fig. 2. The parabolic trough solar collector.

The performance of the concentrating collectors is sensitive to

the optical errors. The optical quality in particular the geometric
precision of the reectors, the support structure and the receiver
placement have a signicant impact on the efciency and thus
on the overall performance of the collectors plant. Any deviation
from the optimum shape can lead to optical losses, therefore it is
important to have a tool that can measure surface slope errors with
adequate precision [12]. In practice, there are many techniques of
optical measurement used to analyze the optical errors and hence
determining the intercept factor. Of which we can cite the photogrammetry, the video scanning Hartmann optical test and the

1137

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

deectometry. In this study the shelf photogrammetry was used

for the examination of the parabolic trough solar collectors. The
procedure and results are exposed hereafter.
4.1. Photogrammetry quality tests
The shelf photogrammetry method was used to assess the optical quality of the parabolic trough collectors. The method is based
on the registration of object points in several images taken from
different viewing positions [17]. 3D coordinates of these points
are then reconstructed from their coordinates on images using
the software AICON 3D Studio. It is a valuable tool for the geometric analysis of large solar trough concentrators. This method is useful for determining the surfaces deformations and the slope
deviations from the ideal shape of the collector.
4.1.1. Camera adjustment and calibration
A Nikon D300 camera was used with a Nikkor MF 20 mm f/2.8D
lens and a Sunpak Ring Flash. The camera and lens specications
can be seen in Table 1.
In order to determine accurate camera model parameters (focal
length, principal point position and lens distortion parameters),
the camera must be calibrated. This is could be done by taking several photos convergent to a target grid specially designed for that
purpose. Different distances and camera format positions are used
to avoid coupling between internal parameters [14]. In our case,
scaled bares containing coded targets with known, precise and
xed separating interval were used. Four scaled bares disposed
together and forming a 3D xed shape were used. This process
was conducted in the dark so that the only signicant light source
was from the ash of the camera. Additional adjustments were
conducted using software called DPA CHECK to improve image
sharpness and precision. By running a set of 50 images of the
scaled bares through AICON 3D studio, the calibration parameters
of the camera were determined. However, the measurement system is operated at changing temperatures, usually between 10
and 40 C. Under these conditions, the distortion parameters of
the camera and lens system change. Therefore, one and the same
camera is used to take all pictures instead of using various xed
cameras for the photogrammetric measurement. With this
approach, the distortion parameters can be calibrated within each
measurement leading to reliable high measurement precisions.
The coefcients describing radial-symmetric distortion, tangential
distortion, afnity and shear were determined too, to be used later
for image corrections. The parameters that were used during this
study are listed in Table 2.
Table 1
Technical specications of the camera and lens used.
Camera model name

Nikon D300: NikonD2X-nikkorMF-20 mm

Resolution
Sensor size
Pixel size
Lens type
Lens focal distance
Maximum aperture
Lens construction
Picture angle
Distance scale

4288  2848 (Pixel)

23.584  15.664 (mm)
5.5  5.5 lm
Nikkor 20 mm
19.980 mm
f/2.8
12 elements in 9 groups
94
Graduated in meters and feet from 0.25 m
(0.85 ft) to innity (1)
f/2.8f/22
Fully automatic
Nikon bayonet mount
62 mm (P = 0.75 mm)
62 mm screw-in
Approx. 65 mm  42.5 mm extension from
ange; approx. 54 mm long (overall)

Aperture scale
Diaphragm
Mount
Attachment size
Filters
Dimensions

4.1.2. Collector targeting

The concentrator mirror surfaces are problematic because of
their reective behavior. The absence of natural points must be
solved using targets arranged on a collector surface. These targets
will be detected automatically during the photogrammetric work.
The usual procedure is to attach a sheet of adhesive vinyl on the
surface, with a printed array of targets with appropriate size and
shape [19].
Retro reective targets are xed on the mirror surface. Some of
these targets are coded to be directly distinguished by the software
(Fig. 3).
The targets are distributed uniformly throughout the surface to
ensure greater precision. A reference cross is also xed to establish
a reference point for measurements (Fig. 4). In addition, a number
of coded targets are distributed over the surface to allow the identication by the software. Even, the absorber tube, the ends of the
rotating axis and the reector plane of symmetry are marked.

4.1.3. Image capturing and 3D processing

The images were taken from different positions and different
elevations in order to cover the entire collector area and to assure
high level of recognition of the non-coded targets by the software.
The angular separations between the camera positions of less than
15 or greater than 165 should be avoided. An angular separation
of 90 is optimal for minimizing errors of angular sensitivity of the
camera, but a half-angle of this size is acceptable. Calculating
the coordinates in three dimensions requires that each point of
the object must appear in at least two or more images taken at
angles converging [20]. More than 100 pictures were taken to
ensure greater accuracy during processing. The scale of a picture
cannot be determined from photographs that do not contain
objects of known size, and so the scaling rods (bars of known
length) are included in all scenes (Fig. 5).
The locations and orientations of the camera are automatically
detected when photogrammetric measurements are processed,
and it is not necessary to measure and record these sites while taking photographs.
Digital processing of the captured images was done using a processing software, AICON 3D. The software has interfaces in which
we introduced the camera settings, cross reference, scaling rods
and coded targets parameters (Fig. 6).
A 3D coordinate system is used to restore the digital pictures.
Each target coded and non-coded, identied by a unique number
is assigned by three absolute coordinates (Xa, Ya, Za). A change
of reference is needed to duplicate the representation of the actual
shape of the reector. Thus, it is necessary to identify and readjust
Table 2
Metric calibration parameters of the camera and lens used.

