Development and application of hydrate formation, transportation and bedding models in liquid-dominated systems, The

Abstract

Flow assurance has emerged as a technical discipline that focuses on the design of safe and secure operation techniques for uninterrupted transportation of reservoir fluids in the flowlines. Flow assurance includes various subjects related to the formation of solids, such as asphaltene, wax, scale and hydrate control and management. Gas hydrates are ice-like solid compounds composed of water cavities filled with small gas molecules. Under thermodynamically favorable conditions, gas hydrates can form rapidly in the pipelines and plug the pipeline within a relatively short timescale (i.e. a few hours to a few days), making it one of the major problems in the flow assurance industry. Hydrate plugs are not only a potential risk to safe production, but are also usually costly to locate and remove. Although there are ways to either avoid or control the hydrate formation in subsea pipelines, in order to understand hydrate formation at different operating conditions, a predictive tool for hydrate slurry transportation is critically required. The Center for Hydrate Research in the Colorado School of Mines has devoted large efforts to develop a transient hydrate simulation tool, CSMHyK, which stands for Colorado School of Mines Hydrate Kinetics. CSMHyK is coupled with a transient multiphase flow simulator and contains models for different flow systems. In oil-dominated systems, it is assumed that hydrate growth occurs at the interface of the dispersed water droplets and hydrate agglomeration is caused by the cohesive force between hydrate particles. If the fluid shear is not able to suspend the large hydrate agglomerates, they may bed at the bottom of the pipe. In water-dominated systems, hydrate formation is simplified to form only at the interface of the gas bubbles entrained in the water phase. In gas-dominated systems, hydrates form both on the surface of the gas bubbles in the water phase and on the water droplets entrained in the gas phase. Hydrates can also grow on the pipe wall surface due to water condensation and hydrate film growth. Once formed in the bulk gas phase, hydrate particles could deposit on the pipe wall due to adhesive forces. In this thesis work, CSMHyK has been modified to consider phase inversion from an oil- to a water-dominated system. The effective hydrate slurry viscosity has been modified to consider the contribution from both the hydrate and the emulsion dispersion. Thermodynamic inhibitor concentration changes due to hydrate formation are taken into account inside CSMHyK. CSMHyK models are validated by comparing predictions with the experimental results obtained from the University of Tulsa and ExxoMobil flowloops at different operating conditions, which include different liquid holdup, water cut, oil type, and mixture velocity. The modified CSMHyK has been applied to various field simulations and showed relatively good agreement with field data. Those fields include black oil and gas condensate subsea tiebacks as well as dry tree facilities. When applying CSMHyK to field simulations, it is recommended that multiscale experiments, such as rheology, bottle tests, HP-DSC and HP-MMF are conducted to understand the oil and emulsion properties. From field simulations, it is indicated that CSMHyK can be used for the design and optimization of subsea transport facilities, as well as for the prevention, management and remediation of hydrates in the pipelines. From CSMHyK simulations, it is indicated that the low spots where water accumulates usually have a higher chance of hydrate plugging. High salt concentration in the formation water may limit the maximum amount of hydrates formed in the system. The emulsion stability may play an important role in predicting the hydrate slurry transportation. The maximum hydrate formation volume fraction in a liquid export line could be up to 20 vol.%. Prior to this thesis work, CSMHyK only considered viscosification and agglomeration. However, hydrate plugging is a complicated process which can involve bedding, deposition, jamming and hydrate film growth. A hydrate bedding model for the oil-dominated system is developed in CSMHyK in this thesis. In the bedding model, it is hypothesized that hydrate agglomeration controls bedding and capturing bedding will contribute to pressure drop prediction. In the implementation of the bedding model, it is assumed that hydrate agglomerate sizes follow a log-normal distribution (as indicated from autoclave experiments) and bedded hydrate agglomerates may be re-dispersed to the flow layer based on the force balance. The bedding model is coupled with the hydrate agglomeration and viscosification model. This bedding model has been tested against different flowloop tests with Conroe crude oil at different mixture velocities and shows qualitative agreement with experimental observations both during continuous pumping and ramping tests. In the last chapter of the thesis, a sensitivity study was performed illustrating how the cohesive force between hydrate particles, liquid velocity and water cut could affect hydrate transportability and pressure drop. Recommendations for future development on partially dispersed systems are provided.