Abstract: Quantification of Rock-Petrophysics-Log Correlation Using Digital Rock Classification

Robert M. Cluff, Alan P. Byrnes

Traditional rock description provides textural and bedding characterization which is often non-digital and/or accents properties which are unrelated to (or at a scale larger than) those measured by petrophysical and log analyses. Many consumers of geological descriptions (including engineers, log analysts) are unable to incorporate verbal-textual rock descriptions into their quantitative reservoir analyses.

Correlation between reservoir variables, efficient information transmittal, and computerization can be provided by a digital rock description which describes lithologic as well as petrophysical and log variables of principal interest (porosity/^phgr, permeability/K, grain density/^rgr, capillary pressure/P_c, anisotropy/^egr). A five digit system, where each digit varies from 0 to 9 and the increasing value corresponds to meaningful increases in properties, has been developed and found highly useful for formation evaluation and reservoir characterization studies. The five digit sequence describes the following rock/petrophysics/log variables: 1) major rock type/^tgr (0=coal, 1=sandstone, 2=limestone, etc.); 2) grain size, sorting/K, P_c, ^egr/shaliness (0=clay to 9 conglomerate); 3) consolidation, lithification/^phgr, K, ^rgr, P_c/log ^phgr, resistivity, invasion (0=highly cemented to 9=unconsolidated); 4) rock fabric, sedimentary structures/K, ^egr/resistivity and acoustic ^egr (e.g. 1=burrowed/churned, 4=wavy bedded, 8=parallel laminated, etc.); 5) cementation mineralogy/^rgr, K/grain density (0=sulfide to 9=no interstitial cement).

A typical sandstone could be succinctly described as "14575", equivalent to: Sandstone, fine to very fine grained, moderately well indurated/moderate porosity/unfractured, large scale crossbedding, calcite cemented. Rock codes can be keyed into databases, spreadsheets, and log analysis packages to facilitate predictive correlations between variables, multivariate regression analysis, and to link geologic descriptions to reservoir behavior.

AAPG Search and Discovery Article #90986©1994 AAPG Annual Convention, Denver, Colorado, June 12-15, 1994