The ability to accurately simulate the dynamic phenomenon of slugging, whether hydrodynamic, severe, or terrain-induced, has become a key objective in modeling dynamic multiphase flows in pipelines. Current widely accepted simulation tools can yield errors as high as 100% in the prediction and tracking of slugging in long pipelines. These can be attributed to several factors, including differences in mathematical formulation and numerical approximation methods. These in turn affect the prediction of local transport phenomena such as condensation, entrainment, deposition, slug development, and slug dispersion. The aim of this paper is to compare three methodologies currently available to model dynamic multiphase flow with a specific objective of appraising the accuracy of prediction of such phenomena as discussed above.
An adequate introduction into the analysis of dynamic multiphase flow was provided in [1] where the state of the art was presented in regard to the technology available for modeling such phenomena. As presented in [2], the complexity associated with the modeling of dynamic systems is by far greater than that for steady-state configurations. A return to fundamental fluid mechanics, [3], [4], and [5], has enabled the pipeline engineering community to make progress on the development of the ultimate multi-fluid formulation. However, much work remains to be done in the resolution of challenges related to the modeling of dynamic multiphase flow. Most of these challenges revolve around the ability to describe and model the physics related to the various flow regimes prevalent during the flow of multiphase fluids. In particular, the handling of flow regimes and their respective transitions, [6], provides for a specific dimension of complexity that demands an understanding of the underlying physics and an ability to numerically model it. The advent of dynamic multiphase analysis techniques has brought with it the ability to model hydrocarbon transportation pipeline systems, [7], [8]. The offshore oil and gas exploration environment has driven the development of novel production systems in harsh subsea environments with design engineering and system selection designs being aided by the available modeling technology. With the quest for continued lengthening of flowline and tie-backs that transport multiphase miztures, attention is now driven towards the ability to understand the physics of slugging in pipelines, and in particular, the formation, transport, and dissipation of slugs (or roll-waves) in pipeline systems, [9]. Much research has been conducted in this area, [10], [11], [12], with advanced tools now becoming commercially available. This paper will attempt to introduce some of this technology in a continuing attempt to enlighten the pipeline simulation engineer.
The complexity in the modeling of multiphase flow transportation derives primarily from the numerical characterization of the different flow patterns, Figure1, typically described as:
Stratified
Slug (Intermittent)
Bubble (Dispersed)
Annular
Furthermore, the ability to transition from one pattern to another in manner that is continuous and numerically stable while remaining faithful to the physics of the fluid behavior is a key requirement for robust modeling.
