Analyse des stratégies de contrôle pour un système de chauffage ventilation et conditionnement de l'air (CVCA) complexe à l'aide d'un modèle calibré

Abstract

Improving the energy performance of an existing building can be achieved through recommissioning of Heating Ventilating and Air Conditioning (HVAC) systems. This process makes it possible to detect operational faults and to study alternative solutions. One of these alternative solutions is to modify the control strategies so that they are more adapted to the existing operating conditions. The study of these solutions can be done using a calibrated model that represents the operation of the current existing systems. This makes it possible to test different control strategies without disturbing the occupants. The purpose of this thesis is to analyze the control strategies of the HVAC system of the Maison des Étudiants of the École de Technologie Supérieure, using a calibrated model. The installed system has different levels of air handling that combine several interdependent systems. The calibrated model can be used to analyze energy demand, but also the maximal system power demand and other operating variables such as supply temperatures and airflows. The calibration uses the Bottom-up approach by offering sequencing of the air treatment level and of the HVAC system. The validation of this step is carried out using the Normalized Mean Bias Error (NMBE) and the Coefficient of Variation of the Root Mean Square Error (CVRMSE) on the hourly demand, but also using additional indices, namely the Root Mean Square Error (RMSE), the coefficient of determination (R²) and the Goodness of fit (G) on the instantaneous load, the temperatures and airflows. The results are close to the ASHRAE Guideline 14-2002 criteria (NMBE of 10% and CVRMSE of 30%). The calibrated model is used to test different control strategies that could reduce overall annual demand by 15%.

Date	18 Oct 2018
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Stanislaw Kajl (Supervisor) & Danielle Monfet (Co-supervisor)