Round robin comparison of composition based mineralogical studies of unconventional reservoirs

Understanding the mineralogical composition of reservoir rocks is a vital component of hydrocarbon recovery, especially in unconventional reservoirs, which are the main targets of hydrocarbon exploration today. Obtaining accurate mineralogical data from unconventional hydrocarbon reservoirs has its challenges though. XRD and XRF are commonly applied geochemical techniques that are used in the oil and gas industry in order to obtain mineralogical data, but these techniques have their challenges. This project compares four geochemical techniques (PXRF, conventional XRF using WDS and Uniquant, and XRF powder analysis) for the compositional analysis of unconventional reservoir rocks. The results from the differing approaches are compared against the petrophysical log data taken from the unconventional reservoirs used in this study to test the reliability, reproducibility, accuracy and precision of the techniques and the mineralogical interpretation of logs that are used to interpret sub-surface geology. The reservoir rocks used in this study were taken from forty sidewall cores from reservoirs in the Permian Basin of Southeastern New Mexico. This research concluded that the Matrix Identification (MID) plots, based on petrophysical data, for the three wells predict the dominant minerals present in each reservoir, especially if the correct bulk density values are used. Glass discs analyzed in WDS mode is proposed to be the most accurate and precise of the four XRF methods used in this study, particularly for calcite-poor samples, but for samples that contain >20% CaO or the presence of sulfur, WDS glass is the least accurate of the four methods of XRF. Glass discs analyzed by the Uniquant method is proposed to be the most accurate method for carbonate rich samples unless there is a high abundance of sulfur. PXRF is the most accurate of the four methods for these sulfur rich samples, as the calculated mineral abundances (via normative calculations) from the PXRF data, is the most comparable to the XRD data. And lastly, the XRD data ran in triplicate shows high reproducibility, especially for samples that contain less than 10% clay minerals.