Mechanistic - based models for slug flow in vertical pipes

Slug flow is one of the most common flow regimes encountered in petroleum, chemical and nuclear industries. This flow regime is characterized by pseudo-periodic occurrence of Taylor bubbles and liquid slugs. The approaches for slug flow modeling include empirical methods, mechanistic methods and numerical methods. Numerical methods involve large amount of computation and therefore not suitable for the application in petroleum industry. In contrast to empirical methods that are limited in the range of application, mechanistic models are more applicable because they capture the essential mechanism of the slug flow regime. However, the mechanistic models require some correlations to obtain closure, which bring unexpected error and need to be improved before getting results matching with experimental data. This project established a computational model to simulate the flowing field around Taylor bubble. Boundary fitted structured grid system is built for the simulation model and the turbulence is accounted for by the low Reynolds-number k- model. The previous turbulence model is modified to predict smooth transition from laminar to turbulent flow. To obtain the configuration of the Taylor bubble, circular geometry is assumed at the top of the bubble and Bernoulli equation is applied to estimate the shape of the tail part. The fifth order of polynomial function is used to ensure the convergence of the bubble configuration which has uniform pressure along the interface. Based on the results from the numerical model and previous experimental studies, the currently used mechanistic model is improved to get more accurate prediction for the slug flow in vertical pipes. The improved model is validated by previously reported experimental data. The geometric parameters of slug flow analyzed in this project can provide a basis for studying flow regime transition from slug flow to other flow regimes. The significance of this work is to help us understand more details about slug flow in vertical pipes. The knowledge gained from this work can be used to reestablish existing empirical correlations that are limited to particular conditions but necessary to the mechanistic models. The improvement of mechanistic models can benefit the design of tubing and surface facilities and the determination of artificial-lift methods.