In order to retain the effect of heterogeneities at all scales in the fluid flow simulation, available data on fine scale must be transferred to coarser scales. The tools used for this transformation are the upscaling algorithms, which assign suitable values of porosity, permeability, and other flow functions to cells on the coarse simulation grid. This master thesis describes the behavior of different flow models at various scales, all derived from the same fine-scale geological model. The different flow models were generated using upscaling methods. The first upscaling approach used is the analytical method where the properties were upscaled by algebraic equations. The second upscaling method is the numerical method, where the diagonal tensor method was used specifically to upscale permeability from a fine geological grid to a coarser simulation grid. The 10th SPE Comparative Solution Project was used in all cases to compare upgridding and upscaling approaches and the ability to compare performance of a waterflood. The original model has a simple geometry with more than one-million active cells. It would be hard, though not impossible, to simulate the fine grid model in a reasonable time using conventional fine difference (FD) simulation techniques. PetrelTM software was used in this work to perform the upscaling and Eclipse 100 software was used to perform the dynamic flow simulation of the models at different scales.