Section 2: Pattern Recognition of Lithologies from Log Overlays and Crossplots
Up to this point, the explicit identification of lithology from logs has been almost entirely restricted to the differentiation of shales from non-shales. In some cases, lithologies or minerals with unusual properties can be recognized, such as those with anomalously high or low densities, including anhydrite, halite or coal. Otherwise, the distinction of sandstones from either limestones or dolomites on logs has been inferred from porosity character, stratigraphic correlation or drill-cuttings information. The nature of the matrix mineral composition is important to the log analyst, because he or she needs this knowledge in order to come up with accurate estimates of true porosity. The calculation of porosity from either the neutron, density or sonic logs is keyed to the log response of the matrix. The identification problem becomes even more complex, if the matrix is a mixture of several minerals, such as in a cherty, dolomitic limestone, further aggravated by compositional changes with depth. Log overlays and crossplots of porosity logs were introduced to resolve this problem, primarily as a means to derive true porosities. However, the lithological information from these methods are invaluable byproducts.
The three porosity logs (neutron, sonic and density) are recorded in radically different units of equivalent porosity units, microseconds per foot, and grams per cubic centimeter. If the logs are to be overlaid in a meaningful way, then they must be related to a common reference framework. A scale of equivalent limestone percentage porosity is the most commonly used reference. Because this is a common scale for neutron logs, they generally do not require to be rescaled. Bulk density readings may be converted to this scale by equating the grain density of calcite (2.71 gm/cc) with zero porosity, the porosity fluid density (approximately 1.00 for fresh water mud-filtrate) to 100% porosity, and interpolating other values between these two extremes (see Fig. 30).