This document discusses comparisons between hydrodynamic models using different grids on test cases and a case study of the Afferden-Deest side channel on the Waal river. The test cases showed large differences in results due to bed and friction settings, which were resolved by using WAQUA settings in Flexible Mesh. For the Afferden-Deest case study, initial results found higher water levels in Flexible Mesh compared to WAQUA, with more accurate representation of flow in the side channel. Grid refinement studies will assess the impact of grid resolution on model convergence.

1. www.hkv.nl
Case side-channel Afferden-Deest
Impact of grid refinement
Deltares Software Dagen
Universiteit Twente: Erik ten Hagen, Suzanne
Hulscher, Jord Warmink
HKV lijn in water: Freek Huthoff, Joana Vieira da
Silva, Andries Paarlberg
Deltares: Aukje Spruyt, Arthur van Dam,
Herman Kernkamp

2. www.hkv.nl
Contents
Part 1
• Test models
– Differences WAQUA - Delft3D - Flexible Mesh
Part 2
• Case Afferden-Deest
– Comparison Flexible Mesh - WAQUA
– Impact grid refinement

3. www.hkv.nl
test models
Simple rectangular model
– Comparison with analytical estimation

4. www.hkv.nl
test models
Simple rectangular model
– Comparison with analytical estimation
Test model Waal
– Differences WAQUA - Delft3D - Flexible Mesh in
complex model

5. www.hkv.nl
Rectangular test model
• 2000x200 meter
• Horizontal bed (h=5.0 m) with sill (h = 5.3 m)
• Constant discharge upstream (1515 m3/s)
• Constant waterlevel downstream (11.70 m)
• Uniform roughness (Colebrook-White, k = 0.2 m)

6. www.hkv.nl
Rectangular test model
Analytical estimation:

7. www.hkv.nl
Rectangular test model
Chézy roughness using Colebrook-White:
Used in Delft3D
and Flexible Mesh
Used in WAQUA

8. www.hkv.nl
Rectangular test model
Chézy roughness using Colebrook-White:
Used in Delft3D
and Flexible Mesh
Used in WAQUA
Larger friction in Delft3D
and Flexible Mesh!

9. www.hkv.nl
Rectangular test model
Results (Default settings)

10. www.hkv.nl
Rectangular test model
Colebrook-White formula of WAQUA in Flexible Mesh

11. www.hkv.nl
test model Waal
Differences WAQUA–Delft3D–Flexible Mesh for a
more complex model
– Schematization of 40 km Waal (Nijmegen – Tiel)
– Constant upstream discharge (16000 m3/s at Lobith)
– Constant downstream waterlevel (10.38 m +NAP)
– Friction
• Uniform Colebrook-White roughness (k = 0.2 m)
• Trachytopes (spatially variable and depth dependent
roughness)

12. www.hkv.nl
test model Waal (Colebrook-White)
Results (default settings)
Maximum water level difference FM - WAQUA = 21.5 cm

13. www.hkv.nl
Conveyance2D
WAQUA
Flexible Mesh

14. www.hkv.nl
Weirs and dry points
Weirs
Dry points

15. www.hkv.nl
test model Waal (Colebrook-White)
Water level difference (Flexible Mesh – WAQUA)
Conveyance2D Colebrook-
White
Overlaten Droge
punten
Default settings FM FM X X
Overlaten FM FM X
Droge punten FM FM X
Bodeminstelling WAQUA FM X X
Colebrook-White FM WAQUA X X
Bodeminstelling +
Colebrook-White
WAQUA WAQUA X X

16. www.hkv.nl
test model Waal (Trachytopes)
Results (default settings)
Maximum water level difference FM - WAQUA = 37.6 cm

17. www.hkv.nl
test model Waal (Trachytopes)
Waterlevel difference Flexible Mesh-WAQUA
Conveyance2D Colebrook-
White
Overlaten Droge
punten
Default settings FM FM X X
Bodeminstelling +
Colebrook-White
WAQUA WAQUA X X
Overlaten WAQUA WAQUA X
Droge punten WAQUA WAQUA X

18. www.hkv.nl
Conclusion test models
• Large differences from bed and friction settings
(Conveyance2D and Colebrook-White formula)
• Differences from modelling energy losses by
weirs
• Use WAQUA setting for bed and Colebrook-
White formula to compare Flexible Mesh with
WAQUA

19. www.hkv.nl
Case Afferden-Deest
• Side-channel at Afferden and Deest (River Waal)
Bron: Rijkswaterstaat

20. www.hkv.nl
Case Afferden-Deest in WAQUA

21. www.hkv.nl
Research plan – comparison
Compare Flexible Mesh results with WAQUA
• Waalmodel without side-channel (reference)
– Run Flexible Mesh with WAQUA-settings for design high
water wave (16000 m3/s at Lobith) and 1995 flood wave
– Compare Flexible Mesh with calibrated WAQUA model and
1995 measurements
• Waalmodel with side-channel
– Run Flexible Mesh with WAQUA-settings for design high
water wave (16000 m3/s at Lobith) and 1995 flood wave
– Compare Flexible Mesh with WAQUA model

22. www.hkv.nl
Research plan – study area

23. www.hkv.nl
First results (comparison)
Waalmodel without side-channel (Pannerdense Kop)

24. www.hkv.nl
First results (comparison)
Waalmodel with side-channel (Pannerdense Kop)

25. www.hkv.nl
First results (comparison)
Waal without side
channel
Waal with side
channel
Pannerdense Kop 13.8 cm 17.6 cm
Nijmegen 9.1 cm 7.5 cm
Tiel 2.9 cm -0.8 cm
Water level difference Flexible Mesh - WAQUA

26. www.hkv.nl
First results (comparison)
Flow velocity (m/s) in side channel
WAQUA
Flexible Mesh

27. www.hkv.nl
Research Plan – Grid refinement
Assess impact grid refinement
• Waalmodel without side-channel (reference)
– Refine grid Waal-channel at Afferden-Deest
– Determine when convergence is reached
• Waalmodel with side-channel
– Refine grid side-channel Afferden-Deest
– Determine when convergence is reached

28. www.hkv.nl
Grid refinement

29. www.hkv.nl
Grid refinement

30. www.hkv.nl
Grid refinement
Example allignment
Bron: Herman Kernkamp, 2014

31. www.hkv.nl
Summary
• Differences from bed setting and Colebrook-
White formula  WAQUA settings used in
Flexible Mesh
• Higher water levels in Flexible Mesh
• Flexible Mesh solution for representation
complex geometry
More accurate results for side-channel in
Flexible Mesh??