Optimal control for building heating: an elementary school case study

Due to the development of energy performance contracting and the needs for peak electric demand reduction, the interest for optimal building control is renewed. In this context, the real time prevision and optimization of building heat demand can help the manager to reduce the energy bill and to propose peak shaving offers. Our study aims to illustrate such heat control strategies on a one floor elementary school. The building is modeled through a second order inverse “grey box” model. The inverse model identified during a short learning period is first validated on its ability to forecast heat load and indoor temperature. Then it is used for optimal control and for that purpose two strategies are proposed. The first one consists in optimizing the night setback period with a constant electricity price. The second one aims to set a varying indoor temperature set point in a context of peak and off peak hours. The results show about 5% off electricity consumption for the first strategy and 4% off electricity bill for the second strategy. For a very cold week it appears that the optimization could lead to an over-consumption to improve the comfort.