Solar Energy, Vol.

153, 1 September 2017, Pages 679–690

DOI:10.1016/j.solener.2017.06.010

“Theoretical and experimental analysis of an innovative dual-axis tracking

linear Fresnel lenses concentrated solar thermal collector”
Simoni Perini: Lark Energy, Larkfleet House, Falcon Way, Bourne, Lincolnshire, PE10 0FF
Xavier Tonnellier: Cranfield University, Bedford, Bedfordshire MK43 0AL
Peter King: Cranfield University, Bedford, Bedfordshire MK43 0AL
Christopher Sansom: Cranfield University, Bedford, Bedfordshire MK43 0AL
Corresponding author: Christopher Sansom: [email protected]

Abstract
Linear concentrating solar thermal systems offer a promising method for harvesting solar energy. In
this paper, a model for a novel linear Fresnel lens collector with dual-axis tracking capability is
presented. The main objective is to determine the performance curve of this technology by means of
both experiment and theoretical analysis. A mathematical model including the optical model of the
concentrator and the heat transfer model of the receiver pipe was developed. This tool was validated
with experimental data collected using a proof of concept prototype installed in Bourne, UK. The
performance curve of the collector was derived for temperatures between 40°C and 90°C. The results
show that the global efficiency of the collector is limited to less than 20%. The energy losses have
been analysed. The optical losses in the lens system accounts for 47% of the total energy dissipated.
These are due to absorption, reflection and diffraction in the Fresnel lenses. Furthermore
manufacturing error in the lens fabrication has to be considered. One third of the solar radiation
collected is lost due to the low solar absorptance of the receiver pipe. Thermal radiation and
convection accounts for 6% of the total as relatively low temperatures (up to 90 °C) are involved. In
order to increase the performance of the system, it is recommended to install an evacuated receiver
and to insulate the recirculation system. Considering data from manufacturers, these improvements
could increase the global efficiency up to 55%. Utilising the results from this work, there is the
intention of building an improved version of this prototype and to conduct further tests.

Keywords: Solar Energy, Fresnel lens, Two-axis tracking system

Highlights:
 The optical and thermal models of a novel linear Fresnel lens collector with 2D tracking are
presented and validated using experimental data.
 The performance curve shows that the global efficiency of the collector is limited to 20%.
 Optical losses in the lens system accounts for 47% of the total energy dissipated. 33% of the
solar radiation is lost due to low solar absorptance of the receiver pipe.
 Installing an evacuated receiver and insulating the recirculation system could increase the
global efficiency up to 55%.

Table of Contents
1 Introduction:.................................................................................................................................... 3
1.1 Solar Thermal energy potential and linear concentrated collector systems ............................ 3
1.2 Linear Fresnel lenses............................................................................................................... 4
1.3 Linear Fresnel lenses for thermal applications........................................................................ 5

1
Published by Elsevier. This is the Author Accepted Manuscript issued with:
Creative Commons Attribution Non-Commercial No Derivatives License (CC:BY:NC:ND 4.0).
The final published version (version of record) is available online at DOI:10.1016/j.solener.2017.06.010.
Please refer to any applicable publisher terms of use.
 1.4 Objectives: .............................................................................................................................. 5
2 System Description ......................................................................................................................... 6
2.1 Dual axis linear Fresnel concentrator experimental prototype................................................ 6
2.1.1 Dual-axis Tracking system.............................................................................................. 7
2.1.2 Measuring system ........................................................................................................... 8
3 Mathematical models ...................................................................................................................... 9
3.1 Optical model .......................................................................................................................... 9
3.2 Heat transfer model ............................................................................................................... 10
4 Models Validation and Performance curve ................................................................................... 13
4.1 Validation of the Optical model ............................................................................................ 13
4.2 Validation of the Thermal model .......................................................................................... 14
4.3 Performance Curve and losses analysis ................................................................................ 17
5 Conclusions and future works ....................................................................................................... 19

List of Figures
Figure 1: Principle of Fresnel lens .......................................................................................................... 5
Figure 2: Experimental prototype scheme installed in Bourne, UK. ...................................................... 6
Figure 3: Description of the three section of the optical system. The yellow line represent the axis of
symmetry. ............................................................................................................................................... 7
Figure 4: Schematic diagram and measuring system of the dual-axis linear Fresnel lens collector ....... 8
Figure 5: Prims geometry and refraction principle ................................................................................. 9
Figure 6: Thermal resistance network for the absorber cross section ................................................... 10
Figure 7: Wind speeds during testing ................................................................................................... 11
Figure 8: Prototype set-up for experiment (21st April 2016) ............................................................... 13
Figure 9: a) Concentrated beam during experiment (21st April 2016) b) Flux distribution according to
ray tracing simulation ........................................................................................................................... 14
Figure 10: Measured and calculated temperature profiles for the 4th of June 2015 .............................. 15
Figure 11: Measured and calculated temperature profiles for the 11th of June 2015 ............................ 15
Figure 12: Comparison of experimental and simulation results for the 11th of June 2015 .................. 16
Figure 13: Irradiance approximation..................................................................................................... 16
Figure 14: Performance Curve .............................................................................................................. 17
Figure 15: Thermal losses breakdown for the experiment on the 4th June 2015.................................. 18
Figure 16: Thermal losses analysis for the experiment on the 4th June 2015 ...................................... 18

List of Tables
Table 1: List of heat transfer coefficient used in the model .................................................................. 12

Nomenclature
A Aperture area [m2]
Cp Heat capacity at constant pressure [kJ/kg K]
D Diameter [m]
f Focal distance [m]

