Tesis Doctoral

Ingeniería en Electrónica Industrial y de Telecomunicaciones

Caracterización del rendimiento de los tubos receptores de

tecnología solar termoeléctrica cilindroparabólica
 Tesis Doctoral
Ingeniería en Electrónica Industrial y de
Telecomunicaciones
 Solar Energy 136 (2016) 268–277

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Parabolic trough solar receivers characterization using specific test

bench for transmittance, absorptance and heat loss simultaneous
measurement
J.L. Navarro-Hermoso a,⇑, G. Espinosa-Rueda a, C. Heras c, I. Salinas c, N. Martinez b, M. Gallas a
a
Abengoa Research, Soland A-472, Sanlúcar la Mayor, 41800 Sevilla, Spain
b
Abengoa Research, Paseo de la Castellana, Madrid, Spain
c
Universidad de Zaragoza, Grupo de Tecnologías Fotónicas, Zaragoza, Spain

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

Article history: Parabolic trough is the most extended solar thermoelectric technology. Solar radiation is converted into
Received 2 February 2016 heat and transferred to the heat transfer fluid in the solar receiver tubes. The thermal energy obtained
Received in revised form 13 May 2016 feeds a conventional Rankine power cycle. The performance of the receiver tube can be broken down into
Accepted 7 July 2016
three single components: the optical transmittance of the outer glass envelope or capability in transmit-
ting the radiant energy; the optical absorptance of the metal tube or capability to absorb the radiation;
and heat losses of the tube or capability to retain the heat which depends of inner tube coating emissivity
Keywords:
and the vacuum between both tubes. In this paper a novel test bench implementing both thermal and
Parabolic trough solar receiver
Measurement test bench
optical measurement systems is described and compared with other systems referred in the literature.
Absorptance and transmittance The results obtained from the evaluation of three different solar receivers with different diameters are
Heat losses presented. Optical measurements of transmittance and absorptance parameters are carried out in the
wavelength range of 300–2500 nm. Optical evaluation of trough receivers at operating temperatures
up to 450 °C is feasible. The receivers can be heated using a high intensity electric current flowing
through the internal tube. Uniform Joule heating results a reliable heat losses measurement method at
temperatures up to 650 °C.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction tubes with vacuum in the inter-annular space. Vacuum reduces

Solar thermoelectric technology has emerged as an alternative
to conventional energy sources. Major players in this sector have
made big efforts in research and development to achieve perfor-
mance increase and cost reduction.
Parabolic trough represents the most common STE technology.
Solar radiation concentrated by mirrors is converted into heat in
receiver tubes. According to Fernandez-García et al. (2010) and
Price (2003), receiver tube represents a key component and its per-
formance influences decisively the overall plant efficiency.
Receiver tubes require maximum solar energy absorption and
minimum heat losses. Receivers are configured with two coaxial
Abbreviations: STE, solar thermoelectric; R&D, research and development; PT,
parabolic trough; IR, infrared; ASTM, American Society for Testing Materials; SRC,
spectral reflectance curve; STC, spectral transmittance curve; SD, standard devia-
tion; OH, hydroxyl group.
⇑ Corresponding author.
E-mail address: joseluis.navarro@abengoa.com (J.L. Navarro-Hermoso).
conductive and convective thermal losses and thus increases the
efficiency of the energy generation process. The inner metal tube
or absorber tube uses a high absorption and low emissivity coating
(Kennedy, 2002) to reduce losses caused by thermal radiation in
the far infrared. Coatings have an essential role in the performance
of the receiver. The outer tube or glass envelope is covered with an
antireflective coating to increase the amount of energy transmitted
inwards.
In order to characterize the thermal and optical efficiency of
receivers, appropriate systems are required. Some portable devices
to evaluate operating receivers in the solar field exist. For the opti-
cal properties, portable spectrophotometers (Espinosa and
Martinez, 2014; Navarro-Hermoso and Martinez, 2015) allow mea-
suring the transmittance and reflectance (and thus the absorp-
tance) of receivers at some wavelengths along the solar
spectrum. Regarding the thermal performance, Navarro-Hermoso
et al. (2016) presents a non-contact and simultaneous measure-
ment of the inner and outer tubes temperatures estimating the
thermal performance of operating tubes.

http://dx.doi.org/10.1016/j.solener.2016.07.012
0038-092X/Ó 2016 Elsevier Ltd. All rights reserved.
A novel portable device to measure the
temperature of both the inner and the outer
tubes of a parabolic receiver in the field
Cite as: AIP Conference Proceedings 1734, 130015 (2016); https://doi.org/10.1063/1.4949225
Published Online: 31 May 2016

J. L. Navarro Hermoso, Guillermo Espinosa-Rueda, Noelia Martinez, Carlos Heras, and Marta Osta

ARTICLES YOU MAY BE INTERESTED IN

Low-cost small scale parabolic trough collector design for manufacturing and deployment in
Africa
AIP Conference Proceedings 1734, 020016 (2016); https://doi.org/10.1063/1.4949040

Numerical simulation of a parabolic trough solar collector for hot water and steam generation
AIP Conference Proceedings 1734, 070013 (2016); https://doi.org/10.1063/1.4949160

Accelerated aging tests on ENEA-ASE solar coating for receiver tube suitable to operate up to
550 °C
AIP Conference Proceedings 1734, 030003 (2016); https://doi.org/10.1063/1.4949055

AIP Conference Proceedings 1734, 130015 (2016); https://doi.org/10.1063/1.4949225 1734, 130015

© 2016 Author(s).
 A Novel Portable Device to Measure the Temperature of
Both the Inner and the Outer Tubes of a Parabolic Receiver
in the Field
J.L Navarro Hermoso1, a), Guillermo Espinosa-Rueda1 , Noelia Martinez2 , Carlos
Heras3 and Marta Osta3
1
Abengoa Research, Soland, Sanlúcar la Mayor, 41800, Sevilla, Spain
2
Abengoa Research, Paseo de la Castellana, Madrid, Spain
3
Universidad de Zaragoza, Grupo de Ciencias Fotónicas, Zaragoza, Spain
a)
joseluis.navarro@abengoa.com

Abstract. The performance of parabolic trough (PT) receiver tubes (RT) has a direct impact on Solar Thermal Energy
(STE) plant production. As a result, one major need of operation and maintenance (O&M) in STE plants is to monitor the
state of the receiver tube as a key element in the solar field. However the lack of specific devices so far has limited the
proper evaluation of operating receiver tube´s thermal performance. As a consequence non-accurate approximations have
been accepted until now using infrared thermal images of the glass outer tube. In order to fulfill this need, Abengoa has
developed a unique portable device for evaluating the thermal performance and vacuum state of parabolic trough receiver
tubes placed in the field. The novel device described in this paper, simultaneously provides the temperature of both the
inner steel tube and the outer glass tube enabling a check on manufacturers specifications. The on-field evaluation of any
receiver tube at any operating temperature has become possible thanks to this new measuring device.
The features and usability of this new measurement system as a workable portable device in operating solar fields
provide a very useful tool for all companies in the sector contributing to technology progress. The originality of the
device, patent pending P201431969, is not limited to the CSP sector, also having scientific significance in the general
measuring instruments field. This paper presents the work carried out to develop and validate the device, also detailing its
functioning properties and including the excellent results obtained in the laboratory to determine its accuracy and
standard deviation. This information was validated with data collected by O&M teams using this instrument in a
commercial CSP plant. The relevance of the device has been evidenced by evaluating a wide sample of RT and the
results are discussed in this paper. Finally, all the on field collected data is used to demonstrate the high impact that using
this unique portable device will have on a parabolic trough solar power plant.

