An approach for upscaling the flow effects of multiple deep-water genetic units
MetadataShow full item record
Developing a deep-water basin raises many issues and challenges; one of the most significant and necessary issues is overpressure prediction. Basin modelling to construct full scale models is an effective way to find out overpressure distribution in many conditions. However, with current modelling and other methods, it is difficult to obtain data about highly heterogeneous complex structures, which involve coupling between hydraulics and geo-mechanics (compaction) in a system with spatial complexity and temporal evolution of geological bodies, termed as genetic units (GUs). The data collection is difficult and costly in both economic and time aspects. Understanding the role of the interactions between GUs could play a major role in helping to simplify the highly heterogeneous complex structures. The aim of this thesis is to develop the understanding that can underpin the creation of a workflow to be used to assess the role of interactions between GUs, in relation to predicting overpressure in deep-water sedimentary basins. Tilted sandy aquifers enclosed in muddy sediments (block rotation) are a good reference case which is not uncommon in deep-water basins worldwide. This thesis shows that, by applying basin modelling and response surface methodology, not only is a parameterised prediction possible but also the uncertainty of the parameters can be taken into consideration at the same time. A tilted aquifer, however, rarely exists alone within a ‘featureless’ mud background, but occurs along with other geological architectures of sediment units of the same genetic origins, deposited later. These genetic units may be channels and levees. The GUs could allow the fluid energy to dissipate more easily and therefore can reduce the overpressure at the crest. However, the existence of additional GUs complicates the prediction of the overpressure, as the dimensions of parameter space increases dramatically. This poses a big challenge to extend the prediction, and therefore calls for development of appropriate parameterised overpressure-prediction techniques. This thesis reports the development of parameterised overpressure-prediction techniques in the presence of multiple channels. This result forms the basis for follow-on research that can seek to further generalise the approach to a wider set of systems and their associated descriptive parameters.