2
G Irradiance [W/m2]
H Prisms height [m]
h Convective heat transfer coefficient [W/ m2 K]
i Incident angle on the back surface [°]
k Thermal conductivity [W/ m K]
L Pipe length [m]
𝑚̇ Mass flow rate [kg/ s]
n Refractive index
Nu Nusselt number
Pr Prandtl number
𝑞′̇ Linear heat transfer [W/ m]
Re Reynolds number
R2 Coefficient of determination
T Temperature [° C]
v Fluid velocity [m/s]
x Thermal efficiency parameter
Greek Letters
α Material absorptivity
β Slope angle [°]
β’ Refraction angle [°]
ε Material emissivity
η Thermal efficiency
π Pi
σ Stefan-Boltzmann constant [W / m2 K4]
Subscripts
abs Absorber
amb Ambient
ave Average
cond Conductive
conv Convective
DNI Direct Normal Irradiance
in Inlet section
lens Lenses
loss Losses
out Outlet section
rad Radiative
trans Transmitted

1 Introduction:

1.1 Solar Thermal energy potential and linear concentrated collector systems
Heat represents 76% of the global industrial energy consumption and 57% of this demand is required
at temperatures below 400°C (Rawlins and Ashcroft, 2013). Solar thermal energy can be an extremely
convenient and sustainable source of heat, but only concentrated system can reach temperature levels
up to 250 °C, as required in most industrial processes. In several industry sectors, such as food,
textiles and chemicals the share of heat demand required at temperatures below 250°C is around 60%.

3
The United Nations Industrial Development Organization (UNIDO) estimated the potential of
concentrated solar technologies in industry to be 500 million GJ in 2050 (Taibi et al., 2010).
Flat plate solar collectors are designed to deliver lower temperature levels (up to 150°C) (IRENA,
2015). Concentrated solar collectors can produce temperatures up to 3000°C in central receiver
concentrated solar power (CSP) plants. Concentrated systems (linear or point focus) have higher
efficiency at high temperatures because the ratio of thermal energy loss to total incident solar
radiation on the receiver decreases significantly (Zhai et al., 2010). Current medium temperature
collectors that can meet the industrial process heating requirement are based on concentrated
technologies developed for power generation purposes. Parabolic trough collectors are based on thick
(3-4mm) glass second surface mirrors which have high cost and are difficult to integrate with existing
industrial processes (Kumar et al., 2015). Therefore, there is a need for the development of innovative
concentrated solar thermal collectors for applications in industrial processes.
Linear concentrating solar thermal systems offer a promising method for harvesting solar energy
(Zhai et al., 2010). There are three main linear concentrator technologies: parabolic trough solar
concentrators, linear reflected Fresnel concentrators and linear refracted Fresnel lens arrangements.
Most published studies on linear concentrating collectors are related to parabolic trough systems with
evacuated tubular absorbers, illustrating the emphasis on power generating applications (Fernández-
García et al., 2010; Xie et al., 2011). PMMA Fresnel lenses have the disadvantage of transmittance
loss through the lens, but also have advantages in comparison with small-scale reflecting
concentrators, which tend to mitigate this radiation disadvantage. These include (Kumar et al., 2015):

• Ultra high temperatures can be generated with Fresnel CSP technology, which is suitable for
various thermal energy applications.
• A compact and maintenance free concentrator.
• Fresnel-CSP based Direct Steam Generation (DSG) systems do not need a heat transfer fluid
or heat exchanger.
• A high thermal efficiency, compact plant size, low cost, and a shorter payback period.
• 30% of industrial energy consumption required as process heat is in the low to medium
temperature range i.e. 80°C–250°C, and PMMA Fresnel lens technology is ideally suited to this
range.

1.2 Linear Fresnel lenses

The flat linear Fresnel lens is considered to be a suitable concentrator for both photo-thermal and
photovoltaic conversion of solar radiation (Al-Jumaily and Al-Kaysi., 1998; Boyd et al., 1976; Franc
et al., 1986). A Fresnel lens is an optical component which can be used as a lightweight alternative to
conventional continuous surface optics. Fresnel lenses also possess the advantage of higher flexibility
in optical design when compared to conventional mirror concentrators (Leutz and Suzuki, 2001). The
refractive capacity of standard lenses depends mainly on the curvature of the surface. Removing as
much of the optical material as possible while still maintaining the curvature will create the same
effect. This is obtained by means of a series of parallel grooves with different angles. The solar
radiation passing through each groove is refracted at a slightly different angle and converges on the
focal line of the lens (Figure 1) (Davis and Kuhnlenz, 2007).

4
 Figure 1: Principle of Fresnel lens

1.3 Linear Fresnel lenses for thermal applications

Research into Fresnel lenses is mainly focused on point focus concentrating photovoltaic systems and
there are limited studies available about Fresnel linear concentrating solar collectors for hybrid
photovoltaic and thermal systems.
A first theoretical performance analysis of a linear Fresnel concentrating collector with cylindrical
blackbody absorber was carried out in 1976. The study demonstrated that the concept output
temperatures were between 130°C and 370°C (Boyd et al., 1976). Franc et al. (1986) studied and
tested a stationary linear Fresnel lens system with movable absorber. The experiments showed a
yearly gain of 360 kWh/m2 under climatic condition in Central Europe. AL-Jumaily and Al-Kaysi
(1998) studied a solar collector which consisted of two flat linear Fresnel lenses made of glass. The
system was designed to track the sun in two directions. The peak measured efficiency was 58% in
Iraq. The two lens system reached an outlet temperature of 37°C.
Zhai et al. (2010) conducted a theoretical and experimental analysis on a flat linear Fresnel lens being
integrated with an evacuated tube receiver. It was reported that the performance of the Fresnel lens
collector was higher compared to a simple non-concentrating collector and that the system efficiency
can reach as high as 55% at working temperatures of 90 °C. Zheng et al. (2014) designed a cylindrical
compound Fresnel lens system and they produced a small scale prototype to measure the optical
efficiency. An optical efficiency of up to 84% was measured. The collector was able to deliver heat at
the average outlet temperature of 27°C. In 2011, the European project Efisol demonstrated the
integration of a Fresnel concentrator with an Organic Rankine Cycle engine and thermal storage
[Linaric, 2013]. This fully functional prototype was developed and achieved temperatures up to 150
°C.
More recently, researchers have focused on the analysis of the receiver for linear Fresnel lens systems.
Lin et al. (2014) investigated the performance of a linear Fresnel lens solar collector using different
cavity receivers. They found that the receiver design with a triangular cavity reached the highest
efficiency. More recently, Manikumar et al. (2014) performed a numerical simulation on a trapezoidal
cavity receiver; aiming for a better design with maximum heat transfer rate as well as optimal outlet
temperature. The overall heat loss coefficient simulated was in the range of 8–11 W m-2 K-1 at
operating temperatures of 130–180 °C.

