This paper reviews different methods that have been proposed to characterize thermal stratification in energy storages from a theoretical point of view. Specifically, this paper focuses on the methods that can be used to determine the ability of a storage to promote and maintain stratification during charging, storing and discharging, and represent this ability with a single numerical value in terms of a stratification efficiency for a given experiment or under given boundary conditions. Existing methods for calculating stratification efficiencies have been applied to hypothetical storage processes of charging, discharging and storing, and compared with the rate of entropy production caused by mixing calculated for the same experiments. The results depict that only one of the applied methods is in qualitative agreement with the rate of entropy production, however, none of the applied methods is in agreement with the rate of entropy production and also able to distinguish between the entropy production caused by mixing and the entropy changes due to heat losses. (C) 2009 Elsevier Ltd. All rights reserved.