Distributed utility planning using probabilistic production costing and generalized benders decomposition

Abstract

Regulatory changes and advances in distributed resources (DR) technology have lead utilities to consider DRs as alternatives to central station generation and T&D investments. This paper presents a comprehensive planning and production simulation model that simultaneously evaluates central and local investments to determine the optimal mix for long-term expansion. The model can also be viewed as optimizing DRs while simulating a perfectly competitive wholesale power market. The model is a mixed integer linear stochastic program that enforces Kirchhoff's current and voltage laws and is solved using generalized Benders decomposition (GBD). The formulation includes multiarea probabilistic production costing as a subproblem. DRs and local distribution reinforcements are modeled as integer variables, while transmission and central generation options are represented as continuous variables. The model is applied to a ten-year multi-area example that suggests that DRs are able to modify capacity additions and production costs by changing demand and power flows.