Faulting and fault sealing in production simulation models: Brent Province, northern North Sea

Faults can severely compartmentalize pressures and fluids in producing reservoirs, and it is therefore important to take these effects into account when modelling field production characteristics. The Brent Group fields, northern North Sea, contain a complex arrangement of fault juxtapositions of a well-layered sand-shale reservoir stratigraphy, and fault zones containing a variety of fluid flow-retarding fault rock products. It has been our experience that these fault juxtapositions impact the ‘plumbing’ of the faulted layering system in the reservoirs and the models that are built to mimic them – and are, in fact, a first-order sensitivity on compartmentalization of pressures and fluid flow during production simulation. It is important, therefore, to capture and validate the geological feasibility of fault- horizon geometries, from the seismic interpretation through to the static geocellular model, by model building in conjunction with the interpretation. It is then equally important to preserve this geometrical information during geocellular transfer to the simulation model, where it is critical input data used for calculation of fault zone properties and fault transmissibility multipliers, used to mimic the flow-retarding effects of faults. Application of these multipliers to geometrically weak models tends to produce ambiguous or otherwise potentially misleading simulation results. We have systematically modelled transmissibility multipliers from the upscaled cellular structure and property grids of geometrically robust models – with reference to data on clay content and permeability of fault rocks present within drill core from the particular reservoir under study, or from similar nearby reservoirs within the same stratigraphy. Where these transmissibility multipliers have been incorporated into the production simulation models, the resulting history matches are far better and quicker than had been achieved previously. The results are particularly enhanced where the fault rock data are drawn from rocks that have experienced a similar burial–strain history to the reservoir under study.