Experimental study on flexible energy use feature of multi-split air-conditioning system with different control schemes in summer

Abstract

Demand response (DR) serves as the primary avenue through which building air-conditioning (AC) flexible resources engage in power grid regulation. Considering a dynamic outdoor environment and the constraints of indoor comfort and system mechanical limits, specifying the response accuracy and flexibility potential of the system in the actual response process is beneficial to enhance the response effect and improve the energy efficiency. This study focused on two flexible regulation strategies of AC in DR: global temperature adjustment (GTA) and direct power control (DPC). An experimental investigation was conducted to clarify the flexible regulation characteristics of multi-split AC systems under these two strategies and quantitatively evaluate their flexible potential. The results showed that the DPC strategy achieved a superior response accuracy of 90.4 %, while the GTA strategy’s response accuracy was 23.8 %. And there was a discernible upward trend in flexibility potential with rising outdoor temperatures and increased adjustment steps for GTA, with an associated 1 to 6 W/m² load reduction for each 1 °C increment in temperature set-point. Considering the mechanical limits of the equipment, the theoretical available regulation capacity of the DPC strategy for high, medium, and low outdoor temperature conditions were determined to be 75 %, 50 %, and 25 %, respectively. Moreover, narrowing the temperature gap between the baseline temperature and the upper comfort temperature limit shortened the available response duration. The study further explored the balance between flexible responsiveness and operational efficiency, suggesting that lowering the set-point temperature for valley filling can help improve system performance. These insights offer an exhaustive perspective on anticipating and managing flexible capacity of multi-split AC systems within the context of DR initiatives.