Hostname: page-component-7b7567c8c-tnkh4 Total loading time: 0 Render date: 2026-11-16T10:26:24.907Z Has data issue: false hasContentIssue false

Primary cementing of horizontal wells. Displacement flows in eccentric horizontal annuli. Part 2. Computations

Published online by Cambridge University Press:  19 March 2021

P. Sarmadi
Affiliation:
Department of Mathematics, University of British Columbia, 1984 Mathematics Road, Vancouver, BC V6T 1Z2, Canada
A. Renteria
Affiliation:
Department of Mechanical Engineering, University of British Columbia, 2054-6250 Applied Science Lane, Vancouver, BC V6T 1Z4, Canada
I.A. Frigaard*
Affiliation:
Department of Mathematics, University of British Columbia, 1984 Mathematics Road, Vancouver, BC V6T 1Z2, CanadaDepartment of Mechanical Engineering, University of British Columbia, 2054-6250 Applied Science Lane, Vancouver, BC V6T 1Z4, Canada
*
†Email address for correspondence: frigaard@math.ubc.ca

Abstract

We present results of three-dimensional computational simulations of horizontal eccentric annular displacement flows as studied experimentally by Renteria & Frigaard (J. Fluid Mech., vol. 905, 2020, A7). We show the simulations are able to effectively capture detailed dispersive features of the flows, not resolved in simpler models. Top side flows, slumping flows and narrow side residual layers are predicted reliably and we are able to understand the main mechanisms that drive the flow into one of these classifications. Both experiments and simulations agree in their classification of steady and unsteady displacements. However, when the dispersion is properly resolved the concept of a steady state displacement becomes questionable. We study the axial spreading of the displacement front via one-dimensional simplifications, both advective and diffusive. While these do not fully describe the flows, they give some estimates that are useful for practical application.

Information

Type
JFM Papers
Copyright
© The Author(s), 2021. Published by Cambridge University Press
Figure 0

Table 1.Dimensionless parameter ranges of our experimental study.

Figure 1

Figure 1.Schematic of the displacement flow domain.

Figure 2

Table 2.Apparatus dimensions.

Figure 3

Figure 2.The mesh size study of a displacement flow in an eccentric annulus with $(e,Re,b, \hat {\mu }_2/\hat {\mu }_1)=(0.73,341.9,60.31,1.45)$. Parameters taken from one experiment. (a) The displacement flow: blue shaded region represents the iso-surface $C=0.5$ at $\hat {t}=15$ s. (b) Four different mesh sizes of, mesh 1: $10\times 80 \times 200$, mesh 2: $12\times 96 \times 240$, mesh 3: $15\times 120 \times 300$, mesh 4: $20\times 160 \times 400$ in $\hat {r}, \theta , \hat {x}$ directions. (c) Concentration variation across wide annular gap (shown by a red dashed line in (a)) at $\hat {x}=1.2$ m for different mesh sizes. (d) Axial velocity profile across wide annular gap at $\hat {x}=1.2$ m for different mesh sizes.

Figure 4

Figure 3.Displacement flow near the interface for $(e,Re,b, \hat {\mu }_2/\hat {\mu }_1)=(0.46,701.49,0.16,0.943)$ at $\hat {t}=5$ s. (a) Iso-surface for $C=0.5$ is shown by the blue shaded region. Normalized velocity vectors $(\hat {U},\hat {V})/\hat {W}_0$ and streamlines coloured by vorticity magnitude: (b) at $\hat {x}=2$ m and (c) $\hat {x}=2.5$ m; the colour bars represent normalized velocity and vorticity magnitude. The displacing and displaced fluids are presented by black and white colours. (d) Concentrations in a vertical plane along the narrow side. Vectors show the normalized velocity $\hat {W}/\hat {W}_0$ and colour bar presents normalized velocity magnitude.

Figure 5

Figure 4.Classification of displacements using fully developed flow behaviour. Buoyancy number $b$, increases from top to bottom in the series of figures. Each parameter set is classified as top side or slumping, for both experiment and simulation (see legend). Thus, a $+$ over a diamond, or a $\times$ over a circle indicate predictive agreement between experiment and simulation.

