EURO 2025 Leeds
Abstract Submission

1289. Forecast demand and scenarize energy production in district heating networks

Invited abstract in session MB-38: Optimization in contexts with multi-media signals or data security, stream Data Science meets Optimization.

Monday, 10:30-12:00
Room: Michael Sadler LG19

Authors (first author is the speaker)

1. Slawomir Pietrasz
Computer Science Artificial Intelligence, ENGIE LAB CRIGEN
2. Marinette LOISY
Computer Science Artificial Intelligence, ENGIE LAB CRIGEN

Abstract

Imagine you operate a district heating and cooling (DHC) network. Your goal is to constantly improve its global performance. What asset would you run to produce the daily required energy: gas boilers, cogeneration plants, geothermal heat pumps or biomass boilers?
External temperature variations and network extensions significantly influence fuel supply to DHC networks. The challenge is threefold: satisfy the customers’ heat demand, keep low operating costs and contribute to the ENGIE Group's Net Zero Carbon ambition.
The networks have been equipped with sensors that collect more than 500 million pieces of data every day. A decision supporting platform "Predity for Asset" uses this data as an input for Artificial Intelligence and Operation Research tools. "Predi Demand", forecasts thermal demand over the next 15 days, based on the energy consumed on the network over the past year, and a forecast of outdoor temperature for the coming period. It is based on a parametric multiple linear regression model. "Predi Scenario", relies on the forecast computed by "Predi Demand" to determine the optimal heat production plan including heat recovered from the combustion of municipal solid waste. Boiler plants are thus supplied with the best available local and renewable energy.
The platform is currently operational in France and is being deployed overseas. It will help reduce CO2 emissions by 45 million tons per year by 2030, with annual savings of 500 K€ throughout France.

Keywords

Status: accepted

Back to the list of papers