This report outlines progress in the fourth quarter of the first year of the DOE project High Resolution Prediction of Gas Injection Process Performance for Heterogeneous Reservoirs?. A new approach for qualitative evaluation of the impact of numerical ispersion in compositional finite difference simulation is described. The approach, based on solutions to one-dimensional (1D) flow equations obtained by the method of characteristics (MOC), results from current research on automatic generation of analytical solutions for 1D-gas injection problems. The assessment of the effects of numerical dispersion is demonstrated in a 2D case study of a miscible flooding using multicomponent fluid descriptions. The case study demonstrates the utility of combining streamtube methods with the analytical theory of 1D-gas injection processes.