ASPERITY AND JOINT FAILURES, OVERALL SURFACE RUPTURES, IDENTIFICATION OF AND ROLE IN THE INTERPRETATION OF DISCRETE FRACTURE NETWORKS

Typically, monitoring of hydraulic fracture stimulations is carried out with high frequency geophones temporarily placed close to reservoir depth. Inherit with these arrays are bandwidth limitations that limit the size distribution of observed fractures to failures associated with pre-existing joint sets or asperities/ barriers associated with larger faults. To accurately portray the discrete fracture network, recording with supplementary low frequency geophones or accelerometers can be used to extend the observed fracture size distribution to include larger coalesced fracture/fault sizes. Capturing the various size scales provides opportunities to more completely identify activated structures during hydraulic fracture stimulations. The overall distribution of fracture sizes with magnitude, often referred to as scaling relations, can be closely examined for differences or similarities in scaling behavior, namely whether size distributions follow models of constant stress drop. Departures from self-similarity may be related to source and/or site effects, such as the presence of lower driving stresses resulting from fluids in the development of fracture zones as compared to shear stress driven failures commonly associated with faults.

In this study, we examine the scaling behavior of fractures initiated during hydraulic fracture stimulations with data recorded utilizing both near surface three component Force Balance Accelerometers, 4.5Hz geophones and downhole 15Hz geophones. Based on these data, we observed self-similar behavior for larger magnitude events whereas smaller events exhibited, in cases, non-similar behavior. The observed scaling relationships suggest differences in behavior related to underlying conditions of event generation. Questions as to whether the transition from non-similar to self similar behavior is related to moving from fluid induced fracturing to stress driven failures as compared to moving from asperity/barrier failures to overall slip on faults are considered. Additionally, we will examine the failure characteristics of the events ands their relation to the observed scaling behavior and consider what role these observations have in assessing the discrete fracture network used to optimize stimulation design.