Treating a Stream as a "Pond"

[TheEnginerd]

For years, i have used Hec-Ras to produce flood elevations for streams including those that have culverts. I simply model the culvert in Hec-Ras.
I have just recently come across a project where a very renowned civil engineer received an approval from our state's DEP (NJ) by treating the stream and contours upstream of a culvert as a "Pond" and performed a pond routing.
He basically calculated the drainage area to the culvert and produced a peak flow (as one would have to do for a Hec-Ras analysis), and then modeled the pond using HydroCAD or similar to come up with a peak elevation.
The only thing i see wrong with this method is that it will not tell you a DEPTH of water at a point along the stream at say 1000 feet back from the culvert... but it will tell you a conservative number at close proximity to the culvert which is all that matters for this particular project.
The time of a Hec-Ras stream analysis would be say 80 hours.
The time to do this pond routing would be say 8 hours..... so there is a huge time savings.
Any thoughts? has anyone ever done this?

 

[beej67]

I've done it both ways. It really depends on how detailed an answer you need. If you're just trying to prove the ponding at the culvert isn't adversely impacted by upstream land development, then modeling it as a pond in both the predevelopment and postdevelopment conditions should be plenty good enough for your analysis, and the peak stage elevations at the culvert itself should be good enough to set your adjacent FFEs. As you say, they should be conservative.

If you're actually doing flood mapping, then obviously that analysis isn't enough. But then again, you're being paid a lot more for drawing a flood map than you are for checking a culvert near your land development project.



Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[Twinkie]

How far up the stream do you cut the countours off and create the "pond" or are you chasing this all the way upstream?

 

[cvg]

a typical pond routing would attenuate the flow a little bit. a more conservative approach would be to just use the peak discharge as the design flow and then use culvertmaster or HY8 to calculate the ponding level at the culvert. really not sure why you would go to any trouble with a stage - storage - discharge pond routing unless this was a very small culvert / very high embankment (such as a dam)

 

[beej67]

When I've done streams as ponds, they're typically in fairly hilly watersheds, so it's no real problem to cut the contours off where they close naturally. I suspect that modeling a stream as a pond in very flat watersheds might produce more error.

Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[LincolnPE]

The only thing i see wrong with this method is that it will not tell you a DEPTH of water at a point along the stream at say 1000 feet back from the culvert...

Depth = your peak WSEL - Ground Elevation at point of interest

What am I missing?

 

[beej67]

1000 ft back from the culvert, you need to at least do a direct step method calculation (or something more robust like HEC-RAS's standard step method) to compute the equations of gradually varied flow. The pond approximation presumes no significant flow velocity upstream of the culvert.

Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[LincolnPE]

I thought the premise of this exercise was to ignore the use of HEC-RAS and simply use the headwater elevation computed from the culvert analysis.

We don't know his stream gradient and headwater elevation, so how do we know his headwater elevation does not reach back 1000' under this culvert analysis only premise?

 

[cvg]

who cares if the headwater goes back 1,00 feet. according to the OP, he doesn't care about what happens 1,000 feet away. if he did care, than this simplified culvert analysis only approach would not be appropriate and RAS would definitely be the way to go.

 

[LincolnPE]

..If you're just trying to prove the ponding at the culvert isn't adversely impacted by upstream land development, then modeling it as a pond in both the predevelopment and postdevelopment conditions should be plenty good enough for your analysis..

We don't know this because the OP hasn't given any stream gradient or headwater information. For example, how far is upstream land developement?

What does the topo look like? Is it more channelized or does it look like an aerial of Lake Powell, with coves that can inundate at the headwater elevation?

 

[cvg]

the OP didn't mention anything about upstream development, so that is really not at issue.

 

[LincolnPE]

Obviously both the experienced engineer and the State DEP saw merit in this analysis approach.

HY8 is approved by FEMA for floodplain modeling.

OP: What was the upstream length of ponding at the headwater elevation for the approved DEP study?