The objective is to develop methods to exploit viscoelastic rock and fluid properties to greatly enhance the sensitivity of surface seismic measurements to the presence of hydrocarbon saturation. Researchers plan to use well log, lithology, production, and 3-D seismic data from an oil reservoir in north central Texas. The proposed program provides a methodology for processing and interpreting surface/subsurface data to determine reservoir properties and for applying the information to predict hydrocarbon locations and volume estimates in complex geological environments.
Southwest Research Institute
San Antonio, TX
Seismic attenuation is a physical process governing the reduction in amplitude of seismic energy as it propagates through the ground. Seismic attenuation is caused by a number of physical mechanisms, some of which are poorly understood. These include viscous and viscoelastic wave damping by internal frictional forces in the rock and its saturating fluid, as well as wave scattering at lithological boundaries and fractures or joints. The influence of rock pore structure and saturating fluid on attenuation give some hope that attenuation may be useful as an indicator for various reservoir properties. At the same time, understanding its effect on seismic wave propagation can lead to processing for improved surface seismic images. Since seismic amplitude measurements also are affected by other parameters, including geometric spreading and rock stiffness, measurements of seismic attenuation are most commonly performed using the spectral ratio method. This method exploits the fact that higher frequencies in seismic waves tend to attenuate more rapidly because of their smaller wavelength. The spectral ratio method has been fairly successful in extracting attenuation values from transmitted seismic data, such as those recorded in vertical seismic profiles or cross-well experiments, but has had less success in analyzing seismic reflection data. This is because seismic reflection data is more strongly affected by the presence of closely spaced layers in the rock, whose combined reflections can appear to change the spectral content of the wave.
Acoustic borehole wave attenuation is a power attribute that can be used as indicators of lithology, pore structure, fractures, and clay and fluid content in a reservoir characterization program. Since acoustic attributes from full waveform sonic logs can be easily related to petrophysics and core data in a single borehole, the intrinsic and scattering effects at the borehole scale can be determined. Since the early 1980s, various researchers have attempted to develop techniques to determine Q from single-hole full waveform acoustic log data. It is, however commonly recognized that existing Q processing techniques have major limitations and pitfalls. In particular, these techniques have serious difficulties when the lithology and acoustic properties are highly heterogeneous in the depth direction. Unfortunately, in the sonic signals, the influence of the formation heterogeneity is coupled with the interaction among internally reflected rays of the borehole head wave.
Algorithm to extract Q-logs from full waveform sonic logs was developed and applied to a Waggoner oil reservoir. The application of Q -logs for detecting oil saturation at well locations was successful. The results were presented at the 2005 SEG International Meeting in Houston, TX. The results of the Q-logs were published in The Leading Edge magazine in February 2006. The algorithm was also applied to extract Q from a carbonate rock formation. The results show that Q is sensitive to vuggy and matrix porosities, which suggests that Q logs can discriminate between lithology and provide information on the pore structure when integrated with cores and other logs. These results were published in the Journal of Applied Geophysics in March 2006.
In addition, software to extract Q from 3-D surface seismic was developed and applied to a Waggoner oil reservoir. This software was applied to examine the effects of lateral variability on Q measurements from surface seismic data. This study show that Q measurements are horizontally averaged over region analogous to the Fresnel zone. However, we also found that abrupt lateral changes in seismic medium properties can introduce diffractions and other waves that closely follow the direct and reflected waves. This often results in interface which manifests itself as spectral distortions.
Finally, the ultrasonic experiments on scaled realistic-to-borehole acoustic models were completed. Systematic ultrasonic tests were conducted on a series of borehole specimens to produce a library of scaled sonic micro-seismograms to assist on the interpretation of full waveform sonic logs to better interpret Q -logs. A paper in this subject was submitted for presentation to the SEG 2006 International Meeting.
As an important byproduct, researchers have implemented an imaging and processing algorithm for the event/pattern recognition of micro-seismograms. They can extend this algorithm to the interpretation of existing and/or new full waveform sonic data.
The results of this project represent an advance in the state-of-the-art for providing Q measurements from the field. The lack of reliable methods for field measurement of seismic attenuation was cited at the 2005 SEG Development and Production Forum as being an obstacle to the further development of the technology. Many researchers are hopeful that accurate measurements of seismic attenuation can lead to improved seismic processing; saturation indicators; and indirect signs of pore structure, lithology, and moveable fluids. Knowledge of these parameters in turn leads to improved reservoir characterization and more-efficient production practices. Since the technique focuses on direct detection of hydrocarbons, it will decrease the number of dry holes. Thus fewer wells will be required to extract the hydrocarbons, which will reduce environmental impacts. Not only will the results of the project will be of immediate interest to small and large regional producers, but the methodology developed is general enough to be applied to a range of domestic petroleum production sites, including deepwater and overthrust regions.
