Traditional Approaches to Stratigraphic Interpretation

Stratigraphic interpretation from wireline logs is typically drawn from multiple log traces across multiple wells. The multiple curve display is often supplemented by crossplots of log data. In areas of mixed lithologies and complex stratigraphy such as the thinly bedded mixed carbonate siliciclastic units typical of shallow shelf reservoirs, well to well correlation can be difficult to interpret. Increasing the number of types of logs may improve the accuracy of correlations, but often reduce well cross-sections to a mass of spaghetti-like wiggle traces. In contrast crossplots make compositional end members readily apparent, but the crucial information on relative depth is lost. As a result both multiple curves and crossplots can readily depict single well vertical changes in lithology or reservoir quality, but lateral relationships even among a small number of wells are not readily visualized.

As mentioned in the introduction analysis on a well-by-well basis is a time consuming technique that is useful for examining a single well or a small area. However, for large fields or basins one can get bogged down in the problems of handling the large volumes of data on a more or less ad hoc basis. This is certainly true for manual methods where paper logs can cover walls and form large piles in the corner. Even using computer assisted stratigraphic analysis for larger data sets, the data handling burdens remain a significant obstacle for interpretation and visualization. For example tying correlation loops on multiple horizons through even 1,000 wells can take several months of effort. Trying to visualize such a data set to verify correlations is even more of a challenge. One is limited to displaying only a very small subset of the wells at one time.

As result of these constraints typical studies of large fields or basins are limited to the petrophysical analysis of a small number of wells, construction of selected geologic cross-sections and mapping on only a few "key" horizons. If undertaken with care and knowledge, such a traditional approach has been proven to be productive. The use of the computer workstation has further increased the productivity of this approach to subsurface geologic analysis. However, the traditional geologic approach differs from a typical computer assisted seismic analysis in that the geophysicist examines seismic data as a volume (or section) not on a trace-by-trace basis. The methodologies of the geologist and the geophysicist vary in terms of data format, handling, analysis, and display. A geophysicist with even a rudimentary computer workstation can interpret across 12,000 traces before lunch. Can one adapt some of these techniques and approaches to treat well data as a "seismic volume" so that the geologist can examine 12,000 wells with the same efficiency? This question has lead to the development of what we refer to as the pseudo-seismic approach to stratigraphic interpretation.