On the Importance of Renewable Data's Spatial Dependence for Planning of Distribution Systems

Abstract

By the development of distribution network planning problems, random variables have become an important consideration as their correlation might be ignored for the simplicity. In this study, the impacts of spatial dependency/correlation in terms of distribution network planning problem are investigated by using a Monte Carlo Simulation based optimization framework. This model minimizes the investment and maintenance costs of substations, transformers, feeders, renewable generators while satisfying the chance constraints related to feeder capacities, voltage constraints, and substation capacities. In order to apply this model, an integer genetic algorithm is utilized along with the linearized load flow equations. The case studies applied through the modified 34-nodes test system show that the consideration of spatial dependence yields different investment results.