HEC-RAS Encroachment Analysis Elevations Need Help 1

[davidpro]

Hey All,

I am modeling a stream in HEC-RAS that crosses under a large 4 lane highway with dual 8'x10' box culverts, and then further upstream crosses a dirt road with an old masonry arch bridge. The stream has been modeled by FEMA before, but since then the stream has been realigned. The re-aligned stream was studied by another engineer in the area but a LOMC was never finalized with FEMA.

The problem is this: when I run the encroachment analysis on the stream (started with method 4, but have tried all methods), the downstream 1/2 of the model runs normally( b/t 0.5-1' increases), but on the upstream side of the box culverts, the encroachment elevations are between 4.5-5 feet higher than 100yr, regardless of the target WS values I use. Even when I manually make the encroachment wider by specifying stations, the encroachment elevations stay the same! It's like the model isn't calculating something correctly.

I have some experience with HEC-RAS but far from an expert. The downstream cross sections are relatively narrow (~200 ft), and the sections upstream of the culverts are much wider (~1000 ft), could this have something to do with it?

Thanks for any and all help!

David

 

[sknight701]

There are several reasons that your model may not be working, and it would be useful for you to post your entire model so we can see it.

But I will try a few guesses to start: Is all culvert geometry entered properly? Do the results yield vertical "walls" at the outer boundaries of the cross sections? Is the proper range of cross sections entered in the encroachment data? Does the perspective plot of the results yield a smooth transition, or is it erratic?

Generally, I have found that it is best to use Method 1 to do encroachment analyses.

 

[davidpro]

I think I have attached the full model, a zip file with the geometry, flow, plan, and project files.

The plan file uses encroachment method 1 (imported from method 4), but I get the same results regardless of encroachment method.

Culvert geometry appears to make sense, and the 100 yr flood is captured within all of the cross sections. Encroachment data is entered for every cross section.

Perspective plot looks confusing, I think because of the way the cross sections are laid out on the schematic. Which actually leads to a different question, the cross sections on the schematic plot appear to cross each other, but they are not geo-referenced, so is this an issue or not? Cross sections on my CAD drawing do not cross each other, but I guess due to the large width of some cross sections they cross on the HEC-RAS schematic.

 

[sknight701]

If your cross sections, channel, etc., are not georeferenced, then the perspective plot can really get weird. This in no way affects the quality of the model itself, so don't worry about it. Just don't use the perspective plot in any reports.

I'll take a look at the files and get back to you later. With any luck I'll spot something.

 

[sknight701]

I just opened those files you sent, but they do not provide the information that I need. I suggest that you do the following. Open up the plan file in RAS. Go to File/Generate Report and generate the report. Print out the report, scan it as a .pdf file, and upload the .pdf to this website. The RAS report will give me all the geometry, settings, etc. so I can evaluate your model. Also, upload a .pdf of your site plan showing RAS cross sections, etc., so I can visualize things. You should also remove all references to your client for confidentiality purposes before uploading the .pdf files.

 

[davidpro]

Thanks for all the help, attached are the site plan and report, I included some output info and error messages as well.

I'm glad to hear about the schematic and perspective plots, it made sense that it shouldn't matter but it still had me nervous.

 

[amb2002]

One thing in the report that stands out to me: both culverts have the following note: "The culvert flow is blocked (either by a flap gate or the depth filled option."

Could that blocked flow be what's causing your observed increase?

Depth Blocked for both culverts is 0, which would seem to indicate you're using a flap gate? I've never utilized that functionality in HEC-RAS, so I'm not sure where you even set that.

 

[sknight701]

I am going to look at these during lunch today, and possibly take home printouts over the weekend. I'll get back to you ASAP

 

[sknight701]

I concur with amb2002. Your 100-year profile shows flow thru both culverts, but your floodway profiles show zero flow thru both culverts. That is probably your problem. Be sure you are not blocking the culvert with your encroachment data (essentially burying the culverts).

There may also be issues with the ineffective flow areas, or the adjacent cross sectional geometry may be blocking the culverts, but you should have spotted that by reviewing the cross section graphical output. Also, there is a sudden change in the Manning's n values just downstream of the downstream culvert, but I doubt that would cause such a large jump in water surface elevation.

 

[sknight701]

I checked the report over the weekend, but could not find anything else. Something seems to be blocking your culverts. Hope this helped.

 

[davidpro]

That did it! I had one downstream cross-section whose bottom elevation was blocking a portion of the culvert. I also adjusted the encroachment and cross section stationing to make sure it didn't block what the culvert stations were at, and immediately got more reasonable results. I now have encroachment WSE of close to +1 foot, so that must have been the problem.

Thanks to both of you for the input.

On a different topic, I haven't used ineffective flow areas before, I've read the manual on what they are, but are they necessary at every culvert or bridge crossing or will it depend on the geometry and topo? If they are necessary are there any rules of thumb as to where they should go in the cross-sections adjacent to the bridges?

Thanks again sknight and amb! My first time using this forum and you guys were great.

 

[sknight701]

I noticed that ineffective areas were not part of your model, and was about to inquire when I read your latest posting.

Oh yes, ineffective areas are almost always necessary and there are some rules as to where to place them. First, set up your bounding cross sections at the structure using the procedure on pages 6-25 thru 6-29 of the users manual. Set the contraction and expansion coefficients for these cross sections to 0.3 and 0.5 respectively to account for the fact that these are no longer gradual transitions. Your model has them set at 0.1 and 0.3 in the cross sections immediately adjacent to the culverts, which is not correct.

Make note of the expansion ratio chosen from page 6-27 when setting the cross section locations. Then run the model without any ineffective areas at the bridge or culvert. Check the profile plot to be sure the profiles look reasonable. If something funky happens, like having the 100-year WSE at a lower elevation than the 50-year WSE at any cross section, then you have a problem with geometry. Fix it before proceeding.

For the upstream cross sections, set the ineffective flow boundaries outside the upstream abutments using a 1:1 ratio. For example, if your upstream cross section is 10 feet from the bridge, set the ineffective boundaries at stations 10 feet outside of the abutments. Of course, this may be modified based on topography. With regard to the elevation of the ineffective area, they should be set at the overtopping elevation of the roadway for the appropriate side of the bridge.

For the downstream cross sections, set the ineffective flow boundaries outside the downstream abutments using the expansion ratio. Again, topography may require modification of this rule. For the downstream ineffective elevations, set them at an elevation approximately midway between the downstream bridge low chord and the overtopping elevation of the roadway.

The rest is an iterative technique. If things look OK in the profile plot, then determine when the bridge is overtopped, such as between the 25-year and 50-year storm elevations for example. You then reset all your ineffective flow elevations to between the 25-year and 50-year WSE's at each cross section, for this example, and re-run the model. Keep doing this until things seem to stabilize.

What this all does is account for how the water behaves as it tries to move downstream and is blocked by the roadway embankment.

Other ineffective flow areas located further away from bridges and culverts should be set based on the topography. A common example of this would be a ponded area to the side of the main channel which stores water but where water does not flow downstream.

With regard to your current project, your encroachment stationing will not be valid unless you account for the ineffective flow areas at your culverts. I recommend you add the ineffective flow areas to your "existing conditions" model, and then do the encroachments with the ineffective areas in place. This will give you much more accurate results for your encroachment stationing.

Good luck.