Focal distance
Principal point offset in x-image
coordinate
Principal point offset in x-image
coordinate
3rd order term of radial distortion
correction
5th order term of radial distortion
correction
7th order term of radial distortion
correction
Coefcient of decentering distortion
Coefcient of decentering distortion
Differential scaling between x and y
Non-orthogonality between x and y
axes

Value

Standard
error

19.9820 mm
XH = 0.2097

0.0002 mm
0.00029

YH = 0.2428

0.00034

A1 = 2.8421e004

2.11e007

A2 = 6.6295e007

2.622e009

A3 = 5.9024e010

1.0396e011

B1 = 1.8631e005
B2 = 1.1269e005
C1 = 1.257e004
C2 = 3.9660e005

2.048 e007
1.922 e007
1.82 e006
1.84 e006

1138

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

Fig. 3. Collector targeting.

Fig. 4. Reference cross and scaling rods used as reference points for measurements.

Fig. 5. Samples of targeted collector photos taken from different locations for photogrammetry processing.

targets previously glued to mark the axes coordinates. Usually, one

of the axes is selected to be the rotation axis of the parabolic
trough (in our case it is the Y axis), the origin is chosen on one of
the reector ends symmetry axis and a second axis (Z-axis) will
be marked with targets attached to the absorber tube. Considering
that the tracking axis (Y axis) of the parabola is the eastwest axis,
the Z-axis was oriented to the south and the positive X coordinates
corresponded to the lower part of the reector. After the acquisition and processing of images, the software will generate a graph
as a scatter plot.

4.2. Analysis and readjustment

4.2.1. Real reector surface representation
The real surface of the reector generated by the AICON 3D
processing software was provided as set of points recopying the

real shape of the parabola and related to a 3D referential (x, y, z)

(Fig. 7).
In order to identify the reector surface errors, it must be compared to an ideal surface of the parabola dened by the following
Eq. (1):

X2
X2

2p 4f

To distinguish the surface errors, a comparison of two criteria is

needed: the height deviations regarding the ideal collector DZ and
the slope deviations Da (Fig. 8).
The slope errors are the deviations of the surface normal
directions away from the ideal normal directions. They could be
determined using Eqs. (2) and (3):

a tan1



X1  X2
Z1  Z2

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

1139

Fig. 6. Interfaces of the digital images processing software, AICON 3D.

Fig. 7. Representation of the real surface of the reector generated by the AICON 3D processing software as set of points recopying the real shape of the parabola.

Da areal  aideal

where a is the surface curvature and Da is the slope deviation.

4.2.2. Height deviations
The knowledge of Z coordinate deviations or the height deviations provide information on the magnitude and location of the
concentrator surface differences in respect to the ideal parabolic
shape, and can be useful to improve the manufacturing process
and the collectors mounting and adjusting [16].
The height deviations are obtained using a Matlab code that
shows the difference between the actual and the ideal surface area
of the reector (Fig. 9). By examining gure, we noticed that the
height distribution has almost two areas with an antisymmetric
shape, this observation allowed to identify the problem which
was an error in the identication during the photogrammetric
analysis. There was a shift in the marking of the focal plan relative
to the correct position which caused an angular reposition around
the Y axis.
After readjusting of the referential by a rotation of an angle
U = 0.019 around the Y axis and analyzing the new coordinates,
a representation of the distributions of the surface height errors
in 3D was generated by a MATLAB code (Fig. 10). The graph shows
that the surface deformations increase going from the center to the
edges of the reector.

These deformations are probably caused by the wind impacts

and the weight of the metallic structure. The most intense deformations were located in the edges of the trough, with a maximum
value of 7.5 mm. Although a bigger part of the surface has a deformation value with an average of 2.5 mm.
4.2.3. Slope errors
The optical performance of a trough concentrator depends on
the slope errors, which tend to deviate the reected rays away
from the ideal focal line. Slight variation of the position where light
is reected is less important than the change in the direction of the
reected rays. The coordinates of the targets were used to estimate
the differences of local slopes. The slopes (Sx)m, given by Eq. (4) are
calculated along the x-axis using the coordinates of two adjacent
targets (Pi and Pj).

