Energy consumption analysis of a medium-size primary data center in an academic campus

Experimental, numerical, and simulation studies were performed to evaluate the energy efficiency performance, develop energy conservation measures (ECMs), and conduct overall energy analysis of a medium-size data center at the campus of the University of Maryland, College Park. Based on the analysis, the PUE (power usage effectiveness) of the data center was found to be 2.73, suggesting ample opportunity for energy saving measures. The IT, cooling, and electrical loads consume 36.6%, 32.9%, and 21.7% of the total data center energy consumption, respectively. Four ECMs are suggested to reduce energy consumption by optimizing the thermo-fluid flow in the data center: (1) eliminate unnecessary CRACs (computer room air conditioning units); (2) increase the return air temperature at the CRACs; (3) add cold aisle containment; (4) implement fresh air cooling. In addition, a transient analysis was performed under a total power failure scenario of all cooling systems, as well as failure of individual CRACs as a separate analysis, to predict the corresponding temperature increase with time in the data center and electronics.