Development of Raman Spectroscopy as a Thermal Maturity Proxy in Unconventional Resource Assessment

Abstract

Quantifying thermal maturity (TM) is crucial for properly evaluating an unconventional resource. However, the most widely used methods–vitrinite reflectance and programmed pyrolysis–may be inadequate for shales. Reflectance of solid bitumen (BR_o) has wider applicability in shales, because of greater abundance than vitrinite. However, additional techniques that can accurately and objectively characterize the TM of a sample are still needed. Raman spectroscopy has been proposed as a viable method for this purpose due to its ability not only to characterize molecular structure but also to delineate the changes occurring in organic materials as a result of maturation.

This presentation shares results from the completion of a two-part study designed to develop a Raman spectroscopy-based TM indicator. Immature shale samples from the Bakken and Duvernay formations were subjected to hydrous pyrolysis for 72 hours from 280°C to 360°C. Rock residues from both series were polished for analysis via optical microscopy reflectance at the U. S. Geological Survey (USGS) and confocal-laser scanning Raman microscopy at WellDog. After reflectance characterization, the series from the Bakken was randomized and anonymized for Raman microscopy in Phase I of the study. Here, for the Phase II Duvernay study, BR_o and T_max values determined by USGS were shared with WellDog.

Multiple hyperspectral maps were collected from each sample, resulting in tens of thousands of spectra per sample. Map areas were 5,000 µm², with a spectrum collected from every square micron. The organic carbon G and D bands in Raman spectra were fit algorithmically to a 2-peak model. This is admittedly an oversimplification of the complex underlying peak structure in the G and D bands, and the adequacy of the 2-peak model for these analyses will be addressed in future work.

Peak fitting results, as well as other parameters extracted from Raman spectra, were plotted against previously determined BR_o% values to determine which parameters best correlate with TM. Plotting two of the anonymized spectral parameters vs BR_o% in the Bakken series reveals an exponential trend with strong correlations (R²>0.8). Similar strong relationships occur in the Duvernay samples. These promising results will act as the foundation for a future Phase III study, in which the same methods and analyses will be applied to a series of samples from the Huron Member of the Ohio Shale that have been naturally matured at varying depths.

AAPG Datapages/Search and Discovery Article #90323 ©2018 AAPG Annual Convention and Exhibition, Salt Lake City, Utah, May 20-23, 2018