Relationships between the physical properties of Triassic sandstones in the West Netherlands Basin and short-wave infrared reflectance spectroscopy.

Abstract

Short-Wave Infrared Reflectance (SWIR) spectroscopy is a powerful and effective tool for identifying phyllosilicates, carbonates and other mineral groups. Hyperspectral imaging allows for quick, consistent and spatially continuous data analysis. This study investigates whether correlations between the hyperspectral data from short-wave infrared spectroscopy and the physical rock parameters, such as porosity, permeability and mineralogical composition, can be established. Hyperspectral image acquisition was performed from two cored intervals of Triassic sandstones in the West Netherlands Basin, the Naaldwijk-Geothermal (NLW-GT-01) and the Barendrecht-Ziedewij (BRTZ-01) borehole. The slabbed cores were specifically chosen due to their similar lithology yet significant difference in porosity and permeability values. For each pixel (256 μm), the wavelength positions and depths of the three deepest absorption features in specified wavelength intervals were calculated. Calculations were generated with Hyperspectral Python software (HypPy3) and the wavelength mapping technique described by Van Ruitenbeek et al. (2014). These wavelength maps and other hyperspectral parameters that served as proxies were used for spectral algorithms to classify the pixels. The results provided semi-quantitative mineral abundancy maps, detected compositional variation between equivalent minerals and successfully distinguished phyllosilicates, which other methods would not have been able to. However, non-hydroxylated minerals (e.g. silica, feldspar) are non-responsive within the SWIR region of the electromagnetic spectrum. The results advocate that the cored interval of the NLW-GT-01 was extracted from a higher metamorphic intensity zone and was subjected to more significant amounts of carbonate cementation than the BRTZ-01 core, both being mechanisms for porosity loss. The findings in this research are in accord with other studies that have examined the regional geology and/or these boreholes (e.g. De Hoo 2021; Maniar 2019; Ursem 2018; Koenen et al. 2013). Therefore this research concludes that SWIR spectroscopy is of added value to the current standard of core logging. However, the current methodology could not produce quantifiable parameters that correlate correctly with physical parameters. Future research should 1) produce larger hyperspectral datasets which could be directly linked to known physical properties, 2) investigate using a combination of multiple regions within the electromagnetic spectrum, and 3) the usability of this technique on core cuttings.