An Efficient Data Driven Model for Generation Expansion Planning with Short Term Operational Constraints

TL;DR

This paper presents a computationally efficient generation expansion planning model that incorporates essential unit commitment features, specifically ramp constraints, using historical data, enabling fast solutions for large-scale systems.

Contribution

A novel GEP model with minimal UC features, focusing on ramp constraints derived from historical data, reducing computation time for large-scale planning.

Findings

Solves large-scale GEP in under an hour on modest hardware

Uses historical data to estimate ramp limits accurately

Provides credible solutions with simplified UC features

Abstract

Generation expansion planning (GEP) models have been useful aids for long-term planning. Recent growth in intermittent renewable generation has increased the need to represent the capability for non-renewables to respond to rapid changes in daily loads, leading research to bring unit commitment (UC) features into GEPs. Such GEP+UC models usually contain discrete variables which, along with many details, make computation times impractically long for analysts who need to develop, debug, modify and use the GEP for many alternative runs. We propose a GEP with generation aggregated by technology type, and with the minimal UC content necessary to represent the limitations on generation to respond to rapid changes in demand, i.e., ramp-up and ramp-down constraints, with ramp limits estimated from historical data on maximum rates of change of each generation type. We illustrate with data for the province of Ontario in Canada; the GEP is a large scale linear program that solves in less than one hour on modest computing equipment, with credible solutions.