An experimental study on three-phase viscous oil flow has been conducted at IFE's Well Flow Loop in a 25 m long 10 cm I.D. pipe at +1 degree. In the experiments the gas phase is a dense gas (Sulphur hexafluoride, SF6) at 7.5 bar pressure, the oil phase is a mixture of Primol 352 /ExxsolD60 with different compositions to achieve an oil viscosity between 69 and 160 mPa.s., tap water is used as the water phase. Both a fast response Xray tomography system and different gamma densitometers were utilised to measure holdups and phase distributions. In this work, a detailed analysis of time series of measurement from these two types of instruments for three-phase flow is performed. The focus is on the volume fraction of gas in the liquid slug. A comparison of the measured results from these two instruments is conducted, and the effect of water cut and oil-water flow regime on the void-in-slug is also analysed. The performance of some physical models for void-in-slug is assessed against the current data.
Long distance transport of offshore heavy oil production fluid requires an accurate prediction of thermo-hydraulics of multiphase flow in pipes. The possible scenarios are: single phase oil flow (under-saturated oil), two-phase gas-oil flow (saturated oil, or with gas production, or gas-lift in well), oil-water flow (under-saturated oil) and three-phase oil-water flow. In situations with presence of gas and water in the flowline, the flow regime is dominated by three-phase slug flow.
Many studies of three-phase slug flow with viscous oil have been focussing on the generic behaviour of hydrodynamics – oil-water flow regime, pressure gradient, and phase holdup (1) – (10). All of these studies were conducted with relatively small pipe diameter (<0.07m I.D.) and low gas density. According to these studies, there has been a relatively good understanding on the slug front velocity under three-phase flow conditions. Little work has been found in the open literature on the gas entrainment in the liquid slug, the main reason for the lack of experimental data on the void in slug is simply due to expensive instrumentation needed to get time series of three-phase volume fraction with relatively high time resolution.