1.4 Objectives:
From the current literature review, limited published works were found related to the performance of
linear Fresnel lenses made of plastic materials with a dual-axis tracking system for thermal
applications. In this study, a model for a dual-axis tracking flat acrylic linear Fresnel lens concentrator

5
is developed and validated. The main objective of this work is to determine the performance curve of
this innovative collector by means of both experiment and theoretical analysis.

2 System Description

2.1 Dual axis linear Fresnel concentrator experimental prototype

The experimental prototype used in this study consists of an array of acrylic (PMMA) linear Fresnel
lenses which focuses the solar radiation onto a receiver pipe (Figure 2).

Figure 2: Experimental prototype scheme installed in Bourne, UK.

The prototype consist of three modules of 36 Fresnel lenses having a total aperture area of 75 m2. The
aluminium structure hosting the lenses is maintained normal to the sun rays throughout the day using
a 2D tracking system. Hydraulic pistons are responsible for the vertical movement of the frame. An
electric slew motor provides the torque to rotate the system around the central pivoting point. An
automatic dual axis tracking system controls the axis motions and ensures alignment with the sun. The
working fluid is circulated in a copper piping system using two circulation pumps. Water is used in
the current version of the prototype. A buffer tank is installed close to the outlet section of the
receiver. The tank is not insulated. Excess heat is dissipated through a fanned air heat exchanger to the
surroundings.
The system operates in recirculation mode: the water is heated in the absorber exposed to the
concentrated solar radiation and it is recirculated back to the inlet section. When the temperature at
the outlet section of the absorber approaches 100 °C, the water can be diverted to the heat exchanger
for cooling. The current setting of the prototype does not allow the generation of steam for safety
reasons.
The receiver is a 13 meters carbon steel pipe (schedule 40). Each Fresnel lens consists of a panel of 1
m in length and 0.7 m in width with grooves facing inwards. The focal length of each lens (Lens 1,
Lens 2, and Lens 3) is 2.6 m, and the prism size is 0.5 millimetres. The lenses were fabricated using
acrylic resin (polymethylmethacrylate, PMMA) through moulding. They have been manufactured in
three symmetrical sections. Lenses 1 and 2 are aligned on the same plane, while lens 3 is tilted by 10
degrees to ensure that the concentrated solar radiation is focused on the receiver tube. The current
work considers a sub-section of the total experimental rig that encompasses only lenses 1 and 2
(Figure 3). Lens 3 has not been included in the modelling or experimental work because optical
analysis and tests have proven that the Lens 3 section has been manufactured incorrectly. Tests show
that Lens 3 section is not focusing energy on the target (the receiver).

6
 Figure 3: Description of the three section of the optical system. The yellow line represent the axis of symmetry.

2.1.1 Dual-axis Tracking system

Solar tracking methods can be divided into two broad categories, namely passive tracking and active
tracking. The active electronic systems generally exhibit improved reliability and tracking accuracy.
These can be further subdivided into open-loop, closed-loop and mixed control system. In this study,
a closed-loop control method is used. The control unit (Horner XLT with special Firmware) compares
the calculated position of the sun with the actual position of the prototype. The sun position is
obtained using known sun algorithms based on site location and solar time (Reda et al., 2008). The
geographical position of the rig is acquired from a GPS system installed on site. The actual position of
the prototype is determined by the elevation angle and horizontal angle (angular distance from
geographical North). The elevation angle is measured using an inclinometer with a 0.1 degree
resolution attached to the tilting frame supporting the lenses. The horizontal angle (azimuth) is
measured using an encoder installed on the pivoting mast of the prototype. When the difference
between the tilting angle and the sun elevation is larger than 0.5 degree, the control unit moves the
prototype to re-focus the lenses. The accuracy for the azimuth angle is 1 degree. The motor is
controlled by an inverter drive while the control unit is connected to forward/ reverse digital outputs
of the inverter. The pistons are controlled by digital outputs which operate the pump solenoids via
relays, whilst a second set of relays simultaneously start the hydraulic pump. The control unit

7
compares the measured position and the sun angles every 20 milliseconds. The control unit is supplied
by a single phase 230 Volt input. The hydraulic system is supplied by 12 Volt direct current. The
electric motor needs a three phase 230 Volt input.

2.1.2 Measuring system

The measuring system installed on the experimental prototype consists of the following equipment
(Figure 4):

Figure 4: Schematic diagram and measuring system of the dual-axis linear Fresnel lens collector

1. A vortex flow meter. This instrument is used to measure the volumetric flow rate in litres per
hours. The accuracy of the device is 0.75% for a Reynolds number between 4000 and 20000
(turbulent flow). It is supplied with a thermo-resistor device (Pt1000) installed in the vortex
sensor. The outlet temperature is measured using this sensor which has a maximum error of
1°C.
2. A resistance thermometer sensor with an accuracy of ±0.3°C is used to measure the fluid
temperature at the inlet section of the receiver.
3. A pressure transmitter with an accuracy of ±3.0% is used to measure the inlet pressure of the
fluid.
4. A digital secondary standard pyranometer is installed to measure the solar radiation incident
on the lenses surface.
The data is recorded every minute and transmitted to an online platform. The pyranometer installed on
the lens frame has been converted into a pyrheliometer to measure the Direct Normal Irradiance
(DNI). A collimator tube has been designed and installed in order to eliminate the diffuse radiation
and to reduce the circumsolar sky radiation. The collimator has been designed following the
procedure presented by Mujahid and Alamoud (1998) who designed a collimator tube for a

8
photovoltaic pyrheliometer. They reported that the error in measurements of direct normal radiation
can be reduced by 2.3% using this system. The error band for the pyrheliometer is ±20 W/m2. The
reliability of collimation tubes with respect to the “sun and shade” method (indirect measurement of
DNI shading the pyranometer) has been studied by Ijima (2010). The measurements of the converted
pyrheliometer have been compared and validated against the data available on the SolarGis platform
for the selected location (Solargis, 2010).

