NRCS TR-60 Rainfall Curve 4

[RWF7437]

NRCS TR-60 (2005) Earth Dams and Reservoirs

I am attempting to design an earth dam near Salem, Oregon, USA using the National Resource Conservation Service’s TR-60.

TR-60 includes a dimensionless rainfall curve ( Fig. 2-4, page 2-12) which is prescribed to be used in the design of the dam spillway and to calculate required freeboard. TR-60 does NOT, however, provide any information on the origin of this curve or the range of its applicability.

Any experienced dam designers:

1. Where does Fig. 2-4 come from?
2. Is it applicable to all 50 United States ?
3. Is it the most conservative method available; assuming a 50 year design life and a desired confidence interval of 95% ?
4. Is it based on the NRCS “standard” 24 hour storms such as Types I, II, III and IA described in TR-55 ?
5. Is it necessary, or desirable, to route a series of storms through the dam / reservoir in order to determine the critical storm duration for design ?

 

[psmart]

Your page/figure reference doesn't match up with TR-60 Rev. 1985 Amended 1900.

Can you post (or link to) the rainfall table?


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[RWF7437]

http://www.wsi.nrcs.usda.gov/products/w2q/H&H/docs/TRs_TPs/TR_60.pdf

 

[cvg]

I notice there is an update to the 1985 version of TR60, however my link to NRCS site is not working. how about posting the new version of TR60?

in my past experience with NRCS dams, we ran not only a 24-hour storm but also the 10-day. NRCS has a specific requirement for the emergency spillway hydrograph which is different than the dam freeboard hydrograph. you don't mention whether this is the principal or emergency spillway, both have completely different criteria. For instance, the principal spillway is designed to handle the 100-year or the 10-day storm without discharge through the emergency spillway and should be adequate to drain the reservoir in 10 days. storm distribution is per chapter 21 of the NEH-4.

the emergency spillway is designed for the PMP and freeboard is calculated using the PMP also. however stability of the emergency spillway is determined at a safe velocity using the spillway design flood which is the 100-year flood plus 26% of the (PMP - 100-year) flood

 

[psmart]

TR-60 figure 2-4 appears to be the same as the Emergency Spillway Hydrograph from TR-20 (ESH rainfall table 6)


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[RWF7437]

Thank you Mr. Smart.

Now, if I knew what "TR-20 (ESH rainfall table 6)" was and had a copy of it was I might have an answer to one of my questions. I can get a copy easily enough but will it tell me if ESH 6 is derived from some particular geographic area rainfall records? Might it tell me if this particular storm pattern is applicable throughout the 50 States ? Would it tell me the probable error of estimate in the peak flow derived from this pattern so that I could calculate the level of confidence to be expected from this calculation ?

Thank you again...I'll keep asking.

 

[cvg]

I doubt that the rainfall distribution recommended by NRCS has a large effect on how conservative the design is for 50-years, probably larger is the requirement to pass the 100-year, 24-hour and 10-day storms without spilling through the emergency spillway and to pass the PMP without overtopping the dam. But if you wish to, you can also use the alternative method for rainfall distribution which is documented: In areas without applicable NWS reference for temporal distribution, the dimensionless auxiliary and freeboard storm distribution shown in figure 2-4 may be used. Alternately, the 24-hour storm can be constructed by critically stacking incremental rainfall amounts of successive 6-, 12-, and 24-hour durations as described in Hydrometeorological Report 52 (HMR52). The maximum 6-hour rainfall should occur in the second 6-hour quadrant. The next highest 6-hour incremental rainfall should occur in the third 6-hour quadrant, the next highest in the first, etc.

 

[psmart]

I was referring to the TR-20 software manual, which includes (what appears to be) the same ESH storm distribution in rainfall table 6. Unfortuntately, the manual doesn't seem to include any discusion of the storm.

There is some discusion in NEH-4 under "Emergency Spillways" starting on page 21.49. Figure 21.2b on page 21.81 appears to show the same 6-hour ESH storm used in TR-60. In fact, the figure is marked "Use Revised Figure in TR-60".

The source for figure 21.2 is "Standard Dwg No ES-1003". I don't have any further info on that document.



Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[RWF7437]

Thank you both for your replies.

If I understand them correctly they are as follows. Please correct me if I have misunderstood you:

“1.. Where does Fig. 2-4 come from?”

