Robust Estimation of Reactive Power for an Active Distribution System

TL;DR

This paper introduces a robust optimization method to accurately estimate the reactive power potential of active distribution systems with DERs, ensuring reliability despite uncertainties in network conditions.

Contribution

It proposes a two-stage robust optimization approach with a tractable algorithm to reliably estimate reactive power potential considering uncertainties.

Findings

The method provides a completely reliable reactive power potential estimate.

Active distribution systems can effectively serve as reactive power prosumers under uncertainty.

The approach outperforms existing methods in reliability and robustness.

Abstract

Increasing distributed energy resources (DERs) may result in reactive power imbalance in a transmission power system (TPS). An active distribution power system (DPS) having DERs reportedly can work as a reactive power prosumer to help balance the reactive power in the TPS. The reactive power potential (RPP) of a DPS, which is the range between the maximal inductive and capacitive reactive power the DPS can reliably provide, should be accurately estimated. However, an accurate estimation is difficult because of the network constraints, mixed discrete and continuous variables, and the nonnegligible uncertainty in the DPS. To solve this problem, this paper proposes a robust RPP estimation method based on two-stage robust optimization, where the uncertainty in DERs and the boundary-bus voltage is considered. In this two-stage robust model, the RPP is pre-estimated in the first stage and its robust feasibility for any possible instance of the uncertainty is checked via a tractable problem in the second stage. The column-and-constraint generation algorithm is adopted, which solves this model in finite iterations. Case studies show that this robust method excels in yielding a completely reliable RPP, and also that a DPS, even under the uncertainty, is still an effective reactive power prosumer for the TPS.