3 Mathematical models

In order to analyse the performance of the system, a mathematical model was developed and
validated. The Fresnel lens system (Lens 1 + Lens 2) was characterized using an optical model
implemented in Matlab. The performance of the lenses was evaluated using ray tracing analyses
carried out with SolTrace. The heat transfer in the receiver was modelled and simulated using Matlab.

3.1 Optical model

The geometry of the Fresnel lenses was analysed using a code implemented in Matlab. This program
calculates the inclination of each surface of the prisms with respect to a global reference system. In
this case, the origin is considered to be in the middle section of the prototype (see Figure 3).
Assuming that the incident angle (i) on the back surface is zero (as in the case of Lens 1 and Lens 2
with negligible tracking error), in each prisms the slope angle β is equal to the incident angle on the
front surface (See Figure 5).

Figure 5: Prims geometry and refraction principle

Furthermore, it can be demonstrated that the refraction angle β’ is:

𝐻 (1)
𝛽′ = tan−1 ( ) + 𝛽
𝑓
Where H is the height of the prism with respect to the reference system and f is the focal length.
According to the Snell’s law (Leutz and Suzuki, 2001):
𝑠𝑖𝑛𝛽 𝑛′ (2)
=
𝑠𝑖𝑛𝛽′ 𝑛

9
The parameters 𝑛 and 𝑛′ represent the refractive indices and they can be defined for each wavelength
in different materials. In this study, 𝑛′ and 𝑛 are equal to 1 (refractive index of air) and 1.49 for
PMMA respectively considering the entire spectrum of light (Leutz and Suzuki, 2001). Therefore, it is
possible to calculate the slope angle using the following non-linear equation:
𝐻
𝑛 ∙ 𝑠𝑖𝑛𝛽 − 𝑠𝑖𝑛 (tan−1 ( ) + 𝛽) = 0 (3)
𝑓
Finally, the code defines the geometry of the lenses: the position of the prisms is calculated to be used
as an input in SolTrace. The ray tracing software also requires the definition of the optical properties
of the material. In this study, the PMMA average transmittance for the visible part of the solar
spectrum for the lenses is assumed to be 0.92 (Leutz and Suzuki, 2001). The slope error of the prisms
have been considered 5 mrad and the Sun Shape has been approximated with a pillbox distribution
and the half angle is 5 mrad (Lovegrove and Stein, 2012).

3.2 Heat transfer model

The heat transfer model is based on the one dimensional steady state energy balance within the
absorber (i.e. the steel pipe) developed by Forristall (2003). Figure 6 shows the thermal resistance
network for a cross-section of the absorber.

Figure 6: Thermal resistance network for the absorber cross section

Note that the optical losses have been omitted from the resistance model: Q trans represents the
energy transmitted by the lenses. The optical losses in the concentrator are due to absorption and
reflection in the plastic material of the lenses. Part of the solar radiation is absorbed by the receiver
and transferred to the heat transfer fluid by convection. The rest of the energy is lost to the
environment through convection and radiation.
In order to simplify the analysis, the following assumptions are made:
1. Tracking error is negligible.
2. The temperature distribution is uniform around the circumference of the absorber.
3. The variation of the surface temperature along the axial direction of the collector is neglected.
4. The thermal properties of component materials do not vary and those of air and fluid are
curve fitted from tables of thermal properties (Holmgren, 2006)
5. The solar absorption is treated as heat flux term. In reality, the solar absorption in the
absorber is a volumetric phenomenon.
6. The wind speed was measured close to the collector in Bourne on the corresponding dates.
The wind speed data is shown in Figure 7 below.

10
 7. The pressure of the fluid is constant in the system and estimated as the average of the value
measured during the experiments.

2.5
wind speed [m/s]

1.5

0.5

0
30

220

270
0
10
20

40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210

230
240
250
260

280
290
300
Time [minutes]
4th June 11th June
Figure 7: Wind speeds during testing

The energy balance in the receiver pipe is determined by applying the energy conservation principle at
each surface of the system (Forristall, 2003). In the equations that follow the numerical subscripts 1-5
refer to locations 1-5 in Figure 6.
2𝜋𝑘23 (𝑇2 −𝑇3 ) (4)
ℎ1 𝜋𝐷2 (𝑇2 − 𝑇1 ) = 𝐷
ln( 3⁄𝐷 )
2

̇ 2𝜋𝑘23 (𝑇2 −𝑇3 ) (5)

𝑘1 𝑞′ 𝑡𝑟𝑎𝑛𝑠 𝜂𝑎𝑏𝑠 𝛼𝑎𝑏𝑠 = 𝑘2 ℎ2 𝜋𝐷3 (𝑇3 − 𝑇5 ) + 𝜎𝜀3 𝜋𝐷3 (𝑇3 4 − 𝑇4 4 ) + 𝐷
ln( 3⁄𝐷 )
2

𝑞′𝑙𝑜𝑠𝑠̇ = 𝑘2 ℎ2 𝜋𝐷3 (𝑇3 − 𝑇5 ) + 𝜎𝜀3 𝜋𝐷3 (𝑇3 4 − 𝑇4 4 ) (6)

