CFD-DEM coupling and machine learning for predicting cuttings bed erosion dynamics in riserless pipelines

Abstract

Owing to the narrow diameter of the riserless lifting pipeline, the large size and irregular shape of the particles being transported, and the significant concentration of cuttings present, the formation of a cuttings bed within the pipeline is highly probable, which seriously affects the operational safety of the lifting pump. Understanding the cuttings erosion mechanism of lifting pipelines has become an important research topic. To investigate the influence of various parameters on the erosion mechanism of pipeline cuttings bed, the erosion process was simulated using the computational fluid dynamics–discrete element method. The effects of the initial cuttings bed thickness, drilling fluid inlet velocity, pipe inclination, and drilling fluid viscosity on the erosion rate of the cuttings bed were analyzed. The simulation results showed that the flow rate of the drilling fluid had a significant impact on the erosion rate of the cuttings bed and that the rheology of the drilling fluid changed the erosion form of the cuttings bed. A relationship model of the cuttings erosion rate was established using machine learning, and the calculation results of the model were in good agreement with literature data. These findings provide a theoretical foundation for the hydraulic design of riserless drilling–lifting pipelines and hole-cleaning processes.