Data centers are extensively used for the storage and retrieval of large amounts of data. Efficient removal of the energy dissipated by the electronic circuitry is a major concern since the temperature must not rise beyond acceptable levels that affect the performance and since the energy required for cooling is generally quite substantial, with typical data centers requiring hundreds for megawatts for heat removal. Though steady operation has been of interest in most studies on data center cooling, transient behavior is also of considerable interest since the load is generally not a constant but varies with time. Also, distribution of load between various data centers may be used effectively to reduce the overall power consumption. The environmental conditions may be used advantageously in the selection of a particular data center in order to reduce the cooling load. All these considerations make it imperative to study both the steady and time-dependent behavior of data centers and to obtain results that may be used to optimize the cooling systems and reduce the overall energy consumption. This paper presents an analytical and numerical study that focuses on load distribution, as well as data center location that determines the relevant environment, in order to obtain characteristic results on the resulting temperature distributions for a range of operating conditions. Such results could form the basis for optimizing the cooling strategy and the system. (C) 2016 Elsevier Ltd. All rights reserved.