1 ̇
𝑚̇ [𝑐𝑝.𝑎𝑣𝑒 (𝑇𝑜𝑢𝑡 − 𝑇𝑖𝑛 ) + 2
(𝑣𝑜𝑢𝑡 2
− 𝑣𝑖𝑛 )] = (𝑞 ′ 𝑡𝑟𝑎𝑛𝑠 ̇ )𝐿
𝜂𝑎𝑏𝑠 𝛼𝑎𝑏𝑠 − 𝑞 ′ 𝑙𝑜𝑠𝑠 (7)
2

Where :
 𝑇1 is the fluid bulk temperature calculated as the average between the inlet and outlet
temperatures
 𝑇2 and 𝑇3 are the internal and external surface temperatures of the receiver
 𝑇4 is the equivalent sky temperature approximated as 𝑇4 = 𝑇5 − 8 (Forristall, 2003)
 𝑇5 is the ambient temperature.
 𝐷2 is the inner diameter of the steel absorber tube
 𝐷3 is the outer diameter of the steel absorber tube

11
The heat transfer coefficients are defined in Table 1. The emissivity 𝜀3 of the receiver tube is assumed
to be 0.2 (Martinez, 1995). In this model, the absorptance 𝛼𝑎𝑏𝑠 for carbon steel is assumed equal to
0.38. 𝜂𝑎𝑏𝑠 represents the interception factor determined with the ray tracing analysis.

Table 1: List of heat transfer coefficient used in the model

Heat Transfer coefficient Nusselt Number Range

𝑘 𝑓2 /8 ∙ (𝑅𝑒𝐷2 − 1000) ∙ 𝑃𝑟1 𝑃𝑟1 0.11 2300 < 𝑅𝑒𝐷2 < 4000
ℎ1 = 𝑁𝑢𝐷2 ∙ 𝐷1 𝑁𝑢𝐷2 = 2/3
∙( )
2 1 + 12.7 ∙ √𝑓2 /8 ∙ (𝑃𝑟 − 1) 𝑃𝑟2
1 0.5 < 𝑃𝑟1 < 2000
𝑓2 = (1.82 ∙ log10 (𝑅𝑒𝐷2 ) − 1.64)−2

𝑁𝑢𝐷2 = 4.36 𝑅𝑒𝐷2 < 2300

𝑇2 +𝑇3 Carbon steel heat conductivity
𝑘23 = −0.0309 ∙ 𝑇23 + 60.8989, where 𝑇23 =
2
𝑘35
ℎ2 = 𝑁𝑢𝐷3 ∙ 𝑃𝑟5 0.25 𝑅𝑒𝐷3 𝐶 𝑚
𝐷3 𝑁𝑢𝐷3 = 𝐶 ∙ 𝑅𝑒𝐷3 𝑚 ∙ 𝑃𝑟5 𝑛 ∙ ( )
𝑃𝑟3 1-40 0.75 0.4
0.7 < 𝑃𝑟5 < 500, and 1 < 𝑅𝑒𝐷3 < 106 40-1000 0.51 0.5
1000-200000 0.26 0.6
200000-1000000 0.076 0.7
𝑛 = 0.37, for 𝑃𝑟 ≤ 10
𝑛 = 0.36, for 𝑃𝑟 > 10

Note that 𝑓2is the friction coefficient at the inner surface of the pipe. 𝑃𝑟1 is the Prandtl number at 𝑇1
and 𝑃𝑟2 is the Prandtl number at 𝑇2 . 𝑅𝑒𝐷3 is the Reynolds number of the air calculated at the absorber
pipe external diameter. 𝑘35 is the thermal conductivity of air at 𝑇35 , where 𝑇35 = (𝑇3 + 𝑇5 )/2.
The irradiance incident on the lenses has been reduced by 10% (the parameter 𝑘1 is equals to 0.9).
The reduction takes into consideration the losses due to manufacturing error in the prisms (Davis and
Kuhnlenz, 2007). Furthermore, the optical characteristic defined in the model refer to a fixed
wavelength of the light. In reality, the solar spectrum is larger and lenses cannot be optimized for each
wavelength. Similarly, the model uses the theoretical value of the transmittance for the PMMA, but
the material behaves differently in real conditions. The shadowing effect due to the frame are also
considered at this stage.

The parameter 𝑘2 has been assumed to be 0.95. The front pipe is heated up more than the back surface
and therefore the cooling effect of the convection on the pipe is reduced with respect to the model
assumptions.

Given the fluid inlet temperature at the receiver at each time step, equations (1) – (4) form a set of
nonlinear algebraic equations to solve the four unknowns of the model: 𝑇𝑜𝑢𝑡 , 𝑇2 , 𝑇3 , 𝑞′ℎ𝑒𝑎𝑡𝑙𝑜𝑠𝑠 . The
model implemented in this study considers also the thermal losses occurring in the recirculation
system. An empirical equation has been developed to calculate the temperature of the fluid at the inlet
section of the receiver pipe:
𝑇𝑖𝑛 (𝑡 + 1) = 0.6 ∙ 𝑇𝑖𝑛 (𝑡) + 0.4 ∙ 𝑇𝑜𝑢𝑡 (𝑡) − 1 (8)
Equation (8) approximates the effect mixing of the heated fluid in the tank and the temperature drops
in the recirculation system as previously described. The equation has been selected to be linear rather
than logarithmic in order to simplify the model. The thermal inertia of our system including the tank
does not require a more complex equation at this stage. However part of our further work will be to
focus on that aspect, which will become relevant as we scale up and increase the volume of liquid
used in the thermal loop. The values of the parameters have been determined by fitting experimental
data.

12
4 Models Validation and Performance curve

The prototype was tested in June 2015 in Bourne in the UK (Latitude 52° 46' 20.8056'' N). The data
gathered during the testing campaign was used for the validation of the thermal model. The
experimental measures were compared to the results of the thermal model. The model has been used
to define the performance curve and evaluate the thermal losses in the system. A specific experiment
has also been conducted to understand the performance of the lens array. The results of the tests have
been compared to the ray tracing simulations to validate the optical model.

