The goal is to develop better ways to explore for natural gas in tight sand reservoirs by demonstrating the applicability of a new exploration technology that identifies areas of high density natural fractures (and, thus, high gas well productivity) in otherwise low-permeability (low productivity) geologic settings.
The purpose of this project is to demonstrate the applicability of a geomechanical exploration model using three-dimensional seismic data and local/regional stress data to predict the location and character of natural fracture clusters. This will be done by performing three field demonstrations in geologically diverse tight gas reservoir settings to establish workflows and techniques and illustrate the method?s applicability in stress sensitive reservoirs.
Advanced Resources International – Project management and all research products
Arlington, Virginia 22201
The geomechanical approach has proven most effective in areas of subtle structure where the reservoir behaves elastically and has proven to be less effective in areas where deformation has been more extreme and reservoir permeability is dominated by stress-insensitive shear fractures. Since most reservoirs contain a mixture of fracture types and scales, no single approach (viscoelastic modeling, geomechanics, or discrete fracture network modeling) is sufficient to answer all questions. While the application of geomechanics has grown increasingly complex, the increased complexity is yielding improved results as the role of fault-related stress in the reservoir is better understood. Integration of these techniques with some level or combination of seismic attributes (within a mutually supportive framework) is presently viewed as the most likely path to yield a true ?breakthrough? in naturally fractured reservoir exploration.
A set of tools and techniques has been developed for applying boundary element stress modeling to project stress distribution (and related permeability) within a reservoir. This technology combines 3-D seismic delineation of structures and faults with a computer model (Poly 3D from Stanford University, now available from IGEOSS [external site]) to describe the distribution of fault related stresses in a reservoir. Viscoelastic and discrete fracture network modeling have also proven to be important to the overall methodology. The potential for using such stress modeling to estimate in-situ bulk permeability within naturally fractured reservoirs was first observed in the Rulison area of the Southern Piceance Basin. There, the technology showed considerable promise as a tool for high grading locations to increase gas recovery (on a per well basis) within a development drilling program. However, the modeling software required field calibration, testing, and demonstration to achieve industry acceptance.
The primary objective of this project was to demonstrate the geomechanical technology in an exploratory setting by performing three field demonstrations. To date, two of the three field demonstrations have been performed. While specific results remain confidential, valuable experience in the appropriate application of geomechanics has been gained during these demonstrations. Geomechanics, as applied in this multi-site project, involves modeling the elastic stresses generated by local faulting in the reservoir interval and identifying drilling prospect areas by the nature and magnitude of the calculated stresses. Significantly, geomechanics has emerged as only one of several core technologies useful in effectively characterizing permeability related to natural fracturing at depth in a reservoir setting.
The initial field demonstration site was a deep Frontier Formation test in the Wind River Basin of Wyoming. The operator made the 3D seismic depth volume available for research and supplied horizon pick information. The project performer (ARI) interpreted the fault plane geometries and built a Poly3D model of the faulted structure. Coincidentally, a deep exploratory test was drilled during the modeling phase of the field demonstration. A large aperture shear fracture was encountered in the objective horizon and, fortuitously, cored. Post-drilling structural modeling and core interpretation indicated the well location was in an area of high potential for shear fractures and that the well had penetrated a small fault near the limits of seismic resolution. The large aperture of the fracture proved capable of flowing large amounts of water as well as gas and was an early indicator of practical problems associated with the basin-centered gas hypothesis. A second exploratory well also encountered significant amounts of water in a shear fracture setting.
The second field demonstration site was in the Anadarko Basin where the target was a deformed Paleozoic reservoir. Existing core from around the field site was described and incorporated into a tectonic synthesis for use in model construction. Faulting within the demonstration area was interpreted and a discrete fracture network model was built based on core data, mechanical modeling, and seismic interpretation. A combined permeability-porosity thickness map was built and calibrated to existing wells in the area. Twelve locations were identified and presented to the operator. Operating rules in Oklahoma create an intensely competitive business environment where production interpretations and drilling results can be used by offset leaseholders to the detriment of the operator. As a consequence, the detailed results of this field demonstration remain confidential for the protection of the operator?s business interests.
To date, the project has significantly advanced the general level of understanding regarding the nature, origin, and distribution of natural fractures in reservoirs. Burial and uplift history, lithology, temperature, and gas generation, along with local, fault-related elastic stress, are all now seen as influencing the distribution and type of fracturing and associated permeability in a reservoir. Geologic processes, as they affect a reservoir over time, tend to be irreversible in nature. Geomechanics, as implemented in this project, is a modeling scheme focused on the distribution and impact of elastic (reversible) stresses in the reservoir. It may not reflect the amount of fracturing in the reservoir as much as, for example, the apertures of existing fractures and their associated permeability. Thus, a discrete fracture network model of Type I extension fractures, based on available statistics, may be the best representation of the fracturing distribution (location, geometry) within the reservoir. Modeling the modification of the fracture network apertures by the distribution of stress in the reservoir is a second step that is necessary to approximate the distribution of stress sensitive permeability within the reservoir and achieve the true objective of the project. Such added steps were not originally envisioned in the project SOW. The application of geomechanics in field demonstration projects outside the Pieance (where its effects were first noted) is allowing the true scope of the problem to be delineated.
The third demonstration project was initiated as part of the Colorado School of Mines, Phase X Reservoir Characterization Project (RCP) [external site] study in Rulison Field. During this third demonstration phase the geomechanical modeling approach developed previously was used to establish location specific estimates of probable recovery (EUR) for wells within RCP?s field demonstration area. Initially, wellbore productivity was correlated against specific geomechanical attributes. Early analysis demonstrated a clear relationship between areas of higher productivity and the penetration of reverse faults. Subsequent work has further substantiated these early findings, however, the geomechanical modeling strategy shifted from a 2-D horizon based approach to a 3-D volumetric approach. This allowed researchers to use depth slices on the geomechanical modeling results to better match the seismic data and further refine the preliminary correlation between faults/fracturing and geomechanical stress regimes.
Current Status and Remaining Tasks:
This project is complete and the final report is available on-line below under "Additional Information".
Project Start: October 1, 1999
Project End: March 31, 2006
DOE Contribution: $1,293,363
Performer Contribution: $9.857 MM (including in-kind)
NETL – Kelly Rose (firstname.lastname@example.org or 304-285-4157)
ARI – Randy Billingsley (email@example.com or 303-295-2722)
Final Report [PDF-7.29MB]
Images extracted from Appendix A - Core to FMI Photos of Core
Colorado School of Mines Reservoir Characterization Project Website [external site]
Maps of test sites [PDF-89KB]