Flood Zone Calculation 3

[uncivilized]

I am trying to set a Base Flood Elevation for a subdivision. This is my first endeavor so please bear with me. A few questions:

1. Hydrologic method? I am using the SCS method. The peak that I am getting is a lot higher than the USGS regression equation (due to storage effects?). Should I use the more conservative?

2. Software? I have XPSWMM which I think is approved. But I also downloaded QUICK 2. The drainage area is 979 acres. At the lower end of the drainage area are (4) 36" RCPs which will cause the area to pond.

any help as far as outlining a solution would be most helpful.

 

[sam74]

Using that large of a drainage area I would use a regression equation. Either USGS or local DOT equations which would factor in more local variables than USGS. Our DOT has different equations for different ranges of drainage areas and rural/urban land use.

I have never used SWMM but I believe I have read that it is FEMA approved software. I have used HEC-RAS which is also FEMA approved and free.

 

[RWF7437]

You should probably use the same flow as published in other flood studies in your geographic area. That is probably the regression equation flow but check other studies in your area.

As for software, FEMA's website has a list of all approved software for flood studies. Go to their website. My own preference would be HEC-RAS. It is free and also available for download at the Hydraulic Engineering Center (HEC) website.

Google these things to get exact web addresses.

Good Luck

 

[uncivilized]

Thanks to sam74 and RWF7437 for your advice. Ive got the flow now and the flow creates road overtopping. The road acts like a dam and creates a ponded area. This ponded elevation will be my base flood elevation, correct? However there are parts of the property more upstream of the ponded area. How does one determine the base flood elevation here. Normal depth? Wouldnt the ponded area cause some sort of tailwater effects on the normal depth? Any help would be appreciated.

 

[gbam]

A step backwater model (HECRAS) is necessary to determin the extent of the road's influence and flooding. Also, is the road truely ponding the water or is the water flowing over the road? I suggest that you run HECRAS or HEC2 to determine the water profile through this reach.

 

[francesca]

A drainage area of that size will produce flooding that is significant for a residential subdivision. You need to produce a model that tells you the flood elevation along the entire length of your subdivision alongside or affected by the stream. A detention basin-type calculation will not be adequate because it won't take into consideration velocity or the bottom profile. You will need to plat the 100-year flood zone on the affected lots, and it is advisable to add a few feet to your answer, particularly if it does not affect your lots (e.g. if there is a bluff or a 50' stream bank buffer requirement).

I have used both SCS method and regression equation method for determining flows; it depends on the basin size. You're not trying to create a FIRM, you're trying to set a flood elevation (or a finished floor elevation) that is going to keep water out of the houses and prevent you being sued or losing your license. Use FEMA methods as a minimum, and be as conservative as your judgement tells you to be. I used SCS where I was sure of my land use, soil types and TR-55 times of concentration. I had recent aerial photos, a county-wide soil map and contours accurate to 5'. I used regression equations when I had a basin that was of the order of square miles, across a state boundary and where topo maps were my only source of contours for basin delimiting and slope estimation.

I have done this exercise several times over the past year and I recommend HEC-RAS. You will need to be very careful with the road; the culvert calculations are tricky in HEC-RAS and you need to read the Hydraulic Reference Manual (or the help!) to make sure you are assuming the right variables for the right calculation method. Make sure you check the "pressure flow" option if the road is being overtopped, which it almost certainly is.

HEC-RAS needs a few extra downstream cross-sections to allow the assumed downstream water depth to settle to the "true" value. (For instance, if you assumed normal depth, but the velocity is low, then the "true" depth will be deeper; HEC-RAS requires a few extra downstream cross-sections to settle the water elevation calculation.) If the road is your downstream limit, you might consider including a few cross-sections downstream of the road. The water surface elevation upstream is controlled by the road/culvert hydraulics only, but HEC-RAS will need at least one downstream cross-section, so you may as well give it a few more. It's much easier to plan for more cross-sections now than to add them later when you've set your stationing, etc. (Adding extra cross-sections within your modeled area is easy thanks to the interpolation feature.)

