A multi-zone building energy simulation of a data center model with hot and cold aisles

In this paper, a building energy simulation code, EnergyPlus, was used to study the effects of wall boundary conditions, climatic locations, supply air temperatures, and volumetric flow rates, on the energy consumption and thermal performance of a popular data center model. The data center model having 1120 servers distributed in four rows of rack was investigated under two major climatic conditions-hot and humid (Miami, FL), and cool and humid (Chicago, IL). A multi-zone modeling approach was proposed to resolve the hot and cold aisles in the data center, and was compared to existing well-mixed single-zone model. Using the multi-zone approach that is believed more reasonable, both monthly and annual overall energy consumptions as well as cooling load were analyzed under various boundary conditions. In addition, monthly thermal behavior in the zones for hot and cold aisles within the data center was analyzed. The simulation results show that thermal performance of the data center is significantly affected by locations or climatic conditions. The effects of location and wall boundary conditions are particularly appreciable during the summer and winter seasons. An optimal supply temperature of 11.8 °C, and air flow rate of 2.5 m3/s were found to be most preferred selections for the data center model.