Sx m

Z j m  Z i m 
X j m  X i m 

The subscript m here designates measured data while the subscript d designates design data.
The slope calculations were performed in the plane (X, Z). The
ideal slopes (Sx)d were calculated with Eqs. (5) and (6) using the
coordinates Zi of the Z-axis and Xi along the X-axis and the focal
length f.

1140

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

The slope error DSx was dened as the deviation of the measured slope from the design slope (Eq. (7)):

DSx Sx d  Sx m

The standard deviation rslope of the zero-mean Gaussian distribution of the local slope errors was calculated using equation (Eq.
(8)) by where n is the number of calculated slopes.

rslope

Fig. 8. Scheme of the height deviations regarding the ideal collector DZ and the
slope deviations Da.

Sx d

Z i

Z j d  Z i d 
X j m  X i m 

X i 2d
4f

s
Xn DSx 2
1

The slope errors were performed using a separate post processing step in MATLAB based on the Delaunay triangulation [21]. The
results yielded a maximum absolute slope error of 27.8 milliradian,
and a standard deviation rslope = 11.9 milliradian.
Fig. 11 shows the distributions of the slope errors on the reector surface. The sign convention for the slope errors has been
dened positive for reected rays passing above the focal line
and negative for reected rays passing below the focal line. It can
be drawn that the lower edge along the trough length is deviated
from the ideal surface with negative deviations; the vertex of the
concentrator is slightly positive-deviated. There is a zone in right
side of the lower part of the collector showing the largest deviations from the ideal slope of up to 27 milliradian and there is also

Fig. 9. Distribution of the height deviations of the reector surface before readjustment of the focal plan marking.

Fig. 10. Distribution of the height deviations of the reector surface after readjustment of the focal plan targeting.

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

1141

a small zone in the top-right corner with important deviations. The

slope errors go from 3 to +27 milliradian.
Using these results, the concentrator geometry could be
improved by readjusting the reector assembly on the structure.

centerline from the design focus of the parabola. The average of this
error was estimated to 6.8 milliradian. Considering rslope of
11.9 milliradian, we get an optical errors roptical of 24.75 milliradian.
The standard deviation of the total errors (combined optical and
sun shape) of the collector is given by Eq. (11):

4.3. Uncertainty analysis

r2tot r2optical r2sun

The characteristics of the camera and the calibrations parameters presented in Tables 1 and 2 were used by AICON 3D Studio
for image corrections. The software calculates the threedimensional coordinate values dening the positions of the targets
in space and the relative uncertainties. The precision of each target
coordinate were provided during the digital processing of the
captured images. Fig. 12 gives a sample of the photogrammetric
measurement uncertainties of the target positions. The largest
uncertainty of all single distance measurements is below
0.4 mm while the standard deviation is 0.12 mm.
Similarly to the measurement of slope deviation, the calculation
of the slope deviation uncertainties were carried out using an algorithm implemented in MATLAB. A detailed uncertainty analysis
was performed according to GUM 08 [20]. The maximum measurement uncertainty in the surface slope is 1.0 mrad with a standard
deviation of 0.6 mrad.

A Gaussian distribution approximation for the errors caused by

the sun shape rsun was adopted to be 2.6 milliradian as expected
by Rabel [22]. Therefore the total error rtot was 24.9 milliradian.
5. Intercept factor
The geometric accuracy of parabolic trough collectors is
described by the intercept factor, which includes the optical effects
of reector shape and receiver absorber alignment among others.
The intercept factor is dened as the fraction of the rays incident
on the aperture surface of the reector that are intercepted by
the receiver. It refers to the question if the rays hit the absorber
or not. Two optical analysis methods have been used to characterize the intercept factor of the concentrator: the sum-of-squares
approximation or the total errors and the camera-target-method.
5.1. Method of the total optical errors

4.4. Total optical errors

In a real collector there will be several statistically independent
sources of optical error: lack of perfect specularity, macroscopic
surface deviations in position and slope, displacement of the receiver, and tracking errors. Averaged over time and over the entire
collector or array of collectors, all of these errors can be assumed
to be approximately Gaussian [22]. The standard deviation
accounts for all optical errors and could be calculated in Eq. (9):

r2optical 4r2slope r2specular r2displacement r2tracking

This method uses the plot of Bent et al. [23] giving the intercept
factor at normal incidence angle cn versus the product of the total
optical error rtot and the geometric concentrating factor C for different rim angles W (Fig. 13).
In our case the rim angle or the opening angle of the parabolic
trough is about 75, the geometric concentrating factor C is 19 and
the product rtotC is 473.1 milliradian. Accordingly the intercept
factor determined is cn = 0.62.

