Evaluation of solar panel performance in near space

Research Group: Wes Tatum, Chad Thomas, Linnea Zavala

Launch: Whitworth Fall 2014

This experiment evaluates the performance of monocrystalline and polycrystalline silicon based solar cells with respect to altitude and temperature. The solar cells were attached to a weather balloon and released to rise to a height of 100,000 ft before returning. The output of the cells was measured on a volts per gram basis as functions of altitude and temperature. Thermocouples were attached directly to the surfaces of each type of cell to accurately record the cell's temperature, as opposed to ambient temperature. The monocrystalline cells were found to more effectively produce energy than the polycrystalline solar cells with respect to both altitude and temperature. This information was used to determine that the optimal height for pseudo-satellites, powered by monocrystalline or polycrystalline solar cells, is 40,000 ft.

Pseudo-Satellites
In the past few decades, space has become more of a residence than a goal. Satellites and the International Space Station have proved that humanity can and is utilizing the upper reaches of earth's atmosphere. In addition to these more common and longer lasting examples there has been a recent surge of interest in high altitude drones. These drones, or "pseudo-satellites" are ultra light at are meant to fly around in the upper regions of the lower atmosphere for weeks at a time. These high altitude pseudo-satellites, space stations, and normal satellites all need energy sources that will not compromise their weight efficiency while fully supplying their vital technologies and payloads. Solar cells are a light-weight means to continually fulfill the energy quotas required by the high altitude technologies. However, there are many different types of solar cells, each with varying efficiencies. What is more, these different solar cell technologies have different physical characteristics and reactions to physical environments. These cells can be compared easily at ground conditions, but there is no way to predict how each type of cell will behave at high altitudes, which has little to no atmospheric shielding of light. Similar projects from other universities have found peak voltages for solar panels at about 50,000 ft, although the weather balloon reaches around 100,000 ft. Different altitudes have varying temperatures, which lead to different energy efficiency levels and can hurt some types of materials, while helping others out.

Atmospheric Temperature
The Earth's atmosphere is made up of four different layers, each with different temperature behaviors (see Fig. 1). The troposphere, approximately 0 to 12 km, is the inner-most layer and is where the majority of weather occurs. The temperature inside the troposphere is linear, decreasing with the increase in altitude. The stratosphere, ranging from 12 km to approximately 50 km, contains the ozone layer. Because the ozone layer absorbs ultraviolet rays, temperature increases with altitude starting around 20 km up until the next atmospheric layer. The mesosphere ranges from 50 km to 80 km and temperatures drop linearly with increasing altitude. The outer-most layer, the thermosphere, ranges from 80 km to approximately 750 km.

A decrease in temperature leads to a decrease in resistance of electrical components. Because of this, solar panels tend to be more efficient in energy production as the temperature decreases. Knowledge of the atmospheric layers and how the temperature changes is crucial in determining the optimal altitude for the efficiency of solar panels.

Objective
The objective of this experiment was to compare monocrystalline and polycrystalline, silicon based solar cells and their energy efficiency in altitudes reaching as high as 100,000 ft. The primary metric for the comparison of these two types of solar cells was volts generated per gram of the cell as altitude increased and temperature varied. This metric is because of the movement towards lighter and more efficient means of energy production that are required by the developing space and near space pseudo-satellites.

We hypothesize that the voltage output will increase as temperature decreases and altitude increases because electronics typically function better at lower temperature and because of thinning of atmosphere. Second, we hypothesize that the monocrystalline solar cells will have more efficient voltage production because of their single crystal structure, which has less boundaries to be crossed by the electrons that in polycrystalline cells.

Equipment
Two types of solar panels were chosen for comparison. As the first solar panel experiment for Whitworth Near Space we decided a simple comparison between two durable solar cell types would be most effective. Polycrystalline and monocrystalline cells were chosen. Panels from Futurlec were chosen with criteria of similar peak voltages and durable construction (see Equipment List). The 3.5 V, 5.40 cm x 4.30 cm polycrystalline cells were chosen as well as the 3.88 V, 4.00 cm x 2.50 cm monocrystalline cells.