4.1 Validation of the Optical model

The optical model was validated using experimental data. Since the design is symmetric and to
simplify the experiment procedure, only a small section of the frame was considered (Figure 8).

Figure 8: Prototype set-up for experiment (21st April 2016)

In this case, the optical system consisted of three Lens 1 components. For experimental purpose, one
of the Lens 1 has been shadowed. A flat panel was installed at 2.5 metres from the lenses on the
optical axis plane which coincide with the receiver tube axis. This setup was implemented in SolTrace
using the optical model previously developed. Figure 9 reports the measurement of the concentrated

13
light beam and the flux distribution on the panel calculated with the ray tracing analysis. It should be
noted that the flux distribution was not measured experimentally. Rather, for validation purposes, the
beam width was compared to the width of the beam generated using the ray – tracing software.

a)

b)

Figure 9: a) Concentrated beam during experiment (21st April 2016) b) Flux distribution according to ray tracing
simulation

The two Lens 1 concentrated the normal incident light into a 50 mm wide beam, symmetrical with
respect to the horizontal axis of the target (see Figure 9a on the left). The single Lens 1 top component
is able to refract a 25 mm concentrated beam above the horizontal axis (see Figure 9a on the right).
The light beam simulated with the ray tracing analysis is reported in Figure 9b. It can be seen that the
model predicts the behaviour of the lenses.

Note that there is a gap between the two beams. This interval is caused by the shadowing effect of the
metal frame between two lenses (see Figure 8). The distance between the lenses has been simulated in
SolTrace. As a result, the model predicts a less intense flux between the beams (Figure 9b).

The SolTrace model has been used to calculate the percentage of solar radiation incident on the lenses
which reaches the receiver pipe. The parameter is named interception factor (𝜂𝑎𝑏𝑠 ) (Forristall, 2003).
The interception factor for the system analysed in this study is 53%.

The shadowing effect has also been analysed using the model. Referring to the experimental set-up
(Figure 8), simulations without gap in the lenses give an interception factor of 40.43%. When the
frame is considered, the calculated interception factor is 39.33%. The shadowing effect contribute to a
reduction of less than 1% in the light refracted onto the target. The shadowing effect have been
considered in the thermal model.

4.2 Validation of the Thermal model

The system has been tested in June 2015. Two days have been considered for the validation of the
thermal model when the solar radiation was more constant during the testing period. The current
thermal model has been used to simulate the heating phase of the system. Figure 10 reports the
comparison with the measured and calculated values for the 4th of June. This day the system was
tested for 116 minutes with an average DNI of 898.5 W/m2. The model was tested also on the 11th of
June for 320 minutes when the average direct normal irradiance was 793 W/m2. The results are
reported in Figure 11.

14
 100 1200

90
1000

80
800

70

[W/m2]
[°C]

600
60

400
50

200
40

30 0
65
1
5
9
13
17
21
25
29
33
37
41
45
49
53
57
61

69
73
77
81
85
89
93
97
101
105
109
113
Time [minutes]

Inlet Temperature Outlet temperature

Outlet Temperature calculated Inlet Temperature calculated
Irradiance

Figure 10: Measured and calculated temperature profiles for the 4th of June 2015

100 1000

900
90
800
80 700

600
70
500
60
400

50 300

200
40
100

30 0
31

51

71

91

111

131
1
11
21

41

61

81

101

121

141
151
161
171
181
191
201
211
221
231
241
251
261
271
281
291
301
311

Inlet Temperature Outlet temperature

Inlet Temperature calculated Outlet Temperature calculated
Irradiance

Figure 11: Measured and calculated temperature profiles for the 11th of June 2015

15
The absolute error is defined as the absolute value of the difference between the measured and the
calculated value. The relative absolute error is defined as the absolute error as a percentage of the
measured value. On the 4th June 2015, the average absolute error in outlet temperature calculated by
the model is 2.13 °C and the relative absolute error is 3%. For the 11th of June the average absolute
error in outlet temperature is 2.50 °C and the relative absolute error is 2.9%.

The model is oversensitive to the variations in the irradiance, compared to the response of the real
system. This is related to the assumption of the model of a steady state condition.
Figure 12 represents the analysis of the error with respect to the irradiance for the 11th of June.
10 1000
900
5 800
700
Error [°C]

0 600

W/m2
500
-5 400
300
89

243
1
12
23
34
45
56
67
78

100
111
122
133
144
155
166
177
188
199
210
221
232

254
265
276
287
298
309
320
-10 200
100
-15 0
Time

Err Tout Abs [C] Err Tout Abs [C] Irradiance [W/m2] Irradiance 2b [W/m2]

Figure 12: Comparison of experimental and simulation results for the 11th of June 2015

The model underestimates the losses when the irradiance is constant and overestimates the value of
the outlet temperature when the irradiance is more variable. In particular, the model is inaccurate for
radiation level below 400 W/m2. Therefore the curve of the irradiance for the 11th of June has been
approximated to reduce the effect of the variability. In practical terms this means that the model does
not cope well when clouds obscure the sun, since it does not take into account the thermal inertia of
the system. As a result the curve of the irradiance for the 11th of June has been approximated
(smoothed) to eliminate sharp reductions in solar irradiance. This is shown in Figure 13, where the
smoothed irradiance plot for the time period up to 141 minutes is shown as “Irradiance 2b”. Both the
original Irradiance and Irradiance 2b are also plotted for the total duration in Figure 12.
1000
900
800
700
600
W/m2

500
400
300
200
100
0
46

91
1
6
11
16
21
26
31
36
41

51
56
61
66
71
76
81
86

96
101
106
111
116
121
126
131
136
141

Time [minutes]

Irradiance [W/m2] Irradiance 2b [W/m2]

Figure 13: Irradiance approximation

16
Future work will focus on the optimization of the model. Further testing will be needed to
understand the losses the receiver pipe and in the recirculation system.

