Jump to content
TUFLOW Forum
mostlyatnight

Representation of Individual Buildings in 2D

Recommended Posts

I was interested to know what generally accepted methods are currently used to represent individual buildings in a 2D domain.

I have seen various SFRA models using a Manning's n value of 1.0 for buildings and remember hearing discussions at the last TUFLOW meet in Bristol about setting the ZPoints of the building to the finished floor level and using a similarly high roughness coefficient.

Does anyone else have any different experiences?

Share this post


Link to post
Share on other sites

Your method sounds good, although we have refined it because of the impact on the results with some models.

We use two methods depending on whether we bring the flow into the model at ground level or via the direct rainfall method.

If the flow comes into the model at ground level (via a series of hydrographs) we use a polygon layer representing the building outlines and use the Read MI Zpts [ADD] command. Each polygon has an arbitrary value of say 3 (3 metres will be added to zpts within each polygon), and we ensure that the polygons are narrower than the buildings. This attempts to create a gap between buildings if one exists. Generally we use this method for small scale models with a 2m grid. For gaps between buildings that have a highly sensitive impact on results, we close the building gap in 2D and use a 1D channel to represent the gap. This method allows for fine tuning of the gap width, and is independent of the alignment of the grid to the buildings. Take care with the location of the sx lines to ensure that they are away from the raised zpts of the buildings.

This method has the advantage that the buildings appear as obstructions on the result maps, and the flood level at the building can be determined from the water surface level at the building edges. Obtaining floor level data is therefore not on the critical path for modelling. You can collect floor level data just for selected buildings where you need a freeboard check.

For models using the direct rainfall method (Read MI RF command) we generally keep all buildings at ground level and select a roughness based on whether the building is within a floodway or not. If all buildings are roughened the rainfall applied to the buildings is held back, and can produce hydrograph attenuation downstream. Depending on the catchment this attenuation can have a significant effect on the peak flow. This approach requires an initial run to define the floodways, and we only roughen small buildings to minimise the attenuation effects. For large buildings we keep the building smooth, at gound level, and use a levee at the upstream edges (using the Read MI Z line [ADD] command). This allows for the rainfall within the building to runoff unimpeded, but flowpaths are deflected around the building.

We have decided that raising buildings in a direct rainfall model is not a good option. The mass balance errors increase to an acceptable level.

This is a viewpoint for models in residential Sydney. I hope it is useful for your modelling.

Share this post


Link to post
Share on other sites

We once modelled a floodplain with fairly flat terrain using SA_RF option. The used grid size was 2m square . The model went smoothly under existing condition. A pretty large proposed building partly encroached downstream of the overland flow path. We used "Read MI Zpts " polygon to represent building pad 5m above ground. The model then went unstable. We tried to fix the problem with two approaches:

1/ Raising Zpts adjacent to building (using 2d Zpts check layer) to create a smooth transition as recommended in Tuflow manual. This approach produced several warnings only. Still the graphical results were a bit odd with pretty deep ponding along one side of the building (no major depression in ground)

2/ Reduce timestep to 1/6 of cell size ( 0.3s to 0.4s ). This approach fixed the problem with only a single warning and no ponding along the building

We only experienced this in one model so it may not work for other models

Share this post


Link to post
Share on other sites

I was interested to know what generally accepted methods are currently used to represent individual buildings in a 2D domain.

I have seen various SFRA models using a Manning's n value of 1.0 for buildings and remember hearing discussions at the last TUFLOW meet in Bristol about setting the ZPoints of the building to the finished floor level and using a similarly high roughness coefficient.

Does anyone else have any different experiences?

Has anyone used the porous buildings method? It is possible to reduce the flow through a cell to simulate the presence of a building, but I'm not sure how it's done. Does anyone know?

Share this post


Link to post
Share on other sites

Hi

You can use the .tgc "Read MI FC Shape" command to assign a "porosity" to building polygons. Refer to Section 4.7.2 and Table 4.17 of the 2008 TUFLOW manual. I've also attached 2d_fcsh_porous_house.zip which is a simple example of a building with porosity and also additional energy losses (form loss) assigned to it.

For a paper on this topic please click here.

Cheers

Bill

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...