ICMF-2016 – 9th International Conference on Multiphase Flow May 22nd –27th 2016, Firenze, Italy I would to acknowledge the support of my sponsor the Petroleum Technology Development Fund, Nigeria. Study on the removal of low loading heavier phase by lighter phase in multiphase pipe flows Aminu J A Koguna*, Liyun Lao and Hoi Yeung 2 1 Cranfield University College Road, Cranfield Bedford MK43 0AL ajak007@hotmail.com 2 Cranfield University College Road, Cranfield Bedford MK43 0AL l.lao@cranfield.ac.uk Abstract Production fluids from the reservoir are transported in pipelines from the well heads to the platform and from the platform to process facilities. At low flow velocity water, sand or liquids like condensate could settle at the bottom of pipelines that could lead to grave implications for flow assurance. During shutdown the settled heavy liquid (e.g. water), could result in corrosion in pipelines, while following restart stages the settled water could form water plugs that could damage equipment, while settled sand could also form a blockage that needs to be purged. Observations of low water cut in oil and water flows and low sand concentrations in water and sand flows were conducted in four and two inch diameter pipelines. Conductive film thickness probes were used to ascertain structural velocities. Comparisons are made between two cases in order to gain better understanding of the dispersal process of low loading heavier phase in multiphase flows. Keywords: Water/oil flow, Solid/liquid flow, Low water cut, Interface height, Low sand loading, Velocity. 1. Introduction Production fluid transport is fundamental part of the energy value chain. The inherent price volatility due to varying supply and demand entails knowing ideal production flow rates to prevent settling of heavier phase like water in oil transfer pipelines and sand in flow lines. A classification was provided by [1] for dispersed, stratified dispersed, stratified, slug, and annular flow regimes while [2] gave a classification for separated flow, water dominated and oil dominated dispersed flow regimes. Similarly [3] studied water wetting in two-phase oil-water flows and used four main methods of flow pattern visualization, wall sampling, iron concentration monitoring and wall conductance probes to detect phase wetting at different superficial water and oil velocities in large diameter horizontal pipes. Also [4], [5] and [6] discussed velocities required to prevent settling of solids in pipelines while [7] studied low loading solid particles under intermittent flow conditions. However there has not been work studying load loading heavier phase in multiphase pipeline under stratified flow conditions as considered in this study. Experimental setup Oil and water flow loop The Oil and water experiments were conducted on the Three- Phase Test Facility at the Process Systems Engineering Laboratory at Cranfield University. As shown in Fig. 1, water and oil are pumped to the test area through a valve manifold system after measurement by flow meters. The length of the horizontal 4” line, in which the tests are conducted, is 26.3 m. The probe spool used in this study is mounted on the 4” horizontal flow line. This horizontal line exits directly into the 3-phase separator. The water and oil are stored in 12,500 litres capacity tank respectively, and are pumped into the flow loop using two multistage Grundfos CR90-5 pumps. The pumps are identical and have a duty cycle of100 m3/hour at 10 bar g pressure. The flow rates of the water and oil are regulated by their respective control valves. The water flow rate is metered by a 1” Rosemount 8742 Magnetic flow meter (up to 1 kg/s) and 3” Foxboro CFT50 Coriolis meter (up to 10 kg/s) while the oil flow rate is metered by a 1” Micro Motion Mass flow meter (up to 1 kg/s) and 3” Foxboro CFT50 Coriolis meter (up to 10 kg/s). Water injected into 4” test line via a horizontal 2” flow line which is connected to the 4” flow line by a flexible 11mm i.d. PVC tubing. There is a gate valve to supply the water and another ball valve for flow control upstream of an ultrasonic flow meter. From the flow meter, the water is supplied to the four inch loop through a manifold that separates the flow into the three injection points that have mini-ball valves V2, V4 and V5 on the four inch pipe. During a test only one valve of the three is open allowing water injected at a specific location at a time. The three injection points are located 10D, 30D and 50D from the test spool. The flow meter used for precise water injection is an Atrato ultrasonic flow meter model 760 V10 with an accuracy of ±1.5% repeatability of ±0.1%, turn down of 250:1 and 10bar rating. It uses time of flight method and measures a flow range from 0.1 to 20 litre/min. The maximum full scale error or maximum absolute error is 3.75% for the flow meter under pulsating conditions.