Classifying and modelling demand response in power systems

Demand response (DR) is expected to play a major role in integrating large shares of variable renewable energy (VRE) sources in power systems. For example, DR can increase or decrease consumption depending on the VRE availability, and use generating and network assets more efficiently. Detailed DR models are usually very complex, hence, unsuitable for large-scale energy models, where simplicity and linearity are key elements to keep a reasonable computational performance. In contrast, aggregated DR models are usually too simplistic and therefore conclusions derived from them may be misleading. This paper focuses on classifying and modelling DR in large-scale models. The first part of the paper classifies different DR services, and provides an overview of benefits and challenges. The second part presents mathematical formulations for different types of DR ranging from curtailment and ideal shifting, to shifting including saturation and immediate load recovery. Here, we suggest a collection of linear constraints that are appropriate for large-scale power systems and integrated energy system models, but sufficiently sophisticated to capture the key effects of DR in the energy system. We also propose a mixed integer programming formulation for load shifting that guarantees immediate load recovery, and its linear relaxation better approximates the exact solution compared with previous models.(c) 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Automated summary

Demand response (DR) is crucial for integrating variable renewable energy (VRE) sources. Detailed DR models are too complex for large-scale energy models, while aggregated DR models are too simplistic. This paper proposes a classification and modeling approach for DR in large-scale models, using linear constraints and a mixed integer programming formulation to capture the key effects of DR in the energy system.