Data-Driven Distributionally Robust Optimal Power Flow for Distribution Systems

TL;DR

This paper introduces a data-driven, distributionally robust optimal power flow method for distribution systems that accounts for uncertainties in distributed energy resources using a tractable AC power flow approximation.

Contribution

It develops a novel AC OPF formulation based on LinDistFlow and distributionally robust optimization, specifically tailored for radial distribution systems with uncertain injections.

Findings

Model is computationally tractable and effective on IEEE test systems.

Robust decisions improve system reliability under forecast errors.

Open-source Julia implementation available.

Abstract

Increasing penetration of distributed energy resources complicate operations of electric power distribution systems by amplifying volatility of nodal power injections. On the other hand, these resources can provide additional control means to the distribution system operator (DSO). This paper takes the DSO perspective and leverages a data-driven distributionally robust decision-making framework to overcome the uncertainty of these injections and its impact on the distribution system operations. We develop an AC OPF formulation for radial distribution systems based on the LinDistFlow AC power flow approximation and exploit distributionally robust optimization to immunize the optimized decisions against uncertainty in the probabilistic models of forecast errors obtained from the available observations. The model is reformulated to be computationally tractable and tested on multiple IEEE distribution test systems. We also release the code supplement that implements the proposed model in Julia and can be used to reproduce our numerical results.