|dc.description.abstract||The main objective of the work presented in this thesis is to improve our understanding of elastic wave velocities in uncemented sediments, such as sand, clay and mixtures of sand and clay. The focus here has been on experimental work, addressing the effects of stress and stress path on velocities, including anisotropic characteristics. Stresses have been fairly low, typically not more than 10 – 15MPa above the pore pressure, which means that the focus has been on the relevance to shallow petroleum reservoirs, but also to deeper unconsolidated reservoirs that are over pressured. It is further relevant for characterization of shallow aquifers and possible subsurface storage sites for e.g. CO2 or for waste disposal. Furthermore, the role of lithology has been addressed, i.e. how behaviors of sand and clay differ and how sand and clay can be distinguished with seismic techniques.
It is common that researchers compare their experimental results with previous experimental data obtained with almost identical materials. Such comparisons require knowledge of procedure for sample preparation, test setup, methods for data acquisition and analysis, and test conditions, such as stress path and stress history, temperature etc. In this study, tests have been conducted in two setups: These are an oedometer1 (only uniaxial strain conditions) and a triaxial test system (both uniaxial strain and hydrostatic stress conditions). An initial study was carried out to compare results from the two test systems with similarly prepared samples, focusing on both dynamic (derived from wave velocities) and static behavior of unconsolidated sand and clay.
Gassmann's fluid substitution theories (1951) are extensively used in reservoir evaluation to quantify changes in fluid saturation from time-lapse (4D) seismic data acquired during production or injection. The same model is also used in exploration to identify lithological or fluid boundaries in the Earth. Therefore, it is important to understand the basic assumptions of this theory and figure out the validity of those assumptions in the studied case. Knowledge of − relation for sand, clay and other rock forming minerals is important in reflection seismology and formation evaluation.
The present dissertation presents the static and dynamic characteristic of unconsolidated, uncemented clean sands, clays (kaolinite and smectite), and mixtures of sand-clay (kaolinite) and kaolinite-smectite over a range of stress from 0.7-15MPa. It is observed that the static and dynamic behavior of the mixture depends on the amount of clay present in the mixtures. Generally, wave velocities increase and porosity decreases with increase in clay content until clay becomes the load bearing phase. However, pure clay always shows lower velocity than pure sand. Porosity is higher at low stress, but decrease strongly with increasing stress. The velocities decrease with decrease in effective grain size of clay for sandy clay or kaolinite-smectite mixtures. The effective stress coefficient decreases with increase in clay content. Stress sensitivity of the velocities follows the same trend as velocities as function of clay content. However, strain sensitivity decreases with increase in clay content.
Fluid saturation effects on ultrasonic compressional and shear wave velocities and their corresponding moduli are extensively studied using unconsolidated, uncemented clean sands of different grain sizes and shapes under uniaxial strain and hydrostatic stress conditions. These studies also address the anisotropy parameters of the sands, where anisotropy is induced by the applied stress. It is observed that the shear modulus increases upon brine saturation. Biot dispersion is found to explain the observed shear wave hardening. The isotropic Gassmann's fluid substitution mostly underestimates the compressional wave modulus upon saturation, but the prediction can be improved by adding Biot's dispersion effect.
The study reveals that anisotropy parameters are sensitive to fluid saturation. It is surprising that shear wave anisotropy (Thomsen’s γ) is found to be sensitive to fluid. However, the effect of brine saturation on the observed P-wave anisotropy (ɛ) and the anisotropic δ parameter is well described by anisotropic fluid substitution theory. For the relatively weak stress-induced anisotropy that is seen in the presented experiments, the effect of anisotropy is less significant compared to the effect of dispersion to predict the plane wave modulus upon saturation.
Effects of saturating fluid, stress paths, and the state of consolidation on the ratio between P- and S-wave velocities (/), as well as effects of grain size and shape of sands are presented in this dissertation. The − relationship can be useful to discriminate between different lithologies. / ratio is almost insensitive to stress in dry conditions, but shows increase with decrease in net stress upon saturation. Fluid saturation increases the / ratio at any stress level above the / ratio of dry samples. At low stresses / ratio is higher for sand than clay, but shows comparable values at high stresses. The / ratio shows significant hysteresis upon loading and unloading, and the hysteresis is higher for higher amount of clay. It is found that once all the stresses (vertical and confining pressure) are known then one might not observe any difference in / plotted vs. net mean stress due to stress paths differences. But if / ratio is plotted as function of net vertical stress (without knowledge of confining pressure at uniaxial strain condition), then significant difference between the stress paths can be visible. A set of trends are suggested for the − relation of sand-clay mixtures and clay. Two trends are suggested for sand with different grain shapes, that can be applied to both reservoir seismics and sonic or laboratory measurements. All those are valid for unconsolidated, uncemented sediments up to 10MPa net stress||nb_NO