Leveraging Commercial Building HVAC Fans for Sub-hourly Demand Response

PASSCODE: SHIFDR

Commercial building heating, ventilation, and air conditioning (HVAC) systems have been identified as flexible loads that can support renewable generation through demand response. HVAC demand response is often modeled as a virtual battery, where power deviations from normal operation are considered charging and discharging the battery. Past research has found that HVAC fans performing sub-hourly demand response act as inefficient virtual batteries, with disagreements between experiments and simulations, and no clear explanation for the inefficiency.

This thesis explores the nature of this inefficiency through analysis of experimental data and development of a novel mixing air building model to examine the tradeoffs between the demand response characteristics, such event magnitude or polarity, and the power quality provided, room temperature changes, and additional energy required to restore normal building operation. We examine the building physics relating to airflow, pressure, and heat transfer to investigate the root cause of energy discrepancies and find that physical limits and unmeasured phenomenon contribute to virtual battery inefficiency. This thesis questions the readiness of existing HVAC control systems to perform closed-loop demand response without adding sub-metered power measurements, but suggests that smarter control could improve demand response performance.

CHAIR: Professor Johanna Mathieu