Abstract

Seismic curvature analysis of 2-D surfaces is commonly used to locate or predict zones of open and closed fractures. Unlike older seismic curvature attributes, the new multi-trace attributes being developed at the University of Houston operate directly on the entire 3-D seismic volume and do not require pre-analysis interpretation of horizons. These attribute are mathematically independent of, and are interpretationally complementary to, well-established coherence attributes. We have developed in excess of seven multi-trace reflector curvature attributes (Al-Dossary, and Marfurt, 2004) that include strike, dip, mean, positive, negative, and Gaussian curvature (Figures 1 and 2). These curvatures can also be combined into shape indices (indicating ridges, domes, bowls, valleys, saddles). Curvature attributes are amenable to spectral analysis, which differentially highlights features of a few tens to several hundreds of feet (meters). In this presentation, we present insights gained from using curvature attributes to image geologic features in a seismic survey from the Fort Worth Basin (Figure 3) where collapse chimneys vertically extend some 2300 ft (> 760 m) from the Ellenburger carbonates into Pennsylvanian siliciclastics (Figures 4 and 5).

We find that the mean curvature (Figure 6) highlights small-scale lineaments that are extremely useful in interpreting local stress regimes. In Figure 7, and we compare the mean curvature images with more familiar coherence images. The most negative and most positive (also called principal) curvatures are effective in rapid time-slice mapping of lineaments, faults, folds, and flexures (Figure 8 and 9). Long spectral wave Length curvature estimates are of particular value in recognizing subtle, broad features in the seismic data, including field scale joint systems and compaction features (Figure 10). We use time slices and horizon extractions to illustrate the seismic secrets of the Ellenburger, and interpret changes in stress regime through time and temporal constraints on the formation of collapse chimneys. We track the evolution of the morphology of these collapse features from rhomboids formed along intersecting regional joint systems in the Ordovician interval through rounded, sink-hole shaped features in the Lower Pennsylvanian Marble Falls, to hummocky compaction features at the top of the Middle Pennsylvanian Caddo Limestone.