Magister Scientiae - MSc (Earth Science) / The syn-rift succession encompasses the primary exploration target in the southern Pletmos Basin. Several fault-bounded structural traps that contain gas accumulations have been discovered within this succession. Likewise, ubiquitous residual gas shows have been encountered in most drilled wells. Yet, the impact of faults on fluid flow is poorly understood. Therefore, this study aspires to predict, and where possible, quantify fault seal integrity and sealing capacities of some of the major prospect-bounding faults. A multi-disciplinary research strategy was employed in order to fulfil the study objectives. Fault mapping and geo-cellular modelling using geostatistical algorithms were undertaken to provide the basic geometric and structural input for more advanced fault seal analysis applications. Juxtaposition analysis was carried out to identify zones with a high probability to seal (or leak) and as the first-order tool for predicting fault seal potential. Threshold pressures, hydrocarbon column heights, cross-fault permeability and transmissibility were used to estimate the sealing capacities of the faults. In addition to juxtaposition and customary fault-rock properties, the study also analysed parameters that can be deemed to be representative of cross-fault fluid flow (i.e. effective cross-fault permeability and transmissibility; ECFP and ECFT). Finally, modelling of the geo-history facilitated the validation of the properties that underpinned fault seal analysis studies. The Ga-Q and proposed Ga-K prospects along with their main bounding faults formed the foci of the fault seal analysis results. The analysed faults showed excellent initial sealing potential due to either favourable juxtaposition or shale gouge development. Nonetheless, predicted hydrocarbon column heights and threshold pressures were low suggesting that the seal integrity of the analysed faults is predisposed to failure. In addition, high predicted fault permeability and transmissibility values signify the presence of open and permeable fracture networks within the fault zones. Thus, it is proposed that the faults are very likely to have leaked during hydrocarbon migration and filling of traps resulting in empty or under-filled hydrocarbon reservoirs.

The Gulf Coast of Texas has been proposed as a high capacity storage region for geologic sequestration of anthropogenic CO₂. The Miocene section within the Texas State Waters is an attractive offshore alternative to onshore sequestration. However, the stratigraphic targets of interest highlight a need to utilize fault-bounded structural traps. Regional capacity estimates in this area have previously focused on simple volumetric estimations or more sophisticated fill-to-spill scenarios with faults acting as no-flow boundaries. Capacity estimations that ignore the static and dynamic sealing capacities of faults may therefore be inaccurate. A comprehensive fault seal analysis workflow for CO₂-brine membrane fault seal potential has been developed for geologic site selection in the Miocene section of the Texas State Waters. To reduce uncertainty of fault performance, a fault seal calibration has been performed on 6 Miocene natural gas traps in the Texas State Waters in order to constrain the capillary entry pressures of the modeled fault gouge. Results indicate that modeled membrane fault seal capacity for the Lower Miocene section agrees with published global fault seal databases. Faults can therefore serve as effective seals, as suggested by natural hydrocarbon accumulations. However, fault seal capacity is generally an order of magnitude lower than top seal capacity in the same stratigraphic setting, with implications for storage projects. For a specific non-hydrocarbon producing site studied for sequestration (San Luis Pass salt dome setting) with moderately dipping (16°) traps (i.e. high potential column height), membrane fault seal modeling is shown to decrease fault-bound trap area, and therefore storage capacity volume, compared with fill-to-spill modeling. However, using the developed fault seal workflow at other potential storage sites will predict the degree to which storage capacity may approach fill-to-spill capacity, depending primarily on the geology of the fault (shale gouge ratio – SGR) and the structural relief of the trap. / text

