van Batenburg, Diederik W. (Shell Exploration & Production) | Bosch, Marco (Shell) | Boerrigter, Paulus Maria (Shell Development Co.) | De Zwart, Albert Hendrik (Shell Intl E&P Co) | Vink, Jeroen C. (Shell Intl E&P Co)
IOR/EOR processes have moving fronts and require an accurate description at a scale much smaller than typical grid block sizes used in black-oil simulations. Furthermore, they may require extra uid components and phases, or thermal properties, or sometimes chemical reactions between components. However, simulation of these processes on a high resolution grid throughout the reservoir is still not practical with today's computing power. One solution is to start on a coarse grid that is dynamically adjusted to provide sufficient resolution where needed. The challenge in such an approach is in identifying where and when to adjust the grid. This paper presents a semi- implicit scheme, that allows for the evaluation of grid adaption criteria within a time step, and guarantees that grid refinements occur where the fronts have propagated at the end of each time step. No assumptions on front movement need to be made in advance. The scheme is robust and accurate, while still providing strict control on the number of blocks by minimizing the extent of refined zone(s). This is of particular importance for nested refinements as the number of blocks rapidly grows with each extra refinement level. The combined use of changes in properties in space and in time allows for the definition of criteria that always reliably position the refinements along the fronts. Application of this dynamic gridding approach in simulation models with water-flooding, chemical-flooding, gas injection, and in-situ combustion demonstrates that efficient nested dynamic gridding can be implemented in a general purpose simulator to provide sufficient physical and spatial detail in meaningful eld or pattern models for IOR/EOR. The key to fast and accurate dynamic gridding is the implicit evaluation of proper criteria for refining and coarsening.