Risk evaluation of photovoltaic power systems: An improved failure mode and effect analysis under uncertainty

Photovoltaic (PV) power systems are confronted with many failure risks threatening operational security and leading to adverse impacts on the sustainable development. This paper identified the potential PV system failures and extended the traditional failure mode and effects analysis (FMEA) that has been questioned due to its inadequacy in industrial applications. First, the risk factor evaluation system was expanded from the original 3 indicators to 7 indicators associated with PV system failures, including personal safety, power loss, property damage, frequency of failure, visualization, maintenance expense, and technical difficulty. Second, a modified FMEA approach combining rough set theory, cloud model theory and multiplicative Multi-objective Optimization by Ratio Analysis (MULTIMOORA) method was developed. The proposed approach could not only settle various uncertainties (i.e., intrapersonal uncertainty, interpersonal uncertainty, and randomness) by integrating cloud model and rough sets, but also introduce the triple subordinate method MULTIMOORA to obtain a final risk ranking of PV system failures. Finally, the improved FMEA model was adopted to a PV system failure risk assessment case, verifying its effectiveness and superiority. It revealed that hot spots and disconnection of underground cables were the riskiest potential failures in PV systems. Essential risk mitigation measures should be implemented in advance to maintain the safety and promote the sustainable development of PV systems.