Mineral-Chemistry Quantification and Petrophysical Calibration for Multi-Mineral Evaluations

Abstract

There are a number of situations where modeling of mineral composition (as opposed to rock/lithology) adds real business value, for example; complex mineralogy leading to variable grain density means that more accurate estimation of porosity could be achieved through mineral volume analysis; in some reservoirs certain minerals may control permeability (eg diagenetic cements); the geomechanical properties of certain reservoirs may be controlled by abundance of key minerals; and mineral volume interpretations may help with well to well correlation. The mineralogical complexity of mudstone reservoirs has led to the increased usage of multi-mineral optimizing petrophysical models in recent times for estimating porosity, water and hydrocarbon volumes. A key uncertainty in these models is the log response parameter assigned for each log equation related to each volumetric variable. Default parameter values are commonly used and often need to be modified by taking into account subjective local knowledge or intuition in order to achieve a result that is considered acceptable. This paper describes the methods developed at Chevron for calibration of mineral log response parameters using core data (referred to as BestRock). Mineral log response parameters are controlled by the major and trace element chemistry of the individual minerals in the formation rock matrix. BestRock uses a non-linear approach to optimize whole rock chemistry with mineralogy to calculate individual mineral structural formulae and trace element associations from which log response parameters can be determined. Accurate quantitative phase analysis by X-ray diffraction (QXRD) and rigorous sample preparation are especially critical steps in the process. QXRD in combination with aliquot whole rock elemental analyses are processed using Chevron's BestRock optimization software to provide refined quantities of the mineral species present in the formation, their structural formulae, and their predicted wireline log responses. Calibrated petrophysical models are built from the information obtained from the QXRD and BestRock results. The method described herein provides an independent and robust method for determining petrophysical parameters that is independent of the interpreter, quick to implement, and is supported by real data. The application of this method leads to improved mineral volume predictions from petrophysical models.

AAPG Datapages/Search and Discovery Article #90217 © 2015 International Conference & Exhibition, Melbourne, Australia, September 13-16, 2015