INTRODUCTION
In a parabolic trough STE plant, the solar field has two main components, mirror troughs and receivers tubes,
which must be operated at the highest possible efficiency [1] to enhance plant performance. To do so, operating
plants need to characterize the properties of the installed components. Simple, accurate, portable devices are
required. Mirror reflectance monitoring is performed daily in the solar field using specialized portable reflectometers
[2]. Also the receiver tubes optical performance, the transmittance of the outer glass tube and absorbance of the
inner metal tube, is monitored using specialized portable spectrophotometers [3] [4].
However, there are not any specific portable measuring instruments to evaluate the thermal properties of
receivers in the field. All the models [5] [6] [7] which allow prediction of power plant performance and carrying out
sensitivity analysis [8] consider the receiver´s thermal performance a key parameter with a direct impact on facility
performance. Tube thermal performance depends on the maintenance of the vacuum between both tubes [5] [6], that
reduces convective and conductive losses and assures the maintenance of the inner tube coatings [9]. Thermal

SolarPACES 2015
AIP Conf. Proc. 1734, 130015-1–130015-8; doi: 10.1063/1.4949225
Published by AIP Publishing. 978-0-7354-1386-3/$30.00

130015-1
models [5][10] allow obtaining a theoretical glass tube temperature depending on the inner tube and the ambient
temperatures for a specific vacuum quality state, that is gas composition and pressure. So to prove that the vacuum
has lost its proper state, it is necessary to simultaneously measure the temperature of the glass envelope, the inner
tube and the ambient temperature, all under normal operating conditions.
Several instruments have been developed to evaluate tube’s thermal performance by calculating the outer glass
tube temperature using common infrared images from a thermal camera [11] [12]. None of the techniques in the
state of the art fulfills the requirements of simultaneously measuring the temperature of two concentric tubes.
Abengoa in collaboration with Universidad de Zaragoza has developed a specialized portable device to
simultaneously and accurately measure the temperature of the inner and outer tubes of parabolic trough receivers.

SCIENTIFIC MEASUREMENT PRINCIPLE

Every object emits energy as electromagnetic radiation. Emissive spectral power determines the amount of
energy emitted per unit surface and time between the wavelengths λ and λ+dλ. It is given by Planck’s Law:
c
E ( , T )    d    ( ) 5 c /1T , 0   ( )  1
 
 e 2 1
(1)

Where T is the object temperature, ε(λ) is the spectral emissivity of the surface and c1 and c2 are radiation
constants
hc
c1  2 hc02 ; c2  0 ,
k (2)

Where c0 = 3•108(m/s) is the speed of light, k = 1.38 10-23(J/K) is the Boltzmann constant and h = 6.63•10-34
(J•s) is the Planck constant.
During operation the receiver’s temperature takes values between 300 and 400ºC for the inner tube and around
60ºC for the outer tube thanks to the vacuum insulation. Emissive spectral power emitted by a black body at these
temperatures is shown in Figure 1(a). According to the Wien Law, the higher the temperature of the black body, the
larger the amount of emitted energy and the lower the wavelength where the maximum is located.

(a) (b)
FIGURE 1. Power emitted by black bodies (a) and tubes optical properties (b)

The optical properties of the receiver tubes, Fig.1 (b), enhance the solar energy absorption by the inner tube and
consequently the energy transmitted to the Heat Transfer Fluid (HTF) running through it. The inner metal tube
coating has low reflectance in the solar spectrum [13] region (0.4-2 µm), which implies high absorbance, meanwhile
in the medium and high infrared it is highly reflective and lowly absorptive, minimizing radiation losses. The outer
glass tube has high transmittance in the visible and near infrared spectrum (0.4-3.5 µm), coinciding with the solar
spectrum. However for wavelengths higher than 4 µm the glass tube transmittance is null, meaning that the glass is
opaque and radiation does not pass through it. The glass is considered as a black body for wavelengths beyond 4
µm. The emissivity of an opaque black body (T (λ) =0) is given by (3).

130015-2
  ( )  1  R ( )  1 (3)

Where R (λ) is the glass reflectance coefficient in the infrared, being nearly zero (0.03-0.04) beyond 4µm.

DEVICE DESCRIPTION
The device is based on the independent and simultaneous measurement of the electromagnetic power emitted by
the inner and the outer tube. An accurate measurement without contact or interferences is possible thanks to the
combination of: the difference between the temperature and the emitted power spectrum of both tubes; the variable
transmittance of the glass tube depending on the wavelengths; and the selection on the proper measurement spectral
range for each tube.
According to Plank’s law, below 3.5 µm the emission of a black body at the temperature of the glass envelope
during operation is practically null compared with the emission of the inner tube at its operating temperature. This
fact is used to detect and measure a defined part of the whole radiation emitted by the inner tube. Calibrating this
signal coming practically exclusively from the inner tube it is possible to determine its temperature with no
interference from the outer tube, which in addition is highly transparent in these wavelengths.
On the other hand, the glass opacity at longer wavelengths is used to measure the radiation exclusively coming
from the outer tube and determine its temperature with no interference from the inner tube.
The device is formed by two units, the measuring unit and the control unit. The measuring unit, contains the
detectors and its function is to obtain a precise electromagnetic power signal from the evaluated tubes. The
measuring unit is contained in a case adapted to the tube geometry, which keeps constant the parameters which
influence the measurement to assure repeatability. The system configuration avoids any influence from adjacent
bodies or other radiation. The measuring unit respects the glass surface and the antireflective layer by using a
minimum contact system with tiny rubber feet.

FIGURE 2. Control unit (a) and device measuring with a pole (b).

