Investigation on pre-cooling potential of UFAD via model-based predictive control

•A decentralized model was introduced for grey-box building models with measurements.•Presented energy and cost-saving potential of underfloor air distribution system via model-based predictive control with a simulation case study.•The impact of the outdoor environment and plenum design was quantified. This study investigated the operation cost-saving potential of an underfloor air distribution (UFAD) system by applying model-based predictive control (MPC). Based on the measurements from an actual building, the grey-box building model was constructed using a decentralized approach. The simulation case study was then carried out to quantify the saving potential of the MPC compared to feedback control. For the cooling season, the saving percentage of the MPC exceeded 30%, while those of the intermediate season were reduced by approximately 20%. More savings were achieved when the outdoor air temperature and solar radiation were high. The size of the plenum was not sensitive to the saving rate. The large savings were attributed to precooling that utilizes the cheaper utility cost of the off-peak and higher efficiency of the HVAC plant. This suggests that the larger thermal mass of the under plenum, including the air and still structure (bracket), is suitable for maximizing pre-cooling of the system.