No description available.

Independent determination of the heat capacity of the sapphirine used in this study is currently underway. It is expected that extrapolation of this data to the results of this study will result in the derivation of a heat capacity function for natural sapphirine from 0-1573K. Incorporation of this into existing thermodynamic datasets should allow the quantitative determination of the position of sapphirine-bearing reactions in petrogenetic grids relevant to ultra-high temperature metamorphism in natural systems. The Archean Napier Complex of Antarctica is one of the best documented UHT terrains, long recognised as having experienced temperatures in excess of 1000°C followed by a long period of near-isobaric cooling at deep crustal levels. However, the early history of the terrain, the timing of deformation and metamorphism, and the tectonic processes responsible for the generation of the extreme temperatures of metamorphism, have not been resolved (e.g. Ellis, 1987; Harley, 1989; Sandiford, 1989; Hensen and Motoyoshi, 1992). Mineral textural relationships linked to deformation features from a range of localities in and around Amundsen Bay are consistent with peak metamorphic conditions of 900-1100°C at 0.8-1.1 GPa during intense lower crustal extension. Rare decompression textures from widely spaced localities attest to decompression of the whole terrain to depths equivalent to the base of a normal thickness crust after peak metamorphism, while still under UHT conditions, and indicate that intense lower crustal extension and UHT metamorphism occurred synchronously with crustal thickening. Retrograde reactions textures may have been produced either by isobaric cooling, or by later granulite facies metamorphic event/s, or both. Mineral reaction textures and structural features support a tectonic model of lithosphere delamination for the development of UHT metamorphism, in the Napier Complex. This model involves the detachment of the lower part of the mantle lithosphere during continental collision, allowing upwelling of hot asthenosphere material directly beneath the crust, which in turn results in intense extensional deformation of the lower crust and lateral expulsion of melts.

This PhD study uses two data sets to investigate lithofacies predictions via neural analysis. A neural network model is developed to predict lithofacies in the Palaeocene submarine fan deposits of the Lomond Field in the Central North Sea. When the results from the optimum model are compared directly to core in many cases a close match is found, even in intervals where the model interpretation does not match the ‘human’ interpretation. In wells where predictions do not match the core this is shown to be a consequence not of the methodology itself but because of characteristics of the data collected in the training wells. This is due to variations in tool type, interval depth and fluid saturation. When this occurs, prior calibration of logs in the training data, for example through making a correction for fluid saturation, can improve performance. In all cases the observation of the activation levels of all output nodes in the network can qualitatively describe any uncertainty in the results, thus aiding interpretation. Using a second data set of wells form the deep marine West Delta Deep Concession area, Nile delta, Egypt more neural network models are developed to predict ‘image’ facies. Model inputs consist of conventional logs and derived logs from statistical and power spectral analysis of the pad data from the Fullbore Formation Microlager (FMI) tool. Although results are similar, slight improvement is seen if separate networks are trained to predict image facies that belong to specific lithological groups, rather than a single network trained to predict all facies. Where sandy thin beds (<1cm in thickness) occur within thicker shale units further neural networks are needed to discriminate these, trained with different inputs from previous networks.

Illite growth during oil field burial detrimentally effects reservoir sandstones by reducing permeability. Illite diagenesis can be dated using the isotopic K-Ar decay system. The potential value of this data is the paleocalibration of basin models to illite age related geological events. The fundamental control upon illite growth is a likely combination of solute supply and illite thermodynamic stability. The most important source of solutes for illite growth is the dissolution of potassium feldspar. However, petrographic data from reservoir sandstones from both the U.S. Gulf Coast and the UK North Sea indicate that feldspar dissolution is a continuous process, taking place over a considerable depth interval. Feldspar dissolution reactions are currently taking place in many sandstones where thermodynamics indicate that illite is stable, and where illite might be expected to be grown at the present day. There is no zero Ma age illite found in such sandstones. The implication of episodic illite K-Ar ages is that illite growth is an event-triggered process. Existing published illite data for the Northern North Sea are reviewed, and interpreted as showing episodic growth in the late Mesozoic and Cenozoic consistent with the expected timing of hydrocarbon charge to Northern North Sea oil fields. Experimental isotopic analysis of illite sampled from the Penguin field suggests a complex multiphase filling history for the Penguin field. Two dimensional basin modelling of a small petroleum system based on the geology of the Magnus and Penguin fields, related provinces and local embayments, examines this interpretation of the data. This simple model, the Magus model, supports the hypothesis that the onset of hydrocarbon charge controls the growth of illite within oil field sandstones. A calibrated Magnus model is adapted to examine likely migration timings for the Viking Graben and More Basin. In light of this, a simple explanation is proffered for the general population of illite ages published for the Northern North Sea - fibrous illite diagenesis in the sandstone reservoirs the Northern North Sea oil field is a response to the onset of oil migration.

