Spatial Distribution of Oil and Water in Horizontal Pipe Flow
- Arash Soleimani (Imperial College) | C.J. Lawrence (Imperial College) | G.F. Hewitt (Imperial College)
- Document ID
- Society of Petroleum Engineers
- SPE Journal
- Publication Date
- December 2000
- Document Type
- Journal Paper
- 394 - 401
- 2000. Society of Petroleum Engineers
- 4.1.2 Separation and Treating, 1.6 Drilling Operations, 5.4.2 Gas Injection Methods, 4.2 Pipelines, Flowlines and Risers, 1.6.9 Coring, Fishing, 4.1.5 Processing Equipment, 5.3.2 Multiphase Flow, 4.1.9 Tanks and storage systems
- 0 in the last 30 days
- 351 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
In this paper we report a series of experiments to quantify the spatial distribution and identify the flow pattern of liquid-liquid flow in a horizontal 25.4-mm (nominal 1-in.) tube. Experimental results are presented for kerosene (EXXOL D80) and tap water at room temperature. Two different measurement techniques (a high frequency impedance probe and a gamma densitometer system) were applied for measuring the volume fraction distribution across the tube and to obtain tomographic results for phase distribution. The use of the gamma densitometer system to obtain the tomographic results in liquid-liquid cocurrent flow is believed to be the first in this field. These methods are more precise than other techniques such as visualization and help to distinguish certain differences in flow patterns for different superficial velocities and liquid fractions. The two sets of measurements were compared and it was concluded that the gamma densitometer system was a more reliable method by which to measure the volume fraction. Two important phenomena in liquid-liquid flow were observed: oil encapsulation by water at low mixture velocity and droplets concentrated at the center of the pipe in the dispersed flow regime. Some possible explanations are given regarding these phenomena.
The aim of this work was to quantify the spatial distribution and identify the flow pattern of liquids in the cross section of a pipe using a high frequency impedance probe and a gamma densitometer system. These methods allow the identification of key features in the flow patterns and provide comprehensive information on the liquid fractions for different superficial velocities. Most previous studies have identified the flow pattern visually and/or according to variations in pressure drop measurements. Only more recently have extra tools, such as conductivity probes or sampling tubes been used.1-5 However, the use of the gamma densitometer system to obtain tomographic results in liquid-liquid cocurrent flow reported below is believed to be the first in this field. Some of these methods of measurement have been used in multiphase flowmeters to measure the oil and water fractions and it is important to evaluate them. The number of multiphase-flowmeters in operation has reached 150 since 1995.6 Obviously knowledge of the spatial distribution of the two liquids will help in constructing better phenomenological models for dispersed and stratified flow.
Experimental System and Methods
The two-phase oil and water experimental rig (Tower) experimental facility is shown schematically in Fig. 1; it consists of the following items.
Two storage tanks, one for each fluid with a volume of 0.681 m3 each.
A fluid pumping section. The oil and water are pumped to the test section by a flow control system. In this system part of the total flow that comes from the pump discharge is recycled back to the storage section so that the desired flow rates are maintained in the test section.
A liquid-liquid separator vessel, made of PVC reinforced with steel, containing a knitmesh coalescer to promote efficient separation of the fluids.
A horizontal 25.4-mm (1-in. nominal) diameter stainless steel test section, which consists of five stainless steel tubes with lengths of 1, 3.8, 1.91, 1.91 and 1 m connected together in this order, giving a total length of 9.7 m.
A transpalite test section not used in the current work.
Flow mixers not used in the current work.
All the measurements were taken 8 m from the entrance of the test section at three different mixture velocities (1.25, 2.12, and 3 m/s) for a range of water cuts (20, 24, 32, 34, 36, 42, 46, 60, and 70%).
High Frequency Impedance Probe System
The high frequency impedance probe is shown in Fig. 2. It was mounted in the test section between two flanges and the tip was traversed across diameters at different angles to the horizontal to collect the data. Data were collected every 2 mm across the diameter of the tube up to 1 mm from the wall. This process was carried out for three different angles, 15, 45 and 75°, from a horizontal plane. However, the vertical plane splits the pipe into two symmetrical sections, so symmetrical data can be obtained. The data on the 15° line are also used for the 165° line and similarly the 45° for the 135° and also 75° for the 105°. The data were collected at a frequency of 5000 Hz over a period of 12 seconds and were processed using the Fortran program of Angeli.7 A program in Mathematica was written to interpolate the data and to produce a contour diagram of the oil and water distribution.
Gamma Densitometer System
This system measures the average phase fraction along a chord and can be traversed across the tube cross section. The gamma source was mounted on a platform so that the angle of the beam in relation to vertical could be altered from 0 to 45° and 90° (Fig. 3). The distances of the source and photomultiplier tube from the test section tube were kept constant for all angles. The distance between measurement positions was 1 mm. In this series of experiments, a relatively long counting time of 15 seconds was used, so as to minimize the errors in the system. A stepping motorized unislide was used to traverse the entire gamma ray system across the cross section of the tube, so that a hold-up profile could be obtained. Results from multiple scans were combined to obtain tomographic images of the pipe cross section.
A calibration was performed at the start of each series of experiments. The tube was filled with one phase and measurements were taken for each preselected position across the cross section. To check the absolute position, the empty tube was also scanned and the plots of intensity against distance were aligned with those obtained when the tube was filled with oil or water. Points where these graphs intersect each other are outside the tube or within the tube walls and provide a limit on the positions that need to be measured. From the source, the gamma beam passed through a collimator, (a 30-mm long and 3-mm diameter hole drilled in the lead shielding container), diverged after emerging from the collimator, and the part of the beam with significant intensity had a width of approximately 6 mm at the detector.
|File Size||1 MB||Number of Pages||8|