See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269321442

Designing a concentrating photovoltaic (CPV) system in adjunct with a silicon

photovoltaic panel for a solar competition car

Conference Paper in Proceedings of SPIE - The International Society for Optical Engineering · June 2014
DOI: 10.1117/12.2050830

CITATIONS READS

12 2,255

4 authors:

Andrés Arias-Rosales Jorge Barrera-Velásquez

Carnegie Mellon University Universidad EAFIT
19 PUBLICATIONS 144 CITATIONS 3 PUBLICATIONS 16 CITATIONS

SEE PROFILE SEE PROFILE

Gilberto Osorio-Gómez Ricardo Mejia-Gutierrez

Universidad EAFIT Universidad EAFIT
73 PUBLICATIONS 495 CITATIONS 82 PUBLICATIONS 532 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Andrés Arias-Rosales on 31 March 2015.

The user has requested enhancement of the downloaded file.

 Designing a Concentrating Photovoltaic (CPV) system in adjunct with
a silicon photovoltaic panel for a solar competition car

Andrés Arias-Rosalesa, Jorge Barrera-Velásqueza, Gilberto Osorio-Gómez*a, Ricardo Mejía-

Gutiérreza
a
Design Engineering Research Group (GRID), Universidad EAFIT, Medellin, Colombia
* Corresponding author: [email protected], Other authors’ e-mail: [email protected]; [email protected];
[email protected]

ABSTRACT

Solar competition cars are a very interesting research laboratory for the development of new technologies heading to
their further implementation in either commercial passenger vehicles or related applications. Besides, worldwide
competitions allow the spreading of such ideas where the best and experienced teams bet on innovation and leading edge
technologies, in order to develop more efficient vehicles. In these vehicles, some aspects generally make the difference
such as aerodynamics, shape, weight, wheels and the main solar panels. Therefore, seeking to innovate in a competitive
advantage, the first Colombian solar vehicle “Primavera”, competitor at the World Solar Challenge (WSC)-2013, has
implemented the usage of a Concentrating Photovoltaic (CPV) system as a complementary solar energy module to the
common silicon photovoltaic panel. By harvesting sunlight with concentrating optical devices, CPVs are capable of
maximizing the allowable photovoltaic area. However, the entire CPV system weight must be less harmful than the
benefit of the extra electric energy generated, which in adjunct with added manufacture and design complexity, has
intervened in the fact that CPVs had never been implemented in a solar car in such a scale as the one described in this
work. Design considerations, the system development process and implementation are presented in this document
considering both the restrictions of the context and the interaction of the CPV system with the solar car setup. The
measured data evidences the advantage of using this complementary system during the competition and the potential this
technology has for further developments.
Keywords: Concentrating Photovoltaic (CPV), solar car, Primavera, World Solar Challenge (WSC), Fresnel lens,
gallium photovoltaic cell

1. INTRODUCTION
With annual sun radiation as high as 2100 kwh/m2 in certain territories [1], Colombia has a great potential in terms of
solar clean energy. However, solar solutions are not spreading as fast as they could in developing countries, presumably
because most of the products have to be imported raising costs, and due to potential clients unwilling to trust a new
technology. Therefore, a project such as the first Colombian solar car, so-called “Primavera”, competing in the World
Solar Challenge (WSC)-2013 [2] has the background goal of catalyzing local solar industry and inspiring Colombian
people into believing in these sustainable technologies.
The World Solar Challenge is considered to be among the most important sustainable vehicle events in the world. Since
1987, teams from all over the world cross Australia from north to south in solar cars, taking their vehicles beyond the
limits of efficiency. Innovation has always been one of the pillars of the spirit of the competition, especially in the
category Primavera was designed for, so-called Challenger. Being this the most demanding category, the vehicles are
constrained in terms of dimensions, number of wheels, photovoltaic (PV) area and are allowed to use only one initial
direct battery charge, among other restrictions. Even though the World Solar Challenge is a 3000km long race, the high
level of leading edge technological development and competitiveness, implies that every saved race minute is crucial and
could mean the difference among the first positions. In such a long pathway, average speed is the key, and with most of
the top teams being already highly efficient in terms of energy consumption, a suitable approach is to focus on
generating more energy to be spent in higher speeds. Based on that premise, the Primavera team strategy to use
concentrating photovoltaics (CPV) had a main goal of increasing the total solar energy harvested and transformed into
electrical energy, without negatively affecting the vehicle performance and complying with the WSC-2013´s regulations
[2]. This work describes how the CPVs design and development process was conducted correlating optical applications,
solar power technology and manufacturing in order to overcome the drawbacks and reach a positive outcome.