The control unit is the interface between the measuring unit and the user. Signals detected with the measuring
unit are sent to the control unit wirelessly. The control unit does the data processing and conversion and is used for
device control. Two control unit formats are offered to the users. The first option is a custom control unit, Fig.2 (a)
that consists of an ergonomic case with backlighted digital screen and alphanumerical keyboard and equipped with
an easy-to-use interface. It has high battery life, internal memory storage exportable to a computer and an ambient
temperature sensor. The second option is to use a mobile or tablet, installing an available custom application. The
mobile and the measuring unit are synchronized and data management and device control are performed from the
mobile, exporting data to a PC or even to another application using internet connection.
Measuring and control units can be integrated in a unique system, Fig.2 (b). A pole with two connections enables
rapid attachment of both modules. The measuring unit is fixed in a rotating joint and the control unit in fixed within
reach of the user. This configuration facilitates an easy and rapid evaluation of multiples tubes, just placing the
device on the desired tube, temperature values are shown in real time in the screen.
The device measurement range is from 150 to 550ºC for the inner tube, and from 20 to 150ºC for the outer tube.
The temperature range could be displaced up or down, adapting detectors sensitivity. Higher temperatures could be
measured, but in this case accuracy would be lost for low temperatures. The selected range has been chosen in order
to accurately cover current operation modes of parabolic trough technology.
The new thermometer is designed to survive daily use in a solar plant extreme environment: high temperature,
dust, possible shocks, etc. It is a compact portable device with robust mechanical and electronic configuration to

130015-3
minimize damages. It is conceived for use as an operation and maintenance instrument but also provides valuable
data for research and development.

Calibration
The device just requires an initial laboratory calibration, not needing repeating calibrations in the daily use. For
this purpose, a laboratory experimental calibration curve signal-temperature is obtained. A receiver tube is placed on
a test bench equipped with a heating system which allows maintaining a homogeneous and constant tube
temperature which is controlled by k-type thermocouples. The detectors readings registered in the experimental
calibration matched with the theoretical values obtained from Planck’s Law.

Measurement process
Switch on the measuring unit and the control unit. Place the measuring unit above the evaluated tube so that the
device case is correctly positioned around the glass envelope. The analog signal from detectors is processed and
digitized. The digital signal is converted to temperature using the calibration curves. Values of the six temperatures,
three from each tube, are shown in real time at the control unit screen and can be captured and saved.

PERFORMANCE EVALUATION

To determine the new thermometer performance a laboratory test was performed as shown in Fig.4. A
commercial receiver tube was placed on a test bench equipped with a heating system which used an electrical source
to maintain a homogeneous and constant set-point temperature in the inner metallic tube. To do so, K-type
thermocouples with an accuracy of ±0.3% value +1ºC, are introduced along the tube and a PID controller converges
the measured temperature and the set-point temperature. Both values are shown in a real time display. Laboratory
ambient temperature and humidity are continuously controlled and also showed in a display.
During the test, the inner tube set-point temperature was fixed and kept constant, repeating the process for three
temperatures: 200, 300 and 400ºC. Two K-type thermocouples were used to monitor the glass temperature,
connecting then to a Chauvin Arnoux C-A 863 thermometer, resulting a system with accuracy ±0.3% value +1ºC.
One thermocouple was placed on one side and the other on the top of the glass tube.
A Flir Systems Thermacam P25 was used to monitor the glass temperature. It measures the far infrared, in
particular 7.5-13 μm and with an accuracy of ±2ºC or ±2%. This camera required the image center spot to coincide
with the evaluated body and its emissivity had to be manually introduced, being fixed at 0.88 for this test.
And lastly, a Flir SC 7000 thermal camera was used to separately monitor both tubes temperature. It evaluated
the medium infrared, in particular 1.5-5.1 μm with an accuracy of ±1ºC or ±1%. This camera required an electrical
outlet and also connection to a computer from which the camera was controlled and the image analyzed. Different
filters in the camera allow narrowing the analyzed spectrum. Just one filter can be used at a time. The 4.6-5.1 μm
filter was used for the outer tube and the 2.36 μm filter for the inner tube. Images were taken looking towards the
tube from the side at 1.5 meters. Emissivity was fixed at 0.88 and 0.5 for the outer and the inner tube evaluation, and
the temperature was taken as the average of the tube cross section.

(a) (b)
FIGURE 3. Laboratory test comparing multiple instruments (a), and new thermometer (b)

130015-4
 In Table 1 the outer tube temperature measurement with each instrument are shown. Similarly the inner tube
temperature measurements are shown in Table 2.

TABLE 1. Glass temperature value (ºC) obtained with the different instruments for the three evaluated inner tube temperature

Instrument 200ºC 300ºC 400ºC

New thermometer top measure 35.3 51.5 69.5
New thermometer side measure 34.9 48.3 66.1
Thermocouple top 36 50.7 70.4
Thermocouple side 35.6 49.1 67.4
Flir SC 7000 (cross section) 38.5 52.5 72.6
Flir P25 portable pointing to the side 40 49.9 62.1

TABLE 2. Inner tube temperature value (ºC) obtained with the different instruments

Instrument 200ºC 300ºC 400ºC

New thermometer top measure 202.8 298.5 398.3
New thermometer side measure 202.5 298.3 398.4
Thermocouples (*) 200ºC 300ºC 400ºC
Flir SC 7000 (cross section) Out of range Out of range 407
(*) Thermocouples are the same used to control the tube heating system controller.

Uncertainty of the measurement

Quality of the measurement is evaluated by calculating its uncertainty [14]. For that purpose one commercial
receiver is evaluated using 100 measurements with the new thermometer. This evaluation was carried out in a
laboratory with ambient stationary conditions and with no fluid flowing through it. A test bench equipped with a
heating system was used to maintain a homogeneous and constant temperature of 400ºC in the inner metallic tube.

FIGURE 4. Uncertainty test

The distributions of the obtained results are shown in Fig.5 for both temperatures measured. The average values
were T=69.5 ºC for the outer tube and T=398.3ºC for the inner tube. The uncertainty type a (absolute value) of the
device, which related to precision or repeatability, has been calculated from these 100 readings resulting in the
following values for both sensors. The uncertainty type b was calculated from the uncertainty of the thermocouples
used to calibrate the thermometer. This value is 0.3 (absolute value) at a level of confidence of 95% (k=2).

, 0.6 % , 0.04 % , 0.15

The standard combined uncertainty is calculated as:

, , , 0.6 , , , 0.15 (4)

130015-5
 According to these results, the expanded uncertainty expressed in absolute values at a level of confidence of 95%
(k=2) is:
1.2 % 0.3 %

Defining the term accuracy as the difference between the real value of a parameter, and the value provided by a
given measuring device, and based on all laboratory test results, it can be affirmed that for any sensor, both for the
outer and inner tubes, and for any temperature within in the specified working range, the accuracy of the new
thermometer is ±1%value +1ºC.

BENEFITS OBTAINED IN A SOLAR PLANT

Receivers are designed to withstand exposure to harsh environmental conditions and the rigors of daily
operation. However, sometimes, receiver vacuum loss may occur. The vacuum loss can be total due to a glass
fracture or breakage of the glass-metal welding, or can be partial with progressive de-gassing through micro-pores.
According to models [5] [10], the tube thermal losses are directly related with the vacuum state. For a given
inner tube and ambient temperature, the glass temperature is an indication of the vacuum state. Using these models a
threshold glass temperature can be defined to determine if the vacuum conditions and consequent thermal losses
meet a performance threshold. An example of a non-conservative equation obtained from these models is.

