Studying hydrate management

Jerry Lee

March 1, 2016

Jerry Lee takes a look at recent work conducted by the Colorado School of Mines and the University of Tulsa (Oklahoma) to mitigate gas hydrate formation.

Hydrate formation, deposition, and detection experiments performed for RPSEA in the CSM Hydrate Center in collaboration with Paulsson Inc. (A. Majid, H. Qin, T. Charlton). 

Image courtesy of Colorado School of Mines, Center for Hydrate Research.

Gas hydrate formation is a bane for subsea oil and gas production. Composed of ice-like crystalline solids, these masses can cause flow assurance issues and plug up subsea oil and gas flow lines.

To tackle the problem, the US-based Research Partnership to Secure Energy for America has partnered with Colorado School of Mines (CSM) and University of Tulsa (TU) to study hydrate formation.

“When hydrates form in subsea oil and gas flowlines, they can lead to flowline blockages and loss of production, as well as potential safety and environmental risks,” says Carolyn Koh, professor of chemical and biological engineering and director of the Center for Hydrate Research at CSM.

Hydrates form when light hydrocarbons (methane, ethane, and propane), mix with produced water, and are exposed to low temperatures and high pressure, conditions well within the gas hydrate stability zone.

Current mitigation tactics include thermodynamic hydrate inhibitors (e.g. methanol, glycol), which shift the boundaries of the hydrate stability zone to more extreme conditions and prevent the hydrate from forming; low dosage hydrate inhibitors: kinetic hydrate inhibitors, which aim to delay hydrate formation long enough so the fluids can be produced, and anti-agglomerants, which prevent hydrate crystals from grouping together, and keeps the hydrate as a transportable slurry. However, due to water-cut increasing as fields mature, greater inhibitor volumes are required for it to remain effective. This leads to greater costs, making it economically impractical, and increasing environmental concerns. This has led to a shift in ideology from hydrate prevention, to hydrate management.

In order to improve hydrate management strategies, a comprehensive model is needed, which can be used to predict the formation and transportability of hydrates under various oil and gas production scenarios. However, the current models are only for oil continuous (oil dominates with all the water dispersed as droplets) and water/gas systems.

Hydrate adhesion micromechanical force measurements. (E. Brown)Image courtesy of Colorado School of Mines, Center for Hydrate Research. 

To supplement our understanding of hydrates, the joint study investigates water continuous systems and partially dispersed systems, which lack modeling. Water continuous systems are water dominant with oil dispersed as droplets, and partially dispersed systems, are systems where water is both dispersed in oil and is in a free phase.

Before determining a model, researchers need a fundamental understanding of how hydrates form. For phase one1 of the two-phase study, the objective was to study partially dispersed and water continuous systems using a large-scale flowloop test, which is closer to field conditions than lab-scale measurements.

“Specifically, to investigate the hydrate formation process as a function of water amount, velocity of the fluid in the pipe, and pressure (subcooling; driving force for hydrate formation). A plugging mechanism and conceptual model for these systems would be investigated from the tests,” Koh says. “A first-pass, plugging onset correlation for both partially and fully dispersed systems would then be determined.”

Ten experiments were performed using TU’s flowloop and mineral oil (Crystal Plus 350T). Each test involved adding calculated amounts of sodium chloride solution (mimicking salt water), pressurizing the flowloop to 1500psig with Tulsa City gas, varying the velocity from 2.3-5.5 ft/s, and cooling the system with a glycol jacket. Incorporating a colored dye into the water, visual observations were made to determine when free water formed, how hydrates grow and plug, and the pressure drop was tracked to determine when hydrates began to form. From these observations it was determined that partially dispersed systems were more problematic than continuous systems.

Top: Before hydrate formation, water wets the pipe walls. Bottom: Hydrates (white) form and accumulate at the interfaces and occlude water. Source: Rosli, F., Majid, A.A.A., et al, CSM, CHR. Redrawn with permission from SPE, OTC-25661-MS, P. Vijayamohan et al., 2015.