2. STATE OF THE ART

2.1 Unconventional concentrating photovoltaic approaches

The following are several unconventional approaches to solar concentrating systems found in the academic and
commercial state of the art.
Figure 1(a) and Figure 1(b) consist in a low compound parabolic mirrors concentrating approach [3] [4]. As the number
of concentrated suns increases, the angle of acceptance in solar tracking declines. However, this approach can raise the
angle of acceptance concerning the incoming solar rays as shown in Figure 1(b).
Figure 1(c) shows a medium compound parabolic mirrors device [5], which can be designed to have a broader angle of
acceptance compared to the previous approach shown. However, the bouncing solar rays, presented in Figure 1(c),
converge in a suspended line, which can lead to impracticality for photovoltaic cells.
The concentrating device in Figure 1(d), provided by Morgan Solar, can harvest and refract sunlight towards a
photovoltaic cell with an up to 900 suns concentration rate. This device is supposed to remain unaltered by the
temperatures that can be reached and is visually compact [6].
The refraction that takes place in the secondary optic element (SOE) [7] shown in Figure 1(e), redirects the sunlight rays
coming from a range of directions towards the photovoltaic cell underneath. In adjunct with a Fresnel lens, this approach
can keep generating energy even when not fully aligned towards the sun. Nevertheless, there is a considerable increase in
height compared to using the Fresnel lens alone.

(a) (b) (c) (d) (e)

Figure 1. Unconventional concentrating photovoltaic approaches [3] [4] [5] [6] [7].
2.2 Concentrating photovoltaics in solar cars
In spite uncommon, there have been several teams that have used CPVs in their solar cars during the World Solar
Challenge. The most representative examples are briefly described next.
Below on Figure 2(a), Michigan 2007 Continuum solar car is shown using its CPVs exposed in the upper surface. These
CPVs consisted of reflecting mirrors, presumably linear parabolic mirrors, that could work during the static stages and
while the car was moving due to an active tracking system that could pivot the CPVs to face the sun within a certain
range. On Figure 2(b), the 2009 Michigan car Infinium can be seen using static CPVs, which could work only when the
upper bodywork was removed while the car wasn’t moving, and in order for them to face the sun, the entire lower
bodywork had to be tilted.
Figure 2(c) consists of the 2007 Solar Team Twente car tilting its upper bodywork while moving for the main panels to
face the sun as much as possible, and for the CPVs to work. The CPV system uses linear Fresnel lenses to focus sunlight
into rows of solar cells underneath. Moreover, the rows of solar cells could move by using electric actuators for them to
follow the focused lines when the tilting system reached its functional limits [9].
 (a) (b) (c)
Figure 2. Previous implementations of CPVs in solar competition cars [8] [9].

3. DESIGN PROCESS
Following the CPVs main goal, the design process focused on maximizing the net income of energy measured in Watts
into the battery pack. Showing the related variables and the way they affect the final result, the authors summarize the
approach into the following equation, serving a as a guide for the entire design process.

( ) (( ) ( )) (1)
With net income of energy [ ]; PV area [ ] of a single solar cell; Number of concentrated suns; R
incoming solar radiation [( ⁄ )⁄ ]; Solar cells efficiency; Optical technology efficiency, System
efficiency (taking into account electric connections and MPPT efficiency among others); Number of CPVs;
Time [ ] using the CPVs; Power losses [( ⁄ )] due to the CPV system weight increase in the rolling
resistance; Time [ ] moving the vehicle by its own power supply; Direct power consumption [( ⁄ )] due
to the CPV system energy consuming components, such as fans or pumps; and Time [ ] using energy consuming
components of the CPV system.
After choosing the concentrating technology according to the design requirements, the physical variables were defined in
a progressive process including the size of the solar cells, the total PV area, the optical devices area, the CPVs height, the
number of concentrated suns, and the number of CPVs. Afterwards, the detailed design was carried out and improved in
terms of weight, manufacturability and performance through three prototypes. Finally, the CPVs were designed as a
system, comprising the distribution, the electric circuit, the tracking and cooling strategy, and the safety considerations.
3.1 Design requirements
Designing a CPV array for a solar car, meant to compete in a world class race, is a challenging and complex process due
to all regulations and specifications that it must meet. In the first place, the system must produce more energy than the
energy consumed due to its weight and its electronics. In order for this to be accomplished, early simulations were
performed in a Strategy-Simulation-Software (3S)1 in which different weights for the CPV array were assumed and
possible total power generations were calculated. The balance, on which the CPV viability relies on, depends on the
trade-off on energy harvested by the CPVs versus the extra-energy that the CPV system will demand principally because
of the extra weight added to the vehicle, which increases the rolling resistance. If the energy harvested by the CPVs is
lower than the extra energy consumed by the vehicle because of the CPVs system, then the system is not viable. A
preliminary analysis performed with the 3S program showed that the extra energy obtained because of the CPV system
could save approximately 20 race minutes. Thus, the entire CPV system shouldn’t exceed 20Kg taking into account that
every extra Kg increments in a minute the time necessary to finish the race.
Even if the CPV system reached a positive ratio of cost/benefit, it should also meet all requirements stated in the World
Solar Challenge 2013 regulations [2]. WSC-2013’s regulations constrain the maximum photovoltaic (PV) area, but don’t
limit its usage exclusively for the main solar panel, making it possible for the PV area to be distributed among the main
solar panel and a CPV system. Moreover, the CPV design process was carried out always taking into account all other
relating regulations, such as not interfering with the pilot performing an emergency exit.

