Including a DC network approximation in a multiarea probabilistic production costing model

A multiarea power system consists of several areas (subsystems) interconnected by a transmission network. In estimating expected generation costs for such systems, transmission capacity limits of the network should be recognized. Transportation network models have generally been used because of their simplicity, but they only enforce Kirchhoff's current law. AC power flow modeling of the transmission network, which recognizes thermal, voltage and stability constraints, is theoretically best, but is too unwieldy for assessing expected costs. The so-called "DC" linearized network model is adopted here as a compromise, as it enforces both Kirchhoff's current and voltage laws while its linearity facilitates incorporation in probabilistic production costing models. In this paper, we generalize a bounding-based multiarea probabilistic production costing model to include loop flow and resistance losses based on the DC network model. This is the first multiarea model based on efficient convolution methods for production costing that also includes loop flows and resistance losses. Computational examples are presented to highlight the modeling and solution procedures.