Development of a New Approach of Total Cooling Load Based Chiller Sequencing Control To Deal with Uncertainties Systematically

Total cooling load based chiller sequencing control has been developed to stage chillers on or off in multiple-chiller plants by matching the maximum cooling capacity provided by the chiller plants to building instantaneous cooling load. It optimizes the number of operating chillers to satisfy thermal comfort requirement and simultaneously achieve energy efficiency. This research will investigate uncertainties associated with this type of chiller sequencing control which significantly affect the control performance, and explore a systematic strategy to deal with those uncertainties. A laboratory test-rig will be set up and two types of uncertainties will be considered: measurement uncertainties associated with the cooling load estimation and model uncertainties associated with the online estimation of chiller maximum cooling capacity. Data fusion will be investigated and developed to reduce the measurement uncertainties. A model-based method will be employed to online estimate the chiller maximum cooling capacity; and uncertainties associated with the model will be evaluated using data collected from the test-rig. A method of online tuning the model parameters will be developed to reduce the model uncertainties. Decision making under uncertain information will be reviewed and a suitable approach will be developed for chiller staging operations. Case studies including simulation-based and laboratory-based tests will be carried out for validation.