Performance Evaluation of Silicon-Based Irradiance Sensors Versus Thermopile Pyranometer

Abstract

There are several sensors available in the market to measure the plane-of-array irradiance for photovoltaic applications. The prices of these sensors vary according to the design, calibration procedure, and conducted characterization. In this article, two types of silicon-based sensors with and without temperature correction capabilities are compared with a high-accuracy thermopile pyranometer to check their performance. The obtained results showed that silicon-based sensors deviate from the output of the pyranometers. The tested silicon-based pyranometers overestimate the irradiance with the median bias deviations of around 1.43% (with the average measured irradiance of 256 W/m2). For temperature-corrected silicon pyranometer, the bias deviation is 0.07% with the deviation range of -6.5%-10% (with the average measured irradiance of 257 W/m2). A working-class reference cell was also tested, resulting in a bias deviation of -1.74% and the deviation range of -13%-7% (with the average measured irradiance of 304 W/m2). The effect of air mass on the performance of cost-effective sensors was additionally analyzed. Within the measurement time window, the result also showed that for the silicon-based sensors under tests, the effects of the environmental conditions have the following qualitative order of influence: angle of incidence > red-shift > temperature. The performance of silicon-based sensors also showed seasonal dependence, being more accurate during summertime and wintertime, respectively, for the silicon pyranometer and the working-class reference cell. Finally, using the statistical evaluation, simple linear correction functions are introduced for silicon-based sensors.