Invited Presentation of Special Session 4: An AC-Feasible Linear Model in Distribution Networks with Energy Storage

Abstract

With the increasing deployment of distributed energy resources (DERs), dispatching DERs subject to operational constraints in distribution networks draws much attention. One challenge is the non-convexities in 1) system-wide AC power flow constraints and 2) the individual complementarity constraint of energy storage. To resolve this challenge, this presentation studies an AC-feasible linear model in distribution networks with energy storage, including its formulation, analysis and application. Firstly, an AC-feasible linear model is formulated as a set of linear constraints on controllable DERs and uncontrollable power demand by 1) converting the non-convex system-wide constraints into linear constraints constructed based on the Brouwer's fixedpoint theorem and the second-order Taylor expansion, and 2) replacing the non-convex individual complementarity constraint of energy storage with one designed linear constraint. Furthermore, to analyze the power demand level at which the proposed linear model can provide a solution, this presentation proposes an examination-based projection method under the Monte Carlo framework to handle projections of thousands of dimensions of the linear constraints over time periods. Finally, the potential applications (e.g., zero uplift payments in markets with ACfeasibility) are discussed. Numerical experiments are conducted in the IEEE 33-bus and 136-bus test systems to demonstrate the proposed methods.