| Summary: | 2015 Fall. Includes illustrations (some color). Includes bibliographical references. In this doctoral work, the viscosity and transportability of gas hydrate slurries were investigated using a high pressure rheometer and an industrial-scale high pressure flowloop. A model water-in-oil emulsion was developed specifically for this study and consists of mineral oil (350T and 70T), a surfactant mixture (sorbitane monooleate, Span 80 and Aerosol OT, AOT) and de-ionized water. The water volume fraction (water cut) that will form a stable water-in-oil emulsion depends on the type of mineral oil used. Mineral oil 70T can form a stable emulsion for water cuts in the range of 10 – 70 vol.% water, while mineral oil 350T can form a stable emulsion for water cuts in the range of 10 – 40 vol.% water. Characterization tests were conducted on these model emulsions, and the results show that the emulsions are relatively stable (no phase separation, sedimentation and coalescence) for a period of one week. In addition, the average water droplet size was determined to be in the range of 2 – 5 μm. Finally, high pressure autoclave tests were conducted using the model emulsions, and showed that the emulsions have similar properties (i.e. relation of motor current versus hydrate volume fraction) to that of water-in-crude oil emulsions. In order to understand the effect of hydrate particles on the viscosity of the system, viscosity measurements of the emulsions (prior gas hydrate formation) were conducted at various temperatures, pressures and water cuts. A generalized equation that is a function of temperature, water volume fraction and saturation of the oil phase was developed. This generalized equation is able to predict the viscosity of the emulsion fairly accurately (within ± 13%) at low temperature (≤ 10 °C). In-situ gas hydrate formation and hydrate slurries viscosity measurements were also performed in this work. Measurements were made using a high pressure rheometer connected to a high pressure ISCO pump. A four-blades vane impeller was used to mix the slurries. Experiments were conducted using the two model emulsions that were developed in this work. Viscosity measurements were conducted at a constant temperature of 1 °C, constant pressure of 1500 psig and constant mixing speed of 477 RPM. The water cut was set to be between 5 – 30 vol.% water. In addition, for mineral oil 70T emulsions, experiments were also performed near the emulsion inversion point. The results of this work shows that the relative viscosity of gas hydrate slurries can be modeled as a function of the hydrate volume fraction of the systems. In addition, emulsion breaking after hydrate formation was observed for hydrate slurries tests near the emulsion inversion points. Next, gas hydrate transportability was also investigated in an industrial-scale flowloop. Investigations were made at two different flow conditions (fully dispersed and partially dispersed systems). The different flow conditions were achieved by changing the water cut, as well as the flowloop pump speed (fluid mixture velocity). Results of the tests shows that the relative pressure drop, ΔPrel decreases with increasing pump speed. In addition, there is a higher tendency for hydrate plugging to occur at low fluid mixture velocity. Similar to gas hydrate studies in the high pressure rheometer, emulsion breaking was also observed in the high pressure flowloop for tests near the emulsion inversion point. In addition, initial investigations comparing the results obtained in the high pressure rheometer with results obtained in the industrial-scale rheometer were also conducted. Results of the comparison shows that even though the systems studied were different (different oil and different flow system), there are relatively similar increases in the relative calculations (relative ΔP and relative viscosity) when hydrates have formed in the system. Lastly in this doctoral work, the droplet size distribution in water-in-oil emulsions was determined using Diffusion-Transverse Relaxation (T2) using low field Nuclear Magnetic Resonance (NMR). The proposed method provides several advantages over the traditional optical microscopy image analysis, such as giving a better representation of the droplet size in flowlines, since the measurement uses the entire sample to determine the droplet size. In addtion, the method is relatively fast and has a low cost compared to high field NMR tests. In the low field NMR method, knowledge of the composition of the oil is not needed, which is not the case for the corresponding high field NMR method. Results from low field NMR tests were compared with the results obtained from optical microscopy image analysis. Both methods show that the average size of water droplets is relatively similar across all water volume fractions investigated, but a minimum in size was observed by both methods at 50 vol.% water, which is close to the emulsion inversion point. The research conducted in this doctoral thesis has made several contributions towards both advancements in academic research, as well as industial flow assurance research. In the area of academic research, the model emulsion that was developed in this doctoral work (discussed in Chapter 2) has been applied in a wide range of research including wax deposition studies by several other researchers. In addition, the research conducted in this doctoral thesis has been acknowledged by the hydrate flow assurance community as a breakthrough in the hydrate slurry rheological characterization. The work performed here was the first work that decoupled the effect of emulsion/gas saturation/oil swelling on the viscosity of hydrate slurries. In summary, the new findings and model/method developments presented in this thesis collectively are critical to advancing the understanding of gas hydrate slurry properties and transportability. The latter is required in the development of new hydrate management strategies during oil/gas production.
|
|---|