Josianne L. Pafeng Tschuindjang
Published: 2017
Total Pages: 146
Get eBook
This dissertation addresses a seismic reservoir characterization study and time-lapse feasibility of reservoir monitoring of carbon dioxide using seismic data, via rock physics models, global and local nonlinear inversions. It also aims to investigate the value of data integration, the relative impact of elastic and electrical rock physics model parameters on inverted petrophysical properties, and the feasibility of using resistivity data from time-lapse electromagnetic survey to monitor the displacement of carbon dioxide in the subsurface. This study focuses on the identification of target storage and sealing lithologies for a future carbon dioxide (CO2) monitoring project at the Rock Springs Uplift (RSU), Wyoming, USA. Seismic reservoir characterization aims to estimate reservoir rock and fluid properties such as porosity, fluid saturation, lithology, which are important properties for hydrocarbon exploration as well as carbon dioxide sequestration and monitoring projects. These petrophysical properties affect elastic attributes which in turn, affect the seismic response. Estimating reservoir properties therefore constitutes an inverse problem. Geophysical inverse problems are challenging because of the noise in recorded data, the nonlinearity of the inverse problem, the nonuniqueness of the solutions, etc. Depending upon the complexity of the problems, we can either use a local or a global optimization scheme to solve the specific problem. In this dissertation, we use a multilevel parallelization of a global prestack waveform inversion to three-dimensional seismic data with sparse well-information, to estimate subsurface elastic attributes like P-, S-wave velocity and density. This study contributes to the inversion of 3D large seismic data volume in an efficient computational time while providing high-resolution structural images of the subsurface compared to amplitude-variation-with-offset/angle (AVO/AVA) inversion. Following prestack waveform inversion, we use rock physics models to relate elastic attributes to reservoir properties and apply a local nonlinear least squares inversion scheme based on the trust-region algorithm, to invert elastic attributes for petrophysical properties like porosity and volumetric fractions of minerals. We apply this approach on well log data to validate the method, followed by applying it to the volumes of inverted elastic attributes obtained from prestack waveform inversion, to provide reservoir characterization away from the well. Because a carbon dioxide sequestration project is planned at the Rock Springs Uplift, we also investigate the feasibility of a time-lapse reservoir monitoring for the area using seismic data, by simulating the pressure and fluid effects on elastic velocities and synthetic seismograms. In the final part of this dissertation, we investigate the value of data integration by combining elastic and electrical attributes in a joint petrophysical inversion for reservoir rock and fluid properties. We illustrate the methodology using well log data sets from the Barents Sea and the Rock Springs Uplift, and show that the estimation of reservoir properties can be improved by combining multiple geophysical data. Despite the geological information we might have on a study area, there is often uncertainty in the choice of an adequate rock physics model and its input parameters not only at the well location, but also in areas with sparse well control. This study therefore helps understand the impact of such model parameters on inverted petrophysical properties and how it could affect reservoir interpretation. Next, we use a simple sharp interface model in order to provide a preliminary assessment of the extent of the CO2 plume, and thus address potential leakage risks. We also simulate the spatial distribution of CO2 after injection and compute corresponding resistivity datasets at different spatial resolutions, which we invert for water saturation. This synthetic study helps investigate the ability of monitoring the CO2 displacement using geophysical data.