“The free water enables hydrate growth and accumulation to be far more rapid than when free water is not present, with the water being occluded in the crystallites so the effective volume of solids is much greater,” Koh says. “Capillary bridging between hydrate crystals is also facilitated by the free water.”

With free water at the bottom of the pipe, the inner pipe surface could be wetted by oil pushing some water to the pipe wall or by a water slug, creating a thin film of water. Once conditions reach the gas hydrate stability zone, hydrate films can form rapidly around the pipe walls. As the oil pushes more water towards the pipe wall, hydrate thickness will increase, eventually forming stationary or moving beds. This combination of water wetting and hydrate bedding is proposed as the mechanistic conceptual model for hydrate formation/plugging for partially dispersed systems.

Expanding on these findings, phase two2 of the study, currently ongoing, aims to validate and test the conceptual model and correlation developed in phase one with further flowloop tests.

“Flowloop tests were initially performed in phase two to understand the effect of liquid loading and oil properties on hydrate formation and plugging onset,” Koh says.

Initial work for phase two began with 18 hydrate formation experiments performed using the TU flowloop. This study investigates the effects of water cut (65% and 80%), mixture velocity (2.3 ft/s and 5.5 ft/s), and liquid loading (50%, 70%, 90%). To determine whether the results are oil-specific, tests were done using three oils (MO 350T, MO 70FG and kerosene); effects on hydrate transportability, plugging tendencies, and the effects of viscosity would also result.

Experimental procedures similar to those used in phase one were used for phase two.

Partially dispersed systems are more complex and problematic than fully dispersed systems. Source: A. Majid et al., Colorado School of Mines, Center for Hydrate Research 

From these experiments, a new conceptual mechanistic model has been developed. Significant portions of the pipe are wetted by the free water layer (bottom) and pockets of water carried and spread by oil (top). Under hydrate forming condition, hydrates will develop from the thin water layer and at the oil-water interface. These hydrates occlude (soak-up) water. Further water occlusion contributes to hydrate growth and agglomeration, eventually leading to water depletion. The combination of water depletion, pipe roughness and increasing viscosity of the oil phase, due to the hydrates, will lead to stationary or moving hydrate beds forming.

If stationary beds form, the reduced area of flow, increased friction, and greater oil viscosity can contribute to plugging.

The experimental data showed that hydrate transportability is a stronger function of liquid loading and water cut than mixture velocity. However, this is only based on experiments with a single oil (MO-350T). To test the findings, and evaluate the effects of viscosity on hydrate transportability, further experiments with MO-70FG and kerosene were performed.

“Also developed was the plugging onset correlation to indicate when plugging will occur, and the extent of plugging risk for different systems/conditions,” Koh says.

The next set of flowloop tests will further develop the plugging onset model and examine methods to mitigate partially dispersed systems, including the use of anti-agglomerants. The study will also consider the effects of advanced robust coating with Oceanit. In support of the study, new fiber optic acoustic sensor technology, provided by Paulsson, will be used to detect and quantify hydrate accumulation.

When the study completes (ca.September 2016), the model will be available to aid production strategies development to reduce hydrate plugging related risks. “The mechanistic models and plugging risk assessment provide new understanding on systems that have been previously underexplored,” Koh says. “Such information can provide advanced assessments of the safety risks associated with hydrate formation where free water is present. These assessments can also be used to provide field case studies on partially dispersed systems.”

Works cited

1 Vijayamohan, P., Majid, A., Chaudhari, P., Sloan, E. D., Sum, A. K., Koh, C. A., … Volk, M. (2014, May 5). Hydrate Modeling & Flow Loop Experiments for Water Continuous & Partially Dispersed Systems. Offshore Technology Conference. doi:10.4043/25307-MS

2 Vijayamohan, P., Majid, A., Chaudhari, P., Sum, A. K., Koh, C. A., Dellacase, E., & Volk, M. (2015, May 4). Understanding Gas Hydrate Growth in Partially Dispersed and Water Continuous Systems from Flowloop Tests. Offshore Technology Conference. doi:10.4043/25661-MS