The discharge rate of coarse powders (mean particle size 500 ¿m) from bunkers without aeration can be described by both empirical relations and theoretical models. In the case of small particles the discharge rate is largely overestimated. As the powder dilates during flow a negative pressure gradient develops near the hopper outlet, inducing an air flow into the hopper. This extra drag force decreases the discharge rate for fine particles. Aeration of the hopper through a porous cone section will create an opposite pressure gradient, and thereby increase the discharge rate. The aim of this investigation was to incorporate the dilation in an ad hoc way into the model of Altiner in order to improve its predictive power. To test the modified model we carried out experiments with a fluid catalytic cracking powder to study its discharge as a function of aeration. As the improved model needs a dilation parameter as input, the local bulk density was measured during flow at the outlet and at the bin/hopper junction using gamma-ray absorption. At the bin/hopper junction the bulk density was found to be independent of the discharge rate and equal to the bulk density at minimum fluidisation. At the outlet the bulk density goes through a maximum when the amount of aeration gas is increased. Without aeration gas a large dilation, i.e. a 15¿35% lower bulk density, was observed. With these data the model predictions improved from 600% overestimation error to 25¿90% underestimation for pure gravity discharge, and from 100% to 0¿20% error for aerated discharge. However, the bulk density at the outlet cannot be predicted from the powder compressibility, as it seems to depend on dilation at fluidisation.