Data center performance is often focused on energy efficiency. A recent survey found that the only metric for data center cooling performance in common use is PUE. This focus has had significant benefits in reducing energy consumption per kW IT in many modern installations. CFD based studies will show how a singular energy focus may not be the best approach for the business. By ignoring risk to capacity and resilience there is no guarantee of effective cooling. Focusing on effective cooling, on the other hand, is likely to facilitate an energy efficient data center in operation.

A novel method is proposed to reduce the power consumed by data centers cooling infrastructure, blowers and chillers. This method is applied to enclosed aisle configurations in which the CRAH is partially bypassed by inducing hot air into the plenum. This slightly depresses the plenum pressure and eliminates wasteful leakage of cold air. The proposed approach decreases the CRAH blower power and leads to a substantial reduction in the combined power consumption of blowers/fans and chillers.

Non-uniform airflow distribution through perforated tiles can result in inefficient cooling of servers mounted in racks in data centers. The application of strategies to maximize airflow uniformity is therefore very important because of its direct impact on power density capacity. This presentation examines how computer room air handler (CRAH) position and number of operating units as well as perforated tile types affect the airflow uniformity in selected data center applications. Also, the use of mixed tile types to strategically control airflow to accommodate greater rack heat loads is evaluated.

Shipping-container size enclosures now often house modular substations and data centers promising higher quality, lower risk, and better scalability than their stick-built counterparts. The electrical or electronics equipment housed in such enclosures may lack fans or be more like IT equipment of traditional data centers. Often a spatially-detailed thermal analysis is required to ensure that air temperatures are adequately controlled throughout the entire enclosure. Such modular applications introduce several modeling challenges related to the type of equipment housed, the cooling systems employed, and the extreme environments in which they must be designed to operate.