SPLTRAK Abstract Submission
Solar Power Analysis and Simulation for Shade Detection
David J. Florez Rodriguez1& Bennet E. Meyers1,2
1Stanford University, Stanford, CA, United States
/2SLAC National Accelerator Laboratory, Menlo Park, CA, United States

We model the AC power signal of distributed residential solar systems with missing system specifications and meteorological data. The models estimate losses due to shading. The goal is validating an algorithm that can automatically quantify these losses. We infer the necessary specifications and model necessary irradiance values from site coordinates. This information, along with the available power signal allows the construction of a `clear sky' model of the residential power signal via the PVLib python library. We select only `clear sky', error-free days from the data to compare to the modelled power signals. On this data, we detect and correct for soiling losses. Free of weather, malfunctions, and soiling losses, the data’s primary difference from the model are the shading losses. This difference is our estimate for a site’s present shade. This metric can then estimate the potential energy available from a site and provide an empirical approximation for the annual energy lost to shade at an installation.