In this two-part paper, we describe the construction, validation, and application of a multiscale model of entrained flow gasification. The accuracy of the model is demonstrated by (1) rigorously constructing and validating the key constituent submodels against relevant canonical test cases from the literature and (2) validating the integrated model against experimental data from laboratory scale and commercial scale gasifiers. In part I, the flow solver and particle turbulent dispersion models are validated against experimental data from nonswirling flow and swirling flow test cases in an axisymmetric sudden expansion geometry and a two-phase flow test case in a cylindrical bluff body geometry. Results show that while the large eddy simulation (LES) performs best among all tested models in predicting both swirling and nonswirling flows, the shear stress transport (SST) k-omega model is the best choice among the commonly used Reynolds-averaged Navier-Stokes (RANS) models. The particle turbulent dispersion model is accurate enough in predicting particle trajectories in complex turbulent flows when the underlying turbulent flow is well predicted. Moreover, a commonly used modeling constant in the particle dispersion model is optimized on the basis of comparisons with particle-phase experimental data for the two-phase flow bluff body case.