Use artificial intelligence to optimize energy consumption.
This is the new direction of the city of Courbevoie, which recently presented the main lines of its new energy performance contract (EPC) signed with Dalkia.
The municipality's stated objective is to reduce its gas consumption by 14%, that of heating by 10%, that of electricity by 27% and that of water by 29% over the next seven years.
This concerns the 135 buildings of the municipality.
“The challenge to be met is crucial, insists the mayor (LR) Jacques Kossowski.
And the results of the first childcare center, with a 6% reduction in electricity - which at the scale of a city with bills of several hundred thousand euros is very important -, confirms us in this desire to 'go further.
"
With investments of 824,000 euros over seven years, the city must be able to save 545,000 euros.
Otherwise, Dalkia will pay the difference.
Hence the concept of a performance contract.
In addition to the project to install solar panels on certain municipal buildings, by offering residents the opportunity to become shareholders of a cooperative company, the city has also opted for the latest technologies in terms of energy saving.
Sensors on a building to reproduce its behavior
In particular the process that specialists call the digital twin.
"It is about the possibility, thanks to sensors placed in a building, of reproducing in a digital way its behavior according to various criteria, whether it is the weather forecast or the use made of it by the occupants", explains Philippe. Chevalier, who co-founded the company Kipsum in 2017, selected for this part of the CPE.
This Rueil-Malmaison-based start-up also has offices within the Pep It business incubator in Nanterre.
Concretely, the company has already installed sensors on eight buildings in the city of Courbevoie: the schools and school groups André-Malraux, Anatole-France and Logie, Molière and Lamartine, Louis-Hubert-Allié as well as on the events center, the 'Espace Carpeaux, the Jean-Blot stadium and finally the town hall. In all, seventy-two small boxes, which sense temperature and humidity and transmit data every fifteen minutes, creating a model of buildings in real time.
Data first analyzed by an engineer, who makes recommendations to optimize consumption, only for heating and air conditioning.
"A good digital twin requires an important field phase to have the best possible transcription of the building", underlines Philippe Chevalier, estimating that this first step takes a short year.
A tool to better manage heating
"Each part of each building is then reproduced at home, not in 3D, but in the form of graphics and data", specifies the engineer, who has already installed this type of technology on buildings in Rueil-Malmaison, Garches, Gennevilliers and Clamart.
"Whether they are elected officials and agents in town hall, spectators in a theater or children at school, we can optimize energy consumption according to the use of the building", details Philippe Chevalier.
The example of schools is particularly topical.
In the midst of a health crisis, these buildings are subject to new constraints, such as the repeated opening of windows to comply with sanitary protocol and renew the air.
“The digital twin makes it possible to better manage the heating,” explains the founder of Kipsum.
Because it is useless to heat to the maximum when the windows are open.
However, it is necessary to provide for the restarting of the heating system before they are closed.
"