83. A Benchmark Synthetic Dataset for Fractured Reservoir

A benchmark synthetic fractured reservoir dataset is built as an initial step towards evaluating methodologies for using deterministic Discrete Fracture Network (DFN) models from an explored area in predicting performance in relatively underexplored adjacent zones. In this exercise we create a robust synthetic dataset comprising about seven million grid cells with details on geology, geomechanical and geophysical properties. It starts with a simple three-layered subsurface geology reflecting aeolian, fluvial and coastal environments and four major sealing faults that dissect the domain into a “core”, “graben” and a “horst” area. The entire reservoir is populated with relevant facies properties, porosity and permeability. Fracture intensity and orientation distributions are computed from geomechanical constraints. A subset of this data within the middle-layer of the core region is considered to be the “area of interest”. This region is populated with fractures invoking a DFN model by talking into account fracture intensity and orientations from geomechanical constraints. The influence of these fractures on seismic responses is evaluated based on computation of the effective elastic stiffness tensor. The seismic response can then be used as templates to interpret the characteristics of the fractured reservoir in the unexplored areas. Finally, the feasibility of production within the “area of interest” is evaluated by adding a set of injector and producer wells and simulating flow over a 10-year time period.