Abstract
This work presents a study on the design of three-phase separators. It summarises the theory of operation with specific emphasis on the influence of droplet size, frequency, and distribution on separator efficiency (defined as the percentage of oil in the oil outlet).This research aims to improve the basic understanding of the underlying physics of the separation process via rigorous experimental and computational analysis. This knowledge is then used to develop a demonstrator and associated learning materials to teach this theory to further and higher education students.
Current design methods assume a constant 500 μm droplet size and hence produce conservative designs. The investigation set out to define a methodology that would account for droplet size effects. It started with experimental work on a transparent laboratory separator to determine the actual droplet size and associated properties. Data was carefully collected using advanced photography and image processing techniques. The results indicated the significant impact of oil flow rate and oil pad thickness on droplet size, distribution, and separator efficiency.
The study then moved on to an investigation of how temperature affects the efficiency of the separator. The temperature is increased to decrease fluid viscosity and increase settling rates. It was discovered that raising temperature improves separator efficiency but it has an inverse effect on droplet sizes, decreasing the diameter while increasing droplet frequency. This implies separator efficiency should decrease. However, it was found that the temperature increases coalescence which is the dominant process and this accounts for the increased observed efficiency.
The study then investigated the impact of salinity on separation. (Shafiei M., 2022) found that increasing salt concentration droplet size increases the viscosity of the continuous phase. A full factorial statistical design method was employed to investigate the effect of salinity alongside changes to flow rate, oil layer and temperature. It was found that increasing salt content (12%) from freshwater the salt concentration becomes the dominant factor. No significant binary interactions were found for the independent variables considered.
The study then considered the impact of the slenderness ratio at constant retention time on separator efficiency. It found for increases in slenderness ratio of 2:1 to 4:1 that the efficiency increased by up to 30% for the thinnest oil layer of 9 mm. No change in efficiency was observed for 18 mm oil layer thickness. For oil layers thicker than 18 mm the efficiency is reduced by up to 15 %. It can be concluded that for very thin layers the slenderness ratio would improve performance but this effect is negated by oil layer thickness which becomes the dominant factor.
The experimental work is limited to a fixed range of parameters constrained by the configuration of the separator. CFD Modelling offers a way forward to extend the range of applicability of the theoretical insights gained from the experimental work Completion of the experimental work enabled several CFD models to be developed for the different experimental scenarios this work aimed to identify the best CFD method to model the separator and define its limits of accuracy. 2D models based on the Discrete Phase Model (DPM) and Population Balance Model (PBM) were developed. The PBM was found to be the best for modelling this type of equipment under the different scenarios investigated.
Finally, a demonstrator was designed based on the lessons learned from the experimental work undertaken in this study. The optimised design addresses critical factors such as droplet size calculation, temperature-salinity interaction, slenderness ratio effects and CFD simulations. The design is combined with experimental activities that have been designed to meet the different learning outcome-driven objectives of all levels of university study.
| Date of Award | 21 Jun 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Paul Russell (Supervisor) & Faik Hamad (Supervisor) |