Effect of Load Models on Probabilistic Characterization of Aggregated Load Patterns

Abstract

This paper presents a comprehensive approach to probabilistic characterization of aggregated load pattern of low-voltage consumers. Different from previous methods, consumer load patterns obtained from load survey are converted to empirical cumulative density functions. The functions are then used to address stochastic nature of load pattern in any given point of distribution network. The proposed approach is adopted to investigate the effect of load models on the characterization of aggregated load patterns. In addition, a goodness-of-fit analysis has been carried out to show that the load models can significantly affect the accuracy of aggregated load modeling. It is demonstrated that the constant power (real and reactive) load model which is normally adopted in most distribution network management studies may lead to misleading results compared to the actual network. Case studies are presented and discussed with reference to a real distribution network. The results verify that the proposed method is accurate and flexible, and the voltage-dependent load model is the most promising solution for the aggregated load modeling.