Seismic exploration methods for hydrothermal dolomite reservoirs : a case study of the Trenton-Black River Groups

This thesis discusses 3D seismic exploration for hydrothermal dolomite (HTD) reservoirs. During this project, I utilized some of the existing algorithms as well as developed my own methods. I mentioned that HTD reservoirs constitute an important hydrocarbon play-type the world over, and are particularly important within the Ordovician aged Trenton-Black River Group in eastern North America. I noted that 3D seismic is the best tool to explore for HTD but it is adversely affected by random noise and acquisition footprint which must be attenuated. I designed a post-stack processing flow that can be applied to attenuate noise, highlight fault terminations, and increase reflection events continuity.By predicting porosity using seismic attributes, I examined the relationships between porosity distribution and faulting, and between porosity and seismic attributes. Results show that porosity development depends on basement-related faults, and that hydrocarbon production rate depends on the presence of faults, fractures and porosity. The results show that the sag observed to be associated with HTD reservoirs on seismic sections is due to velocity pull-down and that the best porosity is developed in these areas. The methods and results presented in this project can be used in other places with similar geologic settings.By way of semblance, I developed a dual-parameter scanning algorithm that can recover both NMO velocity (vnmo) and the effective anisotropic parameter (etaeff) at the same time, from which interval eta (etaint) is obtained by inversion. Noise-free synthetic data application show 10% maximum deviation between the extracted and actual flint values, while noise-contaminated synthetic data show >20% deviation. Real data application shows that it is possible to discriminate between HTD and limestone. Further testing is required to establish its validity. A velocity ratio (Vp/Vs) extracting algorithm was developed using AVO prestack amplitude inversion. Noise-free synthetic data application shows <11% maximum error between the scanned and the actual Vp/Vs values, while the noisy data deviation is 22%. Real data application indicates a distinction between HTD and limestone. Testing with other real data sets is required to validate the usefulness of the method