Increasing the thermal efficiency of an operational data center using cold aisle containment

Abstract

Cooling systems play a vital role in the design and operation of Data Centers (DCs), as they consume a considerable amount of energy. Hot air infiltration from the servers' outlets into their inlets, which creates hot spots and flow short-circuiting, is one of the main sources of thermal inefficiency inside air-cooled DCs. This inefficiency increases the total energy consumption of DCs especially when the allocated workload and heat dissipation of the servers are increased. Therefore, efficient thermal management of DCs is considered as one of the main challenges of the DC industry. In order to highlight the importance of this inefficiency, an operational DC has been considered in this study. The thermal behaviour of the DC has been evaluated by conducting a numerical analysis of the flow and temperature fields based on the experimental measurements in a real DC. Numerical simulation highlighted a number of undesirable hot spots near the racks. Cold aisle containment, which is one of the efficient and industrially affordable methods of thermal inefficiency reduction, has been considered. The effectiveness of this method has been evaluated using three non-dimensional metrics known as Supply Heat Index (SHI), Rack Cooling Index (RCI) and the Coefficient of Performance (COP) of the cooling system. In this case study, by applying the proposed method, SHI, RCI and COP of the cooling system have been improved by more than 0.45, 17% and 19.5%, respectively. These results demonstrate the effectiveness of cold aisle containment for energy efficiency enhancement of DCs.