Spray drying is a well-established method used in the food industry for production of powders, such as dry milk, coffee, tea, and soup. The operating conditions of spray dryers depend on the product and chamber design. Thus scaling up the process is complex, and simulation tools are needed to reduce time and cost, and to enhance quality of the final product. In this study a validated distributed-parameter model for predicting drying of single particles was combined with a CFD simulation model of an eight-meter pilot dryer to investigate drying kinetics of skim milk powders. A qualitative assessment of the effect of water diffusivity values on preconditions for agglomeration and a multi-scale analysis of coalescence and agglomeration regions were performed. Results showed that accurate water diffusivity values and sticky conditions have to be implemented when modelling spray drying to investigate preconditions for coalescence and agglomeration. The results also showed the need to simulate surface conditions of particles during spray drying to predict probable regions of coalescence and agglomeration. In conclusion, the applied methodology allows understanding of the stickiness ability of a product and size of the chamber. The methodology can be used to support preliminary design of spray dryers.