Time-lapse seismic is proving to be a valuable method for monitoring subsurface pressure changes and fluid movements in producing hydrocarbon reservoirs. This case study aims to show the importance of multicomponent seismic data in time-lapse monitoring as shear wave splitting proves to be a highly sensitive diagnostic for monitoring hydrocarbon production. The investigated 4D, 3C land seismic data are from Vacuum Field, New Mexico, USA with a fractured dolomite forming the reservoir of interest. In 1995,<i> </i>over a period of two months, two seismic surveys were acquired before and after a pilot tertiary recovery program of CO₂-injection in a single well. The pore fluid pressure in the reservoir increased significantly from a ‘normal’ pressure of 10.6 MPa to an ‘overpressure’ of 17 MPa. CO₂is a highly compressible fluid at these pressures and therefore decreases the bulk modulus of the reservoir fluid. After the application of a data processing sequence that aims to preserve anisotropy and maximise repeatability, the interval time analysis of the reservoir interval shows a significant 10% change in shear-wave anisotropy and a 3% decrease in the <i>P</i>-interval velocities to the south and east of the CO₂-injection well. The fast shear wave becomes the slowest after the injection. Further, the rms-amplitudes of the shear-wave parallel to the maximum horizontal stress direction decrease in the area of the anisotropy anomaly after injection. In 1997 and 1998 two further surveys were acquired one year apart in Vacuum Field during a major CO₂-injection program. Due to repeatability issues these data are analysed separately. Small pore fluid pressure increases of up to 1.4 MPa in the reservoir led to velocity decreases of generally less than 2% in <i>P </i>and in shear. In order to interpret the observations several 1D models incorporating both saturation and pressure changes are matched to the data in the vicinity of the injection well. Due to the high stiffness of the dolomitic reservoir rock the saturation changes have little effect on the seismic velocities.

Currently, a controversy exists as to whether oil charge in reservoirs has an inhibiting effect on quartz cementation. The Brae Formation sandstone reservoirs in the Miller and Kingfisher oilfields display in some areas anomalously high porosity (25-30%) preserved to depths in excess of 4km (~120°C). The high porosity in these sandstones is directly linked to low quartz cement volumes (<5%). In the Miller Field, the crestal areas of the reservoir have porosities of up to 25% and a quartz cement content of <5%. Towards the oil-water contact (OWC), and in the aquifer of this same reservoir, porosity decreases and quartz cement content increases to 14% and 15% respectively. In the Kingfisher Field, two reservoirs separated by a mudstone layer are present. The lower reservoir, which connects with the reservoir in the adjacent Miller Field, displays porosities ~14% and quartz cement contents of ~10%. The upper reservoir is of overall high porosity (25-30%) and exhibits low quartz cement contents (<5%). Combined results from fluid inclusion and basin modelling studies in the Brae Formation reservoirs show that the duration of quartz cement precipitation is linked to hydrocarbon emplacement. In the Miller Field. quartz cementation in the high porosity (up to 25%) parts of the reservoir continued until the sandstones were buried to 95-105°C. This temperature corresponds to the burial depth (3km) and time (40Ma) at which oil entered the reservoir. Results for the Kingfisher Field indicate a similar oil emplacement time in the upper reservoir with 25-30% porosity and <5% quartz cement content, but a more recent (l5Ma) oil-fill for the lower reservoir with ~14% porosity and ~10% quartz cement content. Kinetic modelling of the quartz cementation process reveals that quartz cementation rates in the oil-filled parts of reservoirs examined (10^-22 moles/cm².s) are at least two orders of magnitude smaller than in the aquifers (10^-19 moles/cm².s).

This thesis presents the background, methodology, and results of an experimental study of the roles of stress, fault zone micro-structure, and environmental conditions on fault sealing. Synthetic fault gouges were tested in a suite of controlled experiments designed to elucidate the effects of a range of physical, structural, and chemical parameters on gouge reactivity. The experiments began with a series of ‘batch’ tests, designed to calibrate the purely chemical reactions. These were followed by a series of hydrostatic and non-hydrostatic pressure tests, designed to investigate the effect of burial depth and tectonic stress (confining pressure and deviatoric stress). The latter suites of experiments were conducted using a new experimental rig, designed and built specifically for this study. Each experiment used either a controlled size fraction, or a bulk mixture of particle sizes produced by crushing a rock sample to form a synthetic gouge. Highly reactive ultra-fine particles were removed to allow more accurate measurement of gouge surface area and hence reaction rates. Dissolution and precipitation rates were determined for each experiment by analysing dissolved silica concentrations of the pore fluids, using High Performance Liquid Chromatography, and applying them to a kinetic model. The non-hydrostatic tests also employed new apparatus that allowed the simultaneous measurement of fault-normal permeability and compaction throughout each experiment. The observed reaction rates have a strong dependence on temperature, particle size distribution, and pressure, in decreasing order of influence. This is consistent with field observations from a range of hydrocarbon reservoir rocks around the world. Theoretically, the finer particles would be expected to have higher specific surface energies and greater total reactive surface areas, and hence faster reaction rates. This has been confirmed by the results.

