Integration of XRF and shades of grey profiles for in-depth characterization of shale properties

Abstract

Integration of XRF and of shades of grey profiles has been tested on cores and on several hundreds of drill cuttings samples pertaining to several Duvernay vertical and horizontal wells.

Very high resolution images have been used to extract a shade of grey profile for complete Duvernay and Montney cored intervals. These profiles have helped define detailed sequence stratigraphy as minima and maxima clearly showed trends that could be associated to basin related changes occurring in neighboring wells at the same stratigraphic levels.

The study also looked at differences in output and usefulness of shades of grey profiles between various core photography set-ups, i.e. between high resolution (200 microns per pixel) traditional lab photo equipment and very high resolution photos (35 microns per pixel) taken on a rolling bench. The results integrated with work on other images from various laboratories on the same cores have delivered a series of criteria and patterns useful to recognize lighting issues that would minimize the usefulness of shades of grey profiles.

For a test study, shades of grey have also been extracted using a line scanner at a resolution of 200 microns per pixel on several extensive series of drill cuttings. Ten successive XRF measurements for each cutting vial allowed for statistical analysis of 26 elements and gave a solid data set to evaluate the shade of grey approach. These ten XRF measurements per vial were compared to single XRF value from a handheld XRF device to outline similarities and differences linked to the tools. Prediction of total organic carbon was then tested using the two sets of XRF mentioned and Leco TOC done on many of the samples.

Additionally, the study generated a series of workflows that successfully tested frac placement prediction using XRF data (elements, ratios and multilinear regression). It is important point to strengthen the analysis by analyzing frac stages that would have limited facies changes as evidenced by XRF; thus, stages that would saddle two very different lithologies should not be incorporated in the initial learning data set.

Both XRF and shades of grey can be extremely useful to geologists and engineers to better understand what will make a good hydraulic frac, but calibration and quality control are essential.

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