Northeastern Section - 53rd Annual Meeting - 2018

Paper No. 15-3
Presentation Time: 2:10 PM

STRAIN LOCALIZATION CONTROLS ON FRACTURE INTENSIFICATION IN THE ORDOVICIAN UTICA GROUP AND SCHENECTADY FORMATION IN NEW YORK STATE: QUANTIFYING FRACTURE FREQUENCY VARIATIONS USING LINEAR PIECEWISE REGRESSION AND THE AKAIKE INFORMATION CRITERION


O'HARA, Alex P., Department of Geology, University at Buffalo, UB Rock Fracture Group, 126 Cooke Hall, Buffalo, NY 14260 and JACOBI, Robert D., UB Rock Fracture Group, University at Buffalo, EQT Production, Pittsburgh, PA 15222

We present a statistics-based methodology for analyzing fracture frequency variations using a linear piecewise regression (LPR) analysis and the Akaike Information Criterion (AIC). The LPR/AIC technique can accurately calculate the width of fault damages zones by fitting multiple slope segments to cumulative fracture frequency (CFF) data. This analysis suggests that higher frequency fractures overprint the pre-existing background set during strain localization.

Fracture and fault data were collected from 38 sites in the Utica black shale and overlying clastics of the Mohawk Valley in eastern New York State. A total of 3678 fracture and vein measurements were taken using scanline, scangrid, and abbreviated methods. The piecewise function in the R package “Segmented” performed the calculations for the LPR/AIC analysis. Additionally, individual CFF slopes were reorganized as new, standalone datasets, and replotted to analyze potential clustering of segments. Average slopes and 95% confidence intervals were calculated for each group of linear segments, producing three discrete CFF slope intervals that are used to classify background, transition, and fracture intensification domain (FID) frequencies.

Background frequencies are defined by an average CFF slope of 8 and an average fracture frequency of 2.4 fractures/m. Transition frequencies have higher CFF slopes of 111 and a higher average fracture frequency of 12.3 fractures/m. FIDs (including fractures in fault damage zones) are defined by the highest average CFF slope of 1649 and highest average fracture frequency of 44.6 fractures/m. The changes in slope among the CFF intervals differentiate damage zone fractures from the background fracture set. The three CFF slope intervals are hypothesized to represent multiple strain localization events. Each successive overprinting event results in a reduction in the active fracture zone width and an increase in fracture frequency, concluding with the formation of a fault surface.