Iman Goroohi Sardou

Network expansion planning (NEP) is usually executed by power system planner to determine the capacity and location of the new substations and lines throughout the future years. In this study, an effective bi-objective robust model is proposed for the NEP problem considering the integration of the microgrid aggregators. The objectives include minimisation of both the expansion cost and transmission lines loading index. Forced outages of system components (transformers and lines) are taken into account as the system uncertainties. A hybrid method as the combination of gravitational search algorithm (GSA) and primal-dual interior point (PDIP) method is employed to solve the non-linear programming (NLP) problem of the NEP. In the proposed hybrid method, the NLP operation sub-problems are solved by the PDIP method under the worst-case single component contingencies, while the expansion plan is defined as scenario independent variables (first stage decision variables) obtained by the GSA. To detect the worst-case single component contingencies with the severest effects on the system security, a subsidiary optimisation problem is solved for each load level of the system. A realistic network of Qom as a part of Tehran Regional Electric Company, Iran, as well as IEEE 118-bus test system is analysed to evaluate the efficiency of the proposed method. A key conclusion is that the stochastic model may not provide sufficient security level to find a more economical solution, and a robust model based on the worst-case single component contingencies is required to ensure the system security against the probable severe contingencies.