1
The Strategy-Simulation-Software is a custom made application developed in Python to predict the overall energy management
system of the vehicle, considering the energy harvesting (Panel+CPVs), storage (Battery) and consumption, in order to strategically
plan all race decisions such as speed and stops, among others.
Since CPV performance is subjected to direct sunlight and a precise sun tracking, it was considered neither practical, nor
strategic to locate them exposed in the upper surface subtracting area from the main solar panel. Instead, the Primavera
concentrators were planned to remain under the main exposed solar panel until the static stages of the race, when the
upper bodywork is removed to be tilted towards the sun and thus, exposing the CPV systems beneath. Therefore, the
entire system should fit inside the slim and aerodynamic bodywork, making the CPVs height crucial since it is inversely
proportional to the number of CPVs that could fit. Likewise, the system needed to be rigid and stiff enough to resist
being in a moving vehicle, its materials ought to resist the desert temperatures (up to 40°C) and temperature raise due to
concentrating sunlight (Over 70°C) and the system should not compromise the security by, for instance, starting a fire.
Furthermore, a tracking and cooling strategy was necessary for the system to be efficient and competitive. In terms of the
CPVs as individual units, manufacturability and design simplicity were essential to ensure meeting all the requirements
with precision and reliability, given that the budget, workforce and implementation time were limited.
3.2 Chosen concentrating approach
Having analyzed all conventional and unconventional approaches, the chosen concentrating technology were the
Punctual Fresnel Lenses. As shown in Figure 3, this optical element consists in a lens that has a stepped curvature which
refracts the incoming rays (perpendicular to a central plane) into a common focal point.

Figure 3. CPV scheme with a Fresnel lens-solar cell configuration. [10]

The advantages compared to other concentrating approaches are the following:
 Compared to regular lenses, Fresnels are lighter, less expensive, more compact and easier to manufacture.
 There is a wide commercial availability of Fresnel lenses of all sorts of sizes and focal lengths contributing to the
design flexibility.
 From a general perspective, this solution had the potential of getting to a simpler and more manufacturable overall
design.
3.3 Physical variables definition
The process of defining all physical variables of the design was a progressive development from two different
approaches. From one side, there was an approximate process of preliminary calculations and CAD (Computer Aided
Design) volume simulations based on general assumptions, provided that some variables depended on each other such as
height and number of CPVs, so there was not a single possible solution. From the other side, there was a process of
selection between available and suitable commercial elements. The convergence between these two approaches defined
the physical variables of the three developed prototypes shown in Figures 7(a), 7(b) and 7(c).
3.3.1 Areas and number of CPVs
The WSC-2013´s regulations stated a mathematical relation between Silicon (Si) and Gallium (Ga) photovoltaic area,
where

( ( )) (2)

with =Gallium Area and =Silicon Area. Since the Ga is more efficient, the regulations limit its use and the bigger
the Ga area, the smaller the total photovoltaic allowable area. The main Silicon panel had a 5.796m2 area, limited by the
maximum photovoltaic area (6 m2), the aerodynamic design and the regulated maximum dimensions. This meant an
allowable extra 0.102m2 Ga area, or 0.204m2 Si area that could be used in the CPV photovoltaic cells. For CPVs, Ga was
chosen because of its superior efficiency (usually >30% efficiency) [11], for its capability to increase its efficiency
(under a physical limit) as the number of concentrated suns rises and for it being able to resist higher concentrations
(>1000 suns) [12] as well as higher temperatures.
After analyzing available commercial Fresnel lenses, the research focused on the Edmund Optics supplier [13], being a
reference of cost/quality. A potential height for the CPVs was assumed to be of 10-13 cm and, starting from that, a
preliminary CAD volume simulation was performed. The image in Figure 4(a) shows how some sections were trimmed
from the volume simulation to avoid collisions with other systems, the bodywork and the cockpit. Since the Fresnel
lenses are flat and should be in parallel planes, the upper curved surface of the simulated volume was flattened and
measured. Figure 4(b) shows the resulting preliminary available area for the CPVs of 3, 56 m2. Since the widest elements
of the CPVs are the Fresnel lenses, this area can be taken as the maximum area for all lenses together.

