Optimum Control Method of DC Load Placement and Air Conditioners Considering Fluctuations in Electricity Cost

High-performance and energy efficiency are now becoming more important requirements for datacenters (DC). Due to the increasing power density and heat generation of newer equipment, cooling and air-conditioning energy costs account for a large proportion of power consumption costs in DCs. On the other hand, the use of renewable energy is expanding with the goal of reducing greenhouse gas emissions, so the efficiency of renewable energy use needs to be maximized by increasing demand to coincide with peak electricity supply. Since the amount of load in the DC also fluctuates throughout the day, the overall electricity rate efficiency can be maximized by applying an air conditioning control level that takes into account both the amount of load arriving at the DC and fluctuations in electricity rates. In addition, there is a strong relationship between flexible load placement within a DC and cooling efficiency. In this paper, we propose a new control method that maximizes energy cost efficiency by combining two patterns of load placement between DC rooms and within the DC room, and air conditioning control that optimizes electricity cost efficiency that follows electricity rate fluctuations and load. Experimental results including those measured in an actual DC environment showed that our proposed method has valid electricity cost saving effects following the fluctuations in the electricity price and workload in the DC.