EURO 2025 Leeds
Abstract Submission

1553. Worksites scheduling for Distribution System Operator

Invited abstract in session WC-46: Optimal operation planning in energy systems, stream Energy Economics & Management.

Wednesday, 12:30-14:00
Room: Newlyn 1.07

Authors (first author is the speaker)

1. Simon Bayle
Sia IA,
2. Jean Jodeau
Sia AI
3. Pierre Mordant
Sia AI
4. Saoussen Abidi
Data Science, Sia Partners
5. Tristan BASLER
Sia AI
6. Germain Francois
7. Nicolas Blandamour
Sia AI
8. Lucas Selini
Sia AI
9. Olav LAFOURCADE
R&D, Sia
10. Paul Javal
Direction Technique, Enedis

Abstract

Efficient scheduling of planned works within electricity distribution networks is crucial for maximizing efficiency while ensuring that all safety and operational constraints are met. This work introduces a model for scheduling network works over a multiple year time horizon, aiming to optimize the number of projects that can be undertaken while maintaining rigorous adherence to safety standards and operational requirements.
The worksite planning process faces several challenges. First, the scale is large, as the objective is to ensure the feasibility of every worksite through their entire durations. We provide a fast electrical constraint analysis tool to speed up the solving process. Second, uncertainties are high: network conditions and forecasts evolve, requiring stochastic optimization. Lastly, with an increasing number of worksites, concurrent worksites are more common and can be used to the DSO’ advantage. The proposed model introduces pre-processing and local search techniques to minimize the computational burden.
The model and its design will be presented, before introducing numerical experiments on real data. A comparison to a historical rule-based approach will also be conducted on a French DSO’s data. Finally, the opportunity of using proxies in our framework, as machine-learning-based proxy-load flow, aiming at a global solving speed-up will be discussed.

Keywords

Status: accepted

Back to the list of papers