While geophysical reservoir characterization has been an area of research for the last three decades, geophysical reservoir monitoring, time-lapse studies, have recently become an important geophysical application. Generally speaking, the main target is to detect, estimate, and discriminate the changes in subsurface rock properties due to production. This research develops various sensitivity and feasibility analyses to investigate the effects of production-induced time-lapse changes on geophysical measurements including seismic and controlled-source electromagnetic (CSEM) data. For doing so, a realistic reservoir model is numerically simulated based on a prograding near-shore sandstone reservoir. To account for the spatial distribution of petrophysical properties, an effective porosity model is first simulated by Gaussian geostatistics. Dispersed clay and dual water models are then efficiently combined with other well-known theoretical and experimental petrophysical correlations to consistently simulate reservoir model parameters. Next, the constructed reservoir model is subjected to numerical simulation of multi-phase fluid flow to replicate a waterflooding scenario of a black oil reservoir and to predict the spatial distributions of fluid pressure and saturation. A modified Archie’s equation for shaly sandstones is utilized to simulate rock resistivity. Finally, a geologically consistent stress-sensitive rock physics model, combined with the modified Gassmann theory for shaly sandstones, is utilized to simulate seismic elastic parameters. As a result, the comprehensive petro-electro-elastic model developed in this dissertation can be efficiently utilized in sensitivity and feasibility analyses of seismic/CSEM data with respect to petrophysical properties and, ultimately, applied to reservoir characterization and monitoring research. Using the resistivity models, a base and two monitor time-lapse CSEM surveys are simulated via accurate numerical algorithms. 2.5D CSEM modeling demonstrates that a detectable time-lapse signal after 5 years and a strong time-lapse signal after 10 years of waterflooding are attainable with the careful application of currently available CSEM technology. To simulate seismic waves, I employ different seismic modeling algorithms, one-dimensional (1D) acoustic and elastic ray tracing, 1D full elastic reflectivity, 2D split-step Fourier plane-wave (SFPW), and 2D stagger grid explicit finite difference (FD). My analyses demonstrate that acoustic modeling of an elastic medium is a good approximation up to ray parameter (p) equal to 0.2 sec/km. However, at p=0.3 sec/km, differences between elastic and acoustic wave propagation is the more dominant effect compared to internal multiples. Here, converted waves are also generated with significant amplitudes compared to primaries and internal multiples. I also show that time-lapse modeling of the reservoir using SFPW approach is very fast compared to FD, 100 times faster for my case here. It is capable of handling higher frequencies than FD. It provides an accurate image of the waterflooding process comparable to FD. Consequently, it is a powerful alternative for time-lapse seismic modeling. I conclude that both seismic and CSEM data have adequate but different sensitivities to changes in reservoir properties and therefore have the potential to quantitatively map production-induced time-lapse changes. / text

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第22423号 / 工博第4684号 / 新制||工||1731(附属図書館) / 京都大学大学院工学研究科都市社会工学専攻 / (主査)教授 小池 克明, 教授 三ｹ田 均, 准教授 柏谷 公希, 教授 後藤 忠徳 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

Hydrocarbon deposits in the form of petroleum, natural gas, and natural gas hydrates occur offshore worldwide. Electromagnetic methods that measure the electrical resistivity of sediments can be used to map, assess, and monitor these resistive targets. In particular, quantitative assessment of hydrate content in marine deposits, which form within the upper few hundred meters of seafloor, is greatly facilitated by complementing conventional seismic methods with EM data. The North-East Pacific Time-series Undersea Network Experiment (NEPTUNE) is an underwater marine observatory that provides power and network connection to a host of instruments installed on the seafloor on the Cascadia Margin offshore Vancouver Island. The observatory’s aim is to provide a platform for very long-term studies in which access to data is available on a continuous basis. For this thesis project, a transient dipole-dipole time-domain EM system was constructed and deployed on the NEPTUNE network with the goal of long-term monitoring of a well-studied hydrate deposit in the area. The instrument includes a source transmitter of electrical current and individual receivers to measure small electric field variations. The instrument is powered by the NEPTUNE observatory and data can be collected remotely by connecting to the instrument through the web. Data collected so far from the instrument are consistent with a resistive structure. The best fitting model from 1D inversion is a 36 ± 3 m thick layer of 5.3 ± 0.3 Ωm resistivity, overlaying a less resistive 0.7 ± 0.1 Ωm halfspace. Average hydrate concentration, deduced with the aid of ODP-889 well-log derived Archie’s parameters, is approximately 72% of pore space in the resistive layer, consistent with the very high concentration of gas hydrates (~80%) recovered from seafloor cores. The weekly collection of data from the instrument shows that the resistive structure has changed little since monitoring began in October of 2010.