º
º º 40 (5)

The new thermometer described here allows simultaneous measurement of the glass tube, inner tube and ambient
temperatures of any receiver tube installed in the solar field, which makes it the first specific measuring instrument
to easily and instantly determine a receiver’s thermal performance during operating conditions. The plant average
performance relative to this parameter can be known and used for production planning [7]. Plant operation and
maintenance [15] can be optimized by replacing those tubes which do not meet the performance threshold.
Commercial parabolic trough STE plants are composed of thousands of receivers grouped in hundreds of solar
collector assemblies (SCA). To control and optimize STE plant operation, the temperature of the HTF is monitored
at several points of the plant. This monitoring is done using thermocouples. The more measuring points installed in
the plant, the more detailed and localized the performance information. Commonly thermocouples are placed in the
inlet and outlet point of a SCA.
During operation, a performance decrease can be detected in a specific SCA, which could be caused by different
reasons such as tracking errors, mirror reflectance, tube transmittance decrease or tube vacuum loss. In this case, the
new thermometer would represent a useful tool to analyze tube by tube the thermal performance of a specific SCA.
Also anomalous readings from thermocouples can be caused because they are out of calibration, which could also be
rapidly detected with the thermometer.

MEASUREMENT STUDIES IN THE SOLAR FIELD

The presented thermometer has been tested in STE plants. As an example, results from an evaluation carried out
in a research pilot plant in Sanlúcar la Mayor, Spain are shown.
A total of 45 random tubes from 10 SCAs were evaluated using two instruments: new thermometer and Flir P25
portable camera. Tubes were evaluated during plant operation. While the SCA was being evaluated, it was partially
defocused as concentrated light prevents proper camera operation. For the camera the image center spot was pointed
at the front of the tube from a distance of 2 meters. The thermometer was just positioned around the tube using the
special pole.
During the evaluation, data from the SCA thermocouples were collected to compare them with the inner tube
temperature measured with the thermometer. Note that these sensors are placed at the SCA inlet point so small
differences from the evaluated tube are expected.

130015-6
 FIGURE 5. Results for the solar plant evaluation

It can be seen a high level of correlation between the inner tube temperatures measured by the thermometer and
the SCA thermocouple reading, with differences lower than 4ºC in most cases. It confirms the accuracy determined
in the laboratory.
Regarding the glass temperature, there are sometimes higher differences between thermometer and thermal
camera, and it is more evident at high temperatures, something also noticed in the laboratory test. In spite of the fact
that tubes tested in the same SCA represent the same inner tube temperature, a large variability of the glass
temperature can be seen.
Useful data obtained with the thermometer in this plant evaluation is the detection of tubes with total vacuum
loss. Applying equation (5) to data collected from the new thermometer it can be verified that tubes number 4 and
26, which presents glass temperature of 120 and 122 ºC have lost vacuum.

370 240
30 40 113 º
3

CONCLUSION
A novel portable device to simultaneously measure the temperature of both the inner metal and the outer glass
tubes of parabolic trough receivers has been presented. The designed measuring principles, patent pending
P201431969, based on the combination of the receiver’s optical properties and the emissive spectral power from
each tube at operating temperatures, allow performing non-contact measurement of the temperature of both tubes.
The new thermometer is a unique compact and portable device which no adjustments, external connections, wires or
daily calibration requirements. It integrates the features for which until now two measurement systems were
necessary, assuring for both temperatures an accuracy of ±1% value +1ºC. The working range covers all operating
temperatures for parabolic trough solar plants.
The presented device is the first specific measurement instrument which offers a technical solution for the
evaluation of the thermal performance of parabolic trough receiver tubes installed in a plant. This provides data on
an influential parameter in the solar plant performance model. Therefore, as it has been already proved by use in
solar plants, the new thermometer provides a useful tool for plant operation enabling an accurate characterization of
any point of the plant during operation, measuring its performance or detecting tubes which do not meet a thermal
performance threshold. Also maintenance work to optimize the solar field can be done thanks to the detailed
information obtained.
It has a robust and light weight mechanical and electronic design to make it a suitable instrument to work
outdoors in the field. Measurement is convenient and instantaneous. The measuring unit is wirelessly connected to
the control unit, equipped with battery and memory and with backlighted keyboard and screen where temperatures
are shown in real time. It can be also be controlled using a custom developed mobile phone application.

130015-7
 REFERENCES
1. Andreas Poullikas. Economic analysis of power generation from parabolic trough solar thermal plants for the
Mediterranean region. A case study for the island of Cyprus. Renewable and sustainable Energy reviews. 13
(2009) 2474 -2484.
2. Stephanie Meyen ... [et al.]. Standardization of Solar Mirror Reflectance Measurements--Round Robin Test
Preprint. [Golden, CO]:National Renewable Energy Laboratory, 2010.
3. J.L.Navarro, N.Martinez. Receiver tube performance depending on cleaning methods. (Energy Procedia
Volume 69, May 2015, Pages 1529–1539)
4. G.Espinosa, N.Martinez. A Novel Portable Device to Measure Transmittance and Reflectance of Parabolic
Trough Receiver Tubes in the Field, ASME 2014, ES2014-6500
5. Forristal R. Heat transfer analysis and modeling of parabolic trough solar receiver implemented in engineering
equation solver. NREL Technical Report. 2003, NREL/TP550-34169.
6. Price, H.,. A Parabolic Trough Solar Power Plant Simulation Model. 2003 Report No. NREL//CP-550-33209,
7. Andrea Giostri, Marco Binotti, Marco Astolfi, Paolo Silva, Ennio Macchi, Giampaolo Manzolini. Comparison
of different solar plants based on parabolic trough technology. Solar Energy 86 (2012) 1208–1221
8. Fritz Zaversky, Javier García-Barberena, Marcelino Sanchez,. Probabilistic modeling of a parabolic trough
collector power plant – An uncertainty and sensitivity analysis. Solar Energy. 86 (2012 ) 2128 -2139.
9. Kennedy C.E. Review of mild to high temperature solar selective absorber materials. NREL Technical report.
NREL/TP-520-31267.
10. Frank.P. Incropera ,Fundamentals of Heat and Mass Transfer.
11. M. Pfänder, E. Lüpfert y P. Pistor “Infrared temperature measurements on solar trough absorber tubes” (Solar
Energy, Vol. 81, Is. 5, pp 629-635, 2007)
12. Henry Price, Russell Forristall, Field Survey of Parabolic Trough Receiver Thermal Performance,
10.1115/ISEC2006-99167, ASME 2006 International Solar Energy Conference
13. ASTM G 173-03; Terrestrial Reference Spectra for Photovoltaic Performance Evaluation.
14. BIPM, IEC, IFCC, ILAC, ISO, IUPAC, IUPAP, OIML, The International Vocabulary of Metrology, Basic and
General Concepts and Associated Terms - VIM, JCGM 200:2008 [ISO/IEC Guide 99]
15. Cohen et al, “Final report on the operation and maintenance improvement program for concentrating solar
power plants”, SAND99-1290 Unlimited Release, June 1999

130015-8
 Available online at www.sciencedirect.com

ScienceDirect
Energy Procedia 69 (2015) 1529 – 1539

International Conference on Concentrating Solar Power and Chemical Energy Systems,

SolarPACES 2014

Receiver tube performance depending on cleaning methods

J.L.Navarro Hermoso a, N. Martinez Sanz b
a
Industrial Engineer, R&D Engineer at Abengoa Solar NT. Road A-472, Soland, Sanlúcar la Mayor (Seville, Spain) 41800
b
Head of Materials Department, Abengoa Solar NT. Paseo de la Castellana, 31, 5th floor (Madrid, Spain) 28046.

