PETROPHYSICS-BASED GEOLOGY: INTEGRATION OF MINERALOGY, SEDIMENTARY PETROLOGY, STATISTICS, AND STRATIGRAPHY USING DIGITAL WELL LOG ANALYSIS SOFTWARE

The teaching scenario is college-based geology or earth science majors who have already had introductory courses in physical and historical geology. Specifically, we have used this medium as an extended laboratory exercise within a sedimentation-stratigraphy class. A petrophysical software application (GAMLS), available to educational institutions, is used as the geologic analysis “tool”. Data are digitized well logs (available from some state geological survey web sites).

The theme is rock properties (mineralogy and chemistry), reservoir properties (porosity, permeability, and water saturation), stratigraphic sequences, and correlation. After an introduction to well logging principles, students are requested to import well log data, QC the data, delineate depths with similar character (similar well log signal), and then to determine probable rock types (assuming that no core data is available). This is accomplished using a probabilistic clustering analysis procedure and an expert system that assigns each cluster to likely rock types (sandstones, siltstones, shales, limestones, dolostones, etc.). Students are asked to rationalize the expert assignments with the data and to override the data as desirable.

Emphasis is given to understanding how rock properties (e.g., bulk density, hydrogen content, natural radioactivity, sonic velocity) are determined using well logging procedures and how these properties are used to interpret rock types. Multiwell clustering provides information for generation of cross sections from which depositional cycles and sequence stratigraphy concepts can be interpreted and discussed. The method both permits and requires integration of geologic concepts from different sub disciplines: chemistry, physics, math, mineralogy, petrology, stratigraphy, and tectonics. Exercises can be done individually or as a group.

An example project would be: Analyze well logs from a few wells in south central NYS and create a cross section showing the stratigraphy defined by the clustering analysis and interpreted rock types. Compare this stratigraphy with the known Paleozoic stratigraphy, and then relateto Sloss sequences and sequence stratigraphy.