4.3 Performance Curve and losses analysis

The instantaneous thermal efficiency can be estimated as follow (Zhai et al., 2010):
𝑚̇ ∙ 𝑐𝑝,𝑎𝑣𝑒 ∙ (𝑇𝑜𝑢𝑡,𝑐𝑎𝑙𝑐 − 𝑇𝑖𝑛 ) (9)
𝜂𝑖𝑠𝑡 =
𝐴𝑙𝑒𝑛𝑠 ∙ 𝐺𝐷𝑁𝐼
𝑇𝑜𝑢𝑡,𝑐𝑎𝑙𝑐 is the calculated temperature at the outlet section of the receiver pipe. Six inlet temperature
levels were considered: 40°C, 50°C, 60°C, 70°C, 80°C, and 90°C. The calculated global efficiency
was plotted against the parameter defined as (Zhai et al., 2010):
((𝑇𝑖𝑛 + 𝑇𝑜𝑢𝑡 )/2 − 𝑇𝑎𝑚𝑏 ) (10)
𝑥=
𝐺𝐷𝑁𝐼
Figure 14 shows the performance curve of the current prototype.
0.2
0.18
0.16
0.14
0.12
ηglobal

0.1
0.08
0.06
0.04
y = 0.2078 -1.4791x
0.02
0
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11
(((𝑇𝑖𝑛 + 𝑇𝑜𝑢𝑡)/2 − 𝑇𝑎𝑚𝑏))/𝐺_𝐷𝑁𝐼

Figure 14: Performance Curve

The uncertainty of the global efficiency of the system can be calculated using the following equation
(Zhai et al, 2010).

𝑑𝜂𝑖𝑛𝑠 𝑑𝑚̇ 𝑑𝑇0 𝑑𝑇𝑖 𝑑𝐼𝑏

= + − − 2 (11)
𝜂𝑖𝑛𝑠 𝑚̇ 𝑇0 − 𝑇𝑖 𝑇0 − 𝑇𝑖 𝐼𝑏

Substituting the measured data and the error in the measurement instrument from section 2.1.2, we
determine the maximum relative error to be 16.0%.

Previous prototype based on single axes tracking showed better performance at similar temperature
levels (Zhai et al., 2010). Therefore, a detailed analysis of the losses in the system was carried out for
the reported days. Figure 15 reports the energy losses breakdown for the 4th June.

17
 8575 5995 1132 2543
W

0 5000 10000 15000 20000

qloss_optic qloss_abs qloss_th q_fluid

Figure 15: Thermal losses breakdown for the experiment on the 4th June 2015

The optical losses (qloss_optic) represent the energy lost due to absorption, reflection and diffraction
in the lenses. The effect of the light spectrum and the material properties are also affecting the
refraction process. The heat which is not absorbed by the pipe is considered in the absorption losses
(qloss_abs). The thermal losses (qloss_th) represent the sum of the convective and radiative losses
from the absorber pipe (see Equation (6)). The thermal power actually transmitted to the fluid is
reported as “q_fluid”. The losses in the lenses accounts for the 47% of the total energy lost in the
collector (see Figure 16).

14%
6%
47%

33%

qloss_optic qloss_abs qloss_th q_fluid

Figure 16: Thermal losses analysis for the experiment on the 4th June 2015

The optical efficiency has been calculated to be about 53% in design condition for the current set of
lenses. Note that the effect of dust on the lenses has not been considered in this work. Future study
will be carried out to analyse specifically the impact of uncleaned PMMA surface on the optical
efficiency of the system. One third of the total losses (33%) is due to the low absorption in the carbon
steel pipe absorber. The receiver installed on the prototype is a schedule 40 pipe without selective
coating to improve the absorption of the solar radiation and no glass envelope to reduce convective
losses. The heat transferred to the ambient through convection and radiation from the pipe is lower
with respect to other loss mechanisms (only 6%).

18
5 Conclusions and future works

In this study, the performance of a PMMA acrylic Fresnel lens solar collector, based on a dual-axis
tracking system, has been analysed. A mathematical model, including the optical model of the
concentrator and the heat transfer model of the receiver pipe and recirculation system was developed
and validated using experimental data. The performance curve of the collector has been derived for
temperatures between 40°C and 90°C.
The calculated global efficiency of the collector is limited to less than 20%. The main sources of
energy losses are due to: the optical losses in the lens system (47%) and the low solar absorptance of
the absorber pipe (33%). The thermal losses due to radiation and convection, account for only 6% as
relatively low temperatures (up to 90°C) are reached on the absorber pipe to promote radiation and
convection heat transfer.
While this technology is promising as a low cost concentrated solar collector for medium
temperatures, the results of this study highlighted that it can be improved by reducing the absorption
losses in the receiver pipe. It is recommended to install a solar selective coated evacuated receiver to
reduce the convective losses from the receiver pipe and increase the solar absorption. Previous works
showed that single tracking linear Fresnel lenses collector with evacuated tube receiver achieved a
global efficiency up to 55% (Zhai et al., 2010). Furthermore, the use of insulation will reduce the
thermal losses in the recirculation system. It should be noted that the performance of the experimental
prototype is below expectations partly due to its location and the non-optimal optical design of the
lenses. Linear Fresnel lenses have been proven to achieve higher temperatures, as presented most
recently by Linaric (2013). The target temperature for the next prototype is above 200°C to match the
requirements for industrial heat processes (57% of industrial heat demand requires temperatures
below 400°C (Rawlins and Ashcroft, 2013)).
Both single and dual axis tracking remain an option. The current use of dual-axis tracking is an
additional cost. However, simple and low cost closed loop dual-axis tracking systems can be
implemented with open source software and hardware. Furthermore, additional modules of lenses can
be clustered (as in the current prototype) to be moved using the existing tracking and control system.
An improved version of the prototype is currently be designed and further tests will be conducted. The
goal is to obtain a commercially attractive product based on the acrylic Fresnel lens principle for
applications in industrial process heat processes, with a target cost of €300/m2 or 0.04-0.07€/kWh
(Ivancic et al, 2014).

