Technological advances in equipment and computer software have enabled new approaches to the generation of mineralogical datasets at petroleum wellsites. These datasets are currently utilized by hydraulic fracturing engineers to assist in designing optimized fracture stage intervals in horizontal wellbores, rather than using evenly spaced intervals between treatment stages. Mineralogical data is generated from geochemical logging tools, drill cuttings, conventional cores, and rotary sidewall core plugs utilizing a variety of analytical instrumentation techniques. This paper documents a study to assess mineralogical datasets generated on comparable samples, focused on evaluating analytical limitations and variances, toward obtaining consistent mineralogical results.
Instrumentation typically used to generate these datasets include geochemical logging, x-ray diffraction (XRD), x-ray fluorescence (XRF), scanning electron microscopy-energy dispersive spectroscopy (SEMEDS), Fourier transform infrared spectroscopy (FTIR), and inductively coupled plasma techniques (ICP mass spectroscopy or ICP optical emission spectroscopy). Variables introduced into the analysis, in addition to the different analytical techniques, include sampling methods, sample type and size, sample preparation, drilling mud contaminants, lithological heterogeneity, and depth correlations between cuttings, cores, and logs.
As the study evolved it became clear that wellsite sampling and preparation protocols needed to be properly defined to assure sampling quality and an accurate sampling depth reference. Analytical equipment limitations must be fully understood, as should differences in measurement technologies in the laboratory, at the wellsite surface and downhole. Equipment destined for wellsite surface analysis was evaluated in a controlled laboratory environment using reference mineral standards and standard mixtures to understand testing limitations and refine mineral phase calculations. Mineral terminologies, classifications, compositions, and the resulting databases were reviewed for consistency. Multiple cuttings and core sample sets from conventional sandstones, carbonates, and current mudrock plays were sub-divided and analyzed to enable direct comparisons of generated datasets. Complementary testing was conducted to confirm data quality.
This study yielded increased confidence in wellsite and laboratory analyses, including caveats where necessary, procedural guidelines for each analytical technique, and verification of deliverables appropriate to unconventional mudstone reservoirs. The ability to compare lab, wellsite and log-derived mineralogy is valuable in obtaining confidence in final solutions from all sources. The resultant wellsite datasets, in tandem with additional wellsite analytics, enhance confidence in optimized fracture stage interval decisions.