The search for oil and gas continues to progress towards increasingly hostile environments. Environments such as found in deepwater preclude the ability to efficiently separate the fluid phases prior to export. As a result, multiphase transportation has become commonplace with systems being designed using integrated flow assurance techniques. Additionally, pipelines that had been designed for single phase flow are now expected to cope with the transport of multiple phases. With the continuous variation in production, these pipelines tend to operate under both steady-state and transient conditions. The requirement to model these systems is, therefore, critical for adequate field development planning. This paper will present a tutorial focusing on the complex analysis of multiphase flow in pipelines with an emphasis on the tools currently available for modeling purposes. The tools selected for use during the tutorial include Pipesim and OLGA.
The analysis of multiphase flow phenomena in pipeline systems is usually classified along two levels of complexity. The first is that associated with steady state flow where there are no major changes transgressing the pipeline network. The second related to transient or dynamic flows where the flow behavior is changing on a regular and significant basis. These situations will be dealt with in turn.
Two distinct approaches are available to the petroleum engineer in accounting for the behavior of multiphase flow systems. The first is a global approach that relies on empiricism in developing simplified models that contain parameters which are evaluated from experimental data. The second is a continuum approach in which more complex physically-based models are used to describe the flow phenomena. Some authors of empirical correlations select an appropriate functional form for equation 1, and proceed to back calculate a holdup function that best fits experimental data for two-phase pressure drop. The Hagedorn and Brown, , correlation for flow in vertical pipes is based on such an approach. Other investigators such as Beggs and Brill, , and Dukler, , have correlated both variables (f m, θ l) as functions of gas and liquid flow rates, pipe geometry, fluid PVT and transport properties, among other variables. Empirical correlations have been successful in terms of:
Enabling models for particular flow conditions to be formulated quickly.
Being amenable to tuning to yield results of reasonable accuracy over well-defined ranges of operating conditions.
Being relatively easy to employ as design tools.
Table 1 lists some of the more successful and widely-used correlations with recommended areas of application. The limitations of empirical techniques stem from some or all of the following:
Individual equations tend not to apply with sufficient accuracy to the broad range of flow conditions usually encountered in practice.
The use of a number of different correlations to predict the hydraulic conditions of the dominant flow regimes in a pipeline system can result in numerical difficulties and/or discontinuous predictions.