Loading...
Investigation of hydrate formation and transportability in mutiphase flow systems
Grasso, Giovanny A.
Grasso, Giovanny A.
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2015
Date Submitted
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
The oil and gas industry is moving towards offshore developments in more challenging environments, where evaluating hydrate plugging risks to avoid operational/safety hazards becomes more difficult (Sloan, 2005). Even though mechanistic models for hydrate plug formation have been developed, components for a full comprehensive model are still missing. Prior to this work, research efforts were focused on flowing hydrate particles with relatively little research on hydrate accumulation, leaving hydrate deposition in multiphase flow an unexplored subject. The focus of this thesis was to better understand hydrate deposition as a form of accumulation in pipelines. To incorporate the multiphase flow effect, hydrate formation experiments were carried out at varying water cut (WC) from 15 to 100 vol.%, liquid loading (LL) from 50 to 85 vol.%, mixture velocity (vmix) from 0.75 to 3 m/s, for three fluids systems (100 % WC, water in Conroe crude oil emulsions and King Ranch condensate + water) on the ExxonMobil flowloop (4 in. nominal size and 314 ft. long) at Friendswood, TX. For the 100 % WC flowloop tests, hydrate particle distribution transitions beyond a critical hydrate volume concentration, observed values were between 8.2 to 29.4 vol.%, causing a sudden increase in pressure drop (DP). A revised correlation of the transition as a function of Reynolds number and liquid loading was developed. For Conroe emulsions, DP starts increasing at higher hydrate concentrations than King Ranch condensate, many times at 10 vol.%. Experiments with King Ranch show higher relative DP (10 to 25) than Conroe (2 to 10) performed at the same vmix and LL. Cohesive force measurements between cyclopentane hydrate particles were reduced from a value of 3.32 mN/m to 1.26 mN/m when 6 wt.% Conroe was used and to 0.41 mN/m when 5 wt.% Caratinga crude oil was used; similar values were obtained when extracted asphaltenes were used. King Ranch condensate (11 wt.%) did not significantly change the cohesion force (3.32 mN/m). These measurements prove the importance of natural surfactants in crude oil for particle dispersion. An experimental methodology was provided to determine the effectiveness of asphaltenes as a dispersant. Even though hydrate deposition was inferred from the flowloop tests, it could not be verified from these measurements . Custom-made experimental set-ups (a recirculation liquid system, a rocking cell and a lab-scale mini-loop) were used to isolate the hydrate deposition investigation. Besides water, mineral oil 70T and King Ranch condensate were used in combination with water for the deposition investigation. One of the most important deliverables of this thesis was the construction of a lab-scale flowloop that provides insight on deposition phenomenon in multiphase flow, representing the only set-up, reported in the literature, suitable for this investigation. The miniloop can handle gas-liquid flow (maximum flow rates of 10 Nm3/m for gas and 22 GPM for liquid) through a 10 ft. long straight section (2 in. standard tubing). The testing section (30 in. long) was designed to observe hydrate deposition on the wall. Three mechanisms of hydrate deposition were identified: film growth, particles adhering and particle bedding. The maximum water conversions were: 27.5 ml in the rocking cell, 2400 ml in the miniloop with 100 % WC and 250 ml in the miniloop for dispersed water in mineral oil 70T. The measured DP across to the testing section ranged from 0 to 8 in. H2O. Deposits were obtained for different flow regimes, including 100 % LL, stratified, stratified-wavy and slug flow. The maximum deposit thickness was 1.5 in., obtained in the gas flowing section. When deposits form from particle cohesion, they were easy to slough. From all the experimental worked in this thesis, hydrates accumulated depending on the degrees of subcooling of the bulk fluid, film growth (between 3 to 5 F), deposition from a combination of film growth and particle cohesion/adhesion (>7 F) and bedding (>10 F). An approach for hydrate deposition modeling from film growth has been proposed; which aids in the future development of physical models to evaluate hydrate plugging risks. This investigation has provided insight on hydrate deposition mechanisms.
Associated Publications
Rights
Copyright of the original work is retained by the author.