It is common practice to model shales in dynamic simulation models implicitly by assigning a global reduction factor to the sand's vertical permeability. Such an approach is required due to the complexity and computational limitations of modelling shales explicitly in large simulation models. The validity of this approach in the case of discontinuous shale baffles in a steam flood is the subject of this paper. The study concentrates on understanding the physics of flow in the presence of these discontinuous shales by conducting generic fine scale simulations for a water flood, a gas-oil gravity drainage system and a steam flood. The results show that in general implicit modelling of shales by an effective vertical permeability that is derived from single phase flow simulations cannot capture the flow mechanism and predicts a different oil recovery to models where shales are accounted for explicitly. For steam flooding implicit shale models take longer to heat the reservoir than explicit shale models; the main reason being that steam can percolate around shale baffles and so quickly achieve an elaborate network of continuous flow paths to the top of the reservoir. A parameter evaluation technique is worked-out that captures explicit shales correctly in an effective vertical permeability although this needs also tuning of (pseudo-) relative permeability parameters. Encouragingly, when a good history match is obtained for a bottom-up water flood by modeling the shales implicitly in the described way, these tuned implicit models also give good steam forecasts. The actual appearance of the floods can, however, be very different from the explicit appearance.

You can access this article if you purchase or spend a download.