The other key equipment decision was how to determine the temperature of the actual panel and thus its performance as it cooled/heated. We ended up using two temperature sensors by National Semiconductor which had been used in a previous Whitworth Near Space experiment. Thermoconductivity grease from Omega was used to contact the detectors to the edge of the panels. Details of equipment purchased is provide in the Equipment List table below. Parts whose price is not listed were already owned by the Whitworth University Physics Department.



Mechanical
The panels were attached to the standard, approximately 18 cm x 15 cm x 15 cm styrofoam "pod." Holes were punched into the pod so the wires we soldered to the panels could be threaded to the interior. For additional stability, the cells were also held to the outside surface of the pod with epoxy. The cells were arranged so that 2 polycrystalline and 2 monocrystalline were on each of the 4 sides of the pod (see Fig. 2). By this arrangement, we hoped to evenly distribute the amount of light reaching each type of cell.

The 2 temperature sensors were placed on a master component provided by the Whitworth Physics Department. This master component had a switch that allowed us to turn on the equipment from outside the pod. A 9 V battery connected to the switch and powered the temperature sensors. Holes were cut in the pod so that both switch and temperature sensors could be secured with electrical tape on the outside of the pod. The sensors were placed on the same side of the pod (see Fig. 3). They were placed in contact with their respective panel, secured with electrical tape, and then thermoconductivity grease was applied so contact was ensured and stabilized. The transmitting module and the battery were secured in the interior by zip-ties.



Solar Panels
The solar panels were split into two different circuits, one for the 8 monocrystalline solar panels and another for the 8 polycrystalline solar panels. As seen in the Fig. 5, the solar panels are all connected in parallel to a 220 &Omega; resistor. The voltage produced was measured across the resistor and divided by 8 to receive the average voltage. The circuits for the monocrystalline and the polycrystalline solar panels are the same in layout.

Temperature Sensors
The temperature sensors were both connected to separate terminals in the master component which turned the temperature readings into voltages for the command module (for more information on how the voltage readings were converted to temperature see Testing). This circuit was powered by a 9 V battery and had an on/off switch which was placed on the outside of the box for easy access.

Testing
Testing ensured that each solar cell was working throughout the construction process. After soldering each panel, the polycrystalline cells read about 1.5 V in interior ambient light and 2.0 V when a cell phone light was shined on it. The monocrystalline panels were less consistent with an approximate range of 1.7 V with ambient light and 2.2 V with a cell phone light. This test was completed because of concern that soldering may damage the cells. However, we found reasonable voltage consistency across the cells post-soldering. Next, at several stages during the process we checked that each panel worked by covering one at time and seeing a change in voltage.

Temperature testing was important to develop a way to read temperature from voltage for the two different detectors. Two readings were taken, one at 0.7 ºC and the other at 95.6 ºC. These temperatures were chosen by heating/cooling water until a laboratory thermocouple read a consistent temperature. Then, the voltage values were read off of the weather balloon software. The Channel 7 detector, measuring the temperature for the monocrystalline panels, read 2.7617 V at 0.7 ºC and 3.7295 V at 95.6 ºC. The Channel 8 detector, measuring the temperature for the polycrystalline panels, read 2.7862 V at 0.7 C and 3.8615 V at 95.6 ºC. Thus the temperature can be determined by T=98.1v-270.1 for the monocrystalline solar panels and by T=88.3v-245.2 for the polycrystalline solar panels where v is the voltage read and T is the temperature in degrees Celsius.

