Application of Bayesian networks for risk assessment in electricity distribution system maintenance management

Electricity distribution companies regard a risk-based approach as a good philosophy to address challenges of asset management, and there is an increasing trend on developing maintenance strategies where different aspects of risks are taken into consideration. This paper illustrates the use of Bayesian networks (BNs) to model safety risk as a part of maintenance decision making. A case study is included illustrating the use of BNs to investigate the risk impact of different maintenance intervals of MV electrical switches. The case study uses input from expert judgment to establish the qualitative relations and quantitative figures in the BN. The paper summarises some challenges and benefits of using BNs as a part of distribution system maintenance management.