Methods for Quality Control of Monitoring Data from Commercial PV Systems

The aim of this work is to develop and test new methods for quality control of data from commercial monitoring systems for small and medium sized PV installations. Such installations often have limited or non-existent maintenance of their monitoring systems. Quality issues in e.g. irradiance and temperature measurements will cause errors in the analysis of the PV system performance and might lead to non-optimal maintenance of the system. To determine the condition of the sensors and the monitoring system based on the measured data itself is therefore essential to improve performance analysis algorithms and to understand historical data from these types of PV systems, and consequently this is of significant economical and practical value. In this work, we use data from both commercial and research systems in Norway to assess the robustness of the methods in a real-world scenario. We demonstrate that drift and deviations in the sensitivity of irradiance sensors, in addition to misalignment of the sensors, can be accurately quantified and detected based on comparison with clear sky irradiance modeling. Furthermore, we show that analysis of temperature data potentially can be used to detect snow cover of modules, in addition to identification of detached temperature sensors.