Static and dynamic elastic behaviour of siliciclastic reservoir rocks

Abstract

Siliciclastic rocks are composed of a wide range of minerals; including quartz, feldspar, clay minerals, calcite to name the most common; and various textural properties such as grain size (sorting), shape, orientation and packing. However, most of the rock physics models used to reproduce their elastic behaviour make use of severe idealizations of those rock properties, such as assuming single mineralogy, spherical, uniform grain sizes and shapes, or by representing the pore space by an ellipsoid characterized by one aspect ratio only. Even when those models may have considerable success modelling clean, well sorted siliciclastic rocks; their application faces considerable challenges in the modelling of heterogeneous sedimentary rocks. The objective of this thesis was to develop strategies to parameterize existing models from geologically controlled observations, with the aim of extending their application to wider ranges of composition and textural properties. The study used mainly existing laboratory data, which provided extensive information about composition and texture of the analyzed samples. It was found that a quantitative comparison between the results of various rock physics model is essential when making diagnostics about rock microstructure from elastic measurements. We were able to calculate percolation limits of clay-bearing rocks through an analogy between the critical porosity concept, as used in rock physics, and the liquid limit of clay powders in clay and soil science. Scanning electron microscope (SEM) images of clay-water composites showed that the large scale (>0.1 μm) pore space of kaolinite samples had more elongated and oriented pores (smaller aspect ratios) than smectite samples. However, we needed to include smaller aspect ratio pores, which exist at the scale of the individual clay platelets, to reproduce the observed elastic measurements. We observed a significant correlation between pressure sensitivity, in terms of velocity and attenuation, and the abundance/lack of quartz mineral in siliciclastic rocks. Therefore, by associating the content of secondary minerals to the amount of crack-like pores we modelled the pressure variations of velocity and attenuation using a visco-elastic theory. In addition, we found that pores at different scales such as equant, elongated and cracks-like require different relaxation time constant to honour the observed attenuation values.