Warnings, Calibration questions

[Mogel]

Hello! I am new to this forum, using Hec-Ras for my thesis. I am trying to model (and calibrate) an old channel, 9 km long, in the beginning for the first 2 km very narrow (2-3 m), then quite wide (15-19 m). So the discharge is small at the beginning (0.035) and later around 0.1. There are 12 culverts, 1 bridge and a weir after 8 km. I’ve tried almost everything, changes here and there, different BC, Mannings n, coefficents, steady flow options…. But for the first 3 km water level is way to low, the 3 km before the weir are almost not flowing and it shows only backwater (though I measured a much higher velocity). There is no flowing curve, even though I (and someone else) measured one. (see the Profile Plot with observed WS points).

I always get these warning:
Warning: The conveyance ratio (upstream conveyance divided by downstream conveyance) is less than 0.7 or greater than 1.4. This may indicate the need for additional cross sections.

I put XS' 1 m and 5-10 m before and after each culvert and interpolated every 40 m, more than enough I would suppose?

Warning: During subcritical analysis, with the exit loss set =1.0, the projected WSEL in culvert has a lower energy than the downstream energy. Most likely, the downstream cross section blocks part of the culvert or the ineffective area is set too far in. Instead of projecting the WSEL, the program did an energy balance to get the WSEL inside the culvert at the downstream end.

Most culvert openings are lower than the water bed (official surveying). Should I change that?

Also a total newbie question: Do I have to worry about the warning: geometry is newer than output?

I appreciate every help and hint since I am pretty frustrated already and can’t think of other possibilities to calibrate the model. I also tried higher discharges but it doesn’t change that much.

Thanks!

[Mogel]

Could anyone help me, please!
At least it would be helpful to explain the warnings shortly (right now I just ignore them but I don't know if it affects the results seriously).
Thanks a lot

[jarvus lunt]

There is not enough information to say why your model is not matching the calibration.

Regarding the warning messages:


Warning: The conveyance ratio (upstream conveyance divided by downstream conveyance) is less than 0.7 or greater than 1.4. This may indicate the need for additional cross sections.

>I put XS' 1 m and 5-10 m before and after each culvert and interpolated every 40 m, more than enough I would suppose?


40m sounds like it is probably close enough. This warning message comes up a lot. It should probably be a "note" instead of a "warning". It just means that the flow has changed significantly (faster velocity, etc.).


Warning: During subcritical analysis, with the exit loss set =1.0, the projected WSEL in culvert has a lower energy than the downstream energy. Most likely, the downstream cross section blocks part of the culvert or the ineffective area is set too far in. Instead of projecting the WSEL, the program did an energy balance to get the WSEL inside the culvert at the downstream end.

Most culvert openings are lower than the water bed (official surveying). Should I change that?

The cuvlert outlet is below the ground level of the downstream cross section? That would explain the message. If the culvert is not physically blocked right at the exit, I would adjust the downstream cross section to match conditions right at the outlet of the culvert. If the culvert is actually partially blocked, that is trickier.


>Also a total newbie question: Do I have to worry about the warning: geometry is newer than output?

This means the program thinks that you have made changes to the geometry since the last time you computed results. It is saying, "hey, your results are outdated". However, I have gotten this warning when I should not have. Just make sure you compute the results after any changes to the geometry.

[Mogel]

Thank you very much!

I guess there are several problems.
I set Normal WS (which I measured) as BC but I dont have the actual discharge at the end of the reach (estimated). And thats there HEC-RAS starts computing...
Second: The discharge is very low and the energy slope (hydraulic gradient) is very low, too. And many culverts..

Does anyone have experience with modelling a channel like that?