In our study we consider the case of normal incidence, the

tracking errors can be overlooked. Even the reector material is
the polished aluminum which has a low diffusive reectance, consequently, the specular errors can be neglected in the calculation of
the total errors. Accordingly, Eq. (9) is reduced to Eq. (10):

r2optical 4r2slope r2displacement

11

10

The receiver displacement or misalignment error rdisplacement is

dened as the angular deviation of the location of the receiver

5.2. The camera-target-method CTM

Another method to estimate the intercept factor of parabolic
troughs is based on the indirect CAMERA-TARGET-method
CTM. This method uses a at, diffuse reecting target placed perpendicular to the absorber tube. A cut in the target allows to envelope the receiver tube almost entirely. The solar rays around the
receiver are visualized on the objective, and the rays which miss
the absorber tube are identied. Many pictures at different receiver
locations are captured (Fig. 14).

Fig. 11. Distributions of the slope errors on the reector surface.

1142

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

0.4

The resulting optical efciencies at normal incidence obtained

using the camera-target-method and the method of the total optical errors are respectively 0.514 and 0.48.

0.3

Deviation (mm)

0.2

7. Thermal performances

0.1
0

500

1000

1500

2000

-0.1
-0.2
-0.3
-0.4

target distance (mm)

Fig. 12. Photogrammetric measurement uncertainties relative to the targets

positions.

These pictures were used to determine an approximate intercept factor of the collector at the target location. By integrating
over two half circles; one in front of the absorber tube for the
incoming rays and one behind the absorber tube for the passing
rays [24]. The intercept can be calculated as the ratio of the rays
that hit the absorber to the incoming rays. Fig. 15 shows the estimations of the intercept factor in different locations of the receiver.
The obtained intercept factor varies from 60% to 87% with an average value of 70%.

6. Optical efciency
The optical efciency gO is dened as the amount of radiation
absorbed by the absorber tube divided by the amount of direct normal radiation incident on the aperture area. The optical efciency
when the incident radiation is normal to the aperture is given in
Eq. (12):

gO qm sac cn

12

Accordingly, considering the optical properties of our collector

qm = 0.89 and the effective product transmittance absorbance

sac = 0.85, the optical efciency of the collector can be calculated
by substituting cn by their values obtained by the two methods.

Fig. 13. Determination of the Intercept factor using the plot of Bent et al. [21].

The thermal performance of solar collectors can be determined

by experimental performance testing under controlled conditions.
In general, experimental verication of the collector characteristics
is necessary in order to determine the thermal efciency of the collector. There are a number of standards, which describe the testing
procedures for the thermal performance of solar collectors. The
most well known are the ISO 9806-1:1994 and the ANSI/ASHRAE
Standard 93:2003 [25].
Since the collector will be optimized based on either instantaneous or all-day average efciency, a steady-state thermal analysis
of the receiver will sufce for the design studies. The performance
of the PTSC according to the previously mentioned tests is determined by obtaining values of the useful heat gain, the collector
instantaneous thermal efciency, the energy gained by the storage
tank and the overall efciency, for different parameters of operation; incident radiation, ambient temperature and inlet HTF
temperature.
7.1. Performance testing of the parabolic trough collector
The solar collectors plant was tested during the summer. An
acquisition data system HP Agilent Data Logger was used to
record solar radiation intensities, temperatures at different locations of the receiver and the reector, the HTF mass ow rates
and the wind velocity. 30 thermocouples K, 4 Platinum resistance
thermometers Pt100, two Pyranometer CM 21, two Pyrheliometers
CHP1 and two ow meter were connected to two Agilent HP
34901A 20-Channel Multiplexer for data acquisition and
processing.
The used instrumentation devices, their specications and relatives precisions are presented in Table 3.
The used heating uid HTF in the presented experiments is the
pressurized water. Actually and due to problems of overpressure,
we used thermal oil instead of water to assure more stability of
the installation pressure.
A summer day measurements are presented in Fig.16. The heating uid HTF ow rate were 1.2 m3/h. The direct solar normal
radiation varied from 400 to 950 W/m2 and the maximum HTF
temperature at the collector outlet was 164 C.
Experimental data shows that the solar thermal system normally takes about three to 4 h to heat the system to 160 C, the
nominal operating temperature of the absorption chiller. The
overall system has a high heat capacity. This heat capacity
consumes much useful solar energy and prolongs the warm up
period before solar energy is available to be used by the absorption chiller.
There were thermal losses occurring at the receiver due to the
difference between the glass cover temperature and the ambient
temperature whose values are 49 C and 37 C respectively. Even,
in addition of the losses of vacuum in the annular space between
the absorber and the glass cover, heat losses are due to heat conduction throughout the steel supports, which are in contact with
both absorber tubes and the glass cover.
This constitutes a weak point in the design of the collector. In
fact, the temperature of supports exceeded 60 C. Besides, the
reector surface temperature is higher than the ambient temperature and is equal to 42 C. Accordingly, there were energy losses on
the surface and the reection was not perfect. This could be due to
dust deposit on the reector surface.

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

1143

Fig. 14. Photos of the CAMERA-TARGET-method CTM procedure.

90%
80%

Intercept Factor

70%
60%
50%
40%
30%
20%
10%
0%
0

10

11

12

13

14

Cible positions on the reciever

Fig. 15. Determination of the intercept factor using the CAMERA-TARGET-method.

