Linking measurements and models in commercial buildings: A case study for model calibration and demand response strategy evaluation

• Published in: Energy and buildings 2016-07, Vol.124 (C), p.222-235
• Main Authors: Yin, Rongxin, Kiliccote, Sila, Piette, Mary Ann
• Format: Article
• Language: English
• Subjects: Automated model calibration; CO2 concentrations; Demand reduction; Demand response; DR strategies; Model calibration
• ISSN: 0378-7788
• DOI: 10.1016/j.enbuild.2015.10.042

.•Demonstrated a step-by-step, component-based procedure of building model development.•Proposed an automated model calibration procedure that links the model to a real-time data monitoring system.•Demonstrated an implementation of CO2 measurement for estimation of occupancy profiles.•Evaluated peak load reduction potentials for each demand response control strategy using calibrated model.•Identified optimal DR control strategy for VAV system by integrating thermostat setpoint increase along with minimum airflow rate adjustment.•Achieved about 50% of power reduction from AHUs during the peak hours. This paper describes a step-by-step procedure for using measured end-use energy data from a campus building to calibrate a simulation model developed in EnergyPlus. This process included identification of key input parameters for reducing uncertainties in the model. Building thermal zones were modeled to match the actual heating ventilation and air conditioning (HVAC) zoning for each individual variable air-volume (VAV) zone. We evaluated most key building and HVAC system components, including space loads (actual occupancy number, lighting and plug loads), HVAC air-side components (VAV terminals, supply and return fans) and water-side components (chillers, pumps, and cooling towers). Comparison of the pre- and post-calibration model shows that the calibration process greatly improves the model's accuracy for each end use. We propose an automated model calibration procedure that links the model to a real-time data monitoring system, allowing the model to be updated any time. The approach enables the automated data feed from simple measuring and actuation profile (sMAP) into the EnergyPlus model to create realistic schedules of space loads (occupancy, lighting and plug), performance curves of fans, chillers and cooling towers. We also field-tested demand response (DR) control strategies to evaluate the model's performance in predicting dynamic response effects. Finally, this paper describes application of the calibrated model to analyze control systems and DR strategies with the goal of reducing peak demand.