Electricity Consumption Forecasting

Skip to content

Initializing search

Introduction and objective

From simple baselines to the first autoregressive model

Introducing calendar structure

Meteorological Data: What Information Is Useful?

Modeling Residuals: A Shift in Perspective

Building Thermal Aggregates

Results and Validation

Comparison with RTE Forecasts

Final Results and Interpretation

Conclusion and Future Work

Appendices

Abstract

This tutorial provides a comprehensive guide to building a robust, interpretable model for forecasting daily electricity consumption in France for the next day (D+1).

Moving beyond simple black-box approaches, we prioritize methodological rigor and physical intuition. We demonstrate how to combine classic time-series analysis with weather-driven corrections to achieve performance competitive with official transmission system operator (TSO) proxies.

The Learning Path

The tutorial follows an incremental logic to isolate the impact of each modeling component:

The Baseline: Establishing a starting point with a persistence model (MAE ≈ 56,400 MWh).

Calendar & Autoregression: Integrating one-hot encoded temporal variables and historical consumption patterns (MAE ≈ 26,300 MWh).

The Hybrid Architecture: Using a two-stage approach where a base model is supplemented by a Residual Corrector driven by ERA5 thermal aggregates (HDD/CDD, 14-day lag memory).

Final Evaluation: Achieving a final MAE of approximately 24,300 MWh—a 7.6% improvement over weather-free models.

Key Methodological Pillars

Strict Time-Series Validation: Training on 2012–2021 and testing on 2022–2024 to simulate real operational conditions and ensure the model handles recent energy market dynamics.

Physically Motivated Feature Engineering: Transforming raw meteorological data into spatially aggregated thermal variables that reflect the actual thermosensitivity of the French grid.

Modular Design: Separating the "base" consumption from "weather-driven" corrections to maintain model interpretability and ease of maintenance.

This project is tailored for students, data scientists, and energy professionals looking to master a structured, real-world approach to time-series forecasting.

© 2026 Éric Duhamel