The other side of the coin: power output of PV modules at hot climate

From FSLR's website (http://www.firstsolar.com/en/Innovation/Advanced-Thin-Film-Modules): "In hot climates, solar modules spend more hours at higher temperatures. For example, on an annualized basis in the Southwestern U.S. desert, more than 90% of a PV plant’s generation will occur when module temperatures are above 25°C—where First Solar has a proven performance advantage." There is a graph in the page showing CdTe modules output 11.3% peak power as an example.

In a recent article (http://www.bloomberg.com/news/2012-04-18/solar-thin-film-panels-may-outperform-rival-technology-in-india.html) from bloomberg titled "Solar Thin-Film Panels May Outperform Rival Technology in India", an executive at the nation’s largest contractor on the developments said "the last six months for which we have data show that the performance of crystalline in hot climates is not as efficient as thin film”. The article went on to discuss the degradation issue of FSLR's CdTe panels, on which I wrote a long blog. From the words of the executive, I doubt that he is referring to the degradation issue as half year is too short period to make a claim on degradation, be it for crystalline silicon (c-Si) or thin-film modules.  His words are likely a corroboration of the FSLR's claim that TF modules output more power than c-Si in hot climate.

It is not just FSLR and the India case, there are reports from Germany and China supporting the claim that for the same nameplate systems - the systems using TF module have a higher output than those using c-Si modules. Of course, for TF system to achieve the same design power output, more panels are needed, resulting more space requirement.

All solar cells have lower conversion efficiency when working temperature increases in spite of different mechanisms. The nameplate efficiency and power output of a solar cell are measured at standard test condition (STC). STC has a cell working temperature of 25 degree C. The implied ambient temperature at STC should be lower. If there is no extra heat dissipation (use cooling fin, increase air flow etc) in the test, then the ambient T should be less than 0 degree C. When ambient temperature and solar irradiance go up, the actual cell temperature will increase, leading to a drop in power output (compared to steady cell working temperature). FSLR's claim is actually a known fact that TF cells have a lower temperature coefficient than c-Si cells. According to FSLR, its CdTe cell has temperature coefficient of -0.25 degree/C (negative sign indicate efficiency will drop when T increases), among the lowest. Here is a list of temperature coefficient for different solar cells:

multi-crystalline silicon: -0.5%

mono-crystalline silicon: -0.35-0.4%

amorphous silicon thin film: -0.2%

CIGS thin film: -0.3% (not quite linear)

Sanyo HIT: -0.34%

Readers who are interested in the actual drop of efficiency can read PVCDROM's the "Temperature Effects" section in Chapter 7 "Modules and Arrays" (http://pveducation.org/pvcdrom/modules/pv-module-temperature). One key parameter in calculating the actual cell working temperature is nominal operation cell temperature (NOCT) http://pveducation.org/pvcdrom/modules/nominal-operating-cell-temperature.

The actual power output can be estimated by
P_real = P_m * S / 1000 * [1 - λ(T_cell - 25)]
T_cell = T_ambient + S / 800 * (T_NOCT - 20)

where S - the solar radiation on the panel surface, T_ambient - the ambient temperature, T_NOCT - the Nominal Operating Cell Temperature, and λ - Maximum Power Temperature Coefficient.

As an example, when solar irradiance is 1000 w/m2 and ambient temperature is 40C, a 15% multi c-Si module will have an actual efficiency of 11.7% while a 12% CdTe module has an efficiency of 10.7%. We can see the efficiency gap narrowed at hot climate. Here the assumption is that the cell working temperature for both c-Si and CdTe are about the same. There are studies showing that the cell working temperature for c-Si and TF are similar at the same ambient condition.