Identification and classification of solid–liquid flow patterns in deviated and horizontal annuli

Abstract

During horizontal well drilling, the interaction between drilling fluid and cuttings entering the annulus generates diverse flow patterns. These solid–liquid two-phase flow patterns must be accurately predicted to optimize the determination of hydraulic parameters and improve the efficiency of cuttings transport. Accordingly, this study identified flow patterns and conducted transition experiments under different inclination angles using a visualized wellbore annulus apparatus (120 mm outer diameter/73 mm inner diameter). Through direct visual observations, four primary flow patterns were systematically classified on the basis of the solid–liquid two-phase flow behaviors identified in the experiments: stable bed (SB), sand wave (SW), sand dune (SD), and bed load (BL) flows. The experimental data were then used to construct flow pattern maps with solid/liquid phases as axes, after which the transition boundaries between different flow patterns were established.

The morphological characteristics and transition mechanisms of SB, SW, SD, and BL flows were systematically analyzed to develop three predictive models of the fluid dynamics principles governing these flow patterns’ transitions: (1) A transition boundary model of SB and SW flows was established using Kelvin–Helmholtz stability, for which a stability analysis of solid–liquid two-phase flow in deviated and horizontal annuli was carried out. (2) A transition boundary model of SW and SD flows was constructed through an analysis of the geometric features of sand waves in the annuli, with the critical ratio of the average height of a cuttings bed to its height after erosion being 0.45. (3) A traditional critical velocity model was refined by adjusting the von Karman constant to account for the effect of solid volume concentration, yielding a boundary model for the transition of SW or SD flow into BL flow. All the models were experimentally validated. Finally, we integrated the models to develop a unified method for identifying and classifying the patterns typifying solid–liquid two-phase flow in deviated and horizontal annuli.