Research on systematic analysis and optimization method for ice storage cooling system based on model predictive control: A case study

•Energy performance of ice storage system was field tested and analyzed.•The representative issues leading to poor performance of ice storage system were summarized.•A cooling load prediction model was raised to optimize the system control of ice storage system.•The cost saving rate reached 41.9% with the cooling price decreased to 0.111 Yuan/kWh. Ice storage systems are commonly applied for space cooling in large-scale commercial buildings to realize cost-saving effects. However, the practical operation performance is far poorer than expected. This paper conducted field tests to study the operation performance of ice storage system in a commercial building located in hot summer and warm winter area in China. Where the mismatch between ice storage capacity of chiller plant and cooling demand of buildings was found the typical issue leading to insufficient ice storage in cooling season and over ice-storage in transition season. Thus the annual average COP of Ice-storage chillers and the annual EER of whole chiller plant only reached 3.19 and 2.53 respectively, and the cooling price reached 0.194 Yuan RMB/kWh. To address this, a control strategy based on a cooling load prediction model was implemented. The model inputs the easily obtainable parameters such as ambient meteorological conditions, working/rest days, and previous cooling load data to predict the cooling demand next day, and then to guide the daily ice storage capacity during valley electricity price period, as well as the cooperation of ice melting system and direct cooling system for space cooling. Therefore, significant energy and cost-saving effects were achieved, where the cost-saving rate reached 41.9%, and the cooling price decreased to 0.111 Yuan/kWh.