A Nonintrusive Flowmetering Method for Two-Phase Intermittent Flow in Horizontal Pipes
- Gene E. Kouba (Chevron E and P Services) | Ovadia Shoham (U. of Tulsa) | James P. Brill (U. of Tulsa)
- Document ID
- Society of Petroleum Engineers
- SPE Production Engineering
- Publication Date
- November 1990
- Document Type
- Journal Paper
- 373 - 380
- 1990. Society of Petroleum Engineers
- 5.3.2 Multiphase Flow, 4.1.6 Compressors, Engines and Turbines, 4.4.3 Mutiphase Measurement, 4.1.5 Processing Equipment, 5.1.8 Seismic Modelling, 4.3.4 Scale, 4.2 Pipelines, Flowlines and Risers, 4.1.2 Separation and Treating
- 0 in the last 30 days
- 183 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
Summary. An experimental and theoretical investigation of horizontal two-phase slug flow was performed to develop a nonintrusive method of measuring the in-situ gas and liquid flow rates. The key measurements were the instantaneous liquid holdups at two positions along the pipe. A new drift model was developed to predict the liquid velocity in the slug from the measured slug translational velocity. Both liquid and gas flow rates were predicted within an accuracy of +/-10% of range.
The difficulties experienced in handling multiphase flow have generated widespread interest in in-situ multiphase-flow measurements. Very often it is desirable, if not necessary, to have a means of determining the in-situ phase distribution, flow patterns and individual phase flow rate in a multiphase flowline. In offshore production systems, for example, long multiphase pipelines carrying gas, oil, and water from numerous wells to distant production facilities are becoming more common. Reservoir management and production allocation require the measuring and monitoring of production rates from individual wells. The test separator and its instrumentation represent the state of the art in multiphase-flow-rate measurements. Separation facilities, however, may be impractical for many applications because of their size and expense. Jamieson el al. emphasized that, while phase separation may be impractical, effective well management of undersea and unmanned production systems depends on accurate multiphase-flow-rate measurements. A small, retrievable, accurate multiphase flowmeter is a necessary alternative to separation, Ashkuri and Hill thoroughly discussed the need for such a meter and proposed a set of constraints for a flowmeter suitable for offshore measurements. The most significant constraint is the target mass-flow-rate accuracy of +/-5% of range needed for good reservoir management and field development. Other constraints involved size, operational conditions, response time, and ease of retrieval. In a follow-up study to Ashkuri and Hill's paper, Baker and Hayes discussed the problem of accurate multiphase flow measurements and reviewed several instrumentation options. They concluded that the gamma ray densitometer and impedance device, in conjunction with cross-correlation techniques, were a promising combination.
The problem with any method of in-situ multiphase-flow measurement is that the phase velocities and distributions are difficult to quantify, especially for nonintrusive techniques. Even if only two phases are considered to be present, significant improvements must be made in three important interrelated areas to approach the +/-5% accuracy.
First, accurate local holdup measurements are essential for accurate flow measurement. Most nonintrusive holdup measuring devices are sensitive not only to the amount of each phase present, but also to phase distribution. Therefore, the devices must be calibrated properly for the actual phase distributions encountered.
Second, proper flow-pattern identification and characterization are necessary to determine the proper phase distributions. These tasks directly affect the selection of the appropriate calibration for liquid holdup and determine the choice of flow-pattern-dependent models to use in conjunction with the measurements.
The third area for improvement is the development of models to improve the predictions of those variables that are not measured directly, such as in-situ phase velocity. Cross-correlation methods generally will yield accurate interface velocities, as in slug flow. Using the interface velocity, local holdup, and deduced flow-pattern information, the models must accurately estimate local phase velocity.
The approach taken in this study was to reduce the complexity of the multiphase-flowmetering problem by considering only horizontal, two-phase slug flow. Even with these restrictions, this still represents a broad area of concern because slug flow is generally considered to be the most predominant and complex flow pattern. Extensive data were taken in horizontal slug flow, including reliable instantaneous liquid holdup at multiple points along a flowline section. A new gravity-induced-drift model was developed to allow determination of an average liquid-phase velocity within the slug. Finally, the method enables the prediction of both liquid and gas flow rates within an accuracy of +/- 10 % of range.
Fig. 1 shows the physical model for fully developed slug flow. The bridging of the pipe diameter by the continuous liquid phase (shaded region) is generally accepted to define the slug length, Ls. The mixing region at the front of the slug, m, is determined by the distance necessary to accelerate the slow-moving film to the velocity of the liquid in the slug. The length of the slug unit, Lu, is the sum of Ls and the gas-pocket length, Lp, from one slug front to the next. Underneath the gas pocket is the liquid-film region, which is equal in length to the gas pocket. The average liquid holdups in the slug and in the film/gas-pocket regions are yLs and yLf, respectively.
The slug propagates forward in a wave-like manner with a translational velocity, vt. The velocity of the gas in the pocket, vgp, usually is slightly less than vt, while the velocity of the fluid in the slug, vs, is generally about 80% of vt. The lowest velocity in the slug unit is in the liquid film, which travels at vLf.
Liquid-Flow-Rate Equation. The liquid-flow-rate equation in slug flow is developed by summing the volume fluxes past a point on the conduit, during the passage of a slug and a film, for any number of slug units:
where the first term in the summation is the liquid volume flux in the slug and the second term is the liquid volume flux in the film.
The integration is replaced by using the definitions for a space-averaged variable: (2)
and for a time-averaged variable,
|File Size||1 MB||Number of Pages||8|