(a) (b)
Figure 4. CPVs initial volume emulations in CAD model and available area for the CPVs.
Having defined to use Ga cells, an issue of availability arose. Ga is a material usually reserved for aerospace applications
and high efficiency CPV power stations (solar farms). Additionally, it was experimentally found that cutting a solar cell
and rebuilding its micro-connections implied a level of complexity that should be avoided if possible. Therefore, it was
decided to use available GreenVolts modules [14] which came with 5.5x5.5 mm Ga cells embedded in their own circuit
with diode. With an allowable 0.102 m2 Ga area, it would then be possible to locate 3371 CPVs, each one with a
0.00003025 m2 Ga area. In this configuration, if the whole lens available area of 3,56 m2 were to be used, each lens
would have a concentrating rate of 34,9 suns. However, since Ga cells raise their efficiency considerably in high
concentration rates and the GreenVolts modules were designed to work under as high as 1300suns, the CPVs
configuration was defined for a desired concentrating rate. To leave a margin from operational specifications, but still
having high concentrating rates, the desired concentration was established to be close from 1000 suns and overlapping
that with the information of available commercial Fresnel Lenses, the concentrating rate was set to be of 1107 suns.
Consequently, as

( ) (3)

With X=concentrated suns, =Lens area and =Solar cell photovoltaic area, each lens would then have a 0,03348 m2
area with dimensions of 182,9 mm x 182,9 mm since they would be cut in a square shape. Moreover, this equation means
that with 3, 56m2 of lens area, there would be 0,0032m2 of Ga photovoltaic area distributed among 106 CPVs.
3.3.2 CPVs height
At first, the height of the CPVs was calculated by entering the focal length stated by the supplier in a 2D simulation
using CAD Software. For instance, Figure 5 shows an example of an approximate resulting CPV height of 7,96 mm
providing a focal length of 10,16 mm. The method of entering the focal length stated by the supplier experimentally
validated. Tests showed that the real focal length was in fact different from supplier’s specification.

Figure 5. Height calculation of a CPV using Creo Parametric 2.0.

Figure 6 illustrates a device with variable height designed to find the distance between the lens and the cell which
resulted in the best combination of Amperage and Voltage. This method was used for defining the heights of the three
CPV prototypes developed.

Figure 6. Device designed to experimentally find the most suitable distance between a Fresnel lens and a photovoltaic cell.
3.4 Prototypes
To get to the final design, the CPVs of Primavera evolved through three prototypes shown in Figure 7. The first
prototype shown in Figure 7(a), succeeded in structurality and in accuracy converging the sun rays into the photovoltaic
cell. However, there were two issues involved that consisted in the numerous steps required to manufacture each CPV,
and the high weight. With each CPV weighting 189.7 g, the 106 CPVs would weight 20.11 Kg overpassing the initial
weight limit without even taking into account other elements required for the CPV array, such as the MMPT or CPVs
chassis. Being not possible to reduce the lens weight, which represented the 36.9% of the total weight, the second
prototype shown in Figure 7(b), focused on reducing the weight of the other parts and reducing the manufacture steps.
After reaching an acceptable weight and manufacturability, the prototype-3, shown in 7(c), was developed to increase the
performance solving certain optical issues described in the following subsections.

(a) (b) (c)

Figure 7. CAD lateral views, CAD renderings and photos of prototypes 1, 2 and 3.
3.4.1 Prototype-2 weight reduction
As can be seen in figure 8, the columns were redesigned, as seen in Figure 8(b), inspired by the carbon nanotubes atomic
structure shown in Figure 8(a), which plays a key role in the superlative stiffness of the material while being notoriously
light. The four columns, which represented the 28.62% of the total weight, went from 54.3 g to 26 g, reaching a 52.12%
weight reduction in columns.

(a) (b)
Figure 8. The final column design inspired by the carbon nanotube atomic structure [15] [16].
The screws and nuts were replaced by acrylic resin, going from 9 g to 0.3 g, obtaining a 96.66% weight reduction in
fastener elements. Moreover, the base was reduced as much as possible decreasing its weight in 27.2%. These evolutions
were obtained by improving prototype-1, achieving prototype-2. The overall weight reduction resulting from this step
can be viewed in table 1.
Table 1. Comparison between the weight of prototype-1 and prototype-2 as individual units and as the whole array.
Prototype1 Prototype2 Weight reduction CPVs Array1 CPVs Array2 Total Weight Reduction
189.7g 137.5g 52.2g-27.5 % 20.108Kg 14.57Kg 5.54Kg

3.4.2 Prototype-2 power generation analysis

Having achieved the weight reduction goal, the next step was to assure that the amount of power generated by the CPVs
was enough to justify their use in the vehicle. Several tests with the prototype-2 showed that the CPV was not producing
what it was supposed to, based on control tests as it is evidenced in table 2.
Table 2. CPV prototype 2 and control test comparison in terms of performance.
Prototype-2 Control Test
Open-circuit Short-circuit (VOC)* (Isc) Open-circuit Short-circuit (VOC)* (Isc)
voltage (VOC) current (Isc) voltage (VOC) current (Isc)
2.9 V 0.37 A 1.07 Watts 2.96V 1.5 A 4.44 Watts

