Description |
-
Abstract: Understanding the role of deep fluid migration in geologic processes is essential for exploring natural resources. Modeling multiphase petroleum systems in sedimentary basins is often a difficult task because of the various coupled and nonlinear physics affecting fluid saturation and migration, including effects of capillary forces and relative permeability, anisotropy and heterogeneity ... read moreof the medium, and the effects of pore pressure, composition, and temperature on fluid properties. In addition to those factors, regional fault structures resulting from the tectonic history of the basin also play a decisive role in reservoir distributions, so it is very important to consider both hydrogeologic effects of these structures and their transient nature to characterize the hydrocarbon system more precisely. Physical properties of faults change dynamically due to geologic dependence on the tectonic stresses, fluid pressure, and hydrothermal reactions, and these changes affect hydrocarbon migration and distribution in fault-dominant sedimentary basins located in active continental margins. Despite decades of study, the petroleum migration problem remains still to be a qualitative science. The migration patterns and time are difficult to address quantitatively due to the lack of numerical models that consider rigorous multiphase fluid physics in a basin-scale system. A clearer understanding of fluid migration in the deep basin is not only essential to explore new energy sources such as methane hydrate and shale gas, but also to predict the long-term fate of injected carbon dioxide into the subsurface to contain green house gases. In this Ph.D. dissertation, I have developed a 2-D multiphase fluid flow model, coupled to heat flow, using a hybrid finite element and finite volume method to address these petroleum migration and entrapment issues on spatially large and long temporal scales. The newly developed multi-phase flow program (TUFTS-FV) is verified by comparison with analytic solution, and applied to solve M. King Hubbert's (1953) problem numerically for obtaining new insights on hydrodynamic conditions that control petroleum migration and entrapment mechanisms. I applied this method to solve fundamental issues of long-distance petroleum migration and accumulation in the Los Angeles basin, which is intensely faulted and disturbed by transpressional tectonic stresses, and host to the world's richest oil accumulation. To constrain the model, known subsurface geology and fault structures were rendered using geophysical logs from industry exploration boreholes and published seismic profiles. Plausible multiphase model parameters were estimated, both from known fault permeability measurements in similar strata in the Santa Barbara basin, and from known formation properties obtained from numerous oil fields in the Los Angeles basin. I then developed reservoir-scale hydrocarbon entrapment models to investigate the effects of the episodic fluid flow phenomenon triggered by seismic activities along the Newport-Inglewood fault zone (NIFZ) in the Los Angeles basin, California. Sensitivity tests on fault permeability and the frequency of episodic pulses were performed to investigate their effects on spatial and temporal distribution of hydrocarbons, which mainly accumulated along the fault zone and adjacent reservoir sands. Simulations show that a combination of continuous hydrocarbon generation and primary migration from upper Miocene source rocks in the central Los Angeles basin synclinal region, coupled with subsiding basin fluid dynamics, favored the massive accumulation and alignment of hydrocarbon pools along the NIFZ. According to my multiphase flow calculations, the maximum formation water velocities within fault zones likely ranged between 0.5 and 1.0 m/yr during the middle Miocene to Pliocene (13 to 2.6 Ma). The estimated time for long-distance (~ 25 km) petroleum migration from source beds in the central basin to oil fields along the NIFZ is approximately 90,000 ~ 220,000 years, depending on the effective permeability assigned to the faults (k~ 5 to 50 millidarcys) and adjacent interbedded sandstone and siltstone "petroleum aquifers". With a fault permeability of 20 md (2.0×10-14 m2), the total petroleum volume of oil reservoirs located along the NIFZ (~ 2 billion barrels) would have accumulated likely over 180,000 years or less. The results also suggest that besides the thermal and structural history of the basin, the fault permeability, capillary pressure, and the juxtaposed configuration of aquifer and aquitard layers played an important role in controlling petroleum migration rates, patterns of flow, and the overall fluid mechanics of petroleum accumulation. Episodic flow enhances hydrocarbon accumulation by enabling step-wise build-up in adjacent sedimentary formations due to temporally introduced high pore fluid pressure and permeability during the fault rupture. These episodic migration models could be more preferable than continuous models, considering current hydrocarbon distribution and tectonic settings of the Los Angeles basin. Sensitivity test results suggest that transient fault permeability and pore fluid pressure fluctuation are crucial factors for distributing hydrocarbon accumulations in the fault zone, and they also play important roles to determine time-scale of reservoir formation. Under assumptions that fault permeability and pore pressure fluctuate within the range of 1 ~ 1000 md and 10 ~ 80% of lithostatic pressure, the maximum petroleum velocity at peak of episodic flow pulse is approximately 1.0-2.0 m/yr, and the current level of total oil reserve in the Inglewood oil field (~ 450 million barrels oil equivalent) can be reached in about 24,000 years if the seismically induced fluid flow occurs every 3,000 years.
Thesis (Ph.D.)--Tufts University, 2013.
Submitted to the Dept. of Civil Engineering.
Advisor: Grant Garven.
Committee: Anne Gardulski, Andrew Ramsburg, and James Boles.
Keywords: Petroleum geology, Hydrologic sciences, and Geological engineering.read less
|
This object is in collection