AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Integration of 3-D Surface Seismic and Jointly Inverted Downhole-Shallow Well Microseismic Datasets on a Multi-Lateral Stimulation Campaign: A Marcellus Formation Example


To collect 3D surface seismic data on unconventional plays offers the opportunity to map large-scale features and discontinuities; thus, yielding successful well placement. To reconcile borehole-derived observations (e.g. cores, image logs, etc.) with larger volume-derived datasets (e.g. cross-well, surface seismic, etc.) requires proper upscaling. Passive seismic has the potential to bridge these multi-scale domains. We describe the results of joint inversion of shallow well and deep downhole passive microseismic monitoring in relation to several hydraulic fracture treatments in the Marcellus formation and how results are integrated so to start feed a geomechanical model. Monitoring of several stages along one lateral is performed using one deep twelve-3C-geophone array located in a vertical wellbore and a series of multi-component seismic sensors installed in a grid of 140 shallow boreholes. Initially, we use each array independently to evaluate the pros and cons of each monitoring configuration in this particular geological setting. Then we jointly invert surface and downhole microseismic data to enhance the overall microseismic ‘image’ by improving location accuracy, reducing detection bias, increasing the robustness of the velocity model and its calibration, and minimizing uncertainties of the various source parameters and attributes. We use the optimized velocity model to subsequently process microseismic data acquired during additional stimulations from the shallow wells only. Initial results indicate that along the stimulated laterals, microseismic activity correlates with low acoustic impedance. Overall variations in microseismic activity appear to relate to impedance variations. After superimposing the inverted acoustic impendance and the seismic-derived discontinuity volume with the mapped microseismic data, we observe that preferential fracture directions extracted from the mapped hypocentral locations agree with the natural fracture system oriented according to the dominant-stress tectonics. Microseismic activity correlates with wells intersecting a significant number of surface seismic-scale lineaments. However, mapped microseismic activity negatively correlates with structural features showing significant relief. Initial moment tensor inversion results performed on good quality events detected on the shallow well array provides insight as to distinct failure mechanisms apparently relating to the structural high and lows.