It is important to clarify that (VOC)* (Isc) is not equal to the real power generated due to the power generation curve
inherent to photovoltaic cells, but it is useful to make comparisons and is referred in this work as “Comparative Power”.
The control test was carried out using a GreenVolts module just as the one that Prototype-1 and Prototype-2 used but
using a different Fresnel lens with a bigger focal length and at different distances between the solar cell and the lens. The
control test produced 413.8% of the Comparative Power produced by prototype-2, suggesting that there were optical
issues involved in the prototype design.
Several experimental tests studying laser trajectories revealed the causes of the low Comparative Power generation:
 The used Fresnel Lenses, as they were not specifically designed for solar concentration, didn’t converge light in a
single focal point as it was assumed to, but in two different focal points vertically apart from each other.
Consequently, the light entering the GreenVolts module were the rays being converged to only one of the two focal
points thus, the rest of the rays were not reaching the solar cell.
 The GreenVolts module used a secondary lens over the solar cell which had an angle range of admission. This angle
range was found to be of approximately 26° from the vertical axis and was being surpassed by a portion of the rays
 converged by the used Fresnel lenses. Therefore, all rays converged over the 26° could be reaching the GreenVolts
secondary lens but not reaching the solar cell underneath.

3.4.3 Improvements to obtain the final design (Prototype-3)

The final design, reached in the prototype-3 shown in Figure 7(c), consisted mainly in solving the optical issues found in
the power generation analysis and improving the design for manufacture of the base, in terms of required number of
manufacturing steps and their involved difficulty.
Knowing that the previous Fresnel lenses were not meant to be used for concentration, it was defined to use other lenses
specially designed for CPVs from the supplier Fresnel Factory. These lenses, with the reference CP182-235, were stated
to be capable of converging the perpendicularly incoming rays into a common focal point with a high efficiency.
Nevertheless, they had a longer focal length, going from 70mm in the previous lenses to 182mm, which caused the CPVs
height to increase considerably. Moreover, the secondary lens was removed from the design to avoid the issue
concerning the angle range of admission. This decision was verified with experimental tests which showed an increase in
Comparative Power generation whit the secondary lens removed.
The distance between the lens and the solar cell was recalculated with the experimental method described in the section
3.3.2 (CPVs height). This method led to the definition of the final proportions of the CPVs, where height went from
133.9 mm to 195.7 mm. With the increased height, the columns length increased as well. Also, the secondary lens was
removed and the thickness of the lens changed from 3mm to 4mm. The previous changes resulted in an overall weight of
166.2 g per CPV.
3.4.4 Prototype-3 tests
In table 3, the data shown in Prototype-3 is the result of average measures among several tests performed under the real
sun between 11:00 am – 2:00 pm. The results evidence the effects of the optical solutions implemented, since the
Comparative Power generation increased in a 746.9%.
Table 3. CPV prototypes 2 and 3 comparison in terms of performance.
Prototype-2 Prototype-3
Open-circuit Short-circuit (VOC)* (Isc) Open-circuit Short-circuit (VOC)* (Isc)
voltage (VOC) current (Isc) voltage (VOC) current (Isc)
2.9 V 0.37 A 1.07 Watts 2.96 V 3.07 A 9.09 Watts

3.5 CPVs design as a system

After assuring that the CPVs final design had a suitable weight, structurality and power generation, the next phase
consisted in defining the details required to achieve a functional system for a solar racing car.
3.5.1 Distribution
With a final CAD model of the CPV design, the detailed distribution process was carried out assembling as many CPVs
as possible without interfering with any other system, the bodywork or the cockpit. The following aspects were taken
into account:
 All CPVs should be assembled in such a way that their lenses were all in parallel planes, so all the CPVs could be
aligned facing the sun at the same time if the car were to be tilted towards the sun.
 The CPVs should be grouped in zones that would be rigid themselves but not between them. Each CPV independent
from each other would result in a number of different modules too big to be installed and handled in a practical way,
whereas all CPVs fixed rigidly together would suppose a big and heavy structure to be properly handled. Therefore,
the strategy of dividing them in independent zones.
 Ideally, all CPVs should be positioned in the same height to simplify the design and installation. This was not
possible to achieve due to the changing inner geometries of the vehicle, the heights should be grouped in, as few as
possible, different heights.
Figure 9 shows how it was possible to fit 64 CPVs in 6 zones with 5 different heights plus a “floating CPV”. The floating
CPV consisted in an additional module that could fit in the closed car by lying horizontally and that would then be turned
to be vertical (aided by a rotating mechanism) once removed the upper bodywork. The floating CPV can be seen in the
top view, located at the lower right corner.

