Using modern IT to improve distribution planning substation siting optimization - a new approach

Abstract

Power system distribution substation siting is one of the more complex optimization problems found in practice. In its most complete form, the problem reduces to a nonlinear integer-programming problem. The problem is to determine the optimal siting and sizing of new distribution substations and feeders for a given electric service area knowing the ultimate load density and location/sizes of existing substations. The cost function includes the capital costs of new equipment, the losses in the network and O&M costs. The constraints range from load constraints and voltage constraints to reliability and substation/feeder capacity constraints. The decision variables include installation/upgradation of substations as well as adding feeders at different voltage levels and/or capacities. This is a hard problem and currently no algorithms exist to solve it. The solution is achieved by siting of the substation as a continuous variable using a nine-point substation loading model and quadratic solver. This presentation discusses a new algorithm for the optimal siting of distribution substations to determine the number, location and size of future substations and feeder capacity and/or voltage class. The problem is formulated. The parameters, decision variables and cost functions are explained. One of the crucial parts of the problem is the solution of the I = YV system. Having laid the foundation of the problem, a mathematical programming formulation of the full-blown problem is given. Finally, some results are given for a typical electric distribution system substation siting optimization including 3-D graphical representation of initial and final area voltage profiles, and substation siting/sizing.