Hydrocarbon deposits in the form of petroleum, natural gas, and natural gas hydrates occur offshore worldwide. Electromagnetic methods that measure the electrical resistivity of sediments can be used to map, assess, and monitor these resistive targets. In particular, quantitative assessment of hydrate content in marine deposits, which form within the upper few hundred meters of seafloor, is greatly facilitated by complementing conventional seismic methods with EM data. The North-East Pacific Time-series Undersea Network Experiment (NEPTUNE) is an underwater marine observatory that provides power and network connection to a host of instruments installed on the seafloor on the Cascadia Margin offshore Vancouver Island. The observatory’s aim is to provide a platform for very long-term studies in which access to data is available on a continuous basis. For this thesis project, a transient dipole-dipole time-domain EM system was constructed and deployed on the NEPTUNE network with the goal of long-term monitoring of a well-studied hydrate deposit in the area. The instrument includes a source transmitter of electrical current and individual receivers to measure small electric field variations. The instrument is powered by the NEPTUNE observatory and data can be collected remotely by connecting to the instrument through the web. Data collected so far from the instrument are consistent with a resistive structure. The best fitting model from 1D inversion is a 36 ± 3 m thick layer of 5.3 ± 0.3 Ωm resistivity, overlaying a less resistive 0.7 ± 0.1 Ωm halfspace. Average hydrate concentration, deduced with the aid of ODP-889 well-log derived Archie’s parameters, is approximately 72% of pore space in the resistive layer, consistent with the very high concentration of gas hydrates (~80%) recovered from seafloor cores. The weekly collection of data from the instrument shows that the resistive structure has changed little since monitoring began in October of 2010.

Uncertainty Quantification (UQ) has been an active area of research in recent years with a wide range of applications in data and imaging sciences. In many problems, the source of uncertainty stems from an unknown parameter in the model. In physical and engineering systems for example, the parameters of the partial differential equation (PDE) that model the observed data may be unknown or incompletely specified. In such cases, one may use a probabilistic description based on prior information and formulate a forward UQ problem of characterising the uncertainty in the PDE solution and observations in response to that in the parameters. Conversely, inverse UQ encompasses the statistical estimation of the unknown parameters from the available observations, which can be cast as a Bayesian inverse problem. The contributions of the thesis focus on examining the aforementioned forward and inverse UQ problems for the low-frequency, time-harmonic Maxwell equations, where the model uncertainty emanates from the lack of knowledge of the material conductivity parameter. The motivation comes from the Controlled-Source Electromagnetic Method (CSEM) that aims to detect and image hydrocarbon reservoirs by using electromagnetic field (EM) measurements to obtain information about the conductivity profile of the sub-seabed. Traditionally, algorithms for deterministic models have been employed to solve the inverse problem in CSEM by optimisation and regularisation methods, which aside from the image reconstruction provide no quantitative information on the credibility of its features. This work employs instead stochastic models where the conductivity is represented as a lognormal random field, with the objective of providing a more informative characterisation of the model observables and the unknown parameters. The variational formulation of these stochastic models is analysed and proved to be well-posed under suitable assumptions. For computational purposes the stochastic formulation is recast as a deterministic, parametric problem with distributed uncertainty, which leads to an infinite-dimensional integration problem with respect to the prior and posterior measure. One of the main challenges is thus the approximation of these integrals, with the standard choice being some variant of the Monte-Carlo (MC) method. However, such methods typically fail to take advantage of the intrinsic properties of the model and suffer from unsatisfactory convergence rates. Based on recently developed theory on high-dimensional approximation, this thesis advocates the use of Sparse Quadrature (SQ) to tackle the integration problem. For the models considered here and under certain assumptions, we prove that for forward UQ, Sparse Quadrature can attain dimension-independent convergence rates that out-perform MC. Typical CSEM models are large-scale and thus additional effort is made in this work to reduce the cost of obtaining forward solutions for each sampling parameter by utilising the weighted Reduced Basis method (RB) and the Empirical Interpolation Method (EIM). The proposed variant of a combined SQ-EIM-RB algorithm is based on an adaptive selection of training sets and a primal-dual, goal-oriented formulation for the EIM-RB approximation. Numerical examples show that the suggested computational framework can alleviate the computational costs associated with forward UQ for the pertinent large-scale models, thus providing a viable methodology for practical applications.