Study Area and Dataset

The studied dataset is a 3D dataset that was a significant part of the total data base amassed during a two-year study of the Bend Conglomerate reservoir system in Boonsville Field in north-central Texas. The 3D Boonsville field dataset (Hardage et al., 1996a,b) was made publicly available for educational training purposes for a broad range of industry and academic interests since 1996 by Bob Hardage from the Bureau of economic Geology at The University of Texas at Austin. A special property of this dataset is that it shows interesting karst phenomenon of deep Ellenburger carbonates, which generate collapse structures that have compartmentalized clastic Bend Conglomerate reservoirs 620 to 760 meters above the depths where collapse structures originate. The demonstration of the study area is shown in Figure 1. Figure 2 shows the strati-graphic column of the Fort Worth Basin (Pollastro et al., 2007). The formation of interest is the Ordovician-age Ellenburger carbonate section. The Ellenburger carbonate section creates near-vertical chimneys-like karst collapses.The 3D seismic dataset is shown in Figure 3 and a selected 2D profile (crossline=75) is shown in Figure 5a. It can be observed that around the depth of 1.25s, there are some karst collapses, which is however, not very clear. In the next section, I will show a much better delineation result using time-frequency decomposition.

boons
Figure 1. Middle Pennsylvanian paleogeographic map showing the Fort Worth Basin and other basins related to the Ouachita orogony and the Boonsville project area. The solid rectangle on the Wise-Jack county line designates the site of the 3-D seismic survey, curtersy of Hardage et al. (1996b).

barnett-strat
Figure 2. Strati-graphic column of the Fort Worth Basin, curtersy of Pollastro et al. (2007). The target horizon is the deep Ordovician-age Ellenburger carbonate section which causes numerous karst collapses and significant disruptions in the overlaying strata.

old3d
Figure 3. 3D Boonsville dataset after time migration.

tfsswt1-2d3d-0,tfsswt2-2d3d-0,tfsswt3-2d3d-0,tfsswt4-2d3d-0
Figure 4. 3D karst features delineation result using SSWT with different frequency intervals.

old,lowf-st,lowf-tfsswt2,lowf-stft,lowf-ltft
Figure 5. 2D seismic section (crossline=75) and its karst mapping results. (a) Amplitude section. (b) Low frequency slice of ST. (c) Low frequency slice of SSWT. (d) Low frequency slice of STFT. (e) Low frequency slice of LTFT.

Some structure-related anomalies, such as the faults, discontinuities, karst collapses, are not always easily viewed from the original amplitude-based seismic profile. There are many approaches in the literature which better highlights the structural anomalies. One of the most common methods is to define a new seismic attribute which is sensitive to the structural anomalies (Xie et al., 2015b; Wang et al., 2015). The 3D dataset from the Boonsville field contains typical karst phenomenon, as shown in Figures 3 and 5a. But, from the original seismic reflections, the phenomenon is not obvious.