Integration of clustering analysis and reward/penalty mechanisms for regulating service reliability in distribution systems

This study proposes an approach for improving service reliability in the distribution network based on establishing competition among electric distribution utilities. The idea behind this approach is to categorise the utilities and compare the performance of utilities located in one cluster with the other members of the same cluster. The reward/penalty mechanism (RPM) as a quality regulating instrument is designed for each cluster and used to penalise the utilities with a performance worse than the benchmark and to reward them for providing a performance better than the benchmark. Based on the RPM, utilities located in one cluster compete to make more profit by serving customers in better reliability level. The setting of a RPM in an ex-post manner is also proposed to provide incentives for utilities to improve their reliability in an active manner. The system reliability indices are utilised to measure the utilities performance. The proposed method is applied to 39 electric distribution utilities in Iran. The results show the effectiveness of combining clustering analysis with a reward/penalty instrument in reliability improvement.