Abstract

Parabolic trough collectors employ receiver tubes whose optical performance influence in the whole plant efficiency. In the same
way of the mirror’s reflectance, maintenance of tube’s transmittance at high values is a big challenge in the Operation and
Maintenance of solar plants.
Clean borosilicate glass transmittance is increased from 92% to 96% depositing an antireflective (AR) coating. Until now
periodic measurement of transmittance in the solar field has not been possible. The only options for users were to uninstall the
tubes or even break them to analyze their properties in a laboratory.
In the developed experiment, conservative and aggressive cleaning methods have been applied repeatedly on tubes placed in a
solar field, evaluating their effectiveness regarding transmittance maximization, both short term, attending to soiling reduction,
and long term, evaluating the receiver tubes damaging along their lifespan.
The new portable spectrophotometer, Mini Incus, allows the measuring of the tube´s transmittance in the solar field, and due to
this, the soiling factor and cleaning methods influence on the tubes transmittance evolution can be evaluated.
The experiment evaluates the effects caused on the tube by cleaning procedures, concluding the most efficient method.
Performance maximization and maintenance over tubes working life are considered. In addition, novel transmittance monitoring
is relevant to determine the required cleaning cycles to maintain desired transmittance values and optimize the energy production
in commercial plants.

© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2015 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer review by the scientific conference committee of SolarPACES 2014 under responsibility of PSE AG.
Peer review by the scientific conference committee of SolarPACES 2014 under responsibility of PSE AG
Keywords: Parabolic trough; receiver tube; transmittance; cleaning method; portable spectrophotometer; maintenance

1. Introduction

The biggest challenges for Concentrated Solar Power (CSP) technology are cost reduction and performance
increase. Nowadays parabolic trough collector’s plants are the most common. Being at the beginning state of their

1876-6102 © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer review by the scientific conference committee of SolarPACES 2014 under responsibility of PSE AG
doi:10.1016/j.egypro.2015.03.104
1530 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539

estimated lifespan of at least 25 years, it is crucial to be efficient at Operation and Maintenance (O&M) labors [1, 2]
for the optimum use of the available solar resource in existing plants.

Parabolic troughs use concentrator mirrors like all any other CSP technology, to focus the radiation onto the
receiver tubes [3], which represent the most significant difference from others CSP technologies and are one of the
biggest investments in a plant. To catch the maximum amount of intercepted energy, they are composed by two
concentric tubes. The inner tube is made of steel and coated with a high absorbance layer [4]. The outer tube is made
of borosilicate glass with a solar transmittance [5] of 92%, but an antireflective layer [6, 7] deposited using dip-
coating increases the transmittance up to 96%. Between both tubes the vacuum reduces thermal losses and avoids
the absorptive coating degradation. The performance of the receiver tubes [8] which has a direct impact on the
global plant`s performance depends on different factors, such as geometrical losses, decreased by the tube
displacement from the focus, or thermal losses with high dependence on vacuum maintenance. The tubes optical
performance encompasses the absorbance of the inner tube, which should remain constant within the vacuum and
also the transmittance of the outer glass tube, which present higher dependence on the O&M cleaning activities.

Some parabolic trough power plants dedicate abundant resources to clean the components of huge solar fields [7],
using mirror´s reflectance as the controlled reference. Monitoring of mirror reflectance [9, 10] is a common task
developed daily in commercial plants. However no effort has yet been made to monitor tubes transmittance, which
has also impact on the global plant performance [11]. Due to concentrating factor the loss of an effective surface on
the tube is more critical than losing an equal surface on the mirrors. The main cause of the lack of control over the
transmittance has been the difficulty of its measurement and the absence of portable devices to characterize tubes
placed in the solar field. The only option was to uninstall the tube and measure it in a laboratory, or even break it.

Newest portable spectrophotomer Mini Incus [5, 12], allows the measuring of the receiver tube optical properties
in the solar field, obtaining the outer tube transmittance and the inner tube absorbance values rapidly and accurately.
It offers the possibility to monitor the evolution of the tube´s transmittance in a solar plant, control the decrease of
the transmittance or soiling ratio to determine the cleaning cycles, or measure the efficiency of the applied cleaning
method. It is also relevant concerning tubes guarantee, since the receiver tubes manufacturers use an equation [13]
which includes the transmittance, absorbance and emittance. Therefore, the transmittance is the only property
included in the guarantee on which O&M activities have an influence. Some of the tubes manufacturers fix some
limitations regarding cleaning methods applied on the tubes, which are faithfully followed by users.

2. Test method

The aim of this experiment is to study the dependence of the cleaning methods applied on receiver tubes placed
in the solar field and the performance of the tubes, which influences the global plant performance. In doing so,
different cleaning methods have been defined by varying the factors involved in the process. These methods have
been repeatedly applied to a commercial receiver tube placed in the solar field of Solucar (Sevilla, Spain), in such a
way that it was exposed to common surrounding factors in a solar plant such as wind, dust, rainfalls, vehicles transit,
etc. Using the unique 4 meters long tube, 6 different test samples were established. Each sample is 20 cm long and
they are separated 50 cm. Each of the 5 tested cleaning methods was applied to the correspondence test sample, for
instance, method 1 was applied to test sample number 1. Test sample number 6 kept unalterable during the test as a
reference.

The monitored parameter to evaluate and compare the efficacy of each method and the tube performance is the
solar transmittance [11, 12] of the outer glass tube in each of the test samples. The increase of transmittance
achieved by each method is measured in every single application. The efficacy of the methods is tested at different
levels of dirtiness, even in extreme cases. In addition to the short term efficacy attending to the value reached by the
transmittance with one application, more aspects related to the cleaning methods are studied, such as the long term
efficacy or the capacity of each method to maintain the maximum tube performance over their lifespan, not
removing the antireflective layer or damaging the surface.
 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539 1531

Different tubes manufacturers fix limitations regarding cleaning methods applied on their products. Main
common limitations in their technical specifications [13] were taken into account in order to define the configuration
of the five cleaning methods evaluated in the experiment. Limitations are listed in table 1.

