HEC 2 vs HEC-RAS 3

[krapp]

I am a consultant conducting a riverine floodway analysis to provide updated flood plain elevations. I have the HEC 2 output used by FEMA in the most recent FIS for the waterway. When I put it in HEC 2 along with the updated cross-sections and culvert data, it works perfectly. However, I would much rather work in HEC-RAS.

I've imported the HEC 2 data, made the necessary adjustments to the bridge data, inputted the culvert data, and all of the resulting WSE's are off. Way off.. I've double checked all the cross-sections, manning's n's, and bridge/culvert data everything jives with the HEC 2 input.

Could this really be the difference in how the two programs calculate culverts? Or am I missing something inane?

This forum has been helpful to me in the past, and I thought I'd give it a shot before I give up and do the study solely in HEC 2.

Any insight would be appreciated.

 

[froude]

krapp,

I was wondering how far off the HEC-RAS values were from the HEC-2?

In HEC-RAS, if you change the conveyance calculation method to "HEC-2 STYLE," do you get the same WSE's as in the original HEC-2 model?

 

[krapp]

froude,

Let me explain a little first. The study area is bounded on either end by a culvert. The first downstream cross-section is 100 feet down from the lower culvert, I set downstream boundary condition as the WSE calculated by the original study for that cross-section. Those first two cross-sections (the downstream one and the cross-section immediately downstream of the lower culvert) match exactly with the original study.

The WSE calculated for the next upstream cross-section (the inlet of the lower culvert) is 4.3-feet above the WSE from the original study. The next two upstream cross-sections are 3.7-feet and 1.2-feet above.

The next upstream cross-section is the downstream end of the upper culvert, and is 0.1-feet LOWER than the WSE from the original study. The last two upstream cross-sections (the upstream end of the upper culvert and the last cross-section 85 feet up from that) are 1.6-feet and 1.4-feet lower than the original study.

All of these differences are well outside tolerances suggested in the HEC-RAS reference manual which state that 95.8% of HEC-2 cross-sections are within 0.02-feet when calculated with HEC-RAS. It appears that the two programs definitely calculate culvert capacity differently, but the differences are too far apart to rationalize.

 

[RWF7437]

FEMA has some guidance and policy memos on this problem.

Here's just one ypu can copy and paste:

http://www.google.com/search?client...between+HEC+II+and+HEC-RAS&btnG=Google+Search

Do your own Google search and you may find others. It appears there are indeed considerable differences between the two models and FEMA encourages use of HEC-RAS whenever possible.

There may, of course, be other reasons for the discrepancies you are finding. If possible, try to get an independent check on your work just to make sure you've not made some silly mistake. It's very hard to find one's own mistakes so it may well be worth doing this.

good luck

 

[beej67]

I've never run into this problem before, but I know the more recent versions of RAS had an option where you could toggle between using the old RAS routines for bridge abutment contraction, and the old WSPRO routines, which were different, and more generally accepted by FHWA. I think there's a checkbox in RAS someplace. There's also different guidelines about how far away from the bridge you have to cut your sections for each routine, because of how the models treat streamlines.

That'd be the first thing I checked. I have no idea whether that applies to your problem, though.

It also sounds like you might not have your ineffective flow areas computing correctly. Read about that stuff in the HEC-RAS manual and see if it's right.


These are just some stabs in the dark.

 

[froude]

I agree with what beej67 has said. You may consider creating a revised model which in which you:

1. Re-enter bridge/culvert data (this may be necessary if ground points are used to model piers, etc.)

2. Re-cut cross sections (especially if you have newer topographic information). I've found this to be especially helpful when the HEC-2 model is considerably old.

3. Move or re-cut bounding culvert sections an adequate distance away from the U/S and D/S face of the culvert to properly allow expansion and contraction coefficients.

4. Incorporate the use of ineffective and/or permanent ineffective flow areas where appropriate.

I have found that the above steps often reduce some irregularity in WSE's - but not always. Hope this helps!

 

[RWF7437]

"All of these differences are well outside tolerances suggested in the HEC-RAS reference manual which state that 95.8% of HEC-2 cross-sections are within 0.02-feet when calculated with HEC-RAS."

I looked for, but could not find, this 95.8% / .02' tolerance in the HEC-RAS Reference Manual. Where is it please ?

More important, what is the logic of it ? It sounds completely arbitrary considering that BFEs computed using HEC-RAS are considered accurate only within plus or minus 1 foot.

 

[RWF7437]

Krapp,

The reference you cite is to Appendix C of the HEC-RAS, Hydraulic Reference Manual. IT is NOT a FEMA or NFIP requirement. It is not even a recommendation.

You will find the best information I've been able to find at FEMA's website which was linked to in my August 17 post. There FEMA only says the two studies should agree within 0.50 feet at their upstream and downstream ends. FEMA does not require you to "rationalize" the differences between those two points, which may not be possible. FEMA only expects you to note and explain them.

good luck

 

[RWF7437]

http://www.fema.gov/library/viewRecord.do?id=2292

 

[RWF7437]

To better understand your question, can you upload ( attach ) your system schematic, HEC-RAS Profile, and steady flow (Q100) information ?

If you are trying to establish a floodway, which of the 5 available methods are you using ?

If your client is not FEMA, is your client a local government agency or are they private individuals ?

If water surface elevations computed by HEC-II and HEC-RAS are different, even considerably different, what are the practical consequences of those differences ?