Fuzzy Risk Assessment of Underground Power Distribution Network Cables Based on Geo-analytical Fault Vulnerability

Maintenance prioritization in electrical distribution systems is essential to assure high reliability and security of supply. Previous methods proposed in research do not include inspection data or laying conditions of underground cables which reduces the detail level of their fault vulnerability model, an example could be the inclusion of soil corrosion conditions in the vulnerability which may increase the fault vulnerability of cables. That is why, in this paper, we propose a risk assessment methodology for underground power cables that, also in contrast to current industry practice, models fault vulnerability of cables based on their environmental laying conditions using machine learning and geo-analytics. To realize this we also propose a gradient-based method to convert from relative to absolute fault vulnerability. We then combine fault vulnerability and fault consequences using fuzzy logic to handle modelling uncertainties. Lastly, we demonstrate the value of the proposed method by applying it to a real Danish distribution system.