Table 1. Manufacturers’ cleaning delimitations . Standard method

Limited factor Value
Distance >40 cm
Water pressure <20 bar
Nozzle aperture >25º
Fluid Demineralized water
Contact No contact or brushes

All receiver tubes users follow these parameters so it represents the standard method. Therefore, during the
experiment, conditions from this standard method were replicated on test sample number 5, to compare it with the
other methods. To configure the remaining 4 methods the following factors have been varied

• Number of water jets

• Fluid
• Nozzle aperture
• Application distance
• Contact/ No contact (and contact element)
• Natural/ contact drying

Fig. 1. Water jets incidence directions

For the five evaluated methods there were common factors; a pressure washer with 100 bar of pumping pressure
and 390 l/h of flow was used with demineralized water as cleaning fluid. Pressure washer pistol has a nozzle which
allows increasing the water jet aperture angle (figure 2) and consequently reducing the water jet pressure. Each of
the water jets was applied for 4 seconds, passing along the 20 cm sample. Water jets incidence direction followed a
common criterion, being 30º and 150º with the normal to the tube surface perpendicular to the ground (figure 1).
The reason is because tubes placed in solar plants are cleaned by trucks which pass along the collector and due to
the structural brackets is more difficult to apply the water jets to the part which is looking towards the mirrors (180
to 360º range in figure 1). Methods 1, 2 and 3 entail direct contact with the tube using different elements. For all the
contact cleanings application, 4 passes over the 20 cm sample were done in the longitudinal axis, 2 in each direction.

Standard method which fulfils manufacturer’s recommendations, method 5 in the experiment, respects the tube
and could be considered conservative. For this reason, in the configuration of the remaining 4 methods, factors have
been varied and combined resulting more aggressive methods, which overpass manufacturer limits. In this way it is
desired to make use of this experiment to study the effect of these aggressive methods, and the influence of the
parameters in the efficiency. Established cleaning methods for the experiment are detailed in table 2. All the
methods encompass cleaning and drying, and additionally, methods 1 and 4 have an intermediate secondary rinse.
Method 1 starts with a big aperture diffuse water jet, figure 2(a), to wet the tube, and after contact cleaning using a
microfiber cloth (80% polyester and 20% polyamide), final rinse is done with very low aperture water jets, figure
2(b). Method 2 uses two natural sponges made from vegetable cellulose and high porosity, one for cleaning and
other one for drying. Method 3 uses contact cleaning by a brush with 1mm diameter polypropylene bristle, and two
30º aperture water jets in the two common directions, figure 2(c) and 2(d). Method 4 starts with a diffuse water jet
which includes ammonia as additive fluid diluted in the demineralized water. Low concentration of ammonia results
a basic solution with a pH of 12. Separation between test samples avoids cleaning interaction. Test sample 6 was
covered during cleaning activities to avoid any water splashing.
1532 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539

Table 2. Cleaning methods application definition

Sequential task Parameter Method 1 Method 2 Method 3 Method 4 Method 5
A.Cleaning Number of water jets 1 2 2 1 2
Water jets direction 30º 30º/150º 30º/150º 30º 30º/150º
Additive fluid - - - Ammonia -
Nozzle aperture >90º 30º 30º >90º 30º
Nozzle-tube distance 40 cm 40 cm 40 cm 40 cm 40 cm
Contact cleaning Fiber cloth Sponge A Brush - -
B.Secondary rinse Number of water jets 2 2
Water jets direction 30º/150º 30º/150º
Additive fluid - -
Nozzle aperture <3º 30º
Nozzle-tube distance 40 cm 20cm
Contact cleaning - -
C.Drying Natural Sponge B Natural Natural Natural

a b c d

Fig. 2.(a) Diffuse >90º aperture; (b) < 3º aperture; (c) 30º aperture in 150º direction; (d) 30º aperture in 30º direction

The experiment was carried out until a significant number of applications of each of the 5 methods were
accumulated in their respective test samples. Finally, as is detailed on table 3, the test took 37 days, along which a
total of 9 applications of each method (considering one application the sequence in table 2), were carried out.

Table 3. Experiment schedule

Developed tasks Natural days
Initial state characterization 0
Application 1 3
Application 2 6
Application 3 9
Application 4 12
Application 5 15
Application 6 18
Application 7 21
Application 8 (extremely dirty state) 24
Characterization after rainfalls 31
Application 9 34
Final state characterization 37

Fig. 3. Experiment configuration

 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539 1533

Experiment emplacement and used elements, described below, can be observed in figure 3. Evaluated tube was
sustained by two metallic supports allowing its thermal expansion. The inner tube was heated by non-concentrated
sunlight. The outer examined glass tube temperature was at every time equal to the ambient temperature, due to the
lack of absorbance and the inner vacuum insulation. Climatic conditions during cleaning applications register
temperatures between 27 and 39ºC, and relative humidity between 38 and 67%. From days 27 to 30 abundant
rainfalls took place. After the tube characterization on day 21, all the test samples except the reference one were
intentionally soiled. Directing a high pressure water jet against the ground, the tube was sand spattered, figure 11(a).
This simulated extreme case allowed evaluating the different methods in extreme soiling conditions but it is
necessary to clarify that such high soiling state does not occur in tubes placed in commercial solar plants.

2.1. Measurement

During the experiment, monitoring activities have been periodically developed over the evaluated samples, in
order to characterize the effects caused by each of the cleaning methods applied. The main measurable property
directly affected by the cleaning method is the outer glass tube transmittance, so this property has been used as the
main ratio to compare the efficiency of each of the methods. Mini Incus portable spectrophotomer has been used to
measure the transmittance. The device presents two main optical sets: one for measuring transmittance of the
borosilicate outer tube and one for measuring the reflectance of the absorber tube as it can be seem in figure 4.
Activities in the present experiment do not affect to the inner tube absorbance, which remains invariable during the
study, therefore the device has been used just for the transmittance measurement.

Fig. 4.Measurement mode

The transmittance measurement is taken by emitting a light beam which goes through the borosilicate and is
evaluated. Along its optical path this beam goes through the borosilicate twice at points with different curvature.
The instrument takes this into account when making calculations. Transmittance encompasses the optical range from
UV up to medium IR (365 – 1950 nm) taking measures in 15 different wavelengths: 365, 405, 470, 525, 588, 655,
780, 870, 940, 1050, 1300, 1550, 1720 y 1950 nm. These immediate values measured by the instrument are
processed in order to obtain the final values of solar transmittance, applying the standard ASTM G-173 [14]. This
standard establishes the solar weight of each wavelength from 300 to 2500 nm with a 5 nm step. Therefore, from the
15 wavelengths measured by the instrument, a continuous function from 300 to 2500 nm is obtained in two steps.
Firstly, a spline interpolation method [15] is used to build up a continuous function from 365 to 1950 nm, the range
encompassed by the device. Secondly, the function is completed from 300 to 365 nm and from 1950 to 2500 nm by
extrapolating the remaining values from the curves provided by the manufacturer.

