Quantitative Mineralogy of Reservoir Rocks Using Fourier Transform Infrared Spectroscopy

Abstract

Abstract An inversion scheme was modified to quantitatively determine the mineralogy of reservoir rock samples using transmission Fourier transform infrared spectroscopy (FTIR). This inversion scheme adds apatite and aragonite to the previous inversion and is now capable of recognizing 16 minerals. The other minerals include quartz, calcite, dolomite, illite, smectite, kaolinite, chlorite, pyrite, orthoclase, oligoclase, mixed clays, albite, anhydrite, and siderite. The inversion predicts the mineral concentrations with an average error of +/- 1.1 wt % and within a standard deviation of +/- 2.1 wt %. It also found that 97 % of mineral concentrations are accurately predicted within +/- 5 wt %. These prediction errors are similar to published values for error associated with X-ray diffraction (XRD) which has become the industry standard for mineralogy analysis. The addition of apatite allows rapid and accurate quantification of mineralogy in gas shale reservoirs which have become a major play in U.S. natural gas exploration. Introduction Minerals are the building blocks of rocks. The mineralogy of a rock determines its physical and chemical properties. In a reservoir rock, physical properties such as density, sonic velocity, compressibility, and wetting properties contribute to essential reservoir properties like porosity, permeability, and fluid saturations. The chemical properties of a reservoir rock are especially important in drilling and production operations in order to predict how the formation will react when exposed to foreign materials such as drilling mud and production chemicals. The knowledge of mineralogy also provides information about the deposition and diagenesis of reservoir rocks which further helps in understanding flow characteristics in a reservoir. Detailed reservoir characterization is essential to understanding how to optimally recover oil or gas reserves. The inversion scheme predicts concentrations of quartz, calcite, dolomite, illite, smectite, kaolinite, chlorite, pyrite, orthoclase, oligoclase, mixed clays, albite, anhydrite, siderite, apatite, and aragonite. Apatite was added to the inversion scheme because of its presence in gas shale reservoirs. Shale gas has experienced a recent boom in the U.S. in plays such as the Barnett Shale of Texas and the Woodford Shale of Oklahoma. Aragonite is a carbonate mineral characteristic of young, shallow carbonate reservoirs, and is one of the four most common carbonate minerals. The addition of aragonite to the three previously included carbonate minerals, calcite, dolomite, and siderite, allows for evaluation of a full suite of carbonate minerals. FTIR Theory Mineralogy can be determined both qualitatively and quantitatively in a number of ways. The three primary quantitative methods are thin section analysis, XRD, and FTIR. Thin section analysis is the oldest of these three methods; however, accurate prediction of mineralogy from thin sections requires a well trained specialist, and with advancements in technology, thin section analysis is becoming a lost art. XRD has become the method of choice in industry labs. Despite XRD's prevalence in industry, FTIR offers advantages in that it is a more rapid method of analysis and can be applied at the well site. After sample preparations, it takes one to two minutes to collect an infrared spectrum. In contrast, XRD takes over an hour. With FTIR, many spectra can be collected and inverted in a relatively short period of time, allowing one to make accurate logs of mineralogy for many data points over large depth intervals in the wellbore at the well site in real time. FTIR's time efficiency is reiterated in the findings of Ruessink and Harville (1992). XRD has been noted to have "…some inherent problems in quantitative analysis mostly due to particle size and uncontrolled orientation effects coupled with natural variability in mineral diffraction spectra" (Matteson and Herron 1993). Published error data for both methods, however, show similar accuracy.