Paper No. 4
Presentation Time: 9:35 AM

ASSESSING THE IMPACT OF RECORDING GEOMETRY ON MICROSEISMIC DATA: AN EXAMPLE FROM THE MARCELLUS


HNAT, James S.1, REYNOLDS, Andrea1, LANGIN, William1, LE CALVEZ, Joel H.2 and TAN, Jeff3, (1)Shell Exploration & Production, Warrendale, PA 15086, (2)Schlumberger, 1600 Eldridge Parkway #606, Houston, TX 77077, (3)Shell Canada Limited, Calgary, T2P 2H5, Canada, j.hnat@shell.com

Microseismic data are often collected in unconventional reservoirs to provide an estimate of the stimulated volume of rock produced during hydraulic fracturing. Recording of such data is commonly limited to one monitoring well during a job, resulting in large uncertainties in microseismic event location, although other parameters (e.g., velocity model, processing algorithm, etc.) can also affect positional uncertainties. To test the impact of recording geometry, Shell implemented a robust acquisition geometry in the Marcellus Shale in Pennsylvania. During the stimulation of well "A" in the Marcellus, geophones were placed in three other wells drilled from the same pad to record microseismic events in the following configuration: An array of 10 geophones placed in the horizontal section of an offset Marcellus well (well "B") 1000' away to the southwest of well "A"; an array of 10 geophones placed in a horizontal monitoring well (well "C") placed 800' directly above well "A"; and an array of 12 geophones hung in the vertical section of an offset Marcellus well (well "D") to the northeast of well "A". During stimulation of well "B", only the vertical array in well "D" was used in recording. Events were co-located using Schlumberger's hypocenter determination method. In well "A", events display a distinctly linear pattern, with events developing laterally up to 1500' total. However, while the lateral growth in events in well "B" is similar, the event cloud is much more amorphous in nature, implying a much different model of fracture generation during stimulation. Additionally, event height growth differs significantly between the two wells, with well "B" exhibiting upwards of 250' greater event heights than well "A". These results suggest a considerable impact of recording geometry on microseismic data, which has implications for understanding the stimulated volume of rock, geomechanical modeling and effectiveness of hydraulic stimulation in unconventional reservoirs.