A Deterministic Lithology Model for the Green River-Upper Wasatch Interval of the Uinta Basin

Abstract

Log evaluation of the clastic and carbonate lithologies in the Green River and upper Wasatch formations of the Uinta basin is complicated by the complex mineralogy and thin interbedding of diverse rock types. Simple zoned models smooth out these differences to yield porosities and saturations that may on average be correct, but misleading on a bed-by-bed basis. Probabilistic or stochastic approaches have been advocated for this area, but suffer from a lack of transparency, require specialized software and expert users, and require advanced logging measurements such as elemental capture spectroscopy logs to solve the problem. Although appropriate for an operator with a large formation evaluation budget and full access to all logging data run in a well, probabilistic methods do not work as well with public domain data or for mapping large areas with incomplete competitor data. We present a simple, deterministic 4-mineral solution consisting of quartz, calcite, dolomite, and “mixed clay” without attempting to derive the minor mineral components or individual clay species. The solution only requires the four basic logging measurements that are widely available: gamma ray, density, neutron porosity, and Pe. The solution does require one significant simplifying assumption: there is no dolomite in the shaly fraction of the interval; it only exists in the clean formation fraction. We then solve two triangles in apparent matrix density (RHOMAA) – apparent photoelectric cross-section (UMAA) space: 1) for clean (non-shaly) rocks the quartz-calcite-dolomite triangle; and 2) for shaly rocks, the quartz-calcite-clay triangle. These are solved using a gamma ray filter to separate clean from shaly formation solutions followed by a matrix inversion approach of 3 linear equations for each branch. The results are re-normalized to sum to 1 without any negative components, and are filtered for bad hole or other adverse logging conditions. The proposed approach yields lithology solutions very similar to logging vendor supplied multi-mineral solutions in terms of bulk mineralogy. Mineral endpoints are easily adjusted by the user as the “quartz point” drifts in rocks with high feldspar contents, or if the clay point drifts regionally. Organic matter and in many cases pyrite can also be determined in a subsequent step. The method can be implemented in a general purpose software package (e.g. Petra, GeoGraphix) or in Excel.

AAPG Datapages/Search and Discovery Article #90216 ©2015 AAPG Annual Convention and Exhibition, Denver, CO., May 31 - June 3, 2015