Demand based control of indoor climate in office buildings. Design of local and central feed-forward control systems for high comfort, low energy use and low peak power

Sammanfattning: In this work, a number of novel strategies for controlling the heating, ventilation and air-conditioning (HVAC) systems in office buildings have been designed and evaluated. The purpose of the control strategies is to improve the match between what is supplied by the HVAC-systems and what is required to achieve a desirable indoor climate in the building. This purpose is based on the conjecture that when this match is perfect, both the energy usage and the required peak power of the HVAC-components are minimized. Most of the novel control strategies are based on feed-forward (FF) control, i.e. to utilize information about measurable disturbances by the control system, and the strategies are designed to manage a number of different tasks in the HVAC-systems. When it comes to local control, the work has focused on methods for both thermal climate and indoor air quality control. On a central level, the work has focused on the control of supply air temperature as well as night-mode strategies. These task specific strategies were primary evaluated separately but the most favourable ones were also combined in a consecutive study. Practical issues regarding implementation were also identified experimentally and their impact on the performance of the novel strategies was determined. The studies have been conducted in a simulated environment but the aspect of implementation has always been kept in mind. For that reason, both ideal strategies as well as more practical implementable ones have been evaluated in parallel. The most important finding is that there is a large potential of controlling the ventilation system using the methods that have been designed in this work. However, when it comes to thermal climate control, the difference in performance when the novel and conventional strategies are compared is rather small. In summary, a couple of conditions that are affecting the potential of FF-control have been identified, such as the thermal mass of the building, ambient climate etc. For all possible combinations of these conditions, the most feasible FF-designs resulted in that the control tasks could be performed using substantially less energy compared to when conventional controllers were used. Further, the required peak power of the HVAC-components could at the same time be reduced.