ConocoPhillips School of Geology and Geophysics
Mewbourne College of Earth and Energy
University of Oklahoma

Room 872 Sarkeys Energy Center
100 E. Boyd, Norman, OK 73019-1009

Principle Investigator: Kurt J. Marfurt
Co-investigators: Tim Kwiatkowski, Marcilio Matos
Financial Administrator: Nancy Leonard

Click here to Download Consortium Proposal

Executive Summary

During the past decade, seismic attributes have become not only well-accepted by the interpretation community, but have also become a key component in quantitative reservoir characterization. In addition to enhancing individual faults and discontinuities, geometric attributes help interpreters map axial planes for structural analysis, relate curvature to intensity and orientation of fractures, and map lateral changes in reflectivity to detect channels below seismic resolution. During the 2009 AASPI Consortium research program, we will continue our focus on subtle stratigraphic, diagenetic, and structural features such as thin-bedded turbidites, gas shales, karsting, and fracture systems. Our research is driven by the data provided by our sponsors, such that our primary efforts will be on the calibration of these attributes over mature fields of North America (US, Canada, and Mexico) that have a combination of proprietary 3D surveys, production data, well logs, image logs, and core, within a well-understood geologic framework.  We will extend our calibration to prestack data in two ways - in the selection of processing parameters and velocity model building, and in the application to offset- and azimuthally-limited volumes. We believe that a better understanding of the impact of acquisition, processing, and imaging on seismic attributes is key to quantifying the errors in reservoir characterization and hydrocarbon estimation provided by modern attribute-driven geostatistics, neural nets, and clustering technology.  In addition to research reports, we provide algorithm source code to all sponsors and attribute volumes to those sponsors who provide us 3D seismic data.

A. Scope of Work

When properly calibrated to well-log and production data, high-frequency 3D seismic data play a key role in defining reservoir heterogeneity and compartmentalization. We have found that modern seismic attributes, including coherence, reflector curvature, reflector rotation, coherent energy gradients, seismic textures, and spectral decomposition, greatly improve our ability to visualize stratigraphic features that are considerably below the classical limits of seismic resolution. Recently, we have observed that attribute images computed on offset- and azimuth-limited volumes from North and West Texas have higher lateral resolution than those computed on full offset and azimuth volumes. We have also observed that the illumination of stratigraphic features varies with offset and azimuth. The smearing of lateral discontinuities and subsequent loss of resolution is most problematic on land surveys that are rich in azimuths and subject to heterogeneous shallow surface effects including topography, weathering zones, and stress-induced anisotropy. Our research effort has three major themes: (1) to enhance our ability to map reservoir compartments and delineate fractures, (2) to use attributes to drive seismic processing work flows that will improve lateral and vertical seismic resolution, and (3) to calibrate features seen on seismic attributes in the context of tectonic deformation and seismic geomorphology. 

B. Target Sponsors

Our goal is to serve as a research component for independent and intermediate-sized oil and gas companies, domestic North American business units of large companies, and National Oil companies dealing with imaging subtle structures in land data and shelf environments. In exchange for geologic insight, well control, 3D seismic data, and consortium funding from sponsor companies, we will apply state of the art technology, specialized expertise, and most important, the time necessary to generate and evaluate emerging technologies that can impact costly drilling decisions.

C. Deliverables

We believe that technology is best understood when it is applied to the sponsoring company's own data. Our deliverables will therefore include:

• Generation of a full suite of geometric attributes to those sponsors who provide us with 3D seismic data. Ideally, each sponsor would provide us with a first data volume with good well control that could form the basis of publications and theses. If desired, we will process a 2nd volume that may be of more immediate exploration interest that can be held in confidence.
• Source code for all new and previous developed algorithms. Currently available algorithms including structure-oriented filtering, volumetric dip, azimuth, spectral curvature, coherent energy gradients, various edge detectors, spectral decomposition, volumetric generation of rose diagrams, composite attribute display, post-migration footprint suppression, volumetric GLCM texture analysis, and filtering of dip volumes. Sponsors may use these codes in any way they wish except software resale (which requires a separate agreement), including providing services to others.
• Copies of all AASPI thesis proposals, posters, preprints, expanded abstracts, and technical papers.
• .Generation of geometric attributes or data analysis (having some geological or geophysical research component to comply with OU tax exempt status) on proprietary data at time and materials cost
  

Kurt Marfurt

Kurt Marfurt is a geophysicist with 31 years experience in seismic attribute analysis, algorithm development, data processing, and interpretation. At Amoco, he played a role in developing both coherence and spectral decomposition algorithms and lead Amoco's attribute calibration team. Kurt served as director of first AGL, then CAGE, at the University of Houston. The OU AASPI effort is the current phase of this continuing work.

Tim Kwiatkowski

Tim Kwiatkowski is a geophysicist with a Ph.D. in Physics and 21 years experience in computer science, data processing, AVO, and spectral analysis. Before rejoining OU, Tim implemented software solutions at Fusion..

Marcilio Matos

Marcilio Matos is a geophysicist with a Ph.D. in Electrical Engineering who is working with AASPI as a visiting scholar during the time period January 2008-December 2009. Marcilio has expertise is digital signal processing, spectral decomposition and self-organizing maps. Marcilio is currently on leave of absence from the Brazilian Military College where he teaches and conducts research on behalf of the Brazilian petroleum industry.

Resources

The bulk of the work will be conducted within OU’s Crustal Imaging Facility (CIF) with computationally intensive work being done at the Oklahoma Supercomputing Center (OSCER). Through the generosity of commercial software vendors and service companies, CIF has onsite installations of state-of-the-art software products in seismic interpretation, processing, imaging, modeling, visualization, reservoir calibration, and reservoir simulation that exceeds that available to employees at many independent oil companies.  Through close collaboration with oil and service companies with OU, or through grants with the U.S. Department of Energy, we have been able to obtain licenses to several modern data sets that will allow us to both test and calibrate our new developments in seismic attributes.

Project Leverage

The bulk of this work will be through shared students with other faculty and consortia within OU. At OU, we will draw on expertise on rock physics from Chandra Rai and Carl Sondergeld’s rock physics consortium, on seismic geomorphology from Roger Slatt’s effort on mapping turbidites and calibrating them with outcrop and image logs, on mapping fractures from acoustic emissions with Deepak Devegowda, in addition to geophysical processing and interpretation guidance from Roger Young and Randy Keller.