Figure 6

Figure 5.Example of slumping flow with $(e,Re,b, \hat {\mu }_2/\hat {\mu }_1)=(0.46,52.80,444.18,11.41)$ EXP 103 at ${\hat {t}=29}$ s: (a) depth-averaged front view, and (b) height-averaged bottom view. Dashed black line shows $\bar {C}=0.5$. (c,d) Are pictures from the experiment showing the front and bottom views, respectively. At $\hat {t}=200$ s: (e)  depth-averaged front view, and ( f) height-averaged top view. Dashed black line shows $\bar {C}=0.5$. (g,h)  Are pictures from the experiment showing the front and bottom views, respectively at $\hat {t}=169$ s.

Figure 7

Figure 6.For $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(0.46,52.80,444.18,11.41)$, EXP 103. (a) Height-averaged concentration evolution in time ($\bar {C}_y(\hat {x},\hat {t})$) and (b) dimensionless fully developed front velocities ($\hat {W}_f/\hat {W}_0$) for different height-averaged concentrations ($\bar {C}_y$).

Figure 8

Figure 7.Azimuthal flow near the interface for $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(0.46,52.80,444.18,11.41)$, EXP 103 at $\hat {t}=125$ s. (a) Iso-surface for $C=0.5$, streamlines at: (b) $\hat {x}=2.5$ m, (c) $\hat {x}=2.75$ m, (d) $\hat {x}=3$ m and (e) $\hat {x}=3.25$ m. Streamlines are coloured by vorticity ($\omega _x$). Blue line shows the iso-line of $C=0.5$ and grey scale represents $\hat {W}/\hat {W}_0-1$.

Figure 9

Figure 8.Example of top side flow with $(e,Re,b, \hat {\mu }_2/\hat {\mu }_1)=(-0.07,73.51,-38.17,0.84)$, EXP 208 at $\hat {t}=24$ s: (a) depth-averaged front view, and (b) height-averaged bottom view. Dashed black line shows $\bar {C}=0.5$. (c,d) Are pictures from the experiment showing front and bottom views, respectively. At $\hat {t}=103$ s: (e) depth-averaged front view, and ( f) height-averaged top view. Dashed black line shows $\bar {C}=0.5$. (g,h) Are pictures from the experiment showing front and bottom views, respectively.

Figure 10

Figure 9.For $(e,Re,b, \hat {\mu }_2/\hat {\mu }_1)=(-0.07,73.51,-38.17,0.84)$, EXP 208. (a) Height-averaged concentration evolution in time ($\bar {C}_y(\hat {x},\hat {t})$) and (b) dimensionless fully developed front velocities ($\hat {W}_f/\hat {W}_0$) for different height-averaged concentrations ($\bar {C}_y$).

Figure 11

Figure 10.Azimuthal flow near the interface for $(e,Re,b, \hat {\mu }_2/\hat {\mu }_1)=(-0.07,73.51,-38.17,0.84)$, EXP 208 at $\hat {t}=100$ s. (a) Iso-surface for $C=0.5$, streamlines at: (b) $\hat {x}=3$ m, (c) $\hat {x}=3.5$ m, (d) $\hat {x}=4$ m and (e) $\hat {x}=4.5$ m. Streamlines are coloured by vorticity ($\omega _x$). Blue line shows the iso-line of $C=0.5$ and grey scale represents $\hat {W}/\hat {W}_0-1$.

Figure 12

Figure 11.For $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(0.73,251.75,57.78,1.29)$, EXP 48. (a) Depth-averaged concentration front view at $\hat {t}=7.5$ s and (b) height-averaged concentration bottom view at $\hat {t}=7.5$ s. (c) Depth-averaged concentration front view at $\hat {t}=35$ s and (d) height-averaged concentration bottom view at $\hat {t}=35$ s.

Figure 13

Figure 12.Azimuthal flow near the interface for $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(0.73,251.75,57.78,1.29)$, EXP 48. Streamlines at: (a) $\hat {t}=7.5~s,\hat {x}=0.9$ m, (b) $\hat {t}=7.5~s,\hat {x}=1.1$ m, (c) $\hat {t}=35~s,\hat {x}=4.3$ m and (d) $\hat {t}=35~s,\hat {x}=4.5$ m. Streamlines are coloured by vorticity ($\omega _x$). Blue line shows the iso-line of $C=0.5$ and grey scale represents $\hat {W}/\hat {W}_0-1$.

