Conference article

Dynamic Simulation of Residential Buildings Supporting the Development of Flexible Control in District Heating Systems

Nadine Aoun
L2S, CentraleSupélec, Gif-sur-Yvette, France / CEA, LITEN, Grenoble, France / ADEME, Angers, France

Roland Bavière
CEA, LITEN, Grenoble, France

Mathieu Vallée
CEA, LITEN, Grenoble, France

Adrien Brun
CEA, LITEN, Grenoble, France

Guillaume Sandou
L2S, CentraleSupélec, Gif-sur-Yvette, France

Download articlehttp://dx.doi.org/10.3384/ecp19157129

Published in: Proceedings of the 13th International Modelica Conference, Regensburg, Germany, March 4–6, 2019

Linköping Electronic Conference Proceedings 157:13, p. 10

Show more +

Published: 2019-02-01

ISBN: 978-91-7685-122-7

ISSN: 1650-3686 (print), 1650-3740 (online)

Abstract

Load shifting, peak shaving and night-time setback are key demand-side management measures to make the operation of District Heating Systems (DHSs) more flexible and efficient. These goals can be achieved through appropriate control strategies exploiting the building’s and space heating system’s thermal inertia. To ease the development of such an advanced controller, we programmed a detailed dynamic Modelica simulator representative of French multi-stories radiator-heated residential buildings. We parametrized the simulator to vary the factors influencing the flexibility potential of a building (e.g. envelope properties, additional internal mass such as partition walls and furniture, the heating system…). This helped us designing a reduced-order building model relevant to our application and setting up a robust identification method for its parameters. We finally used the detailed simulator to test an optimal space-heating controller, thereby allowing many incremental improvements without jeopardizing endusers thermal comfort. This simulation work paves the way to considering the actual implementation of our advanced controller on a real building.

Keywords

District Heating System, Optimal Control, Building Simulation, Reduced-order building model

References

No references available

Citations in Crossref