DSD-INT 2018 Slurry modelling: development and application of a non-Newtonian Delft3D-FLOW version - van Maren
1 like446 views
The document discusses the development and application of a non-Newtonian version of Delft3D, targeting the challenges posed by non-Newtonian fluids in hydraulic engineering, particularly in managing oil sands fluid tailings. It highlights the differences between Newtonian and non-Newtonian fluids, the implications of viscosity variations, and solutions for improving consolidation rates in tailing basins. Additionally, it emphasizes the potential for broader applications of this model beyond the oil sands industry to areas such as fluid mud and sediment management.
![Results: non-segregating tailings over beach
11 december 2018
1:100 slope, dX = 10m
Release:
- Q = 0.1 m2/s
- 260 kg/m3 carrier fluid
- 450 kg/m3 sand
U [m/s]
c τ u μ ws
depth](https://image.slidesharecdn.com/dsd-int2018slurrymodelling-developmentandapplicationofanon-newtoniandelft3d-flowversion-vanmaren-181211125529/85/DSD-INT-2018-Slurry-modelling-development-and-application-of-a-non-Newtonian-Delft3D-FLOW-version-van-Maren-21-320.jpg)
![Results: non-segregating tailings over beach
11 december 2018
1:100 slope, dX = 10m
U [m/s]
▪ Shear stress
increases linearly from
surface to bed
▪ Yield stress is uniform
▪ Transition to plug flow
at intersection: du/dz
= 0
▪ Rapid decrease in
apparent viscosity
▪ Results in increase in
ws
▪ And therefore
decrease in C.
c τ u μ ws
depth](https://image.slidesharecdn.com/dsd-int2018slurrymodelling-developmentandapplicationofanon-newtoniandelft3d-flowversion-vanmaren-181211125529/85/DSD-INT-2018-Slurry-modelling-development-and-application-of-a-non-Newtonian-Delft3D-FLOW-version-van-Maren-22-320.jpg)
![Results: non-segregating tailings over beach
11 december 2018
1:100 slope, dX = 10m
U [m/s]
Conclusion:
No segregation of sand &
carrier fluid for this tailing
Next steps: more
complex configurations
c τ u μ ws
depth](https://image.slidesharecdn.com/dsd-int2018slurrymodelling-developmentandapplicationofanon-newtoniandelft3d-flowversion-vanmaren-181211125529/85/DSD-INT-2018-Slurry-modelling-development-and-application-of-a-non-Newtonian-Delft3D-FLOW-version-van-Maren-23-320.jpg)
DSD-INT 2018 Slurry modelling: development and application of a non-Newtonian Delft3D-FLOW version - van Maren
- 1. Slurry modelling: development and application of a non-Newtonian Delft3D-FLOW version 14 November 2018 D.S. van Maren, Luca Sittoni, A.M. Talmon, J.L.J. Hanssen, J.A.Th.M van Kester, J.C. Winterwerp, E. Parent, A. Mourits, Q. Ye, R.E. Uittenbogaard, H. van Es
- 2. What is a non-Newtonian fluid? - Delft Software Days: 138 presentations & courses - 99.3% of these were about Newtonian fluids
- 3. What is a non-Newtonian fluid? Newtonian fluid: a fluid in which the shear rate depends linearly on the shear stress (deformation scales linearly with force) because the fluid viscosity is constant 11 december 2018 viscosity stress Stress(Pa) Shear rate (/s)
- 4. What is a non-Newtonian fluid? Newtonian fluid: a fluid in which the shear rate depends linearly on the shear stress (deformation scales linearly with force) because the fluid viscosity is constant 11 december 2018 Fluid’sresistance Force Force Response
- 5. What is a non-Newtonian fluid? Non-Newtonian fluid: non-lineair relation between the shear rate and the shear stress, because the fluid viscosity is not constant (depending on the shear rate OR shear history) 11 december 2018 Viscosity Stress (Pa) Shear thinning: decrease in viscosity with shear Stress(Pa) Shear rate (/s)
- 6. What is a non-Newtonian fluid? Non-Newtonian fluid: non-lineair relation between the shear rate and the shear stress, because the fluid viscosity is not constant (depending on the shear rate OR shear history) 11 december 2018 Shear thickening: decrease in viscosity with shear Viscosity Stress (Pa) Stress(Pa) Shear rate (/s)
- 7. What is a non-Newtonian fluid? Non-Newtonian fluid: non-lineair relation between the shear rate and the shear stress, because the fluid viscosity is not constant (depending on the shear rate OR shear history) 11 december 2018 Bingham: linear relation with yield stress Viscosity Stress (Pa) Stress(Pa) Shear rate (/s)
- 8. Examples in hydraulic engineering slurries
- 9. Examples in hydraulic engineering Fluid mud
- 10. 11 december 2018 Delft3D slurry o Oil sands industry produce fluid tailings
- 11. 11 december 2018 Delft3D slurry o Oil sands industry produce fluid tailings o These tailings are discharged into basins which should become (in time) nature again. o But: these basins do not consolidate and remain fluid nature does not recover
