Cooling architecture selection for air-cooled Data Centers by minimizing exergy destruction

Abstract

Air-cooled Data Centers (DCs) require effective thermal management of the servers which can be accomplished by implementing new cooling architectures. Nearly 33% of overall energy consumption is attributed to the cooling infrastructure, which indicates the importance of the specific cooling configuration. Our objective is to compare four emerging and traditional DC cooling architectures, (a) in-row cooling, (b) rack-mountable cooling (RMC), (c) underfloor air delivery (UFAD), and (d) overhead air delivery. Since a first law-based energy analysis of a DC cooling architecture seldom considers irreversibility and component level inefficiency, an exergy-based analysis provides an alternate basis of assessment. We propose a methodology that combines computational fluid dynamics simulations with thermodynamic energy and exergy balances to determine the exergy loss in different components in DC. A dimensionless parameter is identified to characterize the exergy loss as a function of the Peclet number and the dimensionless dead state temperature ratio. The architecture containing RMC unit has the lowest exergy loss. The chiller loss constitutes up to 55% of the overall exergy loss. This analysis facilitates better decision making and design choices for air-cooled DCs based on minimizing thermodynamic irreversibility to lower energy waste.