A Data-driven Control Method for Operating the Commercial HVAC Load as a Virtual Battery

Abstract

This paper presents a data-driven control method for operating the commercial heating, ventilation, and air conditioning (HVAC) load as a virtual battery (VB). Unlike an electric battery, a VB charges/discharges by consuming more/less than its scheduled baseline power. Therefore, while acting as a VB, it is critical for an HVAC load to follow a baseline consumption when no control signal is received and increase/decrease its consumption against the baseline if a charging/discharging signal is received. The HVAC load in a commercial building serves many thermal zones, so each zone can be viewed as a cell in a battery. The total power consumption of the HVAC can be controlled by adjusting the airflow rate of each zone. In this paper, models are presented for approximating the thermal dynamic and the operation of different HVAC components based on measurement data. Then, a control algorithm is proposed for adjusting zonal airflow rate in a commercial building to regulate the total HVAC consumption at desired level. A commercial building located in Richland, WA, is modeled in EnergyPlus to test the performance of the proposed control algorithm for operating the building HVAC load as a VB.