Mineralogy-based brittleness index is the simplest and most widely used index in characterizing the brittle property of shale gas/oil reservoirs. Traditionally, mineral compositions are usually determined by laboratory X-ray diffraction (XRD) analysis or interpreted from element logging, among which the former is time-consuming and non-continuous while the latter is costly and not timely. In this paper, a new rapid and economical method is proposed to obtain the mineral composition of shale from a statistical perspective, in the case of the Wufeng-Longmaxi shale, from the southeastern Sichuan Basin, China. The targeted shale cores from one scientific exploration well were scanned in situ to obtain the elemental data by using a non-destructive portable X-ray fluorescence (XRF) and sampled at each scanned site for laboratory XRD analysis to acquire the mineral composition data. Then we established the conversion models for shale with various sub-lithologies (including calcareous shale, sandy shale, and carbonaceous shale) based on a series of multiple regression analyses. These models were verified with the good consistency between predicted and tested mineral compositions of shale cores from another scientific exploration well. The mineralogy-based brittleness characterization results also indicated small differences between the predicted and tested brittleness index values, with the brittleness index curves matching well with the variations of experimental total organic carbon (TOC) with depth and conventional geophysical logging curves. The findings shed light on the application of portable XRF in determining mineral compositions and further the brittleness index of shale, which is essential for fracability evaluation and hydraulic fracturing design of shale gas/oil reservoirs.

Highlights.