Stowa 2017 benchmark (Water Module): Difference between revisions

Revision as of 12:45, 9 April 2019

The Stowa 2017 benchmark was set up in order to find out which tool currently available to the Dutch Water management industry performs reliable in which circumstances. The goal of the benchmark was to provide insight to waterboards and municipalities which water simulation tool would suit their needs best. The benchmark consist of several topics:

Rain water discharge for several types of terrain
Underground water infiltration
Flow direction and speed
2D-1D interaction between surface water and water objects such as weirs and culverts.

For the complete benchmark, see benchmark 2017. Please note that the referenced document contains results made in 2017, with a model that was not yet using the 2D Saint Venant equations.
The Water Module has been much improved since 2017 and is now based on the finite volume method described by Kurganov et al.

The results presented below have been generated on 04-09-2019.

Test case 1

Test case 2

Test case 3

Setup

The lay-out of the terrain has a slanted V-shape. In the middle is a ditch, which is horizontal in the x-direction but slanted downwards in the y-direction. The terrain around the ditch is slanted downwards in the x-direction with a value of 0.05. In the y-direction is is slanted similar to the ditch, with a value of 0.02. For the exact measurements see figure 5.9 above. A graphical impression of the height map in this testcase is given below. Again first the original height map and next the height map adjusted for ensuring a constant slope per grid cell x- and y-direction.

Initial
After slope adjustments

Results

The hydrograph and volume.

hydrograph
volumes

The amount of surface water (in meters), color coded per grid cell, presented for several timeframes.

0h30m
1h00m
1h30m
1h45m
2h00m
2h30m
3h00m

Test case 4

This testcase contains a ponding problem. This testcase is subdivided into 3 sub-testcases in order to investigate several aspects of this ponder problem. Each is described below. See image 5.11 for the layout of the problem. Each of the surfaces in this test case have a Manning roughness value of 0.05. A weir is situated at the bottom of the sloped terrain, causing the creation of a pond during the simulation.

Setup

For each test case the base setup is the same. A graphical impression of the height map in this testcase is given below. Again first the original height map and next the height map adjusted for ensuring a constant slope per grid cell x- and y-direction.

Initial
After slope adjustments

Test case 4A

In this testcase, the amount of rainfall is similar to testcase 3. However, it continues to rain until the end of the simulation. The surface and underground are impenetrable. And the weir is considered to be 'perfect'.

Results

The hydrograph and volume.

hydrograph
volumes

The amount of surface water (in meters), color coded per grid cell, presented for several timeframes.

0h30m
1h00m
1h30m
1h45m
2h00m
2h30m
3h00m

Test case 4B

In this testcase the amount of rainfall is 21.6 mm per hour and lasts for 90 minutes. The underground has an infiltration capacity of 10.8 mm per hour. The simulation runs until the water level has lowered to about +/- 2cm below the crest of the weir. The effects of the groundwater table can be ignored.

Results

The hydrograph and volume.

hydrograph
volumes

The amount of surface water (in meters), color coded per grid cell, presented for several timeframes.

0h30m
1h00m
1h30m
1h45m
2h00m
2h30m
3h00m

Test case 4C

In this testcase, the amount of rainfall and the infiltration capacity are equal to testcase 4B. The porosity of the soil is now set to 0.25. The impenetrable layer is strictly horizontal and 10 meters below the lowest surface height. The groundwatertable at t=0 is horizontal, 50 cm below the height of the weir. The initial surface water level is equal to the initial groundwatertable. The hydraulic condictivity is 5x10^-5 for water infiltration from the surface to the underground. The simulation time is equal to testcase 4B.