Where Did the Sun Go? A Quick Look at Solar Panel Degradation in Massachusetts

Written by: Eric Pinsker-Smith, Senior Analyst; and Toby Armstrong, Principal Analyst

Publish Date: May 9, 2022

Estimated Reading Time: 3 Minutes

It is an unfortunate inevitability that the production from solar projects degrades with time. This does, however, provide a most fortunate opportunity to write another blog. Photovoltaic (PV) module degradation is a well-studied topic and is caused by several factors that include the handling of PV modules during installation, the maintenance and upkeep of solar arrays, the weather and irradiance that panels are exposed to, and chemical degradation of PV cells themselves. However, the overall production degradation that solar projects experience is the product of more than just PV module degradation. Assessing of the “all in” degradation rate of solar projects is an emerging field of study, which has largely revealed that the industry standard degradation rate of around 0.5% (intended to capture module degradation) yields overly optimistic outlooks when compared to in-practice performance.

In support of our Massachusetts Solar Study (which provides fundamental supply/demand price forecasting for the MA SREC I and II markets), we were curious to examine the true degradation rates of projects operating in the real world to refine our supply forecasts in a market where a small change in the supply/demand balance can have a large impact on market prices. So, we started the way we always do…by gathering data. Massachusetts’ Department of Energy Resources provided us with anonymized, monthly production actuals from over 90,000 solar facilities in Massachusetts that were in operation between 2010 and 2019. We cleaned the data and binned it into three categories:

  1. Small – projects with a generating capacity of less than 25 kW;
  2. Medium – projects with a generating capacity between 25 kW and 1 MW; and
  3. Large – projects with a generating capacity exceeding 1 MW.

Each project’s production was tracked over its operating life, less the first year of production to prevent mid-year commercial operation dates from biasing our analysis. We also adjusted each year’s production to account for irregular irradiance (using data provided by NASA’s Prediction of Worldwide Energy Resources project). For example, production in years that had 10% less sunshine (i.e., irradiance) vs. the 20-year average were given a 10% upwards adjustment to their production, to ensure that the analysis was “weather neutral” and was not biased by especially sunny or cloudy years.

The average annual degradation rates we calculated are tabulated as follows:

We then compared these results to the results of a meta-analysis undertaken by kWh Analytics, whose results are tabulated as follows:

Our results were largely consistent with those of the latest meta-analyses: degradation seems to be more pronounced for smaller systems, and degradation estimates that exceed 1% (annually) appear to better represent (if somewhat underestimate) real-world PV degradation rates. These results reinforce that degradation is likely a function of uncontrollable factors (such as PV module degradation) and controllable factors (such as ideal O&M practices), which contribute to differing degradation rates across project sizes.

So, what is one to do with this information? First, if you’re policy and market analysts like us, one would be best served to assume that degradation rates will be higher than 0.5%.  Secondly, one can be a more informed consumer; although several PV manufacturers and installers advertise annual degradation rates of around 0.5%, it does not seem, based on our analysis, that this outcome is likely in practice. Thirdly, one can consider these PV degradation numbers when planning the economics, financing, and system specifications of a PV solar project. It is important to note, however, that as solar technology improves, module degradation rates are expected to fall, but it is unclear how much real-world O&M will improve, so how far all-in degradation rates will recede in practice remains to be seen.

REMO