A statistical methodology to identify imbalance-induced capacity wastes for LV networks

Abstract

Phase imbalance (i.e., load imbalance among the three phases) causes inefficient use of three-phase distribution assets, resulting in additional reinforcement costs (ARCs) compared to if the three phases were balanced. However, two problems remain unresolved when calculating the ARC: 1) previous ARC formulas assume that the three-phase total peak and the single-phase peak current occur simultaneously, thus causing an underestimation of the ARC; and 2) previous formulas calculate ARC as a point value, using one year's peak load. However, the point value is not credible in supporting making investment decisions on phase balancing. This is because the growth of the yearly peak load is inconsistent over the years – the growth could be positive or negative over the years for low voltage (415 V, LV) networks. To address the above problems, this paper originally develops 1) an updated ARC formula; and 2) a customised approach, named as the cluster-wise probability assessment, to deliver a distribution and confidence range of ARCs for any given LV network. This approach only requires the yearly average current and single-phase peak current, making it applicable to the majority of the LV networks that have a minimal amount of data. Case studies also identify a counterintuitive finding of how the yearly single-phase peak current and the yearly average phase current impact the ARCs.