Comparison of marcellus fracturing using azimuthal seismic attributes versus published data from outcrop studies

Using a good quality northern Pennsylvania (PA) Analog 3D survey1, available well data, published outcrop data and subsurface information as well as production data available from the state, we are able to demonstrate that wideazimuth seismic is sensitive to variations in fracturing at the scale of individual pads or even individual wells. This variation in fracturing begins to explain why production varies significantly, even locally, within the Marcellus play. Rose diagrams from quantitative fracture analysis using azimuthal seismic velocity volumes are compared to published data from Appalachian black shale outcrops and subsurface fracture models proposed in various papers in order to validate the results from subsurface data. While it has long been understood that natural fracture systems are essential for achieving the best production in Marcellus shale gas wells, methodologies for understanding the heterogeneities in these fracture systems in the subsurface are less well understood. Analysis of wide-azimuth Pwave seismic velocity attributes at the reservoir level, and for specific laterals or proposed laterals, can provide this insight. Although anisotropy, measured as azimuthal variations in velocity, can reflect rock fabric or stress, we show evidence that the likely source of these anisotropies is the presence of systematic joints. Published data and azimuthal seismic attributes show two primary joint sets, the J1 set and a J2 set, as well as neotectonic J3 joints that affect the Marcellus and other Devonian shales in the Appalachian basin (Engelder et al., 2009.) Evidence suggests that in organic-rich Devonian black shale intervals including the Marcellus, both J1 and J2 joint sets formed in sediments at or near peak burial depth as a result of anomalous pressures generated during thermal maturation of organic matter (Lash et al., 2004.) While authors indicate that the ENE to WSW J1 joint set is generally restricted to the black shales, the younger NNW to SSE J2 joint set is described as being more likely to extend out of the black shales into overlying rock. The late cracking of oil to gas released much larger volumes of gas during the period when J2 joints propagated. Although neither contemporaneous with nor genetically related to folding, J1 and J2 joints in black shales within the Finger Lakes area of New York strike approximately parallel and perpendicular respectively to the Alleghanian orogenic fold axes exposed there. Similar patterns can be seen in seismic anisotropy for the Analog 3D survey and for the Clearfield 3D survey, to the west, in Clearfield County, Pennsylvania. Rose diagrams from azimuthal attributes are similar to those measured for joints in outcrop in southern New York (Fillmore Glen State Park) and in Central Pennsylvania. The J2 azimuths dominate in areas with higher gas estimated ultimate recoveries (EUR) wells, which is attributed to more gas generation as well as longer joint length. High EUR areas also have generally lower "Interval Vfast velocities" and show J2 joints well above the top of the Marcellus in Hamilton Group gray shales. Areas with low EUR wells have a more dominant J1 trend or no dominant trend, along with higher anisotropy, which suggests greater heterogeneity. The seismic velocity attributes show evidence of velocity anisotropy which could be attributed to fractures and are used to identify areas that may be more fractured. Areas showing high azimuthal gradients imply possible reservoir compartmentalization and heterogeneities that correspond with production. These attributes offer a tool to high grade drilling opportunities and improve production estimates for Marcellus wells.