My final bit of advice is this: do not show the flood zone beyond your property boundary. You may set an elevation (or series of elevations), but don't show the area on the opposite side of the creek unless the land belongs to the same developer. Make sure your subdivision won't increase the flood level (local storm water runoff rate may increase, but will be gone before the flood peak arrives); i.e. don't fill in the flood plain, but don't set the flood plain for any property but your client's.

 

[uncivilized]

Thanks to all who replied. I need to get this turned in on Monday morning. I am not comfortable enough with HEC-RAS to sign a plat. I do recognize that this is something that I need to get comfortable with. Is there any way that I can do this by hand. Here is my situation.

1. Property in question is on one side of the creek.
2. Lower end of property is at the road/culvert. Road is at elevation 192.
3. Upper end of property is 1000' upstream of the property. The elevation at the centerline of the creek here is 188.
4. Culvert will not pass flow without overtopping.
5. I have field surveyed cross section at the lower and upper ends of the property and some culvert info.
6. As gbam has stated the creek will be flowing over the road and not act like a pond.
7. I can find the 100-year elevation at the culvert by using the weir equation and culvert nomographs.
8. I need to find the elevation of the 100-year flood at the upper cross-section.

If this is simple enough to outline on this forum I really would appreciate the help. If not, please let me know, and Ill start reading HEC-RAS manuals tonight. Thanks.

 

[RWF7437]

Start reading. There are some approximate methods one can use but we don't know enough yet to say whether you can use them in your case. For a quick review of those methods go to

http://www.pdhonline.org

and review my course Base Flood Elevations by Aprroximate Methods. You don't have to pay to view the course materials and you can download the course in .pdf format to read at your convenience.

Good Luck

 

[RWF7437]

http://www.pdhonline.org/courses/h110/h110content.pdf

 

[uncivilized]

Hope someone is still out there on a friday afternoon. It looks like HEC-RAS is going to require 4 cross-sections for the culvert and one cross-section upstream for the property in question. Is this correct?

 

[francesca]

The HEC-RAS user manual is not worth much, but they Hydraulic Reference Manual is outstanding. It's going to be a long weekend for you, unfortunately! I hope you don't discover that half your lots are under the flood zone and you need to redesign your subdivision.

I guess you could do some quick-and-dirty estimates. You can always replat after doing the HEC-RAS, assuming some lots aren't wiped out by the flood zone. Since you know your flow and your cross-section, and can calculate your average bottom slope so you could come up with a velocity and use Q = v * A to get the depth. (Exclude all area below the elevation of the road, as it's an "ineffective flow area" because the water will be more-or-less stationary.) Be overconservative in your Manning's coefficients, because the slower the water, the deeper the it is.

I wouldn't do that unless I were desperate and I could comfortably add 10' to the answer, however. Remember, it's your license and your reputation on the line, not to speak of potential law suits!

 

[RWF7437]

Its Friday morning here (Oregon)

I would add at least one more cross section downstream of the culvert to define the expansion reach. More sections will increase the final accuracy of your result so don't be stingy. Also, make sure the cross sections are wide enough to include the entire flood plain. HEC-RAS will truncate sections when they are not wide enough leading to errors ( usually small) in the computed water surface elevation. Plot the cross sections to look for this possible error.

Good Luck

 

[sam74]

I'm confused a little on the road/culvert you mention. Is it existing or proposed? We typically only do floodway calculations if we are crossing the channel with a proposed structure (if there is an existing published FEMA FIS).

I know our surveyors have taken a published flood map, scanned to a .tiff, referenced into a .dgn, scaled to fit, and traced over the floodway boundary as provided. I imagine including a reference to the data source. That is what the published maps are for if your not disturbing the floodway.

Of course your out of luck if the stream has not been studied or you are filling/crossing in the floodway.

 

[francesca]

sam74: Would you stake your license on what you suggest? I know I wouldn't. I believe that the road is existing; a proposed road would have to be 2' above the 100 year flood elevation to be accepted, at least in the city I worked in. I think uncivilized's problem is that he has to establish a finished floor elevation to plat on subdivision lots. Even if he's not at the plat stage, he'd likely have to have the flood zone delineated in order to get a grading permit. Scanning the FEMA map might do the trick at this stage, but I worked on a subdivision that had to be redesigned after my HEC-RAS study because the FIRM for the area said "minimal flooding" (zone C).

