As electrification increases pressure on electricity distribution grids, dynamic tariffs could help manage peak demand and reduce the need for costly infrastructure upgrades. However, effective tariff design depends on accurately forecasting how consumers respond to electricity prices.

Research by Lucas Brylle and Henrik Madsen from the Technical University of Denmark (DTU), published in the Energy Informatics conference proceedings, explores a forecasting approach designed to support this challenge. The paper acknowledges support from SEEDS and other research initiatives.

A regime-switching approach to electricity forecasting

The researchers developed a Markov-Switching Seasonal Autoregressive model with Exogenous variables (MS-SARX). The method identifies different consumption patterns and estimates how price responses vary between them.

Using operational smart-meter data from Denmark, the study evaluated the model across three scenarios: industrial consumption, residential electric vehicle charging and rural distribution grids.

Key findings

Compared with the strongest baseline model in each case, MS-SARX reduced forecasting error at primary substations by 6–22%.

The researchers also used the forecasts to simulate revenue-neutral dynamic tariffs. These scenarios indicated potential peak-load reductions of 9–49% and estimated infrastructure-deferral savings of DKK 0.7–47.7 million, depending on the case study.

These tariff and economic results are model-based estimates, not confirmed real-world savings. Further field validation is needed to determine how consumers respond to new tariff structures.

Supporting more flexible grid planning

The study highlights how improved demand forecasting could help distribution system operators identify flexibility opportunities and assess alternatives to conventional grid reinforcement.

By accounting for different consumption patterns and price responses, the approach offers a potential tool for more informed electricity grid planning. The findings contribute to wider research into demand-side flexibility and more adaptable energy systems.

Title: Dynamic Tariff Design for Demand-Side Flexibility: A Regime-Switching Approach to Consumption Forecasting in Distribution Grids
Authors: Lucas Brylle and Henrik Madsen
Affiliation: Technical University of Denmark (DTU), Denmark
Conference: 6th Energy Informatics Academy Conference (EI.A 2026)
Publication: Energy Informatics, Lecture Notes in Computer Science, Volume 16982

Read the paper: Dynamic Tariff Design for Demand-Side Flexibility