Modeling fractures in the reservoir has historically been limited by the sparseness of actual fracture data for a given volume. Good information about fracture orientation sets can be obtained from wellbores by acquiring image logs or oriented core, as well as useful information about fracture intensity relationships to the lithologies intersected by the wellbore. Two very important fracture set characteristics, fracture size and 3-dimensional spatial fracture intensity, cannot be representatively sampled with wellbores. Spatial fracture properties inferred from distributed reservoir properties obtained from seismic attributes provide a useful constraint for stochastic DFN models generation, but the generated DFNs are qualitative representations of possible fracture distributions. Given that microseismic events often illuminate the existing fracture network in a stimulated reservoir, when source mechanism solutions are available as an attribute of the microseismic events, these data provide the actual locations, orientations, and relative sizes of the fracture planes. The actual fracture size can be estimated using the moment magnitude of the source mechanism and typical elastic properties for the rock types so that the resulting DFN can be considered deterministic. The workflow developed to apply this approach results in a DFN containing several of the characteristics of natural fracture distributions observable in geological exposures, such as fault damage zones, bed-bounded fracturing, and fracture/fault interaction relationships. This type of DFN modeling has implications not only for oil and gas reservoir production, but also for geothermal operations, carbon sequestration, and for fractured aquifers intersected by faults.
|Original language||English (US)|
|Title of host publication||2019 AAPG Annual Convention and Exhibition|
|Place of Publication||San Antonio, Texas|
|State||Published - 2019|