MAST [1], a multiphase flow simulator has been validated against a set of laboratory and field data collected from recent R&D activities carried out by TEA Sistemi. A new set of measurements of pressure drops and liquid hold-ups taken at TEA Sistemi Laboratory under conditions of stratified gas-liquid flow has been used to improve the closure relations adopted in MAST. To verify the overall behaviour of the code, a set of field data (mainly pressure and temperature drops) has been selected from the available database. Based on MAST peculiarities (detailed description of slug flow and thin stratified film liquid flow with relevant entrainment), long tieback pipelines, characterized by large diameters, have been considered for validation purposes. The standard version of MAST, with default settings, has been used, and sensitivity studies have been carried out with regard to the numerical discretization and the adopted closure laws. Some statistical error analyses have been performed to evaluate the overall code accuracy. The results obtained showed good agreement between predicted and measured data (with average errors below 20%), which demonstrates that MAST can be used as an alternative tool for advanced flow assurance studies.
Simulation of multi-phase flow in hydrocarbon transportation flow lines is typically carried out using transient one-dimensional flow simulators, among which OLGA [2] is the most widely used by the Oil & Gas industry. In OLGA, the closure laws, required within the framework of one-dimensional simulators, have been validated against the dataset gathered at the Sintef Multiphase flow laboratory [2]. For two-phase flows, six equations are globally solved: three mass conservations equations (for gas, liquid bulk and liquid droplets), two momentum equations (liquid bulk and a mixture equation for gas and liquid droplets) and one mixture energy equation. It is important to remark that the standard OLGA code can simulate the onset of liquid surges due for instance to terrain-induced slugging, but does not predict the flow structure of hydrodynamic slugging.
