Abstract

Gamma-ray logs representing Upper Pennsylvanian strata were correlated using computer-assisted techniques for distances of 220 mi (354 km) along a northwest-to-southeast line of section in western and south-central Kansas. The section crosses a carbonate-dominated shelf of the Virgilian and Missourian Stages that borders the Anadarko Basin. Correlations were based on log data sampled at 0.5-ft (15-cm) intervals. Manual correlation of cyclothemic-scale genetic units approximately 50 ft (15 m) in thickness provided initial constraints to the automated correlation using regionally extensive marine shales (maximum flooding surfaces and condensed sections). These marine-shale markers were correlated previously from a regional study of core, outcrops, and wireline logs. Correlations of natural gamma-ray response within these genetic units were derived using computer-assisted autocorrelation. Stratal units resolved include: (1) cyclothemic-scale genetic units (10's of meters in thickness); (2) longer term sets of cyclothemic-scale genetic units (10's to 100's of meters in thickness); and (3) meter-scale strata] elements including high-frequency cycles within genetic units. Classic onlap and offlap and forward- and backward-stepping stratal geometries are noted. Reliability of correlations was computed. Uncorrelated intervals result in triangular-shaped intervals of no correlation between wells. These are associated with onlap and offlap geometries. This is critical evidence to assess stratigraphic truncations, usually limited in manual correlation unless surfaces are described in core or outcrop or seen in seismic data.

Regional correlation of genetic units and coherent sets of genetic stratigraphic units strongly suggest important allogenic and basin controls on sedimentation. This computer-assisted method provides an opportunity to assist in development of consistent, finer scale correlations that augment traditional geologic analyses. Manual correlation of high-resolution genetic units using wireline logs has been limited by issues of consistency and subjectiveness. Computer-assisted correlation can minimize these issues, if properly calibrated (trained). Integration of manual and computer techniques will help place surface and subsurface studies on a more equal footing to help resolve controls on sedimentation.

Targeting stratigraphic traps is facilitated using digital well logs and computer techniques. Reservoir attributes including porosity and resistivity and computed variables such as bulk volume water, fluid saturations, and even hydrocarbon pay can be introduced into the correlated sections. Furthermore, false-color imaging of the petrophysical properties and computed information within the digital cross sections will help identify anomalies and potential drilling opportunities.