Usage
  • 256 views
  • 584 downloads

Development of Robust Three-Phase Equilibrium Calculation Algorithms for Complex Reservoir Fluids

  • Author / Creator
    Li, Ruixue
  • Three-phase equilibria, such as three-phase vapor-liquid-aqueous (VLA) equilibria and three-phase vapor-liquid-asphaltene (VLS) equilibria, can frequently appear in hydrocarbon reservoirs. In three-phase equilibrium calculations, one of the most prominent problems is caused by the lack of a prior knowledge of the phases that are actually present. Trivial computing results can frequently appear if the equilibrium calculations are improperly initialized. The presence of the aqueous phase or the asphaltene phase (which is nearly a pure phase in most cases) skews the topography of the free-energy surface, leading to that one of the stationary points may appear near the boundary of the Gibbs free-energy surface. This substantially increases the probability of encountering convergence problem in the three-phase equilibrium calculations. In this research, we aim to develop a suite of techniques to improve the robustness of the three-phase equilibrium calculation algorithms for the VLA and VLS equilibria.
    In most cases, the aqueous phase can be considered as a pure water phase (the so-called “free-water assumption”). With this assumption, we first develop a robust and efficient algorithm used for conducting isothermal three-phase equilibrium calculations which can consider single-phase, two-phase, and three-phase VLA equilibria. Subsequently, we develop a three-phase free-water isenthalpic equilibrium calculation algorithm by combining the newly developed VLA algorithm with the energy conservation equation which is used to convert enthalpy to temperature. This three-phase free-water isenthalpic equilibrium calculation algorithm can be applied to reservoirs undergoing thermal enhanced oil recovery treatments (in which temperature dramatically changes and is difficult to be known beforehand). A number of example calculations are carried out to demonstrate the performance of these two algorithms. Testing results prove that the newly developed algorithms are robust and effective. For some reservoir mixtures (e.g., a mixture containing CO2 or H2S), the free-water assumption may not be valid since the aqueous phase can contain a significant amount of other species; to accurately capture the three-phase equilibria for such mixtures, we develop a new initialization scheme as well as a new procedure to improve the robustness of the three-phase isothermal equilibrium calculation algorithm. To test the robustness of this algorithm, it is applied to several fluid mixtures to generate pressure-temperature (P-T) phase diagrams, showing that the newly developed algorithm is both robust and efficient.
    Moreover, we further develop a three-phase VLS isothermal equilibrium calculation algorithm by applying the asphaltene-precipitation model proposed by Nghiem et al. (1993). In their model, they assume that the asphaltenes form a pure phase; this assumption is similar to the free-water assumption. This three-phase VLS isothermal equilibrium calculation algorithm aims to model the CO2 ¬flooding in light oil reservoirs. New initialization methods of equilibrium ratios are provided in this algorithm for both stability test and flash calculation. To test the performance of this algorithm, it is run to generate pressure-composition (P-X) phase diagrams for several reservoir fluids. The new algorithm is shown to be robust as it can always converge to the correct phase equilibrium for all the tested cases. Afterwards, by applying the three-phase VLS equilibrium calculation algorithm, we develop a multiple-mixing-cell (MMC) method to predict the minimum miscibility pressure (MMP) with the consideration of asphaltene-precipitation effect. Example calculations are carried out to predict MMPs between reservoir fluid and pure or impure CO2. The MMPs predicted by our algorithm and those predicted by the MMC algorithm without considering asphaltene-precipitation effect are both compared with the MMPs measured by slim tube experiments. The comparison results show that the MMPs predicted by our algorithm agree reasonably well with the measured MMPs.

  • Subjects / Keywords
  • Graduation date
    Fall 2019
  • Type of Item
    Thesis
  • Degree
    Doctor of Philosophy
  • DOI
    https://doi.org/10.7939/r3-2skd-3s60
  • License
    Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.