Transformer Load Estimation Using Smart Meter Data in Taipower

Abstract

In this paper, an innovative transformer load estimation methodology is proposed to determine the daily power profile of distribution transformers. A two-stage load estimator algorithm that combines customer load pattern from smart meters and state estimation (SE) theory is developed to provide the secondary output power of distribution transformers in the low voltage network at a sound accuracy with customer load measurements. The estimated results that could reflect true transformer loading then are used for transformer load management. A 5-bus test system has been analyzed to show the good convergence characteristics of the developed estimator algorithm and to demonstrate the effectiveness of proposed methodology. The method proposed in this paper can effectively assist electric utilities for the evaluation of loading factors of distribution transformers to prevent the burn out of transformers caused by overloading problem. The operation efficiency of distribution systems can also be enhanced by loading balance of distribution transformers.