Besides transmittance measurement, test samples have been evaluated by visual characterization in order to check
how the surface is affected by cleanings. For this purpose images have been taken using both a common photo
camera and a magnifying camera PCE-MM200. Magnified images with 50 and 200 increases were taken four
different days distributed along the experiment to monitor the evolution An initial characterization of each test
sample was carried out, after taking out the tube protective envelope in a warehouse in order to establish the starting
reference state. During daily test development, characterization was carried out before and after the cleaning
activities. Also an extra characterization was performed after rainy days. For the final characterization, all the test
samples were cleaned using the same method, a simple water rinse.
1534 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539

3. Results

The starting point in the experiment is the initial characterization of every test sample just after remove the
covering envelope from the tube. Transmittance curves of the six samples are showed in figure 5 and solar
transmittance values are listed in table 4. The tube average transmittance was 96.3% and all the samples presented a
very similar curve. It is detected that sample 3 starting value was lower than the others.

Table 4. Initial transmittance

Sample Solar transmittance (%)
1 96.1
2 96.3
3 95.6
4 96.4
5 96.7
6 96.4

Fig. 5. Transmittance of evaluated samples at the initial state

The variation of the solar transmittance value along the experiment in each on the test samples is represented in
figure 6. For each of the days and the methods, 2 values of transmittance are showed, corresponding to the state
before and after cleaning, in order to appreciate the efficiency. The experiment can be divided in two stages, the first
one comprised between days 0-21 where standard soiling conditions occur and the second one starting after the
extreme soiling, since which the transmittance is decreased.

Fig.6. Solar transmittance values registered along the test

 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539 1535

3.1. Reference sample

Test simple number 6 did not suffer any kind of interaction during the experiment, just the exposure to the
surrounding factors, as the rainfalls and the final rinse. The soiling ratio could be determined by the pace of
transmittance decreasing. Obviously this variation is not constant, depending of several factors [5], but in this
experiment this ratio is comprised in 0.2-0.3%/day. The state on day 18 is showed in figure 7.

a b

Fig. 7.(a) Visual image of the reference sample on day 18; (b) Magnified image.

3.2. Contact methods

Three of the five evaluated methods entail direct contact with the tube. It could be expected that these methods
achieve higher degree of cleanness, removing dirt particles that water cannot drag. However according to the
obtained results in this test, non-contact methods have achieved higher values of transmittance in every application.
Hypothetic benefits from contact methods have not been obtained. Nevertheless some of the adverse effects have
clearly appeared. Referring to erosion on the glass surface, method 3 which uses a polypropylene brush has resulted
absolutely damaging. Since firsts applications scratches appeared as it showed in figure 8. In addition, water spots
remain on the tube surface. Scratches on the surface of the glass tube may reduce the structural strength of the tube
leading to breakage, but also it has been proved in this experiment how they cause rapid and notable transmittance
reduction. Even in some applications, contact methods caused a transmittance decrease.

a b c

Fig. 8.(a) Visual image of sample 3 on day 18; (b) 50X magnified image with scratches and water spots; (c) 200X magnified image

In addition of surface damaging contact methods may cause the antireflectance layer removal with the
consequent transmittance decrease. Glass transmittance curve differs from an antireflective coated glass curve,
presenting different morphology and peaks along the solar spectrum. Attending at curves on days 24 (figure 12) and
day 37 (figure 13(a)), sample 3 curve presents quantitative differences but also qualitative with peaks at 470 nm.
Surface deterioration of transmittance curve change has not been detected on samples 2 and 3. Other effect related
with contact method occurred in sample 2. This sample was not dried by natural convection, but using a sponge.
This fact, of absence of a final water rinse has become crucial. Water marks caused by the sponge pass, observed in
figure 9, cause a transmittance decrease, being specially appreciable on days 3 and 12.
1536 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539

a b

Fig. 9.(a) Visual image of sample 2 on day 12; (b) 50X magnified image with sponge marks

3.3. Extreme soiling case. Importance of cleaning directionality

It has to be pointed out that sand spattering was not uniform, with different sand spots density on each area of the
tube. Transmittance value measured is therefore not rigorous as in a common state due to depending on the device
location; the light beam emitted could be blocked by a dust spot. As it can be observed in figure 10, before cleanings
all the samples present transmittance curves lower than the reference sample. After the applications, samples 1, 4
and 5 achieved to recover the transmittance up to 94%, meanwhile method 2 resulted non efficient.

a b

Fig. 10.(a) Transmittance with sand spattered on the tube; (b) Transmittance after 2 applications of each method

To eliminate the extreme soiling in this application it was necessary to apply each methods twice, and not
because of the inefficacy of them, but because of the directional factor and the circumferential geometry. After the
first application of each of the methods, applying the water jets on the described directions, 30 and 150º, half of the
tube was cleaned, lest part in figure 11(b), however the opposite part kept with high degree of sand spattering, as it
can be observed in the figure 11(b), so water jets on 210 and 330º were also applied.

a b

Fig. 11.(a) Extreme soiling state generated by sand spattering; (b) Half cleaned tube after just one application
 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539 1537

3.4. Rainfalls effects

During this test, 3 rainy days took place, with abundant rainfalls. After 8 cleaning applications, the six samples
were exposed to the rain, what allowed appreciating some phenomena. Instead recover the initial state in all the
samples, the result was the on attached in figure 12. Rainfalls increased and homogenize the transmittance curve of
all the samples, except sample 3. This is caused for the deterioration and scratches in this sample, previously
detailed. It can be also be detected how the transmittance in the reference sample is identical to the one in the other 4
samples, both qualitative, looking at curves in figure 12, and quantitative, according to values in table 5. Attending
to this result it can be assumed that methods 1, 2, 4 and 5 have not caused damage to the glass surface, neither
antireflectance removal or scratches.

Table 5. Transmittance after rainfalls

Sample Solar transmittance (%)
1 93.3
2 93.7
3 90.4
4 93.7
5 93.7
6 93.5

Fig. 12. Transmittance measured after rainfalls

Rainfalls verified the high importance of the directionality explained in the previous point. Just before rainfalls
occurred the tube still had sand spots and accumulated dirt from extreme soiling on day 21. Tubes zones between the
evaluated test samples had more soiling. It was expected that abundant rainfalls eliminate this soling. During the
tube characterization after rainfalls a similar appearance of figure 11(b) was observed, but in this case tube´s upper
surface was clean and the lower surface which was looking at the ground during the rainfalls kept soiling spots.