RWF7437: Does HEC-RAS use expansion reaches for culverts? It's been a few months, but I recall only doing expansion reach calculations for bridges. I can see how the principle would hold for bridges and culverts, and my failure to calculate expansion/contraction reaches for culverts would explain my lack of satisfaction with using HEC-RAS for culverts! I looked it up on the online help and I'm still not sure.

uncivilized: Presumably you have a fairly accurate DTM for your subdivision. Cross-sections drawn from your DTM will be accurate enough. The majority of flow will be on your overbanks, not in the main channel; in fact, much of your main channel may be "ineffective flow area" and all of it below the bottom elevation of your culverts will be, so accurate channel x-sections are low-priority.

Other tips:
* I've heard it said that "top of bank is about a 1.5 year storm" -- make sure you extend your x-sections a ways up the slope because it's infuriating to have to extend x-sections at a later date.
* Also make sure that your x-sections aren't straight, but in fact bend so that they're always perpendicular to the contours. A line drawn perpendicular to your x-section needs to point in the direction of flow (which is parallel to the bank contours and perpendicular to the creek bottom contours, assuming a U-shaped ditch contour).

 

[RWF7437]

Francesca,

Yes, of course, HEC-RAS uses expansion reaches for culverts, and bridges, and any other "obstructions" to the natural flow of the stream in question. As you have often said in this forum or these forums, visulize the flow. Then, look at the hydrologic and hydrology manuals and and you will "see" what must be happening.

Thanks for your comments.

Let's both wish "unicivilised" Good Luck

 

[uncivilized]

Sam74 - The approximate method that you are referring to is actually an accepted method provided it meets certain criteria for accuracy and size of subdivision. However, I do not feel comfortable with it. The subdivision in question is only 2 lots with 1000 foot of stream coverage.

Franseca - We are in the flatwoods section of the country. The slope is so mild it is difficult to tell the surveyors to stay perpendicular.

I dont really have enough x-sections for HEC-RAS. I guess I could make my own based on DTM. Is is it typical to make your own, or do you get the surveyor to field survey the x-sections that you need?

I ended up using xp-swmm (fema approved) I used conservative run-off and conservative n values. However I dont quite understand the results. The water elevation at the road crossing is only .2' feet below the water elevation 1000' upstream. There is 1' of difference between the ground elevations in the bottom of the channel. Also the downstream channel cross-section is wider than the upstream which should cause the difference in water elevation to be even more pronounced.

RWF - thanks for the well wishes. Hope that I can you help you one day.

 

[francesca]

Firstly, the road is likely causing the water to back up, influencing your upstream water elevation. Secondly, in a narrow cross-section, your velocity increases, increasing your velocity head and reducing your static head. Look at the bernoulli and continuity equations and it'll make sense that you get deeper water at a wider cross-section.

We found it a waste of time and money to send a survey crew out to conduct the cross-sections. We had a field crew surveying for nearly two weeks taking just stream channel x-sections, but the majority of the flow is on the overbanks. While of course it depends on the accuracy of your DTM, it is better to select cross-sections on paper than in the field and the level of accuracy is sufficient. HEC-RAS reduces your cross-section data to a single point, representing head loss at that location based on x-section geometry and Manning's roughness values.

 

[RWF7437]

I hate to disagree with Francesca BUT, in 45 years of working on this stuff I have never found it a waste of time to select cross sections in the field. No DTM, GIS or other whizbang program can substitute for knowledge gained by viewing the site on the ground. I don't know any program that will give you a reasonable estimate of Manning's "n" values. Nor will such office engineering allow you to clearly visualize the areas of ineffective flow, and possible debris traps in the floodplain fringe. Viewing the stream under study at as many stages of flow as possible will always pay you dividends in understanding how it behaves.

 

[francesca]

I agree whole heartedly that a site visit is a must. What I meant to say was that only selecting your cross-sections in the field is less accurate than from contours because thick woods, for instance, precludes your visibility and it's hard for a survey crew to stay on the correct cross-section line unless the crew has a thorough understanding of the way HEC-RAS requires its cross-sections, which field crews seldom - if ever - have.

 

[uncivilized]

Francesca,
I assume that you construct your DTM from a field survey and not USGS topo maps. Right?