Compositional Multiphase Hydrodynamic Modeling of Gas/Gas-Condensate Dispersed Flow in Gas Pipelines

Abstract

Summary. A nonisothermal ID steady-state compositional two-phase hydrodynamic model describes the formation and flow dynamics of gas condensate in horizontal natural gas pipelines. The two major constituents of the model, hydrodynamics and phase behavior, are coupled through the phase/generation/disappearance-related terms in the continuity and momentum equations. The model is demonstrably capable of predicting the amount, quantity, and distribution of condensate in the pipeline, in addition to the other commonly sought engineering design variables. Parametric studies show that the model is capable of predicting the phenomena associated with gas condensation in pipelines. Introduction For natural gas to be available at the burner, it must be found, produced, gathered, processed, and transported. Gas gathering, a Standard feature of natural gas production, is also a salient component of the exploitational development of any petroleum field. After having been processed, natural gas is transported to the consumers through networks of pipelines that crisscross several developed nations of North America and Europe. Optimal design and operation of such gathering and distribution systems is essential if the ultimate cost to the consumer is to be kept at a reasonable level, especially when the large capital outlay necessary constitutes a sizable share of the total development costs. Quite often, the collected natural gas must be transported over a substantial distance in pipelines of various sizes. These pipelines span between hundreds of feet and hundreds of miles and cross undulating terrain with varying temperature conditions. When such pipelines operate in a single-phase-gas flow regime, design calculations are readily available. Liquid condensation in pipelines, however, commonly occurs because of the multicomponent nature of the transmitted natural gas and its associated phase behavior, as well as the inevitable temperature and pressure variations that occur along the pipeline. Condensate subjects the gas pipeline to two-phase transportation. An additional complicating factor is that neither the point along the pipeline where the condensate is formed nor the quantity formed is) known a priori. Up to 15 vol% of liquid has been reported for transmission lines. This significant liquid volume flowing simultaneously with gas contributes substantially to the overall pressure loss in pipelines. In such situations, the single- phase equations obviously are rendered inapplicable for design and prediction. Furthermore, because the amount of condensation is dictated by both the hydrodynamics and the phase behavior of the fluids, using phase-behavior and hydrodynamic models is critical. The coupling of these two models produces a compositional model. Twophase flow in pipes generally causes a significantly higher pressure drop than the equivalent single-phase flow would, even when the total mass flow rates are the same. This is a result of the interphase interactions between the two coexisting phases, rather than intraphase and wall/fluid interactions. The hydrodynamic behavior of two-phase systems is radically different from that of single-phase systems. The two-phase systems not only exhibit higher pressure loss, but also are subjected to interphase forces and mass-generation-induced forces, all of which are completely absent in single-phase systems. These forces must be adequately accounted for and properly represented in the compositional hydrodynamic model. Other system variables needed to define a two-phase system include the liquid holdup, phase velocity, phase thermophysical properties, and the relative spatial distribution of the phases known as flow regime. These characteristics are mainly responsible for the immense complexity of this problem both conceptually and in terms of susceptibility to mathematical description. Therefore, pipeline designers resort to empirical equations and models rather than the fundamental hydrodynamic approach. The predominantly used empirical and semiempirical approaches invariably do not address the fundamental physics of these systems, but rather develop some lumped empirical groups. In fact, some of the earlier empirical correlations simply lump all these unknown parameters into some multiplication factor. This factor is computed by some empirical correlations depending on the flow regime, which again is predicted by empirical correlations. The approach is justified because of the complexity of the problem, but it usually has limited applications dictated by the data base on which the correlations are based. Quite often, these data are based on a laboratory-scale model, thus making the use of the resulting correlations for scale-up design rather precarious. The inadequacy of these models, especially for gas condensate pipelines, has been discussed recently. Any specific applicability of such purely empirical models to this rather complex system would clearly be over a very limited range of conditions and should be regarded as an exception rather-than a rule. Only models that examine the problem with the fundamental fluid mechanics approach have a chance of success, in terms not only of generality but also of scale-up capability. Brief Survey of Current Approaches Techniques most prevalently used in the design of pipelines carrying two- phase gas/condensate can be classified into two categories: the singlephase-safety-factor (SPSF) and the steady-empirical-two-phase (SETP) approaches. Within each of these groups, however, are subcategories. The categorization is based on the general characteristics of the models in these groups. And because voluminous literature is available, we do not provide an in depth review (see Refs. 2 and 3 for critical analyses). In the early years of wet-gas pipeline design, SPSF approaches were commonly used. These approaches basically rely solely on the well- established design equations for single-phase flow in pipes. Two-phase flow is treated as a simple extension by use of a multiplier, a sort of safety factor to account for the higher pressure drop generally encountered in two-phase flow. This heuristic approach was widely used and generally resulted in inaccurate pipeline design. When the amount of the condensed liquid is negligibly small, the use of SPSF approach could at best prevent underdesign, but more often than not, the quantity of the condensed liquid is significant enough for the SPSF approach to be extremely erroneous. The inadequacy of the SPSF approaches spurred researchers to evolve better design and predictive models for two-phase flow systems. This effort led to the development of several empirical and semiempirical models to describe these rather complex flows. SPEPE P. 85^