Aggregation phenomena are important in food products where many interactions between components are possible. These interactions can occur during food processing: heating, pH change, high pressure process, separation process, etc. Desired or not, they must be well understood to have a perfect control of food quality. Indeed, size and structure of the aggregates influence the texture and taste properties. In this study, a modified version of a classical model for colloidal aggregation (Population Balance Model of Smoluchowski) is presented for heat-induced aggregation of whey proteins. Experimental data were obtained on β-lactoglobulin (β-lg) solution (6% + CaCl2) using process simulator (85°C for 60 s) with different shear conditions. In agreement with literature and experimental observations, three stability factors (W) have been introduced in the model, each one related to an interaction involved in β-lg aggregate formation. Characteristics of the suspensions (residual native fraction, aggregate size distribution and viscosity) were used for fitting between experimental and simulated data. By adjusting the stability factors (W) to the composition, we successfully integrated the role of physico-chemical environment variations such as CaCl2 concentration. The role of shear upon aggregate growth is also well predicted by the model. Prediction of aggregate properties in function of thermo-mechanical treatment can help to better design the processes.