A new parallel and decomposition approach to solve the medium- and low-voltage planning of large-scale power distribution systems

The world energy demand is increasing. Currently, most electricity generation sources are fossil fuels. In recent years, the global energy matrix for power generation is changing, in order to meet the demand with minimal environmental impacts. In this context, a new approach to solve the medium- (MV) and low- voltage (LV) planning of large-scale distribution systems via parallel computing and decomposition techniques is proposed. The mathematical formulation considers the allocation of substations, distribution transformers, MV and LV circuits, support structures, poles, renewable energy sources (RES), and energy storage sources (ESS). In addition, variable costs related to the operation of RES and ESS, power losses in cables, distribution transformers and substations, energy purchased from substations, and greenhouse gas emissions are also taken into account. System reliability and maintenance costs of these devices are also considered in the planning. To evaluate the new methodology performance, tests in a large-scale distribution system with 200 nodes in MV and 1672 nodes in LV is considered. Numerical results show the proposed methodology is able to find good solutions that guarantee the minimization of planning costs considering RES and ESS allocation. Furthermore, system reliability is improved by up to 22% and greenhouse gas emissions mitigation by up to 18%.

Keywords

Energy storage sources, Greenhouse gas emissions, MV/LV distribution system planning, Parallel variable neighborhood decomposition search, Renewable energy sources