Using Microseismic Events to Constrain Fracture Network Models and Implications for Generating Fracture Flow Properties for Reservoir Simulation

Microseismic monitoring of hydraulic fracture stimulation treatments has done much to diminish the expectation of engineers and geoscientists that symmetrical bi-wing fractures extending away from the well bore from as a result of the treatment. Mapping of microseismic event locations indicates that more often, zones of high complexity form which suggest multiple rock failure mechanisms could be in play during the stimulation treatment. The complexity of the failure is further complicated, or perhaps explained, by the interaction of the perturbed stresses with existing fractures in the reservoir relative to the unperturbed stress state of the reservoir. Existing fracture planes favorably oriented for shear will fail at lower stresses than are required to create new fractures. Geologic mapping and regional to local in-situ stress information will allow informed interpretation of the resulting microseismicity patterns as well as providing predictive capability for fracturing patterns of treatments in subsequent area wells and production planning. Correlative to the improved fracture mapping is the use of the fracture interpretation as input to fractured reservoir modeling and fractured reservoir simulation. Utilizing microseismicity data not only to constrain location of fractures, but also fracture size, shape and orientation allows creation of improved fractured reservoir models based on geologic concepts and supported by the real time data. In this paper two examples are presented from a hydraulic stimulation of North American mid-continent wells that were monitored with a surface-based geophone array. The resulting microseismicity patterns in both wells show that the fracture development was strongly influenced by pre-existing discontinuities (fractures or faults), which are easily explained by geologic and in-situ stress analysis. The fracture interpretation and microseismicity data from one example is used to generate a discrete fracture network from which fracture flow properties are created in a geocellular model. The resulting model provides a quantitative framework for production history mapping and reservoir behavior, with hard constraints for the behavior of the dominant fractures in the fracture network. Introduction Mapping microseismic events as indicators of fracture location and extent provides engineers and geoscientists with a tool to evaluate the success of a hydraulic fracture stimulation treatment. Often the microseismic event locations fail to outline an expected trend interpreted to indicate the fracture plane, and in many cases it is difficult to define any trend at all. The ideal case where the fracture induced by the increased fluid pressure forms symmetrical wings on either side of the borehole parallel to the maximum stress direction may in fact be the most uncommon case in the earth, as it requires unbroken brittle rock. As reservoir rock typically has undergone at least one episode of tectonic deformation, natural fractures should be expected to exist in the reservoir. The directions of growth of the induced fractures is critical to well planning, for instance, where the stimulation plan needs to avoid fracture growth into neighboring wells (Maxwell et al, 2002). Increasingly, workers are interpreting that the stimulation treatment in some reservoirs has reactivated existing natural fractures that become the primary contributors for the enhanced flow that results from the stimulation treatment (Rutledge and Phillips, 2003, Gale et al, 2007). Microseismic event character has been described as complex when the microseisms display as a cloud of points rather than having well defined linear trends. Clouds of points are interpreted to result from microseismic events occurring on multiple interacting fracture planes during the fracture treatment, but remain difficult to interpret without supporting data such as cores, image logs, and seismic or wellbore anisotropy analysis. In cases where the supporting data are not available, using