Occupancy-Based HVAC Control : Prerequisites, Problems and Possibilities

Sammanfattning: Humans in developed countries spendthe majority of their lifetime in-doorsbut not in the same building at all times. Despite this, we expect all buildings to be comfortable and create a healthy indoor environment from the point we enter until we leave. The traditional approach to as-sure this has been to condition a building as if it is always occupied.Occupancy-based control (OBC) is a general term for operation of a building’s systems while takingoccupancy informationinto account.In this thesis, the background prerequisites for implementing OBC are dis-cussedwith a focus onoccupancy assessment methods. These are systematically evaluated based on cost, reliability and privacy concern. Special focus is given to CO₂-based assessment methods and a new algo-rithm adapted for variable air velocity (VAV) systems is proposed and tested in a KTH campusbuilding in Stockholm.The assessment method showspromising accuracy but the introduction of VAV flows introducesconsiderable error sources that have to be controlled in order to allow control based on the method. In general, it is shown that detecting pres-ence/absence can be performed easily using several methodsreadily availablebut there is no clear way to count the number of occupants in a zone which limits the viability ofmore advanced OBC methods.Further on, several OBC algorithmsproposed in the literature are out-lined. These are then evaluated using building performance simulation. Results show that reducing the heating setpoints in general leads to a re-duction in energy usage at the cost of thermal comfort. OBC however shows improved thermal comfort per MWh saved, albeit by small amounts. Themaximum discomfort increases exponentially with energy savingsdespite OBC, mainly due to limited heating power.Using occu-pancy predictions showed slightlyimproved thermal comfort and energy usage but the improvements are small compared to setback based on oc-cupancy alone which is easier to detect and simpler from a computational standpoint.A new OBC method was also proposed and tested which utilizes the thermal inertia of the building and information about heating loads to predict cooling needs. Simulation results show that it is possible to reduce the maximum cooling power and energy us-age through this strategy.The viability of OBC with respect to the challengesin implementationand possibilitiesfor improved performanceis finally discussed. What is clearfrom this studyis that there is no guaranteed solution, the trade-offs for implementation of OBC have to stand in relation to the aims of the stakeholdersand the conditions for the specific building.