Fig. 16. Evolutions of the temperatures at different locations of the parabolic

trough solar collector as well as the solar radiation and the ambient temperature.

7.2. Thermal efciency and heat losses

The performance of the parabolic trough concentrator was evaluated according to the thermal tests by obtaining values of the useful heat gain, the available solar energy on the collector and the
collector instantaneous thermal efciency for different parameters
of operation such as the incident beam solar radiation, ambient
temperature and inlet HTF temperature.
The useful heat gain was determined from the measurements of
the inlet and outlet HTF temperatures and mass ow rate. The
available solar power was the product of apparent concentrator
area by the incident beam solar radiation at normal incidence.

Fig. 17 presents the evolutions of the available solar power and

the useful power transferred to the thermal uid. We can see that
the available solar energy reached maximum values between
13:00 and 15:00 with a peak of 30 kW at 14:20 local time. The useful heat gain rst increases, reach a peak value of 12 kW around
14:00 then decreases. Data for the available solar energy and useful
heat gain by the collector was used to determine the instantaneous
efciency.
As expected in Fig. 18, the instantaneous efciency of the parabolic trough solar collector increases, reaches a maximum of 0.43
then declines. Moreover, when the reector becomes dirty by dust
deposition, the efciency decreases notably and does not exceeds
the value of 0.30.
The thermal efciency of the PTSC was also performed under
steady state conditions as proposed by ASHRAE Standard
93(1986) [26]. The periods of tests were chosen when the operating conditions are almost in steady states such as the direct solar
radiations, the HTF ow rate, the wind speed and the ambient temperature. The collector efciency was plotted against the difference
between average HTF temperature in the receiver and the surroundings temperature divided by the direct normal radiation
(Fig. 19). The obtained curve was approximated to a straight line.
The highest efciency of the system is realized when the mean
HTF temperature is equal to the ambient temperature (no thermal
losses).
This efciency represents the optical efciency at the normal
incidence, it is about 0.58. This value is comparative to the values
of the optical efciency determined in Section 6 using the cameratarget-method and the method of the total optical errors. The proprieties of the receiver such as the absorption and the emissivity
have also an important effect on the thermal performances.
Beside, the thermal efciency decreases with the increase of the
temperature difference between receiver uid and the surroundings (Tm  Ta) or the decrease of the incident normal solar radiation
DNI.
The thermal efciency is affected by heat losses caused by the
conductive, convective and radiative exchanges with the surrounding. When the temperatures of the receiver and the glass cover
increase, the radiative losses increase. Even the convective losses
increase due to the temperature gradient between the receiver
and surrounding and wind velocity. Even, the thermal tests have
shown that there were considerable heat losses in the receiver
due to heat conduction throughout the steel supports, which are
in contact with both absorber tubes and the glass cover. Moreover,
the heat losses were extremely high during the night from the solar
collector tubes. This reduces the HTF temperature overnight; leading to a morning temperature close to ambient. These losses were
essentially due to radiant exchanges between the absorber tubes
and the night sky. These night-time heat losses were responsible
for a start-up delay time of the absorption chiller. In fact, the parabolic trough solar collector needs about 4 h to heat up the HTF to a

1144

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

Table 3
Specications and precisions of the used instrument devices.
Instrumentation device

Measured physical parameter

Precision

Thermocouple type K
Platinum resistance thermometers Pt100
Pyranometer CM 21 Kipp and Zonen
Pyrheliometers CHP1 Kipp and Zonen
Flow meter TME/UMC-3
Flow meter OMG
Anemometer NRG #40C

Temperature (C)
Temperature (C)
Global solar radiation (W/m2)
Incident normal solar radiation (W/m2)
Water ow rate (kg/h)
Oil ow rate (kg/h)
Wind speed (m/s)

0.2 C
0.2 C
5 W/m2
1% of the measured quantity (W/m2)
0.3% of the measured quantity (kg/h)
0.3% of the measured quantity (kg/h)
0.1 m/s

Table 4
Comparison with similar designs.
Model

Solitem PTC 1800

IST PTC

NEP Solar Polytrough

1200

NEP Solar Polytrough

1800

Current study

Max. operating temp.(C)

Aperture width (m)
Focal length (m)
Rim angle (degrees)
Geometric concentration ratio
Absorber tube diameter (mm)
Absorber tube material and
coating
Absorptivity
Working uid
Reector material
Cover tube diameter (mm)
Reectivity
Mean detected intercept factor
Peak optical efciency
References

220
1.8
0.78
60
15
38
Stainless steel/Black
chrome
0.94
Press.water/thermal oil
Polished aluminum
65
0.89
0.72
0.53
Weiss et al. [26]
Lokurlu et al. [27]

280
2.3
0.8
72
14.36
51
Steel/blackened
nickel
0.960.98
Press.water
Silvered acrylic
5
0.89
0.76
0.76
Kalogirou et al. [25]

