Development of time-of-use price by clustering techniques

Active demand side response (DSR) from domestic customers can generate benefits in terms of reducing energy costs for customers and shaving peak demand for distribution network operators (DNOs). However, real-time price (RTP) is considered to be too dynamic for customers to response. Also, it is infeasible for most energy storage equipment to response to variable signals, such as RTP, as they can only charge/discharge a few cycles throughout a day. Due to these constraints, time-of-use (TOU) price is a more natural price signal for DMS. This paper proposes a novel statistical method to successfully convert RTP to TOU that captures the most significant price variations without comprising too much accuracy in total energy revenue from customers. The proposed method adopts hierarchical clustering techniques to group RTP into clusters, and each settlement period is assigned to one of the clusters to form a TOU pattern. For each cluster of the TOU tariff pattern, the tariff rate is determined by keeping the total customer revenue unchanged.