This project was devoted to the extraction of more quantitative information by integrating the 3D seismic data with the wireline logs and cores from 10 exploration wells drilled through the Tay system, Gannet South, Central North Sea and by introducing the innovative body and seismic trace shape analysis to the interpretation process. The idea of analysing a seismic data based on variation of seismic trace shapes comes from the assumption that changes in lithology, rock properties and fluid content should affect seismic response in not only amplitude but the whole shape of the trace. This project utilizes the pattern recognition capability of the neural network technology to classify the seismic traces based on their shapes. Both supervised and unsupervised classifications were applied on the Gannet South seismic dataset. Maps produced have revealed subtle geological features only expressed in the shape of the seismic trace and thus substantially enhanced the understanding of the sand geometries of the turbidite system and the structural development and evolution of the basin, provided clues to the timing, nature and extent of factors controlling the sediment transport pathways in the area, and helped in the discovery of hydrocarbon pockets previously gone unnoticed. A set of body shape parameters that would enable a distinctive description of different turbidite sand bodies were established. No single parameter is enough to uniquely describe all shapes but a combination of parameters could be used. The study also showed that the normalised polar representation of any shape can be significant for its recognition as well as for matching purposes. Unfortunately, the body shape analysis was hampered by the lack of accepted general turbidite models in the literature as well as inaccessibility to subsurface seismic datasets on which body shape analysis could have been applied in order to use the results as a database.

The Turku granulite terrain of southern Finland is located within the south-west corner of Svecofennian Schist Belt, part of the Scandinavian Svecofennian Orogenic Belt. The field area is a representative slice of an island arc system which collided against the northern Baltic Shield at 1890 Ma. This resulted in amphibolite-granulite facies metamorphism in the time interval 1890-1870 Ma when two phases of compression (D1 and D2) deformed the Turku rocks into recumbent isoclinical folds, with axial traces striking NW/SE. A second thermal pulse within the Turku terrain was accompanied by a gentler period of compression (D3) and rotation that produced folds with axial traces orientated NE/SW. Maximum pressure and temperature estimates from three garnet/cordierite barometers, five Fe-Mg exchange thermometers and an internally consistent thermodynamic dataset, are 750°C and 5-5.5 kbar. Under these P/T conditions the pelitic supracrustals partially melted, producing felsic segregation and a restite rich in garnet, cordierite and biotite. Two different generations of garnet and cordierite are linked to D2 and D3. The melting started late D2 and was continuous through early/late D3, during which most of the melting occurred. The leucosomes that were produced throughout this period ranged from syenogranites to granodiorites based on their normative quartz and feldspar components. They all have A/CNK values of >1.10 and thus are peraluminous in composition in accordance with melts produced experimentally within pelitic systems. Fe-Mg contents of leucosomes range from 0.31 to 12 wt %, above 3 wt % the leucosomes are also enriched in Zr and Th. This suggests either entrainment of restitic material during inefficient melt segregation, or that leucosomes are themselves restitic after further melt migration i.e. they are enriched in 'cumulate' phases. The composition of the leucosomes and the presence of garnet and cordierite poikiloblasts suggests that melting proceeded via incongruent melting reactions involving the breakdown of biotite and sillimanite to produce garnet, cordierite, K-feldspar and melt. As volatiles liberated through the breakdown of biotite will partition directly into melt and cordierite, the total volatile contents and the XCO₂ (CO₂/CO₂+H₂O) of cordierite were analysed to distinguish between fluid-present or fluid absent conditions and to allow calculation of fluid species activities.

In this thesis, I propose the following: - The P and S-wave velocities can provide a suitable link between reservoir parameters and rock properties using core, log and seismic data. - The pore aspect ratios as key parameters of rock geometry can be used to explain the different responses of elastic properties in clay-sand rocks (especially for thin and varying lithology formations). The use of fixed aspect ratio for physical velocity models will result in obvious errors in the prediction of elastic moduli and velocities (in particular for formations at shallow depth, or in loose and thin layers). - The time-average equation (Wyllie et al., 1956) ignored the effects of pore geometry, degree of consolidation fluid and clay content. It results in a hidden defect in the transformation between porosity (form core and well-log) and velocity (from seismic) when the rock contains clay. - The current models of Gassmann (1951), Kuster & Toksöz (1974) and Xu-White (1995) have some difficulties in calculating elastic moduli for rocks containing aligned pores and minerals in anisotropic formations. To investigate these, I first use method of multiple regression and artificial networks to establish an empirical correlation between reservoir parameters and P and S-wave velocities. This correlation includes porosity, clay content, aspect ratio and velocities, which can be used as an extension of the empirical model of Han et al (1956). Second, in order to overcome the weakness of empirical models, physically realistic theoretical models are established. The first theoretical model is the isotropic dual porosity model (IDP). The aim of the IDP is to develop a general rock physical model that provides a satisfactory integrated approach to the evaluation and prediction of reservoir parameters and rock properties for the purpose of reservoir characterization. Third, because the IDP model does not consider the effects of pore orientation, clay content and velocity anisotropy etc., a refined anisotropic dual porosity model (ADP) is then developed for anisotropic porous media.