Extended-Term Monitoring of Acoustic Emissions Post-Laboratory Hydraulic Fracturing

Oftentimes laboratory fracturing experiments of brittle, high strength rock produce large amounts of high frequency energy from microcracking captured in the form of acoustic emissions. For instance, in granite, it is not uncommon to record more than 100,000 individual waves on a set of 6 transducers throughout a laboratory hydraulic fracturing experiment. The number of waves, and subsequently the number of located AE events, make structures determined within dense point clouds of AE data difficult to interpret. Previous work in this area used several methodologies to determine zones of varying damage and structure of macro-fractures within these large clouds, including 2D and 3D density imaging, weighted density imaging, or statistical collapsing based on error ellipsoids. Although these techniques are fruitful in illuminating features of the dense cloud of microcracks, they are snapshots in time and do not necessarily relate to the state of the rock after the fracturing process is complete or throughout longer-term stress relaxation/redistribution (i.e. snapshot images of fracture networks at the time of fracturing may not illuminate which fractures are active or closed, even moments after failure has occurred). It can be argued at the field scale that the acoustic emission data occurring post-hydraulic fracturing provides a more accurate map of active or contributing fractures relating to production. Although this information is beneficial at the field scale, it is often left unprocessed, or not recorded due to budgetary constraints or availability for long-term array placement. This work will review a previous laboratory hydraulic fracturing experiment in terms of both the AE data collected at the time of fracturing and throughout 24 hours post-test. The results of the laboratory experiment show that differing regions of the fracture network are actively changing throughout the extended term monitoring post-fracture.