unexpected flow in culverts

[ckatsoulas]

I have a 2D model with 2 sets of culverts, one inflow boundary and one outflow boundary. When I run the model and preview results, I find that on the first time step, the flow from the inflow boundary has not yet reached the culverts however there is flow coming out of the culverts. I don’t know where this flow is coming from. Can anyone help. The inflow boundary is the red boxes on the right hand side. The first time step is shown.

[Anonymous]

your culvert must be connected via GIS coordinates to an area far upstream. Update the GIS coordinates to match the true invert and outlet location

[ckatsoulas]

Thanks for reply Luis, the cords were entered by using the control key so they are correct. The long red lines you see are just breaklines for the channel. I had setup some initial conditions for tailwater that I later deleted but this is still happening. I’m thinking there could be a bug with the initial condition not being fully removed?

[Anonymous]

You have a warm up period and ramp up % going on also? Thats the reason for sure. I wouldnt say a bug but when i do run models like that the results in that time frame are not relevant to my peaking location of the real storm and i throw out the results

[ckatsoulas]

No, no warm up initiated. I even checked the culvert property box in case there was anything that would introduce flows but nothing. It really has me stumped. I may have to give the model to HEC for checking.

[Anonymous]

Please show the hydrograph and your initial conditions.

I am left to believe there is a geographic transformation that is occuring connecting the culvert to a wetted cell. I know you said it was taken using the ctrl key but still…..

[ckatsoulas]

Luis you can download the model from the link below to get a better idea. I have sent it to HEC also but no reply yet

https://www.dropbox.com/sh/84rttr0zsku7pk6/AAAC1M8RVlvlKcRzxTY6h_CZa?dl=0

thanks for looking into this

[Lonnie A]

You have a stage hydrograph set as your downstream boundary condition of elev 5.6 and the “use initial stage” box set. This fills the 2D area to that elevation at the start of the simulation (but this isn’t the way your initial problem statement was presented so I assume it was something you’ve tried later).

Also you have advanced time step set up with the maximum courant being less than the minimum value.

I switched the outfall to a normal depth boundary and fixed the time step issue and I too see flow at Culvert 2 the first time step as you were seeing. I tried a handful of things to try and isolate the issue with no luck. I would be interested in what HEC comes back with.

[Anonymous]

He SENT IT TO ME LONNIE.

But yea as he said the stage is your issue. That stage hydrograph sets a WSEL of 5.6 m as you input across your entire terrain from the BC line and that is why you see flow in the culvert prior to the inflow getting there.

STAGE Hydrographs do not work in this software, i made a long post about this like 6 months ago…

Other help info: have more variance in your time input for your hydrograph. Triangle hydrographs will never give you a great answer.

If you change that DS BC to normal depth the model would operate as expected. If your problem is well there is an actual stage of 5.6m at this location, it must be presented and represented in another way

[Lonnie A]

Sorry Luis 🙂
The stage hydrograph is the reason for the flooding out of the 2D area at the start of the model as he sent it. But I don’t think that was the issue that started his post. If you run the model with a time step of say 1s and map at 10s with a boundary set to normal depth (run end time set to 0002) you’ll find flow is being produced at culvert 2 right off the start and well before the upstream flow has had a chance to make it down to the crossing. Only thing I tried that got rid of the flow production at the culvert was removing the upstream flow…even setting it to just 1 cms produced flow creation at the culvert.

[ckatsoulas]

Guys thank you so much for taking the time to look at this in depth, I really appreciate it. The reason for the 5.6m tailwater is that I am trying to reproduce the water level (5.7m) at the subject site from a flood study prepared by another major consultant for the greater flood area. With normal depth the flood level drops to RL 5.4 at the subject site. I had to use a triangular hydrograph as I was not given a hydrograph just a flow. I guess I will have to wait and see what HEC say about this but at least it will help them fix any bugs for the rest of us.

[ckatsoulas]

Lonnie can you please explain in more detail what you mean about the “maximum courant being less than the minimum value” and what you did to fix it.

[ckatsoulas]

Luis, you say you have said before stage hydrographs don’t work, can you elaborate on what is wrong with this in HECRAS. I don’t see an issue in this run with the stage hydrograph.

[Anonymous]

The bug was removed by removing your hydrographs removing your time series data entirely and inputting actual hydrographs.

Let me know what portions or screen shots you want of my test on the model before i delete it.

Oh and dont use dates like 1899 or run full dynamic if you dont have to, that was also removed. No water magically appeared in my culverts.

I dont know what the perfect fix was because i made all these changes at once, but im sure 1899 had to do with it as well as your stage and full dynamic

[Anonymous]

Stage hydrographs or stage DS BC have been the means of an outfall in HEC-RAS for sometime. In 2-Dimensional modeling it should not only represent inflow but the means of an outflow.
I made a post about it awhile back with not much feedback, but me and a former colleague tore hair out as to why it wouldnt work.

We did many many tests trying to simulate free outfall (though i already have another way it just becomes tedious) via a stage elevation. Conceptually you should be able to set a WSEL for which once head builds above that elevation, volume should leave the system; this is not the case whatsoever. Though stage can be used to allow flow into a system, out is another story and the documentation on it, should be able to execute what i have tested but it simply cant.

[Author]

Posts