Gas-liquid separation using axial flow cyclones.

Work to improve the oil-gas extraction processes from the wellhead to basic saleable product has been a consistent area of study for over 50 years. In this project, it is aimed to develop high capacity plant to obtain low liquid entrainment levels by separating oil droplets from the dispensed gas. Commonly used gas cleaning equipment has disadvantages that inhibit its use in separating oil droplets from gas including excess bulk, too low gas handling capacities, poor separation efficiencies and the need for sophisticated maintenance.
The objective of this research is to focus on one of the more recent manifestations of a basic separation technology, the axial flow gas cyclone incorporating drainage slots in the barrel. The work enables quantitative understanding of the performance of the axial flow cyclone separating liquid droplets as an aid to intensifying oil and gas extraction processes. Experimental work was carried out to obtain the pressure drop - flowrate characteristics, data on the onset of re-entrainment and the droplet separation efficiency of the cyclone tested. Modelling work was also carried out using Computational Fluid Dynamics (CFD) and published analytical models to investigate the feasibility of modelling the pressure drop - flowrate characteristics and the grade efficiency of the tested cyclone. The methodologies to integrate individual tubes into a separating vessel and cyclone optimisation have also been covered.
It was noticed that with the centre body swirler as the swirling device, a frothing zone occurred at low air flowrates. With the occurrence of this zone, re-entrainment was bound to occur. This was the deficiency of this sort of inlet design because the airflow was not strong enough to swirl the liquid. Instead it was only enough to prevent the liquid from falling backwards and this increased the system pressure drop significantly. Therefore, tangentially oriented inlet swirl vanes with four of the slots used as additional drainage was employed. The frothing zone was then eliminated at low air flowrates. However, at very high air flowrates re-entrainment still occurred which was due to a liquid creeping mechanism and the stripping of liquid film on the cyclone wall.
CFD and the other analytical models were able to predict the cyclone’s pressure drop - flowrate characteristics well, however, the agreement of the grade efficiency curves with the experimental data was poor. Comparison of the developed axial flow cyclones with the commercially available cyclone indicated that the Sheffield design performs better in terms of droplet separation efficiency, but at the expense of pressure drop.