A nonequilibrium model was developed and simulated by adopting a dual-scale pore network approach. The distributions of temperature, moisture, and pressure in drying air, as well as the temperature and moisture content profiles in grain kernels were obtained from the pore scale model and the kernel scale model, respectively. A small test bin was built to conduct in-bin drying experiments and to validate the model. The results from the experiments and the simulations indicate that the dual-scale pore network approach could explain the drying process of the deep-bed grain drying process well. The influences of grain bulk porosity and cereal particle size distribution were also investigated by numerical simulations.