Reliability-Based Planning and Partitioning of Multiple Micro-Grid Considering Demand Side Response Program

Todays, Micro-Grids (MGs) include various types of distributed energy resources such as Wind Turbines (WT), PhotoVoltaics (PVs), Energy Storage Systems (ESSs), Combined Heat and Power units (CHPs), and demand-responsive loads. The variable nature of WT and PV resources and single contingencies of CHP units jeopardize the reliability of MG's customers during operation periods. Demand-side response program manages the time-consumption pattern of responsive loads (RLs) to overcome these uncertainties. As a distribution network is often divided into multiple MGs therefore this paper proposes a bi-level reliability-based model for the planning and partitioning of them in presence of RLs. At the first level, Tie Switches (TSs) and energy resources placement are determined to meet the annual peak demand to minimize the total costs. Output results of first level feed to next level as input data until the power exchange between MGs and the upstream distribution system and also RLs participation considering the desirable risk for all MGs customers are optimally calculated to maximize multiple MGs benefit. Since the second-level outputs can affect on first-level results, a bi-level model is applied. A genetic algorithm is used to solve each level's problem. For validation, numerical studies are applied to a 25-Bus test distribution network with three MGs and five TSs. The simulation results show that MG planning in islanded mode causes more investment costs depending on the installation of DERs with higher capacities. In addition, the participation of RLs in planning islanded or grid-connected MGs leads to a significant decrease in the system investment and operation, loss, and reliability costs.