Comparative Study of Different Erosion Model Predictions for Single-Phase and Multiphase Flow Conditions

Abstract

Abstract There are various equations and models for assessing erosion for both single phase and multiphase flow scenarios. Consequently, this study was performed to shed light on the issue of which models provide better predictions. For the single phase situation (air-sand flow), Computational Fluid Dynamics (CFD) erosion modeling in a direct impact geometry was conducted, and the performance of four different erosion equations was scrutinized. On the other hand, for the multiphase flow scenario, churn/annular multiphase flow regimes, that can be highly erosive when sand is present within the piping system, were selected to examine the accuracy of six different widely-used multiphase flow erosion models. For both cases, the modeling results were compared to available experimental data. For the first part of the study, ANSYS FLUENT 16.1 was used to perform CFD-based erosion modeling including flow modeling, particle tracking, and erosion calculation. For the other part, several codes were employed to investigate six empirical/mechanistic erosion models including: 1D churn model, 2D mixture model, multiphase model, annular model, DNV GL model, and Salama model. The University of Tulsa SPPS software and the DNV GL Erosion Tool were used for this part. Furthermore, the Salama model was coded. It was revealed that under air-sand flow conditions, Arabnejad et al. (2015a) mechanistic model provided the best results (15% error compared to experimental data). For the multiphase flow part of the study, it was found that regardless of particle size, the multiphase erosion model underpredicts all erosion data, and the 1D churn, Salama and 2D mixture erosion models significantly overpredicted the experimental data. Moreover, the annular and DNV GL models provided the best predictions among all six models. A literature survey reveals that there is a limited number of comparative studies for erosion prediction models. This study, therefore, should be of interest to a broad readership including but not limited to erosion, corrosion, multiphase flow, and flow assurance.