PVSAT: Remote Performance Check for Grid Connected PV Systems Using Satellite Data, Evaluation of one Year of Field-Testing

摘要

The PVSAT procedure provides users of grid-connected PV systems with an individual monthly reference yield. Comparison of this reference yield with measured yield can identify system faults. This paper describes the one-year field test of the PVSAT procedure. The field test was used to test and improve the PVSAT software and organisation; to determine the differences between measured and reference yield under field conditions; to establish the causes of these differences and to determine when a system operator or system supplier should be warned about a possible malfunction within in the system. During the one-year test phase 69 owners of PV-systems in Austria, Germany, the Netherlands and Switzerland reported their measured monthly yield and received a reference yield. The field test enabled us to identify and solve problems in the procedure. Throughout the year the normalised difference between measured and reference yield remained within a band throughout the year, but some outliers occur. An analysis of the data showed that the differences between measured and reference yields could be partially explained by system faults, nearby shading, inaccurate horizon description, inaccurate specifications of components and insufficient accuracy of the PVSAT irradiation model (during one or two months in the year). The results were, where possible, corrected for external effects. This gave a first estimate of the accuracy of the PVSAT procedure under field conditions of 15 to 20 kWh/kWp. This accuracy can be improved if the system installer describes the PV system and surroundings, the true peak power of modules is included in module specifications and a model for the shading of nearby objects is included in the PVSAT model.