When evaluating the energy demand of buildings using stratified hot water storage tanks in an active demand response context, an accurate but low order storage tank model is required. Accurate modelling of buoyancy effects and mixing inside the storage tank as a result of direct inflows is key to obtain valid results. The use of one-dimensional storage tank models is common practice in building simulations, but as buoyancy and mixing are three-dimensional phenomena they cannot be incorporated directly in such a model and special formulations must be used. This paper presents a novel method for incorporating buoyancy and mixing in one-dimensional stratified storage tank models. The model parameters are derived from a series of computational fluid dynamics simulations and correlated with the appropriate non-dimensional parameters. The model is validated against an independent set of charging and discharging experiments. The model is found to represent buoyancy and mixing in the storage tank in a realistic way and to correspond well to experimental results. To allow researchers to assess the uncertainty on performance indicators in building energy simulations, an effective storage capacity ratio is defined and its variation within the model parameter uncertainty range is calculated and compared to other models. The results show a discrepancy between models commonly used in the literature and the presented validated model. When the influence of storage capacity on parameters of interest is known, the effective storage capacity ratio can also be used to estimate the uncertainty on these parameters caused by the storage tank model. (C) 2016 Elsevier Ltd. All rights reserved.