Review and suggestion needed about my HEC-RAS model 1

[philoshophy]

Hi dear all
greetings.
I have attached my HEC-RAS model. The reasult seem to be quite unrealistic. I could not figure out how to handle the problem. If some one have solution please give a favor. The model is preparing for bridge design in Hardin Creek within Napa Watershed. Actually I am dealing with 100 year flow of 1134 cfs.

I will be waiting the clue from you guys.
'
Thanks

 

[RWF7437]

philosophy,

You are asking people on this forum to provide free peer review of your work. It is presumed you are being paid to do this work. You also have a great deal more information than can be conveyed in a HEC-RAS model. Without such information it is doubtful anyone would be willing to undertake this work and liability.

West Consultants in San Diego, and many other California firms, would probably be willing to check your work, for compensation.

good luck

 

[philoshophy]

SORRY RWF7437 IF YOU THINK LIKE THAT
I AM JUST TRYING TO GET SOME SHARING KNOWLEDGES. I AM JUST IN SMALL CONFUSION TO HANDLE.
IF YOU THIK SO, IT' S OK. I APPOLOGIZE IT.

THANKS FOR SUGGESTION

 

[philoshophy]

ANYWAY I WNAT TO KNOW WHY THE SIMULATED PROFILE LOOKS LIKE SHOWN IN FILE RAS.PDF. I THINK THIS WILL BE SUFFICIENT FOR ME. AND REALLY BBE APPRECIABLE.
THANKS

 

[RWF7437]

From just a quick glance it appears you have modeled your Creek BACKWARDS. Channel distances MUST increase from downstream to up stream. If the WSL and HGL are coincident, then velocity is ZERO and therefor flow is ZERO.

Using a flow value with 4 significant figures, such as 1134 cfs is fooling yourself. These flows are seldom closer than plus or minus 30%, usually +/- 50% of you're lucky. Where did you get your design flows ?

 

[philoshophy]

Thanks RWF7437

But I could not understand your concept regarding 4 figures. Actually I got from HEC-HMS with SCS method. Any comment or suggestion for me.

 

[RWF7437]

http://www.asce.org/conted/onsitetraining/course.cfm?layout=detail&c=4

 

[MrNYPE]

Hi philosophy- I think what RWF7437 is trying to say is that calculating a flow of 1134 and assuming that is "accurate" is incorrect, as neither the calculation methods or the field measurement methods are that accurate.

I don't think that entering that exact SCS value into HEC-RAS is wrong, but I do think there is a better way of using that calculated flow. In my company,we apply a safety factor, depending on how accurate we believe the topo and other site data to be. For a project we did that involved a 300+ acre watershed, and we were relying on USGS quads for topo, we used a FS of 2.

Take everything you read on these forums with a grain of salt. I don't know how far along you are in your career, but there is always another engineer to tell you how to do something their way, instead of trying to understand your way of doing it and possibly advancing the profession. He posted the same response, regarding inaccuracy of flow calculations, to one of my questions. This inaccuracy is nothing new, and has been a well known fact at least since I left college (my graduation year starts with a 19).

 

[Crystalct1]

None of you answered the question.

To the person who is quibbling over the significant figures - I say that it does not compromise accuracy to carry these digits throughout your calculations - as long as when you present you findings in report format, you have rounded the values to the reasonable amount, otherwise he's right, you do look like you're fooling someone.

I looked at the model. A few things stand out.

When viewing the geometric data, your cross sections cross one another, making it appear impossible to have a positive reach length between one section and the one downstream of it. Also, this makes your bridge deck look like it is getting skinnier as you go across.
That being said, HEC-RAS doesn't "see" this picture in the calculations and as long as you do have positive reach lengths reflecting the distance downstream to the next section on the left overbank as well as the right, then you'd be fine. But it looks like you didn't reflect this properly:
If the stream curves as it appears to in your picture, then your LOB and ROB reach lengths should not be equal to the channel. What you have modeled is a very straight stream with very simple and common n-values. This is not usually the case and therefore will give you the wrong answers. When stream curves to the right, the LOB reach length should be the greatest, then the channel, then the right.

Next: Boundary conditions. For subcritical flow, you need at minimum one boundary condition downstream. You have used critical depth as both conditions. This is usually because the user doesn't know what else to do and if you use critical depth, you need nothing else. Problem is that the calculations are extremely dependent on the downstream conditions. Only if you have thousands of feet of stream between the downstream boundary conditions and your point of interest, can you put any confidence in your output.

You need a FEMA FIS study or something from you local environmental agency that tells you the elevation of the 100-year storm at your downstream-most point. Without this, you can estimate the elevation using normal depth. You would use the channel slope, which equals the energy slope for uniform flow. Put this in as the downstream boundary conditions for normal depth. Then HEC-RAS will compute the elevation and go from there. Take the upstream condition out unless you think you will find supercritical flow.

If you change those few things, I think your model may give you more reasonable-looking answers.

One more thing: the EGL was not coinciding with the WSE. When you zoom in, they separate.