Temperature
The initial hypothesis was that the efficiency of the solar cells would increase as they became colder. This was based off of the well known fact that electronic components function better at lower temperatures. Additionally, in the p-n junction of the solar cells, there will be more electrons in their ground state, or conduction band. It was also posed that the monocrystalline cells would have a greater increase in efficiency as the temperature dropped. This is because there is only one crystal and so there are fewer grain boundaries for the electrons to travel through, meaning there is less distortion of the semiconductor material. We also thought that the many different crystals in the polycrystalline cells might tend to shrink into themselves, rather than into the center of mass of the cell as the monocrystalline cells would. This would mean that gaps in the cell could occur, decreasing efficiency. As the cells rose through the atmosphere, the temperature followed a linear trend until the balloon reached a height of about 32,500 ft (see Fig. 6). After this height, the temperature increased. This increase is likely because the balloon entered the ozone layer, which is warmer because of ultraviolet light absorption. The efficiency of the solar cells was calculated on a volts generated per gram of solar cell basis and plotted against the temperature (see Fig. 7).





For clarity, the graph in Fig. 7 includes only data up to a height of 32,570 ft because data in the ozone layer was redundant as the cells warmed up again. As is visible in Fig. 6 and Fig. 7, the monocrystalline solar cell array had a much steeper increase in efficiency than the polycrystalline did (to the extent that the voltage went above 5 V, which cannot be read by the module). The array was created with the cells in parallel to each other and the resistor was chosen specifically so that maxing out would not occur, based on the factory supplied specs of the cells. However, the array did max out and thus we know the efficiency of the cells increased so greatly that the maximum output voltage they were capable of according to the distributor increased (refer to Equipment List table for distributor website). The efficiency of the monocrystalline confirms our original hypothesis, though went to a higher voltage than can be read by our equipment.

Altitude
We hypothesized that the output of solar cells would increase as their altitude increased. This was based on the fact that there is less atmosphere between the solar cell and the sun. Atmosphere filters out the light through processes such as Rayleigh Scattering, which elastically scatters light. In near space, where this effect is much less than at ground level, more of the initial energy sent out from the sun should be able to be captured and harvested by the solar array. In looking at the graphs of efficiency versus altitude (Fig. 8), this is partially confirmed.



The direct relationship between output and altitude is apparent in the monocrystalline array, but cannot be completely confirmed, due to the max-out. The polycrystalline array partially also confirms this hypothesis. There is an increase in efficiency of the solar cells up to a height of about 32,500 feet. After this height, there is a small decrease in efficiency as the balloon continues to rise. This means that the negative effects on efficiency from the increasing temperature (as a result of the ozone layer) outweighed the positive effects that could be attributed to the altitude. To reference raw data as well as processed data and various plots see the below excel file.



Conclusions
Our hypothesis was, in general, confirmed. However, we did not predict that there would be significant enough warming in the ozone layer to warm the cells enough to decrease output. Overall, we were correct in that output increased with temperature decrease and with altitude, but the cells were more dependent on temperature changes than altitude, largely because temperature is dependent on altitude. It was also confirmed that the monocrystalline cells were more efficient. The monocrystalline cellse were found to have the best voltage per gram efficiency, which is optimized at about 40,000 ft.

Future Research
One way that this research can expand would be to test the efficiency of the monocrystalline and polycrystalline solar cells with respect to the light spectrum. The semiconductors within these cells have different band gaps, which are characteristic of their conduction and valence bands. These band gaps will absorb different wavelengths of light. As the solar panels reach higher altitudes, the prevalence of different wavelengths of the light will change as the atmospheric filtering changes, possibly affecting the voltage output of the solar cells. This experiment can be done by attaching several photodiodes of different wavelengths and using them to measure the strongest wavelength of light hitting the solar panels. That data could be graphed with the voltage output of the solar cells to compare the relative effect it has.

Now that we have an idea of effects of altitude and temperature, more extensive ground/baseline testing could also be carried out to see the effects that temperature has on the performance of the cells with ground level atmospheric conditions, so that they could be more accurately compared to near space atmospheric conditions. This could include immersing the cells in dry ice or liquid nitrogen while measuring the output voltages of a constant light source. This would more closely simulate the temperatures of space.

Finally, though general solar knowledge predicts that monocrystalline and polycrystalline cells will likely be the most efficient, it may be worthwhile to continue the research and test other types of panels.