Applications of High-Resolution X-ray Fluorescence (XRF) Elemental Data in Mineral Modeling, Brittleness Indices, and Chemostratigraphic Correlation in the STACK Play of Central Oklahoma

Abstract

The multi-target STACK play of central Oklahoma is one of the most actively drilled areas in the continental United States but regional and local facies-scale aspects of the petroleum system, specifically related to stratigraphic correlation, remain unresolved. High-resolution X-ray fluorescence (XRF) spectrometry allows for rapid, non-destructive, quantitative analysis of 25-30 elements and is proven to measure inorganic geochemical variability that aids in modeling mineralogy and total organic carbon (TOC), estimating brittleness, defining rock types with similar elemental composition (chemofacies) for zonation and correlation, and inferring depositional environments. Over 26,000 XRF analyses were acquired at a 2-inch interval from 20 cores covering various strata of the STACK play. The elemental database was integrated with 300 measured X-ray diffraction (XRD) mineralogy results, over 100 TOC measurements, and over 6,400 micro-rebound hammer measurements (proxy for unconfined compressive strength). Robust element-based models were constructed for mineralogy, brittleness, and TOC using these supporting data for calibration and hierarchal cluster analysis (HCA) was used to define chemofacies. Mineralogy is calculated from a deterministic model using stepwise element-partitioning logic and is constrained by the observation that plagioclase is on average 3-5 times more abundant than K-feldspar and chlorite is a minor part of the clay fraction. This model allows for mineralogy to be calculated at any resolution XRF analyses are collected. To estimate brittleness in the STACK, different mineral weightings must be used for cherty intervals, as chert drives the highest brittleness. In non-cherty intervals, carbonate largely controls brittleness. The well-established SiO₂-Zr decoupling relationship is very effective at identifying originally-biogenic silica in cherty intervals. While the presence of self-sourced hydrocarbons in the STACK is in question, there is no strong relationship between TOC and the common organic richness elemental proxies (V, Ni, Mo, U). Combining all these recognized relationships with chemofacies distributions aids in the goal of constructing a robust chemostratigraphic correlation framework for the play, as supplemented by data from historical cored and cuttings wells. The understanding of STACK reservoir compositions is then applied to cuttings from new horizontal wells to generate mechanical properties zones for improved completions.

AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019