Marcellus Shale BEG Natural Fracture Project Final Report

Access full-text files



Journal Title

Journal ISSN

Volume Title



Operators in the Marcellus Shale gas play are aware of the importance of natural fractures, and there has been substantial work on the fracture systems in core and outcrop in the large region covered by this play (Eastern Shale Gas Project reports; Evans, 1980, 1994, 1995; Engelder et al., 2009 and references therein; Lash and Engelder, 2005, 2007, 2009). The most common fractures documented by these authors in core and outcrop are subvertical opening-mode fractures that are broadly strike parallel (J1) or cross-fold joints (J2). Evans (1995) also found strike-parallel veins that post-date the J2 set, and Lash and Engelder (2005) describe bitumen-filled microcracks developed during catagenesis. Gale and Holder (2010) found in a study of several gas-shales that narrow, sealed, subvertical fractures are typically present in most shale cores. In shale-gas plays that are produced using hydraulic fracturing stimulation, these fractures are nevertheless important because of their interaction with hydraulic treatment fractures (Gale et al., 2007). At the scale of hydraulic fracture stimulation, natural fracture patterns and in situ stress can be highly variable, even though a broad tectonic pattern may be consistent over hundreds of miles. Thus, site-specific evaluation of the natural fractures and in situ stress is necessary. Open fractures are observed in a few cases in core. Fracture-size scaling, coupled with a fracture-size control over sealing cementation and a subcritical growth mechanism that favors clustering, suggests that open fractures are likely to be concentrated in clusters spaced hundreds of feet apart (Gale, 2002; Gale et al., 2007). Our goal for this project is to characterize the fractures and identify the characteristic spatial arrangement of fractures, including potential clusters of large fractures.

Our emphasis is on characterizing, quantifying, and modeling fractures that have grown in the subsurface in a chemically reactive environment through a combination of observation at a range of scales, detailed petrographic and microstructural observation of cement fills, and geomechanical modeling (cf. Marrett et al., 1999; Gale, 2002; Laubach 1997, 2003; Olson, 2004). Large natural fractures, open or sealed, are typically sparsely sampled in core or image logs. Yet these are the fractures that would have the most effect in augmenting gas flow or influencing the growth of hydraulic fractures. Our approach overcomes the sampling problem by use of fracture size and spatial scaling analysis coupled with geomechanical modeling. That is, we may make predictions about their attributes without sampling them.

Fracture morphology, orientation, spatial organization, and cementation were analyzed using datasets from the project well-experiment area in SW Pennsylvania. We added a dataset from a field area to evaluate the use of outcrop fracture data in reservoir characterization in the Marcellus, thus expanding the relevance of the study beyond the well-experiment area in SW Pennsylvania.


LCSH Subject Headings