I have found that there is some confusion concerning how to interpret the power rating of Photovoltaic (PV) modules (panels). We often hear about “cost per watt.” It seems that one of the first questions that gets asked about a PV installation is how many kilowatts does it produce. However, there is a bit of a problem with this because there are several different answers to the question of how to rate the output of PV modules. It all depends upon how the watts are measured.
PV modules are rated by PV manufactures in terms of watts per module. This rating is usually based upon what is referred to as “Standard Test Conditions” (or STC for short). This is an industrial test standard performed at a cell temperature of 25°C (77°F), 1000 watts per square meter and a 1.5 Air Mass. The measurement is made by flashing a calibrated light at a temperature controlled module and measuring the output voltage and current, which is used to calculate power (watts). The flash lamp is designed to closely match the spectrum and intensity of the sun on a clear, sunny day at sea level in the mid-latitudes. This is a useful measurement since it provides a repeatable and comparable test for comparison purposes. However, it does not represent an output that would be anticipated by any real system.
A second method of rating PV panels was developed in Davis, California at a large test and demonstration installation called Photovoltaics for Utility Scale Applications (PVUSA). This rating is called the PTC (PVUSA Test Conditions) rating. The PTC rating is designed to represent a "more real life condition" of 1,000 watts per square meter solar irradiance, 1.5 Air Mass, and 20°C ambient temperature measured at 10 meters above ground level and a wind speed of 1 meter per second. The PTC rating is lower than the STC rating.
Both of these ratings are at the module level, and do not include reductions in power caused by soiling, shading, module mismatch, wire losses, inverter and transformer losses, shortfalls in actual nameplate ratings, panel degradation over time, and high-temperature losses for arrays mounted close to or integrated within a roofline. These loss factors can vary by season, geographic location, mounting technique, azimuth, and array tilt. The California Energy Commission (CEC) provides an on-line calculator that attempts to predict system performance based upon the entire system, including derating factors to account for several of these conditions. The CEC rating is used to predict overall system ratings in watts, and annual performance in kWhrs per year, for a specific system design at a specific geographic location. The CEC rating is used to calculate the incentive rebate provided to offset the high cost of new installations. (Currently, the CEC incentive in PG&E’s territory is $0.35 per CEC watt.)
The CEC incentive calculator is based upon a simplified simulation model called PVWatts. The estimations of module performance using the CEC incentive calculator is only marginally accurate, and there are no provisions for adjusting derating factors such as wire size and length. Generic derating factors are used, which gives a quick and easy calculation used to determine the amount of incentive payment, but that does not necessarily predict the actual performance of an installed system. The CEC web site is very specific that the values provided are only intended for determining incentive payments; they are not considered adequate for system design purposes.
There are a number of computer simulation programs available that provide a much better prediction of production in the field. Two of my favorites are “PV Design Pro” by Maui Solar Energy Software Corporation and System Advisor Model (SAM) created by the National Renewable Energy Laboratory (NREL). Both of these models are based upon decades of research by Sandia and NREL. The differences between the two models are mainly in presentation and user interface. Both require detailed test characterization of each type of PV module for maximum accuracy. Unfortunately, not all manufacturers provide the details that are required for these more accurate simulations, meaning that the only choice is to revert to the simplified PVWatts model. PV Design Pro does not provide the ability to select this less accurate model from within the software, whereas SAM does. Both programs provide pull-down pick lists allowing the user to select make and model of PV module to be used. Both allow for a much more detailed selection and optimization of derating factors. Both models provide hour-by-hour simulations based upon local weather conditions available from drop down selection lists or information provided by the user.
The system rating as predicted by these more sophisticated (and accurate) software programs differs from each of the other ratings. So do you want to talk about STC, PTC, CEC or Sandia ratings? It is my opinion that the Sandia ratings are closest to the actual situation when available. If not, then the CEC ratings as calculated by SAM are probably the next best rating. However, in almost all cases people talk about the STC ratings of their systems because that is the larger number and is the value most often quoted when talking about system costs in terms of dollars per watt.
There is one more little trick to this whole rating business, and that has to do with system degradation over time. It is known that PV modules degrade over time, but discussions concerning the amount of degradation to expect vary by more than ten times, depending upon who is making the claim.
Manufacturers typically guarantee there modules to degrade less than 1% a year for the first 25 years. While this might seem like a small number, it ends up being 25% in 25 years (the length of the warranty). That would mean that a 10,000 watt system would only be providing 7,500 watts at the end of the period – a huge reduction in power!
However, that number seems to be wildly exaggerated. In discussions with solar personnel at Sandia and reading a lot of scientific literature on the subject, it appears that the degradation is much less than that. For mono or poly crystalline silicon (the most popular substrates today), the degradation is typically less then 1% the first year, and near zero after that. In other words, there is often a bit of early degradation, but as long as the modules do not become damaged, they tend to be stable for the rest of their service life. The guarantee of less than 1% a year is very conservative, the manufacturers can be assured that they will not have to pay off on this promise.
