A demand limiting strategy by direct load control for air-conditioning systems of buildings towards urgent requests of smart grids

Smart grids enable a bidirectional operation mode (i.e., “two ways” connection with power flow and information flow) to offer better flexibility, reliability and security for grid operation. The control of power demand side in response to smart grid requests (e.g., dynamic price and reliability information) is known as demand response (DR) while DR control on air-conditioning systems in commercial buildings is considered as a most effective means. However, conventional DR control strategies on air-conditioning systems are usually subject to inevitable delay and cannot fulfil the urgent requests of smart grids. Therefore, a direct load control (DLC) strategy by shutting down some of operating chillers directly in an air-conditioning system is developed. The proposed DLC strategy consists of power demand optimization, system alternative settings and online control module. Under specific incentives given by smart grids, power demand optimization is responsible to predict building power limiting threshold during DR events considering thermal comfort of end-users. System alternative settings determine the numbers of retained operating chillers and (primary and secondary) pumps in an air-conditioning system. In online control module, building power demand is adjusted by chiller load regulator to follow predicted building power limiting threshold and cooling distributor is employed to realize required cooling distribution among individual air-conditioned zones in case of the failure of conventional control strategies for air-conditioning systems during DR events. Case studies are conducted on a simulated air-conditioning system to test and validate the DLC strategy. Results show that the proposed DLC strategy is capable to effectively achieve immediate power reduction during DR events without any operation problems occurred. In addition, expected power limiting threshold is realized and accepted indoor air temperature can be guaranteed even existing the uncertainty in prediction process.