Severe Slug Flow in Offshore Flowline/Riser Systems

Abstract

Summary. Small-scale experiments are described that simulate the unstable two-phase flow pattern of severe slugging often found in offshore flowline/riser systems. In particular, the effects of riser-base supplementary gas lift, flowline orientation, and liquid viscosity have been examined. To place the tests in perspective, the physics of severe slugging is discussed, including a method for scaling between model and prototype situations. Test results are compared with a simple hydrodynamic prototype situations. Test results are compared with a simple hydrodynamic computer model. Introduction At an offshore platform receiving an oil/gas two-phase stream from a long flowline, production may be in the form of long liquid slugs and gas surges instead of being smooth. Fig. 1 shows a typical off-shore flowline/riser configuration. When the length of the liquid slug exceeds the length of the riser, severe slugging, as opposed to normal slugging, is said to occur. In normal slug flow, slugs are not very long, on the order of 20 pipe diameters. In severe slugging, however, the liquid slugs build up at the riser base; consequently, the gas is trapped and accumulates upstream until the flowline pressure is high enough to drive the slug out of the riser. Such slugs may be many times the riser length, and the difference between peak and steady-state oil and gas production rates may be considerable. Unless special measures are taken, the processing of long liquid slugs and gas surges may be associated with liquid carryover or liquid- and pressure-control problems at the oil/gas separator. The separator may thus become inoperable at its design capacity, leading to possible production losses. The seriousness of severe slugging increases possible production losses. The seriousness of severe slugging increases when a large number of flowline/riser systems arrive at one platform separator by means of a manifold and when water depths are about 400 m [1,300 ft]. In situations where the oil/gas flow is received from a well with a bottomhole pressure just sufficient to sustain the required production, severe slugging may have a detrimental effect on production, production, severe slugging may have a detrimental effect on production, and even well killing is not unimaginable. Research into severe slugging is aimed at the reliable prediction of its occurrence and of the associated slug length, frequency, and arrival velocity. In addition, techniques should be developed to eliminate or to reduce severe slugging where possible; for example additional backpressure, choking, extra gas lift, and reduced pipeline diameter have been proposed by Yocum and Schmidt et pipeline diameter have been proposed by Yocum and Schmidt et al. These authors carried out comprehensive studies on severe slugging, including small-scale tests, field tests, and hydrodynamic modeling. Nevertheless, there are questions still to be answered. Since this paper was originally prepared, further related articles have been published. Taitel provided a theoretical analysis of severe slugging and its stabilization by an increase in backpressure or by use of controlled choking. Goldzberg and McKees discussed liquid accumulation and severe terrain-induced slugging. To predict severe slugging and to develop elimination techniques for offshore applications, tests have been carried out in a smallscale, 50-mm [2-in.] -diameter, 15-m [50-ft] -high flowline/riser system under near-atmospheric conditions. A long flowline has been simulated by addition of a large gas buffer to a shorter (30-m [ 100-ft]) flowline. In particular, gas injection at the riser base has been examined as a means of reducing severe slugging. The effects of gas and liquid flow rates, flowline inclination, and the presence of flowline undulations, riser-foot geometry, and liquid presence of flowline undulations, riser-foot geometry, and liquid viscosity have also been investigated. Liquid-viscosity effects are of interest with respect to two-phase flow with water cuts where high-viscosity oil/water emulsions can be expected. Further, a simple hydrodynamic model has been developed, and results obtained with it are compared with experimental data. This model has great similarity to that presented by Schmidt et al. Physics of Severe Slugging Severe-Slugging Group. For severe slugging to occur, the flowline gas flow must be completely inhibited during slug buildup (e.g., as a result of a partly declining flowline or the presence of flowline undulations), and the rate of hydrostatic pressure buildup in the riser resulting from the growth of the slug must exceed the rate of gaspressure buildup in the flowline. Under such conditions, the riser becomes filled with liquid before the gas pressure can drive the liquid slug out of the line. Assuming no mass transfer between the phases. PL »pg, and a vertical riser, and neglecting liquid fallback in the riser, we can express the ratio between these pressure buildup rates as (1) If IIss less than 1, then severe slugging could occur. As can be observed from Eq. 1, this severe-slugging group, IIss, depends mainly on the gas-to-liquid mass flow ratio, or gas/liquid ratio (GLR), and the effective flowline length, the latter being the product of actual flowline length and average flowline gas holdup, which represents the volume of the gas buffer. Given the GLR of a field and the flowline length, the severe-slugging group, IIss, is dependent on only the gas holdup. Additional system backpressure would decrease this gas holdup and could eliminate severe slugging, but as Schmidt et al. noted, a substantial increase in backpressure is needed. This would be accompanied by the risk of production losses. As an example, consider a 6-km [3.7-mile] flowline transporting a 90-std-m 3/M [500-scf/bbl] oil/gas mixture with an average flowline gas holdup of 50%. Here, IIss = 1/3 to 1/4, indicating that the occurrence of severe slugging is possible. Because the expected slug length is Ls = LRIIss, the slug lengths could be three to four times the riser length. Longer flowlines, which increase the gas buffer volume, and lower GLR'S, which reduce the pressure buildup rate in the flowline, increase the possibility of severe slugging. Throughout this paper, we use this field case as an example. Further relevant data include the following: pipe diameter is 150 mm [6 in.]. liquid viscosity is 10 mpa.s [10 cp], and riser-base pressure is 2.4 MPa [350 psia]. The field base case has a liquid pressure is 2.4 MPa [350 psia]. The field base case has a liquid flow of 950 m /d 16,000 B/D]. SPEPE