No one seems to know for certain. It is probably derived from rainfall records. It is probably the result of a great deal of statistical manipulation known only to a few people at NRCS or NWS. The amount of scatter in the original data , if known, is not published by either agency. For this reason, it is not known what range of error may be expected in the peak flow calculated using this distribution but it can’t be any more accurate than the rainfall data used. This is probably +/- 30% to +/- 50% in the best of circumstances.

Because it is a dimensionless rainfall curve, it presumably may be used for any storm of any duration or of any Annual Exceedence Probability ( AEP ).

“2. Is it applicable to all 50 United States ?”

Some folks at NRCS believe it is. This belief is based on the assumption that the rainfall pattern ( Fig. 2-4) is reasonably representative of conditions in most, if not all,States. Even if it isn’t representative, the belief is that the temporal distribution of the rainfall would not have a significant effect on the design hydrograph.

Also, in TR-60 the NRCS offers an alternative method , as pointed out by cvg. Unfortunately, this alternative method only permits one to develop a 24 hour storm distribution.

3. Is it the most conservative method available; assuming a 50 year design life and a desired confidence interval of 95% ?

For a 50 year design life, 95% confidence level, it would be necessary to design for an AEP storm of 0.1025 percent ( about a “975 Year storm” ).

It is not known, although it can be tested, what duration storm and what rainfall pattern would produce the highest peak flow for this level of reliability. See 5, below.

4. Is it based on the NRCS "standard" 24 hour storms such as Types I, II, III and IA described in TR-55 ?

No, it is based in something else. It is not clear what that something is.. It may be a 6 hour storm, as suggested by Mr. Smart, or it may be something else; such as the much beloved 100 Year-24 Hour storm. Or, it may be completely arbitrary. Arbitrary is not necessarily unreasonable but it does raise concern.

5. Is it necessary, or desirable, to route a series of storms through the dam / reservoir in order to determine the critical storm duration for design ?

While not required, it is probably desirable to route a series of large storms through the reservoir and emergency spillway ( NRCS calls this the auxilliary spillway).

These questions arise for several reasons.

In the Pacific Northwest major flooding typically occurs after periods of long duration rainfall; sometimes coupled with snowmelt. Such floods have occurred in the late 1800s, in 1964-65 , and in February and November of 1996 in the Willamette Valley. These storms were all of long duration ranging from 5 days to, probably over ten days.

Rainfall and streamflow records in this area are rarely over 50 years.

Regional Regression Equations are considered accurate only to +/- 50%, at best.

For the originally cited design example, the 24 hour PMP is 14 inches, +/- 2 inches.

For these reasons at least, one would wish to be very clear about what design storm is selected and be very conservative if the damage resulting from the hydraulic or structural failure of the dam could cause significant damage.

Thank you both, again.

 

[RWF7437]

For anyone, of the two of you, who may be interested, a close approximation of TR-60 Figure 2-4 is attached.

 

[psmart]

You can also use the numeric values from TR-20 rainfall table 6, which I believe is the same Emergency Spillway Hydrograph (ESH). The 51-point mass-curve values appear below.

This data is also available in HydroCAD as the "Spillway Emergency" distribution. You can use View|Storm to examine the curve and the underlying data.

.0000 .0080 .0162 .0246 .0333
.0425 .0524 .0630 .0743 .0863
.0990 .1124 .1265 .1420 .1595
.1800 .2050 .2550 .3450 .4370
.5300 .6030 .6330 .6600 .6840
.7050 .7240 .7420 .7590 .7750
.7900 .8043 .8180 .8312 .8439
.8561 .8678 .8790 .8898 .9002
.9103 .9201 .9297 .9391 .9483
.9573 .9661 .9747 .9832 .9916
1.000


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[RWF7437]

Thank you Mr. Smart,

I was not aware that this distribution was already "built in" to HydroCad. I did not recognize it by the name "Spillway Emergency". Had I known this I might have saved hours of work recreating it. I have HydroCad 8.0.

This not your fault, of course, but is just one of the many frustrations of trying to deal with the many ways rainfall curves are described and the several H&H programs I use. Not to mention the multitude of reviewing agencies each with their own pecadillos ( sp?).

Strange too, is the fact that only two people responded to the original post. I wonder who designed all those earth dams I see across the landscape.


