From pore to well - identifying clay mineral distribution in sandstones using SWIR spectroscopy

Fluid extraction from geological formations for purposes of subsurface utilization leads to a pore pressure drop in reservoirs and potentially to compaction and seismicity. The geomechanical behaviour, and thus production-related compaction, of siliciclastic reservoirs is governed by the composition of reservoir sandstones, which includes porosity, grain size distribution, and detrital and authigenic mineralogy.

One siliciclastic reservoir which is undergoing compaction related to fluid extraction is the Rotliegend of the Groningen gas field. In this research, we investigate potential approaches to upscale the sandstone composition from the micro-scale (thin sections, plugs) to the well and reservoir scale eventually.

Previous research has highlighted that among the petrographic properties, the presence of authigenic clays tends to affect the geomechanical behaviour the most. Intragranular clays can affect the overall stiffness of the individual grains while the presence of the pore-filling clays  and clay rims may result in grain slip or pressure solution in loaded rock samples. In particular, we are defining the fractions of the clay-coating minerals (illite, kaolinite, etc.) and their effect on the inelastic deformation of reservoir sandstone as well as paying close attention to the presence of chlorite as its distribution corresponds to the areas associated with increased subsidence and seismicity.

We employed Short-Wave Infrared (SWIR) spectroscopy, a non-destructive and time-efficient technique, to obtain the mineralogical composition of core slabs from the Groningen Gas field. The SWIR data, based on a resolution of 200 µm pixels, allows for a detailed analysis of compositional variation within the Upper Rotliegend Group. Verification of the SWIR results against a comprehensive petrographic dataset, including X-ray diffraction, thin section descriptions, and modal point count analysis highlights that SWIR data primarily captures qualitative variations in mineralogy rather than providing precise numerical values. Combination of sedimentary facies, SWIR spectroscopy data and conventional petrographic studies allows to generate descriptive mineralogical trends for each of the studied wells.

Ultimately, the results of our research will serve as a foundation for selecting the samples, designing geomechanical experiments to test the proposed hypotheses, and as the means to select the upscaling and modelling approaches to develop a detailed model of the Dutch subsurface that matches well the existing heterogeneous structures. The derived 3D reservoir composition model of the Groningen gas field will be combined with the results of the deformation experiments to link reservoir composition to geomechanical behaviour. This will enable an updated, more realistic, 3D geomechanical model of the Groningen gas field that can be utilised by other researchers to better predict future compaction and subsidence.