Multi-Period AC-DC Distribution Network Expansion Planning: A Probabilistic Approach

Abstract

In this paper, the probabilistic evaluation of the AC-DC distribution network planning has been discussed. Uncertainty in load demand and the power of renewable distribution generation (DG) leads to the probabilistic behavior of the network. This behavior causes the violation of bus voltage and line loading constraints. On the other hand, violation of the constraints is also the main factor of risk in the problem. Therefore, risk factors have been identified in the problem. The existing uncertainties are modeled through the clustering method. The scenarios with the highest cost are defined as severe events. For risk management, the expected shortfall (ES) index has been used to obtain the cost of severe events. Using the multi-period method, long-term planning has been done. Therefore, the entire planning horizon is divided into several consecutive periods. The optimization problem has been solved concurrently for all periods. Therefore, the type of existing/new buses (AC or DC), the type of existing/new network lines and the optimal scheme of the network have been determined for all periods. The objectives of the problem are the minimization of the total planning costs (investment and operation costs) and the cost of severe events for all periods. Finally, the proposed method has been implemented on a sample expansion network and the results have been analyzed.