3.5. Final state

It can be observed in table 6 how the transmittance of both non-contact samples and the reference sample
concluded the experiment with transmittances higher than 94%. Approximately a 2% decrease from the initial state
was registered in these samples. This phenomenon is caused because the two cleaning applied after the extreme sand
spattering have not reestablished the initial state of a new tube. This is proved with magnified images taken at the
end of the experiment where some spots can be observed on every sample, figure 13 (b). Transmittance decrease
goes up to almost 3% in the case of brushed sample 3, due to the surface deterioration which has even changed the
morphology of the transmittance curve as can be seen in figure 13. Sample 1 and 2 represent an intermediate case,
having decreased their transmittance approximately 2.5%.
1538 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539

Table 6. Final transmittance a b

Sample Solar transmittance (%)
1 93.8
2 93.6
3 92.7
4 94.3
5 94.6
6 94.1

Fig. 13.(a) Transmittance on day 37; (b) Magnified image of sample 6 at the final state

4. Conclusions

It has been developed an innovative study, monitoring the receiver tube performance depending of the applied
cleaning methods. Five different cleaning methods have been tested varying and overpassing parameters limited by
manufacturers. It can be concluded that the most effective method all along this experiment has been precisely the
one which reproduces cleaning conditions indicated by receiver tubes manufacturers. This method is conservative in
comparison with the rest tested methods; however no improvements have been achieved with included changes.

It could be expected that contact methods achieve higher efficacy at the firsts single applications, however the
efficacy of the non-contact methods have achieved higher values of transmittance in every application. Contact
methods have even decreased the transmittance in some applications, due to the non-desired caused effects.
Brushing has showed to be totally counterproductive. Polypropylene brush has resulted absolutely damaging since
first application causing scratches and water spots on the tube surface. Scratches on the surface could lead to
breakage, and also it has been proved in this experiment that they cause rapid and notable transmittance reduction.
Also sponge contact has resulted counterproductive due to pass marks on the surface generate transmittance
decrease. The most effective contact method has been the fiber cloth which cause less transmittance decreasing, and
almost no scratching. Water rinse after contact methods has showed to be a key point to eliminate spots and marks.

The experiment has been developed in a solar field with changing conditions as in a real solar plant. Reference
sample has allowed measuring the soiling ratio or loss of efficiency in the tube, obtaining an average transmittance
decrease of 0.2-0.3%/day. In spite of this soiling ratio, the standard method returned the transmittance almost to the
maximum value in most of the applications before the extreme soiling test. Attending to this result, the standard
method is the most appropriate due to the transmittance maximization and also due to it respects the tube state
favoring a long lifespan. It is also the best attending to the easiness of application and the economical point of view.

No differences have been detected depending on the distance between the nozzle and the tube, neither to the
nozzle aperture or the use of ammonia as additive. The aggressive water jet tested with low aperture angle and high
pressure has not caused damages on the surface but neither has caused differences in the transmittance values
comparing it with 30º aperture angle. The most important effect detected is the directionality, or the influence of the
water incidence direction on the tube. Due to the geometry of the tube, independently of the nozzle aperture angle,
the cleaning is less efficient in the opposite part of the tube. This effect has been detected all along the experiment,
underscored clearly at extreme soiled state and even suffered after abundant rainfalls. The most critical zone of the
receiver tube is the effective part which is looking towards the mirrors and is receiving the concentrated radiation.
Therefore it can be concluded that cleaning trucks used in solar plants should be equipped with water jets looking
directly at the effective part of the tube. And this effect is also relevant in case of rainfalls. So in case that parabolic
trough collector is positioned at zenith position when rainfalls occur, it benefits mirrors cleaning and reflectance
increased, but from the tube point of view the effective part is downward and not against the rain direction.
 J.L. Navarro Hermoso and N. Martinez Sanz / Energy Procedia 69 (2015) 1529 – 1539 1539

Effects detected in this experiment reveal the importance of monitoring the receiver tubes transmittance in the solar
fields, as the same way that mirrors reflectance is controlled. Proper transmittance monitoring has been showed to
be useful to evaluate cleaning cycle’s efficacy and requirements, controlling the state of one of the main components
of parabolic trough collectors and maximizing its exploitation.

Acknowledgements

Rafael Perez Santana and all the Operation & Maintenance staff from Eureka pilot power plant. Guillermo
Espinosa Rueda from Abengoa Solar and Carlos Hermoso Fernandez.

References

[1] Cohen et al, “Final report on the operation and maintenance improvement program for concentrating solar power plants”, SAND99-1290
Unlimited Release, June 1999
[2] Price, H.,. A Parabolic Trough Solar Power Plant Simulation Model. 2003 Report No. NREL//CP-550-33209, NREL, Colorado, USA.
[3] T.Kuckelkrom, Advances in Receiver Technology for Parabolic Trough Collectors- A step Forward Towards Higher Efficiency and Increased
Lifetime 2009 SolarPaces
[4] A.L.Avila Marin, Theoretical Analysis on the Influence of Selective and Antireflective Coating on PT Overall Efficiency SolarPaces 2009
[5] G. Espinosa Rueda, N. Martínez Sanz, D. Izquierdo Núñez, M. Osta Lombardo. A Novel Portable Device to Measure Transmittance and
Reflectance of Parabolic Trough Receiver Tubes in the Field, (2013) SolarPaces
[6] Gombert A., Glaubitt W., Rose K., Dreibholz J., Heinzel et al.Antireflective transparent Covers for solar devices. Solar Energy. 68. (2000) 4,
357 -360.
[7] Forristal R. Heat transfer analysis and modeling of parabolic trough solar receiver implemented in engineering equation solver. NREL
Technical Report. 2003, NREL/TP550-34169.NREL, Colorado USA.
[8] J. Jímenez Huertas, D. Gamez Vela, E.García Ramirez Cleaning vehicle and method for parabolic trough solar collectors, Patent US
20120152281 A1
[9] S.Meyen, A.Fernández-García, C. Kennedy , E. Lüpfert. Standardization of solar mirror reflectance measurements- round robin test.(2010)
SolarPaces.
[10] Crawford et al, “A Comparison of Three Portable Reflectometers for Use in Operations and Maintenance of CSP Plants”, SolarPACES
conference proceedings, Marrakech, Morocco, 2012
[11] Fritz Zaversky, Javier Garcia-Barberena, Marcelino Sanchez, David Astrain. Probabilistic modeling of a parabolic trough collector power
plant – An uncertainty and sensitivity analysis. Solar Energy. 86 (2012 ) 2128 -2139.
[12] G. Espinosa Rueda, A Novel Portable Device to Measure Transmittance and Reflectance of Parabolic Trough Receiver Tubes in the Field.
2014 Boston. ES-FuelCell2014-6500
[13] Siemens UVAC receiver tube user manual and Schott PTR receiver tube user manual.
[14] ASTM G 173-03; Terrestrial Reference Spectra for Photovoltaic Performance Evaluation
[15] A. Kouibia, M. Pasadas. Approximation by interpolating variational splines. Journal of Computational and Applied Mathematics, 218(2)
(2008), pp. 342-349