Figure 14

Figure 13.For $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(0.46,805.69,5.37,1.08)$, EXP 165. At $t=5$ s: (a) depth-averaged concentration front view, (b) height-averaged concentration bottom view, (c) height-averaged concentration bottom view of annulus first section and (d) shows the bottom view picture from the experiment. At $t=7$ s: (e) depth-averaged concentration front view, ( f) height-averaged concentration bottom view and (g) height-averaged concentration bottom view of annulus second section. At $t=15$ s: (i) depth-averaged concentration front view, (j) height-averaged concentration bottom view and (k) shows the bottom view picture from the experiment. Black dashed line shows $\bar {C}_y=0.5$.

Figure 15

Figure 14.Steady/unsteady classifications for slumping displacements according to the threshold. Comparison between experimental data (filled symbols) and the corresponding 3-D simulations (larger white symbols). The colour bar represents the viscosity ratio ($\hat {\mu }_2/\hat {\mu }_1$) of the experiment.

Figure 16

Figure 15.For $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(-0.07,304.55,69.08,1.50)$, EXP 309. (a) Height-averaged concentration evolution in time ($\bar {C}_y(\hat {x},\hat {t})$) and (b) dimensionless fully developed front velocities ($\hat {W}_f/\hat {W}_0$) for different height-averaged concentrations ($\bar {C}_y$).

Figure 17

Figure 16.Azimuthal flow near the interface for $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(-0.07,304.55,69.08,1.50)$, EXP 309 at $\hat {t}=25$ s. (a) Iso-surface $C=0.5$, streamlines at: (b) $\hat {x}=2.5$ m, (c) $\hat {x}=3.5$ m, (d) $\hat {x}=4$ m and (e) $\hat {x}=4.5$ m. Streamlines are coloured by vorticity ($\omega _x$). Blue line shows the iso-line $C=0.5$ and grey scale represents $\hat {W}/\hat {W}_0-1$.

Figure 18

Figure 17.Azimuthal flow near the interface for $(e,Re,b,\hat {\mu }_2/\hat {\mu }_1)=(0.46,299.12,71.74,1.48)$, EXP 313 at $\hat {t}=25$ s. (a) Iso-surface $C=0.5$, streamlines at: (b) $\hat {x}=2.25$ m, (c) $\hat {x}=2.75$ m, (d) $\hat {x}=3.25$ m and (e) $\hat {x}=3.5$ m. Streamlines are coloured by ($\omega _x$). Blue line shows the iso-line $C=0.5$ and grey scale represents $\hat {W}/\hat {W}_0-1$.

Figure 19

Figure 18.Snapshots of displacement flow for different eccentricities for $(Re,b, \hat {\mu }_2/\hat {\mu }_1)=(148.99,0.95,0.87)$ at $\hat {t}=8$ s. Colour map of equally spaced slices represents concentration of the displacing fluid and the blue shaded region shows the iso-surface $C = 0.5$. Dimensionless axial velocity ($\hat {W}/\hat {W}_0$) is shown on slices and the colour map presents its magnitude.

Figure 20

Figure 19.Variation with $e$ of the concentration ($\bar {C}$) averaged across the gap and along the annulus on top and bottom sides of the annulus, for $(Re,b, \hat {\mu }_2/\hat {\mu }_1)=(148.99,0.95,0.87)$ at $\hat {t}=8$ s.

Figure 21

Figure 20.Snapshots of displacement flow for different viscosity ratios for $(e,Re,b)=(0.46,40.14,518.51)$ at $\hat {t}=140$ s; (a) $\hat {\mu }_2/\hat {\mu }_1=0.5$, (b) $\hat {\mu }_2/\hat {\mu }_1=1$, (c) $\hat {\mu }_2/\hat {\mu }_1=1.5$, (d) $\hat {\mu }_2/\hat {\mu }_1=2.5$ and (e) $\hat {\mu }_2/\hat {\mu }_1=5$. Colour map of equally spaced slices represents concentration of the displacing fluid and the blue shaded region shows the iso-surface $C=0.5$. Dimensionless axial velocity ($\hat {W}/\hat {W}_0$) is shown on slices and the colormap presents its magnitude.

