Improving photovoltaic (PV) system reliability and reducing maintenance and operating costs have become important factors in increasing the competitiveness of the PV energy market. Addressing these issues requires diagnostic methods that can detect and identify the occurrence and cause of power loss in the PV system, be it external, such as shading or soiling of the PV modules; or degradation of the solar cells and balance-of-system components. This allows for performing preventive and/or reparative maintenance, thus minimizing further losses and costs.This article proposes and experimentally demonstrates for the first time a complete diagnostic methodology for PV string inverter systems, which takes advantage of the current–voltage (I–V) measurement capability of the string inverter itself to perform the fault diagnosis. This enables the detection of shading, increased series-resistance losses, and potential-induced degradation affecting the PV string, by analysing changes in its I–V curve.The diagnostic methodology is based on parameters that can be easily calculated from the shape of the PV string’s light I–V curve characteristic, making it machine-analysis friendly and suitable for implementation in the string inverter. Moreover, the dimensionless formulation of the diagnostic parameters and the application of fuzzy logic in evaluating the diagnostic rules, make this method applicable to a wide range of conventional crystalline silicon based PV systems.The design and analysis of the diagnostic parameters and logic was performed based on module-level tests on conventional crystalline silicon PV modules, and were optimized to detect even small partial shading and increase series-resistance losses. To demonstrate the practical application and operation of this method, the diagnostic parameters and rules were applied “as is” to a field test PV system consisting of a crystalline silicon based PV string and a commercial string inverter capable of measuring the I–V curve of the PV string, yielding a similar high-detection rate.
Read full abstract