Novel Optimization Model For Long Range Distribution Planning

Abstract

A new model for both static and time-phased long range planning of large distribution systems is presented. The model takes into account both the fixed and the variable costs for all planned facilities (substations and feeders) as well as the cost of the power losses. The planning model incorporates an accurate formulation of the continuous non-linear cost function in terms of the planning variables (time and power flow through the expanding network). The model also gives the possibility of expanding the capacity of an existing feeder or a substation instead of constructing a new one. The formulation of the problem includes constraints on voltage drop in primary feeders, demand satisfaction constraints, power flow conservation constraints and constraints on feeders and substations overloading. In the present non-discrete model the number of the planning variables is reduced compared to the discrete models used hitherto and thus it is applicable to large distribution systems. Also because of the continuous nature of the formulation, it is suitable for non-linear applications such as contingency and reliability studies. The model has been applied to a section of an electric power utility (Ontario Hydro) distribution system. The results are compared to the findings of the utility earlier application of an optimization model.