Multi-dimensional customer segmentation model for power system reliability-worth analysis

• We segment customers using turnover, economic activity and electricity consumption.
• Duration and time of occurrence of power interruptions was included in the analysis.
• A hierarchical clustering technique was used to perform the customer segmentation analysis.
• We examine changes in the coefficient of variation of the final CIC estimates for different customer segments formed.
• The customer segmentation technique results in a reduction of the dispersion of the final CIC estimates and thus allow CICs to be estimated from smaller survey samples.

Accurate assessment of customer interruption costs is essential for making informed decisions in power system planning and operation. This paper presents a multi-dimensional customer segmentation model for reliability-worth analysis of power systems. The proposed model uses a hierarchical clustering technique to cluster electricity customers into customer segments of similar cost characteristics. Three customer parameters – economic size, economic activity and energy consumption are used in the proposed model. The proposed model is examined on two case studies from South Africa and Sweden, and results are compared to the conventional customer segmentation models. The effectiveness of the proposed model is determined based on the coefficient of variation of the final CIC estimates for different duration and time of occurrence of power interruptions. The results show a reduction in the dispersion of the final CIC estimates and thus allow CICs to be estimated from smaller survey samples.