DEVELOPMENT AND VALIDATION OF COMPUTATIONAL MODELS FOR SAND EROSION AND CORROSION PREDICTION IN PIPES AND FITTINGS

The pnmary objective of this research is to develop computational models for
predicting sand erosion and C02 corrosion and their co-action (erosion-corrosion) in
pipelines and pipe components (elbows and tees). The motivation behind this
objective is to replace the sophisticated and time-consuming computational fluid
dynamics (CFD) software used for erosion simulation with an accessible, faster, and
cheaper predictive tool while maintaining the prediction accuracy, and to group the
erosion prediction model with C02 corrosion prediction model in one package to
serve as a comprehensive erosion-corrosion predictive tool. Two models have been
used to build the erosion prediction code. The Salama model, a simple empirical
model, was used to predict erosion rate in elbows and tees, assuming that sand
particles velocity is the same as fluid velocity. This assumption makes the model
more applicable to gas flow and high gas-liquid-ratio flow. An attempt has been made
to increase the accuracy of the Salama model using a set of experimental data from a
number of publications. As a result, three sand erosion models according to gasliquid-
ratio have been proposed ..
Direct Impingement Model (DIM) from University of Tulsa, a semi-empirical
sand erosion model, was selected for application in all kinds of fluids due to its
account of particle velocity instead of fluid velocity. In employing DIM model, a
numerical algorithm was used to solve the simplified equation of particles motion,
proposed by University of Tulsa, to track sand particles within a predetermined length
(so-called stagnation zone). The particle impingement velocity was calculated,
accordingly, and substituted in an empirical sand erosion equation to calculate erosion
rate. An ad hoc equation was used with the DIM model to model temperature
dependency of particle impingement velocity and erosion rate.