220
1.2
0.65
50
15
25.4
Stainless steel/Black
chrome
0.94
Press.water/thermal oil
Polished aluminum
56
0.89
_
0.62
Fernandez et al. [2]
NEP SOLAR [28]

250
1.845
0.65
71
17.28
34
Stainless steel/Black
chrome
0.94
Press.water/thermal oil
Polished aluminum
56
0.89
_
0.68
SPF [29]
NEP SOLAR [28]

200
2.3
0.8
73
19
38
Stainless steel/Black
Nickel
0.96
Press.water/thermal oil
Polished aluminum
90
0.89
0.620.7
0.480.502

32
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
7:55

of 125 C instead of 30 C without the drain back storage. Using

the drain back storage has allowed the outlet solar collector temperature to attain the value of 160 C, required for the chiller
startup, at 10:40 rather than 14:00 in the case of no drain back
storage.
7.3. Comparison with similar designs
The current study is based on optical and thermal performances
of parabolic trough solar concentrator used to drive a cooling
installation under Tunisia environment. This concentrator is classied in the category of medium temperature parabolic trough solar
collector. Comparison between previous and current research are
analyzed. Although this kind of research is new in Tunisia, the
analysis was made on installations over the word focusing on

Collector useful Power

output (kW)
Available solar power (kW)

0.5
0.45

Solar collector effeciency

Power (kW)

temperature of 160 C required by the absorption machine to drive

the cooling process. In a previous work [18], we studied the importance of a drain back night thermal storage system for the parabolic trough collectors. Drain back might signicantly reduce the
loss of heat over night from the HTF contained in the system. An
insulated drain-back tank with a capacity of 400 l was added in
the solar collection loop of the installation. A drain-back system
pumped and drained all the HTF, at about 140 C in the late afternoon and pumped this hot uid back into the system the next
morning. Fig. 20 shows a comparison between HTF outlet temperature with and without the drain back storage. The HTF solution
was drained to the back storage tank at 19:00 local time at a temperature of 142 C, the next day at 8:40, the temperature of the
stored HTF became 129 C. After pumping the HTF back into the
systems, the solar collectors started heating from a temperature

0.4

Clean Reflector

0.35

Reflector with
dust

0.3
0.25
0.2
0.15
0.1
0.05

10:19

12:43

15:07

17:31

19:55

Local time (hh:mm)

Fig. 17. Daily variation of the available solar powers and the useful power of the
parabolic trough solar installation.

0
7:40 8:38 9:36 10:33 11:31 12:28 13:26 14:24 15:21 16:19 17:16 18:14 19:12 20:09

Local time
Fig. 18. Daily parabolic trough solar collector efciency with clean and dirty
mirrors.

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

of 220 C and an optical efciency of 0.62. While the Polytrough

1800 have a maximum temperature of 220 C and a peak optical
efciency of 0.68 [2,30,31].

0.7
0.6

y = -1.1777x + 0.5816
R = 0.9498

Efficiency

0.5

8. Recommendations for the enhancement of the PTC

performances

0.4
0.3
0.2
0.1
0
0

0.05

0.1

0.15

Mean temperature above ambient divided by solar radiation

intensity (Tm - Ta)/G
Fig. 19. The thermal efciency of the PTSC under steady state conditions.

Temperature (C)

1145

170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
7:12

Tcollec_out
Tcollec_out with drain back
storage

12:00 16:48 21:36 2:24

7:12

12:00 16:48 21:36 2:24

7:12

Local me
Fig. 20. Parabolic trough solar collector outlet temperatures with and without
backup night storage.

medium temperature parabolic trough solar collector (80250 C)

used for industrial process heat or solar air conditioning. Table 4
summarizes the main characteristics of some commercial and prototypes of PTCs designed for medium temperature applications and
having similar dimensions to the concentrator subject of this
study. Some additional informations about these collectors are
given below:
 The IST PTC parabolic trough model, founded in the United
States and recently acquired by the Spanish company, Abengoa
Solar. The reector is silvered acrylic or enhanced-polished
aluminum. The absorber is a steel tube with a black nickel
anti-reective coating. The maximum operating temperature
was 280 C and the peak optical efciency was 0.75 [27].
 The Solitem PTC-1800 was developed by Solitem Company
(Turkey, Germany) with the help of the DLR in design and evaluating. It consists of an 1.8 m aperture width polished aluminum reector, a selective coated stainless-steel absorber
having 38 mm in diameter The maximum operating temperature was 280 C and the peak optical efciency was 0.75 [28,29].
 NEP SOLAR Polytrough 1200 and Polytrough 1800 were developed by the Australian company, New Energy Partners Pty
Ltd. (NEP) in collaboration with Australias Commonwealth
Scientic and Industrial Research Organisation (CSIRO). The
two models have polished aluminum reectors and stainless
steel absorber tubes with selective coating (black chrome).
The Polytrough 1200 could attend an operating temperature