Figure 9. CPVs configuration inside of the solar car Primavera seen from top and side views.
3.5.2 Electric circuit
The 64 CPVs were connected by two circuit series each one of 32 CPVs. Both series were connected in parallel in a
MPPT, which function was to regulate and configure the incoming voltage and current in order to deliver the highest
power possible. The used MPPT had a functioning range of 36 V – 144 V, which was suitable for the expected 80 – 96 V
voltage coming from the two series. Furthermore, most of the zones were independent series to simplify connectivity and
zones installation in the vehicle.
3.5.3 Tracking strategy
In order for the CPVs to be able to generate electric power, the lenses must focus light into the photovoltaic cell, which is
only possible when the incoming rays are perpendicular to the Fresnel lenses surface. The CPV system was designed to
reside rigidly fixed to the car’s lower bodywork with all CPVs aligned in parallel planes. This way, the whole lower
bodywork was supposed to be tilted to face the sun aided by an adjustable structure and light perpendicularity measuring
devices. Even though it was possible to align and fix all CPVs with an error margin of 0.6 °, it was found experimentally
that the bodywork rigidity was not enough to keep the CPVs aligned after regular testing of the car in a race track.
Because of this, the tracking strategy changed into aligning each CPV zone individually by hand. Aided by special
welding glasses, the team members could look directly at the focused light and verify if it was reaching the photovoltaic
cells.
3.5.4 Cooling strategy
Under a midday sun, unconcentrated photovoltaic cells exceeded the 60°C during tests. At 1107 suns, the cells in the
CPVs could easily exceed the 100°C. Even though Ga cells can perform well under high temperatures (as high as
100°C), the lower the temperature, the higher the efficiency. Therefore, to increase the efficiency and to reduce the risk
of permanently damaging the Ga cells, a cooling strategy was carried out during the WSC-2013 race. While the CPVs
were being used, dielectric water was atomized directly under the aluminum surface were the photovoltaic cell was. This
alternative was the only option according to WSC-2013’s regulations about externail aids. The strategy was simple, yet
effective.
3.5.5 Safety Strategy
While performing security tests, the focused sunlight coming from the used Fresnel lenses was capable to ignite Wood
almost immediately. This situation can produce smoke from the material of the bodywork (carbon fiber with vinyl ester
resin compound) and even from an anti-fire foam which was stated to resist over the 400°C. However, the CPVs were
not able to degrade a Silica cloth which was capable of resisting 1000°C. Therefore, the temperature reached in the focal
point was likely to be between 400°C-1000°C, enough to represent a real danger of fire inside or outside the vehicle.
There is a 17° safety margin for misalignment, meaning that if the CPVs are tilted in a range of (-17°) to (17°) from total
alignment towards the sun, the focus point will move but will still be focused in the aluminum plaque surrounding the
solar cell, so it would not represent any danger to the car.
Even with the mentioned angle, when the top bodywork was removed, exposing the 64 CPVs, the incoming solar rays
could surpass the 17° depending on the hour. To avoid danger, special reflective coverings were designed to rest over the
CPV zones until they were outside of the car and close to being totally aligned to the sun.

4 IMPLEMENTATION
4.1 Manufacture
Figure 10 shows the main parts of the CPVs and the key milestones in manufacture of 75 units:
 Figure 10(a) shows the GreenVolts original concentrating modules, including a Ga cell embedded in a circuit with
diode, an aluminum heat sink and a secondary lens within an aluminum casing.
 A removal material process was applied. Figure 10(b) shows how each module was milled into the final CPV bases,
reducing volume and weight. Secondary lens and its casing was also removed.
 Figure 10(c) shows columns cut and drilled from 5/16" aluminum tubing.
 Original Fresnel lenses where cut into its final shape by CNC laser cutting, as seen in Figure 10(d).
 Being the assembly precision crucial for the lenses to focus the light into the Ga cell, the CPVs were fixed with
acrylic resin aided by a jig that maintained the parts firmly aligned. Eight of these jigs were manufactured for several
CPVs to be assembled and let curing in parallel as can be seen in Figure 10(e). This method proved to be successful
since only 3 CPVs didn’t passed focus precision tests out of 74 tested.
 Figure 10(f) displays eight finished CPVs, from a total of 75 that were manufactured and assembled.

(a) (b) (c)

(d) (e) (f)

Figure 10. Main CPV parts and key processes concerning manufacture.
4.2 CPV Zones assembly
A lightweight aluminum chassis was developed for each zone to remain assembled together. For every zone, the CPVs
were fixed between them and to their zone chassis using acrylic resin Scotch-Weld DP805. Since all CPVs of a zone
were to be fixed in parallel planes, they were assembled with the resin while being face down over a flat glass with metal
weights pressuring them against the flat surface.
4.3 Installation
64 CPVs were finally fitted into the vehicle, assembled to the lower bodywork and the main chassis (See Figure 11). All
necessary adjustments were carried out for the CPVs not to interfere with other parts and systems, such as reducing the
upper bodywork ribs height.