- 12. Delft3D slurry o Oil sands industry produce fluid tailings o These tailings are discharged into basins which should become (in time) nature again. o But: these basins do not consolidate and remain fluid nature does not recover o Possible solutions: o Mixing of tailing fluid with sand increase consolidation rates o Capping of a tailing basin with sand o Deltares is developing a non-Newtonian version of Delft3D for the Canadian mining industry to quantify potential measures / mitigation efforts
- 13. Delft3D slurry: modified viscosity o Replace the vertical eddy viscosity in the 3D momentum equation by an apparent viscosity 11 december 2018
- 14. 11 december 2018 Delft3D slurry: modified viscosity o Replace the vertical eddy viscosity in the 3D momentum equation by an apparent viscosity o Compute this apparent viscosity from the shear stress and the shear rate a viscosity stress 2 2 u v z z
- 15. 11 december 2018 Delft3D slurry: modified viscosity o Replace the vertical eddy viscosity in the 3D momentum equation by an apparent viscosity o Compute this apparent viscosity from the shear stress and the shear rate o Compute the shear stress from material properties (mainly yield stress & viscosity) and the shear rate a 2 2 u v z z n y stress(Pa) shear rate (/s)
- 16. 11 december 2018 Delft3D slurry: modified viscosity o Replace the vertical eddy viscosity in the 3D momentum equation by an apparent viscosity o Compute this apparent viscosity from the shear stress and the shear rate o Compute the shear stress from material properties (mainly yield stress & viscosity) and the shear rate o Viscosity & yield stress are concentration dependent and material dependent, requiring 1. a choice of rheology model 2. laboratory experiments for input parameters n y
- 17. Delft3D slurry: modified settling velocity o At (very high sediment concentrations, the effective settling velocity is trapped by grain-grain & grain-fluid interactions (hindered settling) Clear water (high ws) Turbid water (low ws) Slurry (no ws)
- 18. 11 december 2018 Delft3D slurry: modified settling velocity o At (very high sediment concentrations, the effective settling velocity is trapped by grain-grain & grain-fluid interactions (hindered settling) o Mud particles do no settle (gelled bed), but sand particles may settle due to shear settling o Shear settling: sand settling velocity resulting from changes in apparent viscosity (in turn resulting from sheared flow) Talmon et al., 2018
- 19. Delft3D slurry: modified settling velocity o At (very high sediment concentrations, the effective settling velocity is trapped by grain-grain & grain-fluid interactions (hindered settling) o Mud particles do no settle (gelled bed), but sand particles may settle due to shear settling o Shear settling: sand settling velocity resulting from changes in apparent viscosity (in turn resulting from sheared flow) 2 ,0, ,18 s ww w s sand w a cf gd w , , ,0,sand _ sa_ max 1 1 wm sa s eff sand s fl cl saw w Concentration terms Apparent viscosity
- 21. Results: non-segregating tailings over beach 11 december 2018 1:100 slope, dX = 10m Release: - Q = 0.1 m2/s - 260 kg/m3 carrier fluid - 450 kg/m3 sand U [m/s] c τ u μ ws depth
- 22. Results: non-segregating tailings over beach 11 december 2018 1:100 slope, dX = 10m U [m/s] ▪ Shear stress increases linearly from surface to bed ▪ Yield stress is uniform ▪ Transition to plug flow at intersection: du/dz = 0 ▪ Rapid decrease in apparent viscosity ▪ Results in increase in ws ▪ And therefore decrease in C. c τ u μ ws depth
- 23. Results: non-segregating tailings over beach 11 december 2018 1:100 slope, dX = 10m U [m/s] Conclusion: No segregation of sand & carrier fluid for this tailing Next steps: more complex configurations c τ u μ ws depth
- 24. Results: segregating tailings over beach Distance along beach 1000 m Elevation10m Sandconcentration SFR SFR in bed Fines Capture ~ 14 hours simulation
- 25. Results: tailings pond 11 december 2018 Yield stress = 2 kPa Density fines = 1400 kg/m3 Density sand = 1520 kg/m3 Distance along pond 500 m Depth5m SandconcentrationSFR Yield stress = 200 kPa Density fines = 1750 kg/m3 Density sand = 1660 kg/m3 sand discharge Depth5m sand discharge Pond with soft tailings: inflowing sandy slurry will mix with tailings pond Pond with strong tailings: inflowing sandy slurry will cap the tailings pond
- 26. Conclusions - Developed a non-Newtonian Delft3D version, specifically optimised for oil sand tailings - But: many possibilities beyond the oil sands industry (fluid mud, turbidity currents, water injection dredging, reservoir failure) - We encourage sharing of beta releases of Delft3D slurry and are open for cooperation!