The evaluations of the parabolic trough solar collector have

revealed signicant optical and thermal losses. The optical analysis
of the collector showed that the relatively low optical efciencies
were caused by the reectors surfaces deformations and the slope
deviations from the ideal shape of the collector other than the displacement of the receiver in the focal line. The thermal losses were
caused by conduction, convection and radiation from the receiver
which include the glass cover, absorber tube, bellows, exible hose
and supports. Beside, considerable heat losses occurs in night-time
due to radiative exchange between the receiver and the night sky
leading to a morning HTF temperature close to ambient and causing a start-up delay time of the solar cooling installation.
Based on these assessments, several recommendations were
given for the improvement of the collectors performances:
 The mounting of the reectors should be revised and some mirrors should be replaced. Even, the supports of the reector
should be readjusted in order to minimize the errors from the
ideal shape of the parabola.
 The receiver and the absorber tube should be altered to avoid
the misalignments from the focal line.
 The absorber tube surfaces should be coated by the right selective coating to get high solar absorptivity and low thermal
emissivity. Moreover, the absorber tube diameter should be
determined by a tradeoff between solar radiation intercepted
by the absorber pipe and its thermal losses.
 The glass envelope should be designed with as small diameter
as possible with material having high transmittance, low
absorptivity and small thickness. In addition the dimensions
and the exposed area of the bellow should be reduced and the
expansion piece should be contained inside the glass cover.
 Using a drain back night storage system will signicantly
reduce the loss of heat over night from the HTF and decrease
the start-up delay time of the solar cooling installation.
9. Conclusion
A medium temperature parabolic trough solar collector used
to drive a cooling installation located at the Center of Researches
and Energy Technologies (CRTEn, Bordj-Cedria, Tunisia) was studied. Optical evaluations of the collectors using photogrammetric
techniques were performed. The procedures of images capturing
and 3D processing are presented. The analysis and readjustments
of the optical results were conducted using a Matlab code
leading to the identication of the slope errors and the height
deviations from the ideal shapes of the parabolic trough collector. Moreover, thermal performance testing of the parabolic
trough collector was presented leading to the thermal efciency
and heat losses evaluations. The main results and relevance are
the followings:
 The most intense deformations were located in the edges of the
trough, with a maximum value of 7.5 mm. But, an average
deformation value of 2.5 mm characterized the larger part of
the surface. The slope errors ranged from 3 to +27 milliradian
with a maximum absolute slope error of 27.8 milliradian, and a
standard deviation rslope = 11.9 milliradian. Accordingly, total
error rtot was 24.9 milliradian and the intercept factor determined was cn = 0.62.

1146

M. Balghouthi et al. / Energy Conversion and Management 86 (2014) 11341146

 The resulting optical efciencies at normal incidence obtained

using both the camera-target-method and the method of the
total optical errors are respectively 0.514 and 0.48.
 The instantaneous efciency of the parabolic trough solar collector increases, reaching a maximum of 0.43 then declines.
However, when the reector becomes dirty by dust deposition,
the efciency decreases notably and did not exceed the value of
0.30.
 The thermal efciency of the PTSC determined under steady
state conditions as proposed by ASHRAE Standard 93(1986)
yielded a highest value of 0.58 when the mean HTF temperature
is equal to the ambient temperature (no thermal losses). This
efciency represents the optical efciency at the normal incidence which is about 0.58. Besides, the thermal efciency
decreases with the increase of the temperature difference
between receiver uid and the surroundings or the decrease
of the incident normal solar radiation.
 The evaluations of the parabolic trough solar collector revealed
signicant optical and thermal losses. Therefore, some recommendations for the enhancement of the PTSC performances
were suggested. They concern the revision and readjustment
of some mirrors, the absorber tube coating and alignment, the
glass envelope diameter and thickness and the expansion pieces
isolation. Even, a night storage tank was proposed.

Acknowledgement
The authors would like to thank the members of the enerMENA
project and the German Aerospace Center DLR for their scientic
and nancial supports.
References
[1] Sansoni P, Fontani D, Francini F, Giannuzzi A, Sani E, Mercatelli L, et al. Optical
collection efciency and orientation of a solar trough medium-power plant
installed in Italy. Renew Energy 2011;36(9):23417.
[2] Fernandez-Garca A, Zarza E, Valenzuela L, Perez M. Parabolic trough solar
collectors and their applications. Renew Sustain Energy Rev 2010;14:
1695721.
[3] Vannoni C, Battisti R, Drigo S. Solar heat for industrial processes, potential solar
heat for industrial processes. IEA SHC Task 33, SolarPACES Task IV. Solar
Heating and Cooling Executive Committee of the International Energy Agency,
Italy; 2008.
[4] Cabrera FJ, Fernandez-Garcia A, Silva RMP, Prez-Garcia M. Use of parabolic
trough solar collectors for solar refrigeration and air-conditioning applications.
Renew Sustain Energy Rev 2013;20:10318.
[5] Minder Stefan. Example of concentrated solar systems (PTC) in the dairy
industry in Switzerland. SHC Workshop on Solar Process Heat for Industry. NEP
Solar AG; 15th of March 2013. <www.nep-solar.com>.
[6] Sagade AA, Satish A, Shinde NN. Performance evaluation of low-cost FRP
parabolic trough reector with mild steel receiver. Int J Energy Environ Eng
2013;4:5.

