Evaluation of Multiphase Flow Models to Predict Pressure Gradient in Vertical Pipes with Highly Viscous Liquids

Abstract

The prediction of pressure drop in multiphase flow for risers is of particular interest for the oil industry and also a critical variable for the right design of surface facilities in offshore fields. Empirical steady state correlations, mechanistic models and dynamic models are available to calculate the multiphase flow pressure drop, holdup and phases distribution.The main purpose of this paper is to evaluate the accuracy of several steady state pressure drop prediction models with two phase flow laboratory data conformed by 108 point using air as gas phase and liquids with viscosity up to 310cP. The models considered in this study, for predicting pressure drop are model, using PIPESIM simulator. The evaluation was based on the comparison between the predicted and the measured pressure drops, demonstrating the performance of each model for highly viscous liquids. Statistical parameters as average absolute percent error and standard deviation have been calculated to find the most acceptable one. The statistical analysis showed that among the evaluated Hagedorn and Brown model has the best performance to predict pressure drop in risers with highly viscous liquids.

Multiple regression-based prediction correlations for enhanced sensor design of magnetic induction tomography systems

“…Multiple regression provides the model that best fits the behaviour of the data. Similarly to the correlation approach taken for data processing from experiments in multiphase flow [23], curve fitting coefficients are used to describe the performance of the MIT systems via correlation models. This paper extends the simulation studies performed in [18] and move from a descriptive analysis to an inferential exploration in order to develop, for the first time, algebraic models to help researchers predict the overall performance of MIT systems for a given coil setup.…”