Abstract
Severe slug flow (i.e., terrain-dominated slug flow) was studied in a simulated offshore pipeline riser-pipe system. Severe slug flow is characterized by extremely long liquid slugs generated at the base of the vertical riser. This phenomenon occurs at low gas and liquid flow rates and for negative pipeline inclinations. Slugging in some offshore platforms has required the use of operating procedures that drastically curtail production. Losses in flow capacity up to 50% have been reported. production. Losses in flow capacity up to 50% have been reported. A hydrodynamic model has been developed for severe slug flow. The model's predictions agree with experimental data. The model can be used to design predictions agree with experimental data. The model can be used to design new pipeline riser-pipe systems or to adjust the operation of existing systems to prevent the occurrence of severe slug flow. Also, a flow-regime map is presented for predicting the severe slug flow regime, where the boundaries are determined analytically. Finally, additional methods are proposed to prevent the flooding of separation facilities by riser-pipe proposed to prevent the flooding of separation facilities by riser-pipe generated slugs. This study is an extension of Ref. 1, in which severe slug flow was introduced and was only partially modeled.
Introduction
Two-phase flow in pipelines frequently involves the formation of liquid slugs. Processing of these slugs with separators can be extremely difficult if the size of the slugs becomes abnormally long. When a long liquid slug reaches a separator, it is possible for the liquid level in the separator to rise faster than the separator can purge the liquid, resulting in possible liquid carry-over into the gas stream. A technique often used for possible liquid carry-over into the gas stream. A technique often used for protecting separators from liquid slugs is to install an additional vessel protecting separators from liquid slugs is to install an additional vessel ahead of the separator, which usually is called a "slug catcher." The combined cost of the two smaller vessels is usually lower than the cost of a single large separator, which must be designed to process liquid slugs. However, the size of the slug catcher and/or separator must increase with increasing expected liquid slug sizes. The cost of installation of large separators and slug catchers, especially in the hostile environments found in Alaska, in swamps, or on offshore platforms, may be prohibitive. Therefore, it is desirable to have a technique that can predict and control both the occurrence and magnitude of liquid slugs so that separation facilities can be designed properly and their size decreased. Recently, studies have been performed that have increased dramatically the accuracy of both slug size and frequency predictions. Earlier studies, performed under laboratory conditions, indicated that slug lengths would performed under laboratory conditions, indicated that slug lengths would be no more than 100 ft [30.48 m]. However, recent studies performed on full-scale pipelines have indicated that slug lengths of more than 2,000 ft [609.6 m] are possible. In addition, it has been discovered that slug flow can be generated by several different mechanisms, each producing liquid slugs with different physical properties. Schmidt et al., in studying slug flow in a simulated offshore pipeline riser-pipe system, found two distinct slug flow patterns: normal (e.g., hydrodynamic) and severe (e.g., terrain-dominated) slug flow. Severe slug flow is characterized by the generation of liquid slugs at the base of the riser pipe, with the remainder of the pipeline in stratified flow. Normal slug flow is characterized by many liquid slugs being generated along the length of the pipeline and occurs at higher gas and liquid flow rates. The liquid slugs generated during severe slug flow were found to range in length from one to several riser-pipe heights, which, at the time this study was performed, generally exceeded the slug lengths associated with normal slug flow. Therefore, riser-pipe-generated slug flow was designated "severe" slug flow, in comparison to "normal" pipeline-generated slug flow. Severe slug flow was found to depend on the geometry of the pipeline riser-pipe system. The pipeline must be in stratified flow, as well as inclined negatively for the liquid slug to be generated at the base of the riser. Also, because of the mechanism by which severe slugs are generated, it was found that the degree of slug aeration for severe slugs was much lower than that associated with normal slug flow. Also, the study showed that the phenomena of severe and normal slug flow are mutually exclusive because normal pipeline slugs and bubbles will flow through the riser pipe nearly unchanged, excluding the possibility of a riser-generated slug. Finally, a hydrodynamic model was developed for severe slug flow. The model was formulated on basic physical principles and was limited to a description of how the liquid slug is generated at the base of the riser pipe. No attempt was made to model the full behavior of the severe slug pipe. No attempt was made to model the full behavior of the severe slug flow cycle. Bendiksen et al. developed a dynamic one-dimensional two-phase flow model for the Norwegian state oil company, Statoil. They gave the mass and momentum conservation equations for each phase, and solved them numerically by using finite difference techniques.
SPEJ
P. 27
