Modelling the degradation of condition indices

To ensure that the electricity distribution network achieves satisfactory levels of reliability and safety it is necessary to have well-founded and justifiable asset management policies in place. For many items of electricity distribution equipment, failures have been rare and inferences about future lifetimes are difficult to make. The paper argues that in such situations, importance should be given to obtaining condition information to aid asset management. A key requirement in using such data is the capability to model changes in an item´s condition and the Markov model is proposed as being particularly suitable. The application of the Markov model to a collection of oil condition data, recorded during maintenance activity, is explained. The impact of such a model in making asset management decisions is then illustrated