Evaluation and optimization of thermal performance and air distribution in raised-floor data center.

Abstract

Raised-floor data centers are the most commonly used facilities for housing computer and telecommunication equipment. The cooling of data centers and network rooms has emerged as a significant challenge as the density of computing equipment increases. The air cooling system will be inefficient when heated exhaust air re-circulation, chilled airflow by-pass and uneven air distribution occur. This paper focuses on the use of performance metrics to evaluate air distribution and thermal performance. Innovative metrics are provided. Thermal Distribution Index (TDI) is a measure of evaluating thermal distribution and cooling uniformity. It is a definition of the layer which is overcooling and undercooling. It can also tell us how many cooling capacity can be utilized. The Supply Air Index (SAI) describes whether the air distribution has achieved the cooling requirements. SAI has a good effect of evaluating the supply air temperature field and air mixing problems. These two indices could be a rule to choose proper cooling strategy. The application of these indices can help reducing energy consumption and improving equipment stability. An experimental data center has been evaluated and relevant optimization measures have been described to lower the energy consumption.