Achieving the Dispatchability of Distribution Feeders through Prosumers Data Driven Forecasting and Model Predictive Control of Electrochemical Storage

TL;DR

This paper presents a data-driven, model predictive control strategy for dispatching distribution feeders using battery energy storage, validated through experiments on a university campus feeder to improve operational flexibility.

Contribution

It introduces a novel two-stage control framework combining day-ahead dispatch planning with real-time MPC, utilizing adaptive forecasting and minimal monitoring infrastructure.

Findings

Successful experimental validation on a real campus feeder.

Effective dispatch of feeder operation using BESS with adaptive forecasting.

Improved operational control with minimal infrastructure requirements.

Abstract

We propose and experimentally validate a control strategy to dispatch the operation of a distribution feeder interfacing heterogeneous prosumers by using a grid-connected battery energy storage system (BESS) as a controllable element coupled with a minimally invasive monitoring infrastructure. It consists in a two-stage procedure: day-ahead dispatch planning, where the feeder 5-minute average power consumption trajectory for the next day of operation (called \emph{dispatch plan}) is determined, and intra-day/real-time operation, where the mismatch with respect to the \emph{dispatch plan} is corrected by applying receding horizon model predictive control (MPC) to decide the BESS charging/discharging profile while accounting for operational constraints. The consumption forecast necessary to compute the \emph{dispatch plan} and the battery model for the MPC algorithm are built by applying adaptive data driven methodologies. The discussed control framework currently operates on a daily basis to dispatch the operation of a 20~kV feeder of the EPFL university campus using a 750~kW/500~kWh lithium titanate BESS.