Self-Thermometry of PV Panels

Kaushal Chapaneri1, Shahzada Pamir Aly1, Jim Joseph John1, Gerhard Mathiak1, Vivian Alberts1, Muhammad A. Alam2 1Research and Development Center, Dubai Electricity and Water Authority, Dubai, United Arab Emirates /2Electrical and Computer Engineering Department, Purdue University, West Lafayette, IN, United States

During field operations, it is an unfortunate reality that over time, the temperature sensors of the PV panels get detached, damaged, or miscalibrated. Since the performance and reliability of a panel are determined by panel temperature (Tmod), a complementary/alternate approach is needed. In this regard, the  Sandia model can be used to predict panel temperature, if the model parameters are calibrated by site-measured historical data. However; this is often impossible because historical data may not be available. In this study, we show that by merely incorporating the panel’s electrical efficiency (degrading over time) into the equation, the panel temperature can be predicted with precision, obviating the need for calibration. Based on an analysis of more than 100 different PV panels installed at the outdoor test facility of DEWA R&D (Dubai, UAE), we show that the generalized model anticipates the actual temperature within 3.1 ˚C (worst case). The approach offers a new paradigm where the panel serves as its own thermometer and flags when the temperature sensor malfunctions.