Acknowledgment
This work has been realised thank to the Knowledge Transfer Partnership programme sponsored by
Innovate UK and Lark Energy (Partnership ID 9630).

References:

Al-Jumaily K. E. J., Al-Kaysi, M. K.A., 1998. The study of the performance and efficiency of flat
linear Fresnel lens collector with sun tracking system in Iraq. Renewable Energy.14, 41–8.

Boyd, DA, Gajewski, R, Swift, R, 1976. A cylindrical blackbody solar energy receiver. Solar Energy.
18, 395–401.

Davis, A, Kuhnlenz, F, 2007. Optical Design using Fresnel Lenses: basic principles and some
practical examples. Optik & Photonik. 2(4), 52-55.

19
Fernández-García, A., Zarza, E., Valenzuela, L. and Pérez, M, 2010. Parabolic-trough solar collectors
and their applications. Renewable and Sustainable Energy Reviews. 14, 1695-1721.

Forristall, R., 2003. Heat Transfer Analysis and Modelling of a Parabolic Trough Solar Receiver
Implemented in Engineering Equation Solver. NREL Technical Report.

Franc F, Jirka V, Maly´ M, Náblek, B., 1986. Concentrating collectors with flat linear Fresnel lenses.
Solar & Wind Technology. 3(2), 77–84.

Holmgren, M., 2006. Xsteam for Matlab. Available online: www.x-eng.com. Accessed: June 2015.

Ijima O, 2010. Calibration of pyranometer by Collimation tube method. Radiation Section,

Atmospheric Environment Division, Global Environment and Marine Department, Japan
Meteorological Agency (JMA), IPC-XI, Davos, Switzerland

IRENA, 2015. “Solar Heat for Industrial Processes: Technology Brief”. Available online:
http://www.irena.org/DocumentDownloads/Publications/IRENA_ETSAP_Tech_Brief_E21_Solar_He
at_Industrial_2015.pdf. Accessed June 2016.

Ivancic A., Mugnier D., Werner Weiss G.S., 2014. Solar Heating and Cooling
Technology Roadmap. Available online (www.rhc-platform.org)

Kumar, V, Shrivastava, R.L., Untawale, S.P., 2015. Fresnel lens: A promising alternative of reflectors
in concentrated solar power. Renewable and Sustainable Energy Reviews. 44, 376–390.

Leutz, R., Suzuki, A., 2001. Nonimaging Fresnel lenses: design and performance of solar
concentrators. Springer, Berlin, New York.

Lin, M., Sumathy, K., Dai, Y.J., Zhao, X.K, 2014. Performance investigation on a linear Fresnel lens
solar collector using cavity receiver. Solar Energy. 107, 50–62.

Linaric, D., 2013. EFISOL Report Summary. Available online:

http://cordis.europa.eu/result/rcn/58185_en.html . Accessed June 2016.

Lovegrove, K., Stein, W., 2012 Concentrating Solar Power Technology Principles, Developments and
Applications. Woodhead Publishing Series in Energy.

Manikumar, R., Valan Arasu, A., Jayaraj, S., 2014. Numerical simulation of a trapezoidal cavity
absorber in the linear Fresnel reflector solar concentrator system. International Journal of Green
Energy. 11(4), 344–363.

Martinez, I., 1995. Available online: webserver.dmt.upm.es/~isidoro/dat1/Thermooptical.pdf .

Accessed June 2016.

Mujahid, A. M., Alamoud, A. R. M., 1988. An easily designed and constructed photovoltaic
pyrheliometer. Solar & Wind Technology. 5(2), 127-130.

20
Rawlins, J., Ashcroft, M., 2013. “Small-scale Concentrated Solar Power: A review of current activity
and potential to accelerate deployment”. Carbon Trust Available online:
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/191058/small_scale_co
ncentrated_solar_power_carbon_trust.pdf . Accessed June 2016

Reda, I., Afshin, A., 2008. Solar Position Algorithm for Solar Radiation Applications. NREL
Technical Report.

Solargis s.r.o., 2010. Online data and tools for solar energy projects. Available online:
http://solargis.info/ Accessed June 2015.

Taibi, E., Gielen, D., Bazilian, M. B., 2010, “Renewable Energy in Industrial Applications: An
assessment of the 2050 potential”. United Nations Industrial Development Organization. Available
online:
http://www.unido.org/fileadmin/user_media/Services/Energy_and_Climate_Change/Energy_Efficienc
y/Renewables_%20Industrial_%20Applications.pdf . Accessed June 2016

Xie, W.T., Dai, Y.J., Wang R.Z., Sumathy, K, 2011. Concentrated solar energy applications using
Fresnel lenses: A review. Renewable and Sustainable Energy Reviews. 15, 2588– 2606.

Zhai, H., Dai, Y.J., Wu, J.Y., Wang, R.Z., Zhang L.Y, 2010. Experimental investigation and analysis
on a concentrating solar collector using linear Fresnel lens. Energy Conversion and Management. 51,
48–55.

Zheng, H., Feng, C, Su, Y, Dai, J., Ma, X, 2010. Design and experimental analysis of a cylindrical
compound Fresnel solar concentrator. Solar Energy. 107, 26–37

21
 Cranfield University
CERES https://dspace.lib.cranfield.ac.uk
School of Aerospace, Transport and Manufacturing (SATM) Staff publications (SATM)

2017-06-21

Theoretical and experimental analysis of

an innovative dual-axis tracking linear
Fresnel lenses concentrated solar
thermal collector

Perini, Simoni
Elsevier

Simoni Perini, Xavier Tonnellier, Peter King, Christopher Sansom, Theoretical and experimental
analysis of an innovative dual-axis tracking linear Fresnel lenses concentrated solar thermal
collector, Solar Energy, Volume 153, 1 September 2017, Pages 679-690
http://dx.doi.org/10.1016/j.solener.2017.06.010
Downloaded from Cranfield Library Services E-Repository