[7] Lei D, Li Q, Wang Z, Li J, Li J. An experimental study of thermal characterization

of parabolic trough receivers. Energy Convers Manage 2013;69:10715.
[8] Li M, Wang LL. Investigation of evacuated tube heated by solar trough
concentrating system. Energy Convers Manage 2006;47:3591601.
[9] Omer SA, Ineld DG. Design and thermal analysis of a two stage solar
concentrator for combined heat and thermoelectric power generation. Energy
Convers Manage 2000;41:73756.
[10] International Renewable Energy Agency IRENA. Concentrating Solar Power:
Technology Brief. IEA-ETSAP and IRENA. Technology Brief E10; January 2013.
[11] Thomas A, Guven HM. Parabolic trough concentratorsdesign, construction
and evaluation. Energy Convers Manage 1993;34(5):40116.
[12] Xiaoa J, Weia X, Lua Z, Yua W, Wua H. A review of available methods for
surface shape measurement of solar concentrator in solar thermal power
applications. Renew Sustain Energy Rev 2012;16:253944.
[13] Garca-Corts S, Bello-Garca A, Ordez C. Estimating intercept factor of a
parabolic solar trough collector with new supporting structure using off-theshelf photogrammetric equipment. Appl Energy 2012;92:81521.
[14] Shortis MR, Johnston GHG, Pottler K, Lpfert E. Photogrammetry analysis of
solar collectors. The International Archives of the Photogrammetry, Remote
Sensing and Spatial Information Sciences, vol. XXXVII. Part B5. Beijing; 2008.
[15] Fernndez-Reche J, Valenzuela L. Geometrical assessment of solar
concentrators using close-range photogrammetry. Energy Procedia 2012;30:
8490.
[16] Pottler K, Lupfert E, Johnston GHG, Shortis MR. Photogrammetry: a powerful
tool for geometric analysis of solar concentrators and their components. J Sol
Energy Eng 2005;127(February):94101.
[17] Shortis MR, Johnston GHG. Photogrammetry: an available surface
characterization tool for solar concentrators, part l: measurements of
surface. J Sol Energy Eng 1996;118:14650.
[18] Balghouthi M, Chahbani HM, Guizani A. Investigation of a solar cooling
installation in Tunisia. Appl Energy 2012;98:13848.
[19] Lpfert E, Pottler K, Ulmer S, Riffelmann K, Neumann A, Schiricke B. Parabolic
trough optical performance analysis techniques. J Sol Energy Eng
2007;129:14752.
[20] Lo SH. Delaunay triangulation of non-uniform point distributions by means of
multi-grid insertion. Finite Elem Anal Des 2013;63:822.
[21] JCGM 100:2008 GUM 1995 with minor corrections. Evaluation of
measurement data: Guide to the expression of uncertainty in measurement,
1st edition; September 2008.
[22] Rabl Ari. Active solar collectors and their applications. New York: Oxford
University Press; 1985. 449 pages.
[23] Bendt P, Rabl A, Gaul H, Reed KA. Optical analysis and optimization of line
focus solar collectors. Golden, CO: SERI/TR 34-092 Solar Energy Research
Institute; 1979. September.
[24] Riffelmann KJ, Neumann A, Ulmer S. Performance enhancement of parabolic
trough collectors by solar ux measurement in the focal region. Sol Energy
2006;80:130313.
[25] IEA-SHC TASK 43: Solar rating and certication procedures. subtask a:
roadmap of collector testing and certication issues. SHC: Solar Heating and
Cooling Program. International Energy Agency; March 2012. 45 pages.
[26] ANSI/ASHRAE 93-method of testing to determine the thermal performance of
solar collectors. New York: ASHRAE Inc.; 1986.
[27] Weiss W, Rommel M. Process heat collectors. State of the art within task 33/IV.
IEA SHC-Task 33 and SolarPACES-Task IV. Gleisdorf: AEE INTEC; 2008.
[28] Lokurlu A, Richarts F, Kruger D. High efcient utilisation of solar energy with
newly developed parabolic trough collectors (SOLITEM PTC) for chilling and
steam production in a hotel at the Mediterranean coast of Turkey. Int J Energy
Technol Policy 2005;3(1/2):13746.
[29] Kalogirou SA. Solar thermal collectors and applications. Prog Energy Combust
Sci 2004;30:23195.
[30] NEP solar; 2013. <http://www.nep-solar.com>.
[31] SPF Testing Institut fr Solartechnik SPF. Solar Collector Factsheet NEP
Polytrough 1800. C1549; 2013. <www.solarenergy.ch>.