Figure 11. Exposed CPVs assembled in Primavera solar car.

4.4 World Solar Challenge 2013
Figure 12 illustrates three CPV zones being used during a control stop with the main solar panel facing the sun as well in
the background.

Figure 12. CPVs being used during the WSC-2013.

According to the WSC-2013´s regulations, the solar cars are allowed to race after 8:00 am and until 5:00 pm. Since the
CPVs could work only while not moving, they were used from the sunrise (around 6:00am on average) until 8:00 am and
from 5:00 pm until sunset (around 6:25 pm on average). Additionally the mandatory control stops, which were of 30min
each, were the opportunity to deploy the CPVs.

5 RESULTS AND ANALYSIS

The entire CPV system including the 64 CPVs, the zones chassis, the electric cables and the MPPT, weighed
approximately 15kg. Considering the initial goal of 20kg, the system represented the 75% of the allowable weight.
The graph shown in Figure 13 depicts the power generated in Watts/h (Y axis) throughout the race days (X axis as time).
The total power generated by the main solar panel and the CPVs, can be viewed in the peaks that rise higher in power,
and the power generated only by the CPVs is represented by the lower line. Although the race lasted 6 days, the power
generation information could not be stored during the first days due to telemetry issues. Therefore, the data showed are
from days 3 to 6.

Figure 13. Power generation chart during race days concerning the main solar panel and the CPV array.
From Figure 13 it can be highlighted:
 Day 3: During the morning, the maximum CPV power generation reached 150Watts/h, which implied 15% of the
power that was being generated by the main solar panel at that time. CPVs were used in the morning, evening and at
a control point around 3:00 pm, when power generation raised higher than in the evening due to greater radiation
being the sun higher in the sky.
 Day 4: Due to logistic issues, it was not possible to use the CPVs during the control stop. The CPV power peak was
reached at 55Watts/h, representing the 7.5% of the power being generated by the main panel at that time.
 Day 5: The cloudy weather affected the CPVs performance, making it not possible to concentrate direct light in the
evening. Without direct light, the CPVs are not capable of generating enough voltage to turn the MPPT on and thus,
whatever power generated couldn’t reach the battery. The maximum power generated was 50 Watts/h, the 9.53% of
the power being generated by the main solar panel at that moment.
 Day 6: CPVs were used during the morning and in a control stop before the arrival to the finish line. The power
generation reached a peak of 100Watts/h, 12% of the main solar panel power generation at that same time.

6 CONCLUSIONS
 With a weight of approximately 15kg, the entire system represented the 5.55% of the total weight of the vehicle
without driver and managed to generate up to 15% of the energy generated by the main solar panel and thus, the
13.04% of the total energy generated by the vehicle. Furthermore, the photovoltaic area of the CPV array of
 0,001936 m2 implied only the 0.033% of the main panel area. The previous relations evidence the benefits of using
concentrated photovoltaics, even though better performance could be achieved based on the individual CPV tests,
which showed a maximum “Comparative Power Generation” of up to 10 Watts/h/CPV. Provided the prior definition
of Comparative Power in this work, it could be assumed a potential 7 Watts/h/CPV, which would reach a total of
448 Watts/h for the entire array, being the 56% of the average power generation of the main panel. Therefore, the
CPV system proved to be not only beneficial during WSC-2013, but also to have the potential of becoming a major
competitive advantage in a further development.
 The design and development process succeeded in weight and resilience initial goals. With an approximate weight
of 15kg, the system weighed 25% less than initially allowable. The CPV zones managed to keep working despite of
more extreme conditions than what they were designed for, such as being manipulated over the dessert sand, held in
the air manually and being constantly taken in and out of the vehicle while being in a hurry.
 Precision in CPV manufacture is crucial, for the lack of alignment between the concentrating optics and the solar
cell can result in sunlight not reaching the photovoltaic surface and thus, in electric power not being generated. With
only 3 out of 74 tested CPVs that didn’t pass light focus tests, the assembly technique of using wooden low cost jigs
proved to be effective and accurate with a 96% success rate.
 The CPV performance results during the WSC-2013 evidenced the main drawback of using concentrated
photovoltaic, which is its weather dependence. For instance, the cloudy weather of day 5 avoided direct sunlight to
reach the CPVs. This resulted in the inability to power on the MPPT and thus, in no electric power generation while
the main un-concentrated panel was generating up to 650Watts. Additionally, with a maximum relation of CPVs-
Main panel of 15%, the CPVs could have theoretically generated up to 97.5Watts/h with direct sunlight.
 The major necessary changes in terms of tracking and main use strategy serve as an instance of the bodywork of a
solar competition vehicle not being stiff enough to be rigidly fixed to a CPV array. Such a lack of stiffness resulted
in the misalignment of the CPVs and, therefore, a tracking strategy independent from the bodywork structurality is
seemingly more suitable.
 The CPVs being grouped in structurally independent zones showed to be highly advantageous for this feature made
the system flexible enough to adapt to a considerably different use strategy. Both individual and single grouped
CPVs would have been too difficult to handle outside of the vehicle in an efficient and practical way and this might
have resulted in the system not being used at all.