Geologic storage of anthropogenic carbon dioxide (CO2) by injection into underground porous sandstone reservoirs has been proposed as a method for the reduction of anthropogenic greenhouse gas emissions. Upwards migration and leakage of injected fluids along natural fault and fracture networks is a key risk factor for potential injection locations. We examine exposed natural analogs to evaluate the impacts of faulting and fracturing on reservoir and top-seal pairs and to evaluate evidence for paleomigration of fluids along the fault zone. We examine the Iron Wash fault, a 25-km long normal fault which cuts Jurassic sedimentary rocks and has throws that range from 20-120 m, to examine how a fault may affect seal integrity. Field mapping, kinematic analysis, petrographic analysis, characterization of the fault zone facies and fault architecture, analysis of altered and mineralized rocks in and around the fault zone, and modeling of fault seal capacity was conducted to provide an understanding of the Iron Wash fault zone. Field data and observations were combined with well log and borehole data to produce three types of models for the Iron Wash fault: 1) geometric model of the fault in the subsurface, 2) predictive models of fault zone behavior and fault seal analysis, and 3) predictive geomechanical models of the response of the fault zone to an imposed stress field and increasing the effective stress on the fault. We conclude that the Iron Wash fault zone has low sealing capacity and will likely not behave as a seal for fluids against the fault zone due primarily to modest throw on the fault and high frequency of fractures associated with the fault zone. Analysis of fluid alteration and mineralization around the fault zone indicates that the fault zone was conduit for paleo-fluids. We conclude that the fault is not likely to develop a sealing membrane and therefore will most likely fail as a seal to fluids moving through the reservoirs modeled here. Modeling results indicate that a reduction in the effective normal stress on fault surfaces may induce failure of faults resulting in earthquakes or increased hydraulic conductivity of fractures.

A new depth-based method of seismic imaging is used to provide insights into the 3D structural geometry of faults, and to facilitate a detailed structural interpretation of the Penola Trough, Otway Basin, South Australia. The structural interpretation is used to assess fault kinematics through geological time and to evaluate across-fault juxtaposition, shale gouge and fault reactivation potential for three selected traps (Zema, Pyrus and Ladbroke Grove) thus providing a full and systematic assessment of fault seal risk for the area. Paper 1 demonstrates how a depth-conversion method was applied to two-way time seismic data in order to redisplay the seismic in a form more closely representative of true depth, here termed ‘pseudo-depth’. Some apparently listric faults in two-way time are demonstrated to be planar and easily distinguishable from genuine listric faults on pseudo-depth sections. The insights into fault geometry provided by pseudo-depth sections have had a significant impact on the new structural interpretation of the area. Paper 2 presents the new 3D structural interpretation of the area. The geometry of faulting is complex and reflects variable stress regimes throughout structural development and the strong influence of pre-existing basement fabrics. Some basement-rooted faults show evidence of continual reactivation throughout their structural history up to very recent times. Structural analysis of all the live and breached traps of the area demonstrate that traps associated with a basement rooted bounding fault host breached or partially breached accumulations, whereas non-basement rooted faults are associated with live hydrocarbon columns. Papers 3 and 4 demonstrate that for all the traps analysed (Zema, Pyrus and Ladbroke Grove), initial in-place seal integrity was good. The initial seal integrity was provided by a combination of both favourable across fault juxtaposition (Ladbroke Grove) and/or sufficiently well developed shale gouge over potential leaky sand on sand juxtaposition windows to retain significant hydrocarbon columns (Zema, Pyrus). The palaeocolumns observed at Zema and Pyrus indicate that there has been subsequent post-charge breach of seal integrity of these traps while Ladbroke Grove retains a live hydrocarbon column. Evidence of open, permeable fracture networks within the Zema Fault Zone suggest that it is likely to have recently reactivated, thus breaching the original hydrocarbon column. Analysis of the in-situ stress tensor and fault geometry demonstrates that most of the bounding faults to the selected traps are at or near optimal orientations for reactivation in the in-situ stress tensor. The main exception being the Ladbroke Grove Fault which has a NW-SE trending segment (associated with a relatively high risk of fault reactivation and possible leakage at the surface) and an E-W trending segment (associated with a relatively low risk of fault reactivation and a present day live column). The free water level of the Ladbroke Grove accumulation coincides with this change in fault orientation. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1339545 / Thesis (Ph.D.) - University of Adelaide, Australian School of Petroleum, 2008