In order to model the complex hydrodynamic phenomena prevailing in industrial scale gas–solid bubbling fluidised bed reactors and especially the macro-scale emulsion phase circulation patterns induced by bubble–bubble interactions and bubble coalescence, a discrete bubble model (DBM) has been developed. In the DBM, the (larger) bubbles are modelled as discrete elements and are tracked individually during their rise through the emulsion phase, which is considered as a continuum. The DBM, originally developed for the description of gas–liquid flows, has been adapted to cope with bubbles with a diameter larger than the size of an Eulerian cell, which is required in view of the large bubble size distribution at higher gas flow rates. Moreover, a new drag model for a single bubble rising in a fluidised bed derived from empirical correlations has been implemented, as well as a simple model to account for bubble coalescence and break-up. The strong advantage of the DBM compared to other models previously reported in the literature for the description of large-scale fluidised beds is that it fully accounts for the two-way coupling between the bubbles and the emulsion phase, which enables direct computation of the emulsion phase velocity profiles. Comparison of the results of simulations ignoring bubble coalescence and simulations taking bubble coalescence properly into account demonstrated the significant effect of bubble coalescence on the large-scale circulation patterns prevailing in bubbling fluidised beds. The simulation results for the lateral profiles of the visible bubble flow rate have been compared qualitatively with experimental results reported by Werther [1974. Influence of the bed diameter on the hydrodynamics of gas fluidized beds. A.I.Ch.E. Symposium Series 70(141), 53–62]. The effect of the superficial gas velocity on the velocity and porosity profiles has been studied. In general, it can be concluded that the DBM is able to capture the salient features of the hydrodynamics of bubbling fluidised beds. However, further research is required to improve the closure equations for the bubble behaviour, bubble–bubble interactions and bubble coalescence and break-up to enable a complete quantitative description.