ACKNOWLEDGEMENTS
The authors wish to thank Universidad EAFIT and EPM (Empresas Públicas de Medellín) as the main sponsors of the
project and for all their support and advice provided during the entire solar car project. A special thanks to all project
sponsors for believing in the project and supporting the team. Also, a special acknowledgement to Andrés Serrano Rico
who provided the measuring of CPV performance during the event and the charts involved in their analysis. Also, to all
team members who made Primavera possible and helped to use de CPVs during the WSC-2013. Special thanks to
Federico Arbeláez, who assisted in the manufacture, installation and testing of the CPVs.

REFERENCES
[1] Alejandro Fula.,"El Mercado de la Energía Solar en Colombia," solarwirtschaft,
http://www.solarwirtschaft.de/fileadmin/media/pdf/Kolumbien_Fula_Sierra_Der_Solarenergiemarkt.pdf (Last
retrieved on February 2014), (2011).
[2] World Solar Challenge,"Overview," Worldsolarchallenge,
http://www.worldsolarchallenge.org/about_wsc_2013/overview (Last retrieved on March 2014), (2013).
[3] Ajitkumar S Gudekar, Atul S Jadhav, Sudhir V Panse, Jyeshtharaj B Joshi, and Aniruddha B Pandit, "Cost effective
design of compound parabolic collector for steam generation," Solar Energy. Papers 90, 43-50 (2013).
[4] Yuehong Su et al., Yuehong Su, Saffa B Riffat, and Gang Pei, "Comparative study on annual solar energy
collection of a novel lens-walled compound parabolic concentrator (lens-walled CPC," Sustainable Cities and
Society. Papers 04, 35-40 (2012).
[5] Brian Norton, [Solar Energy Thermal Technology], Springer, Heidelgurg, 1-279 (1992).
[6] Morgan Solar,"HOW IT WORKS," morgansolar, http://morgansolar.com/how-it-works/ (Last retrieved on March
 2014), (2014).
[7] Pablo Zamora et al., "A high-performance photovoltaic concentrator," SPIE Newsroom, doi.
http://dx.doi.org/10.1117/2.1201008.003105, (2010).
[8] Jeroen Haringman,"Focus on concentrators," Solarracing, http://www.solarracing.org/2013/10/04/focus-on-
concentrators/ (Last retrieved on March 2014), (2013).
[9] Daqpoint Benelux B.V R.H.J. Overgaauw, "Solar Challenge 2007 – 3010 km powered by the sun ," Solar Team
Twente, (2008).
[10] Zhenfeng Zhuang and Feihong Yu, "Optimization design of hybrid Fresnel-based concentrator for generating
uniformity irradiance with the broad solar spectrum," Optics & Laser Technology. Papers 60, 27-33 (2014).
[11] PA Davies, JC Miñano, A Luque, F López, and JM Ruiz, "Calculations of efficiencies achievable using gallium
arsenide and silicon cells mounted together in cavities illuminated by single-stage concentration," Solar energy
materials and solar cells. Papers 26, 203-215 (1992).
[12] Alexis Vossier, Daniel Chemisana, Gilles Flamant, and Alain Dollet, "Very high fluxes for concentrating
photovoltaics: Considerations from simple experiments and modeling," Renewable Energy. Papers 38(1), 31-39
(2012).
[13] Edmund Optics, "Online Catalog of Precision Optical Components," edmundoptics, http://www.edmundoptics.com/
(Last retrieved on February 2014) (2013).
[14] Ghassan Zubi, José Bernal Agustín, and Gian Vincenzo Fracastoro, "High concentration photovoltaic systems
applying III–V cells," Renewable and sustainable energy reviews. Papers 13(9), 2645-2652 (2009).
[15] NANOMEDICINAFISICA, "Los nanotubos," nanomedicinafisica,
http://nanomedicinafisica.wordpress.com/2011/03/31/16/ (Last retrieved on March 2014) (2011)
[16] Takeo Oku, Rikizo Hatakeyama, Takamichi Hirata, and Noriyoshi Sato, "Formation and atomic structure of carbon
mini-nanotube," Physica B: Condensed Matter. Papers 323(1), 284-286 (2002).

View publication stats