Figure 22

Figure 21.Variation with $\hat {\mu }_2/\hat {\mu }_1$ of the concentration ($\bar {C}$) averaged across the gap and along the annulus on the top and bottom of the annulus, for $(e,Re,b)=(0.46,40.14,518.51)$ at $\hat {t}=140$ s.

Figure 23

Figure 22.Average concentration ($\bar {C}$) variation along the well for different viscosity ratios for $(e,Re,b)=(0.46,40.14,518.51)$ at (a) $\hat {t}=140$ s and (b) $\hat {t}=200$ s. Viscosity ratios are $\hat {\mu }_2/\hat {\mu }_1=0.5,\ 1.5,\ 2.5,\ 5$.

Figure 24

Figure 23.Spatio-temporal map of averaged concentration $\bar {C}_y(\hat {x},\hat {t})$ for $(e,Re,b)=(0.46,40.14,518.51)$; (a) $\hat {\mu }_2/\hat {\mu }_1=0.5$, (b) $\hat {\mu }_2/\hat {\mu }_1=1.5$, (c) $\hat {\mu }_2/\hat {\mu }_1=2.5$ and (d) $\hat {\mu }_2/\hat {\mu }_1=5$. Dashed black line and dashed grey line show $\bar {C}_y(\hat {x},\hat {t})=0.3$ and $0.7$, respectively.

Figure 25

Figure 24.Dimensionless fully developed front velocities ($\hat {W}_f/\hat {W}_0$) for different height-averaged concentrations ($\bar {C}_y$) for $(e,Re,b)=(0.46,40.14,518.51)$; (a) $\hat {\mu }_2/\hat {\mu }_1=0.5$, (b) $\hat {\mu }_2/\hat {\mu }_1=1.5$, (c) $\hat {\mu }_2/\hat {\mu }_1=2.5$ and (d) $\hat {\mu }_2/\hat {\mu }_1=5$.

Figure 26

Figure 25.Parameters $(e,Re,b)=(0.46,40.14,518.51)$: (a) $\hat {\mu }_2/\hat {\mu }_1 = 1.50$, advective concentration $\bar {C}_{adv,y}$; (b) $\hat {\mu }_2/\hat {\mu }_1 = 1.50$, spatio-temporal map of $\bar {C}_{y}-\bar {C}_{adv,y}$; (c,d) as (a,b) but with $\hat {\mu }_2/\hat {\mu }_1 = 5$. The black dashed line shows $\bar {C}_y=0.5$.

Figure 27

Figure 26.The result of plotting $\bar {C}_y(\hat {x},\hat {t})$ against $(\hat {x}-\hat {W}_0\hat {t})/\sqrt {\hat {t}}$ for $(e,Re,b)=(0.46,40.14,518.51)$; (a) $\hat {\mu }_2/\hat {\mu }_1=0.5$, (b) $\hat {\mu }_2/\hat {\mu }_1=1.5$, (c) $\hat {\mu }_2/\hat {\mu }_1=2.5$ and (d) $\hat {\mu }_2/\hat {\mu }_1=5$. The solid red line shows a fit using the complementary error function, with diffusivity: (a) $\hat {D}=0.0112$ m$^2$ s$^{-1}$, (b) $\hat {D}=0.0044$ m$^2$ s$^{-1}$, (c) $\hat {D}=0.0022$ m$^2$ s$^{-1}$ and (d) $\hat {D}=0.0008$ m$^2$ s$^{-1}$.

Figure 28

Figure 27.The result of plotting $\hat {D}/{\rm \pi} \hat {r}_o\hat {W}_0$ against $\Delta \hat {W}_f/\hat {W}_0$ for selected slumping classified simulations: $e = -0.07$ circle; $e = 0.46$ square; $e = 0.73$ triangle. The colour bar represents the viscosity ratio ($\hat {\mu }_2/\hat {\mu }_1$) of the simulation.

Figure 29

Figure 28.(a) Value of $\Delta \hat {W}_f/\hat {W}_0$ for experiments and 3-D simulations and (b) dimensionless diffusion value $\hat {D}/{\rm \pi} \hat {r}_o\hat {W}_0$ for selected slumping classified experiments. $e = -0.07$ in red circle, $e = 0.46$ in green square and $e = 0.73$ in blue upside-down triangle.