Researchers successfully applied the surface seismic Q extraction method to a 2-D data set over a Florida aquifer. The presence of numerous thin beds produced tuning effects in these data, which complicated the Q estimation. However, the team was able to overcome this problem by using well logs to correct for the tuning distortion in the field data. In same aquifer, the sonic Q algorithm was applied to process full waveform sonic data. The main accomplishment of this application was to demonstrate that Q can indeed be used as a lithological indicator. The Q-logs can distinguish vuggy porosity zones from those containing single porosity zones. This has allowed us to relate permeability with Q -logs. At the Waggoner Ranch site, researchers applied the surface seismic Q extraction method to a 3-D data set. They were not successful in obtaining consistent estimates of Q from this site. Like the Florida site, the Waggoner site has numerous thin beds that create tuning effects. However, well control at Waggoner is sparse, and none of the few acoustic well logs tied accurately to the seismic data. At the same time, the Waggoner geology has greater lateral variability, and the seismic data had lower frequency and spatial resolution than at the Florida site. For these reasons, these thin, small, north Texas reservoirs appear to be poor candidates for surface seismic Q-attribute estimation at this time. In order to demonstrate the usefulness of the processing seismic method developed in this project, researchers conducted a model study to analyze lateral subsurface variability effects. The study concluded that Q estimations from surface seismic data should be evaluated cautiously in regions containing faulting, abrupt changes in dip, or rapid horizontal variations in saturation or facies. All of these geological features can produce diffractions and other waves that follow closely behind the direct and reflected waves in some propagation directions. It is expected Q estimation from surface seismic to work best in formations with thick, uniform beds; good well control; and no abrupt lateral changes. It will also work best when seismic source has a large bandwidth, and noise is low. Testing our methods in reservoirs of this type would help to establish the technique.
The Q algorithm was validated using sonic logs from Waggoner oilfield. The Q-logs were correlated with the petrophysics, including a sand-shale sequence containing carbonate markers. The high velocity contrast between the carbonate units and shale produced attenuation anomalies that can be easily recognized by an interpreter. This was an important issue, because the interpreter can differentiate them from those attenuation signatures associated with oil saturation that occur in permeable sands.
Ultrasonic experiments were carried to simulate field sonic micro-seismograms using realistic earth and borehole models on a 1/10 scale. They conducted systematic ultrasonic tests on a series of borehole specimens and built a library of scaled sonic micro-seismograms. The signatures of material variations and borehole washout show patterns that resemble realistic borehole data. The results indicate ultrasonic borehole molding can be used to verify the processing algorithm and has potential for being used to interpret field sonic data. The project team found the ultrasonic borehole measurement to be an economic alternative to theoretical/numerical approaches for modeling sonic waves under specific borehole and formation conditions. With the experience we gained, they can design, set up, and perform ultrasonic tests on any new borehole and formation models of interest.
Although, the project was completed on March 31, 2006, they have submitted two papers to Geophysics, and one presentation for the Society of Exploration Geophysicist International meeting that was recently accepted. One of the papers describes the processing algorithm to extract Q-logs from sonic data, and a second paper reports the modeling approach and results of Q estimations from surface seismic data. The presentation will be based on the experimental laboratory-scaled models to interpret sonic data.
$220,000 (20% of total)
Parra, J.O., Haclert, C.L., and P.C-Xu, 2006, “Attenuation analysis of acoustic waveforms in a borehole intercepted by a sand-shale sequence reservoir, The Leading Edge, 25, 186-193.
Parra, J.O., Hackert, C.L. and Bennett, M., 2006, “Permeability and porosity images based on P-wave surface seismic data: Application to a south Florida aquifer, Water Resources Research, 42, W02415, doi: 10.1029/2005WR004114.
Parra, J.O., and Hackert, 2006, “Modeling and interpretation of Q logs in carbonate rock using a double porosity model and well logs, Journal of Applied Geophysics, 58, 253-262.
Parra, J.O., Hackert, C.L., Bennett, M., Jervis, M. and Collier, H., “An integrated approach based on NMR/acoustic techniques to map permeability in carbonate aquifers: from the pore to field scales”, in D.K. Butler, ed., Near-Surface Geophysics, SEG, Investigation in Geophysics Series No 13, 2005, 473-489.
Hackert, C.L., and Parra, J.O., 2004 “Improving Q estimates from seismic reflection data using well-log-based localized spectral correction.” Geophysics, vol. 69, 1521-1529
Parra, J. O., Hackert C.L. , Bennett, B., and Collier, H.A. 2003, “Permeability and porosity images based on NMR, sonic and seismic reflectivity: application to a carbonate aquifer.” The Leading Edge, 22 , 1102-1108.
Parra, J.O., Hackert, C.L.,and Pride, S.,”A double-porosity poroelastic model to relate P-wave attenuation to fluid flow in vuggy carbonate rock,in O. Stephansson, Hudson, J, and Jing, L., eds., Coupled Thermo-Hydro-Mechanical-Chemical Processes in Geo-Systems,2, Elsevier, 2005, 483-488.
Hackert, C. L., and Parra, J. O. 2003, “Estimating scattering attenuation from vugs or karsts,” Geophysics, 68, 1182-1188.