Date of Award

Spring 2019

Degree Type


Degree Name

Master of Science (MS)


Geophysical Engineering

Committee Chair

Marvin Speece

First Advisor

Xiaobing Zhou

Second Advisor

Scott Rosenthal


Fracture characterization is highly important in the oil and gas industry. Knowing the location of fractures allows for the assessment of reservoir quality, aids in well placement and planning, and helps identify locations of possible traps. Fracture locations can be determined using seismic data for attribute calculations and anisotropy analysis. Attribute calculations, such as coherence and curvature, identify subtle changes in the dataset that conventional seismic data interpretation might overlook. Anisotropy analysis looks at directionally dependent variations in the wave propagation velocity. Fractures slow the propagation velocity of a wave if the fractures are perpendicular to the wave direction. A combination of attribute calculations, and anisotropy data can predict locations of faults within a region.

The study region for this thesis project is within Greenbrier and Nicholas Counties, West Virginia. A 36 fold, 128 square miles (333 km2) seismic survey, conducted in 2011, spans two lease regions, the Mead West Vaco, and the Plum Creek South Fork. Ten wells are within the two lease regions, drilled prior to the completion of the seismic survey. Of the 10 wells, a core analysis focusing on the Marcellus Formation utilized data from four wells. In addition, a microseismic analysis used data from three wells.

The Marcellus Formation is the target of the seismic study and the wells. The Marcellus is a middle Devonian, black, shale interbedded with limestone. The formation spans approximately 95,000 mi2 (246,048 km2) in the northeastern United States with thicknesses varying from 50-200 ft (15-61 m). Deposition of the Marcellus occurred in a deep-water, oxygen-deprived environment, which resulted in the accumulation of hydrocarbons. The interbedding of limestone with shale, within the formation, created many traps for hydrocarbons resulting in the reservoirs today. The low permeability of the formation lends to the use of hydraulic fracturing for well completion making the knowledge of the fault systems crucial.

Attribute calculations including semblance, dip steered semblance, and curvature gave information of the potential locations of fractures. Azimuthal anisotropy data mapped at various depths showed the correlation of the velocities with the potential fault locations taken from the attribute results. Results of the mean and dip curvature calculation provided the clearest evidence of faults.

Azimuthal anisotropy is another useful tool in fault characterization. Fractures, when perpendicular to wave direction, slow the rate of propagation of a seismic wave. Methods such as amplitude versus azimuth (AVAz) and velocity variations with azimuth (VVAz) use amplitude and velocity as a function of azimuth. However, migration of data removes the needed azimuthal information. Offset vector tiling (OVT) processing is a solution to the migration problem. OVT groups seismic data by common offsets and common azimuth, which allows the saving of azimuthal information in the migration process.

Attributes, such as dip curvature and mean curvature provide insight to fault locations; however, combining OVT anisotropy data, horizontal slice attribute calculations, and horizon attribute calculations improves the confidence in fault locations. The results of this study show that azimuthal anisotropy maps are in agreement with the major fault locations determined initially from the attribute calculations. OVT anisotropy data shows the fault orientations across the region, with an overall trend of faults in the NE-SW direction.


A thesis submitted in partial fulfillment of the requirements for the degree of Masters of Science in Geoscience: Geophysical Engineering Option