Effect of local variables on crystal breakage rate due to crystal-impeller and crystal-vessel collisions was studied. Recently. Gahn and Mersmann (Chem. Eng. Sci. 54 (1999) 1273) presented a model to calculate maximum fragment size from impact energy. We extended the model by including the tangential velocity of particles and we also introduced local velocities instead of average velocities. Our results for impact velocity in crystal-impeller collisions were in agreement with the experiments of Rielly (Proceedings of the 10th European Conference on Mixing. The Netherlands. 2000, 231). Our results for secondary nucleation showed clearly that crystal-impeller collisions were a dominant source of secondary nuclei. Number density distribution and total number of the fragments generated in each collision were used to describe the material removal from one size group to other size groups leading to birth and death rates of the crystals in each size group. Population balances were used to calculate changes of crystal size distribution against time. The influence of bottom-impeller distance on secondary nucleation was studied. The distance affects on flow velocities. mostly on axial flow and consequently on the impact velocity and the breakage of the crystals. The effect of 45degrees-pitched 6-bladed and 45degrees-pitched 4-bladed impeller at the same rotation speed (1500 rpm) seas studied. Model for 45degrees-pitched 6-bladed impeller was verified with experimental data (Chem. Eng. Sci. 45 (1990) 1405) for secondary nucleation of potassium sulphate in methanol solution. The simulated results agreed well with the experimental results. (C) 2002 Elsevier Science B.V. All rights reserved.