Kansas Geological Survey, Current Research in Earth Sciences, Bulletin 241, part 3
Prev Page--Plattsburg Limestone, cont. || Next Page--Conclusions
Kansas Geological Survey, Current Research in Earth Sciences, Bulletin 241, part 3
Prev Page--Plattsburg Limestone, cont. ||
Next Page--Conclusions
High-frequency GPR can be an invaluable tool for enhancing outcrop studies because, with sufficient control, it allows detailed stratigraphic and lithologic information to be extended into the subsurface beyond the outcrop face in a relatively continuous manner. Detailed lateral and vertical stratigraphic and lithologic information is critical to understanding the complex heterogeneity of reservoirs and can aid in determining production strategies. As our study shows, the shapes of bounding surfaces and channel fills can be imaged in the near surface via high-frequency GPR methods. In addition, our study indicates that vertical imaging resolutions of high-frequency GPR data are sufficient to image many small-scale details associated with these features, such as crossbedding and fractures. Imaging of channel fills and crossbedding can sometimes be used to determine stratal geometries in three dimensions, which can aid in determining paleoflow directions, which are important for establishing three-dimensional geometries of reservoir flow units (Beaty et al., 1997). GPR can be used to image subsurface features and extend outcrop correlations into areas of poor or nonexistent exposure, as long as there is a difference in dielectric constants between subsurface media and there is sufficient control on lithology and geometries from nearby outcrops or cores. The variability in GPR-amplitude responses is related to changes in the media through which the GPR signal passes, and therefore can sometimes be an indication of lithologic changes. Fine-grained material, such as clay and silt, can contain more bonded water molecules than coarser-grained material, thus increasing their dielectric constant if wet. Bounding surfaces between lithologic units may provide the largest dielectric-constant contrasts if they contain finer-grained siliciclastic material. In our study, bounding surfaces were enriched with fine-grained siliciclastics from both depositional and diagenetic processes. In contrast, internal dielectric contrasts within limestone units in our study usually are of lesser magnitude than those of major bounding surfaces because they usually contain thinner, less continuous siliciclastic layers. The massive-bedded and argillaceous units of the Plattsburg Limestone study site illustrate the utility of identifying reflection character and correlating it with subsurface lithology. At this site, the massive-bedded units have almost no internal reflections, whereas the argillaceous units have strong reflections from their upper bounding surfaces and cause significant signal attenuation.
This study was also successful in imaging joints and fractures at both study sites. The characteristics, diffractions, and offsets (sometimes accompanied by slight velocity pull-downs, when soil-filled) of these joints and fractures were confirmed using the data from the outcrop face. Therefore, by identifying reflection characteristics and correlating them to lithology and bedding features at the outcrop, general lithological and sedimentary structural information can be extended into the subsurface beyond the outcrop face. As noted by Knight et al. (1997), knowledge gained from GPR data of the shapes, spatial distribution, and frequency of these features (which may greatly affect reservoir quality) can help create more realistic reservoir models. As shown in the study by Martinez et al. (1998a), small-scale (less than 0.01 m; 0.03 ft) lithologic heterogeneity that affects permeability can be imaged using GPR methods. The results of that study, combined with those of the present study, indicate that laminae, beds, minor and major bounding surfaces, and fractures (all of which may affect fluid-flow characteristics within reservoir strata) can be imaged using GPR and thereby provide additional data for reservoir modeling efforts.
Previous outcrop studies involving GPR have usually used lower-frequency antennas and therefore have much lower vertical imaging resolutions (Beres et al., 1995; Liner and Liner, 1995; Pratt and Miall, 1993). Such resolution may be adequate for imaging relatively large-scale features (major bounding surfaces and faults), but it is not sufficient for imaging the detailed features (thin internal bedding and crossbedding) that were the concern of this study. It is also difficult to determine subtle lithological changes from lower-frequency data because the changes may occur at scales much smaller than the antenna wavelength.
Our study also differs from most previous studies in its detailed correlation of photomosaics of the outcrop face with reflection information. Detailed correlation of GPR data with the outcrop face is critical for understanding the cause of GPR reflections at a study site and recognizing subtle reflection characteristics of the data that allow interpretations to include small-scale lateral and vertical subsurface lithologic and stratigraphic variability.
The successful high-resolution imaging of major bounding surfaces, fractures, and joints indicates that high-frequency GPR may be a useful technique for mapping features associated with sequence-stratigraphic boundaries, including those evidencing paleokarst and paleosol development from subaerial exposure. Many important reservoirs, including those in the subsurface of Kansas, are associated with major sequence boundaries showing extensive subaerial exposure and karst features. GPR studies of analogs may provide an additional tool for quantifying the dimensions and spacing of fractures, caves, and joints associated with such sequence boundaries. Additionally, GPR may also aid in regional correlations of sequence boundaries. Using GPR for three-dimensional mapping of surfaces associated with sequence-stratigraphic boundaries can assist in placing outcrop information within a sequence-stratigraphic framework and help understand basin-scale depositional history, as well as provide important data for reservoir modeling.
Clearly, as shown by the various limiting factors at our study sites, high-frequency GPR is not a panacea for all outcrop studies. Not only is it limited to areas of low surface conductivity (e.g., those lacking clays or shales), it is also very limited in imaging depth because of rapid signal attenuation. For example, the 500-MHz GPR maximum-imaging depths in the limestone units of this study were approximately 3-4 m (9.8-13.1 ft). The penetration-depth limitations can be reduced by using a suite of antenna frequencies to image the subsurface at different penetration depths and resolutions (Martinez et al., 1998b). Depth control of GPR data can also be a problem without adequate outcrop or borehole information to constrain possible GPR velocity values, even when using CDP gathers to determine velocity information. Having a detailed outcrop photomosaic to interpret alongside a GPR profile is critical if highly accurate, depth-constrained interpretations are necessary.
Prev Page--Plattsburg Limestone, cont. || Next Page--Conclusions
Kansas Geological Survey
Web version September 15 1998
http://www.kgs.ku.edu/Current/1998/martinez/martinez9.html
Comments to webadmin@kgs.ku.edu