Natural Fracture Prediction: A Multiscale Integration of Seismic Data, Image Logs and Numerical Forward Modelling

Natural fracture networks are commonly observed in tight carbonate and chalk reservoirs and are believed to have significant impact on the effective permeability and potential fluid flow behavior. For instance, production from chalk fields in the North Sea is believed to be aided by the presence of natural fracture systems. Apart from enhancing production, fractures can also result in channelized fluid flow and early water break-through. In this study, we propose a multiscale and data-driven workflow of automated fault extraction, image log interpretation and both inverse and forward modelling, to characterize and quantify potential inter-well fracture network geometries and densities. The workflow is exemplified on the Ekofisk chalk field situated in the Norwegian North Sea. The seismic and well data show that the Ekofisk fault and fracture system forms a connected system, which can be subdivided into four main structural orientations. Inverse modeling suggests that the orientations of the fracture/fault system can be explained by three separate normal faulting events. By implementing the structural data into a forward simulation, we characterize the potential inter-well fracture/fault network highlighting that fractures occur in clustered zones which follow the four main observed orientations.