Thanks again

 

[psmart]

You will find a complete list of HydroCAD rainfall tables at

http://www.hydrocad.net/rftables.htm

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[cvg]

RWF - not so strange since most of those NRCS dams were built decades ago. Majority of the dams I have worked on were not built by NRCS. For the ones that were, specific design parameters were agreed to up front regarding hydrology methods. These had to be not only acceptable to NRCS, but also to the state and owner. The state or owner frequently have even more conservative methods and requirements for the dams than those required by minimum NRCS standards. By all means, if you feel that a given storm distribution or duration is more appropriate or more conservative "for your particular project", NRCS would likely approve of it if it exceeded the minimum requirements.

 

[RWF7437]

"RWF - not so strange since most of those NRCS dams were built decades ago."

I don't have a national perspective on this and no data upon which to base an opinion. I do know of a number ( less than 10 )of small dams which have been built in Oregon and Washington within the past ten years. In addition, the proliferation of local and regional stormwater detention basins, mandated by State and local government agencies, results in a multitude of small dams. The larger examples of these are not, in my experience, even recognized as dams by most of these agencies. This is a subject for another forum, perhaps ?

"Majority of the dams I have worked on were not built by NRCS. For the ones that were, specific design parameters were agreed to up front regarding hydrology methods. These had to be not only acceptable to NRCS, but also to the state and owner. The state or owner frequently have even more conservative methods and requirements for the dams than those required by minimum NRCS standards."

I've only worked in one large dam project in Sonoma County, CA. That was designed by the Corps of Engineers, hydraulic section, using, as I recall, the Corps' Standard Project Flood criteria, over 40 years ago.

Your comment suggests that NRCS criteria are "negotiable". That may be but the wording of most NRCS documents I've read, especially TR-60, use words like "shall" and "must".

Whether States or Owners "frequently" have more conservative methods I am also unable ( too ignorant) to say.

Still, that is hard to imagine. For the originally cited example I am finding the peak flow for the emergency ( auxiliary) spillway design for a 1.59 sq. mile drainage basin with 7.1 inches of rainfall, 24 Hour storm to be about 8,900 cfs. This is over 20 times the peak flow for a 24 Hour, 1% AEP storm , Type IA. The PMP ( 14 inches/ 24 hours ) storm would be over 50 times larger than the 24 Hr/ 1% AEP/Type IA storm.

"By all means, if you feel that a given storm distribution or duration is more appropriate or more conservative "for your particular project", NRCS would likely approve of it if it exceeded the minimum requirements."

Have I misinterpreted the NRCS criteria ? This would be very easy to do. TR-60 , Table 2-5 makes no mention of the storm duration to be analyzed ! Elsewhere in TR-60 only the 6 hour and 24 hour storms are mentioned.

Similarly, TR-60, in the preface, only says NRCS that they will consider "more conservative" criteria. It says nothing about "more appropriate" methods.

Thanks for your comments and interest.

 

[cvg]

I didn't say or suggest that the NRCS criteria were negotiable, only that they are minimum requirements. Stricter / more conservative criterial can be proposed and used for modifications to NRCS dams. I have not seen NRCS object to using higher flow rates, greater volumes or requiring more dead storage than that required by TR60. Note that I am speaking only of modifications to NRCS dams where NRCS is participating in the funding. Many of these NRCS dams were turned over to local agencies to operate and maintain. These owner agencies may have their own policies and standards which are more conservative.

 

[RWF7437]

Can anyone provide an example of a design using TR-60 ?

 

[RWF7437]

Advanced Search

Google
Web Eng-Tips Forums
Home > Forums > Civil / Environmental Engineers > Activities > Storm/Flood engineering Forum
Reservoir routing - hydrological problem
thread162-236161
Forum Search FAQs Links Jobs Whitepapers Forum MVPs
Read
New Posts Reply To
This Thread e-mail
E-mail It print
Print Next
Thread
SMIAH (Civil/Environme)
28 Jan 09 11:17
Hi,

I have a project involving the determination of a 100-yr storm peak flow. The watershed has the following characteristics:

Area = 8.7 km2 (3.35 mile2)
Average slope = 13.1%
% woods = 67%
% pasture = 3.4%
% residential areas = 12%
% swmamps = 2%
% lake = 15%

I have a CN number for every subbasin.
I'm thinking of using a SCS Type II storm (6 hour, 12 hour and 24 hour) and compare the results.

My problem is the 15% of lake... Actually the lake is at the downstream end of the watershed (see the file attached).

I have the volume of the lake (1 250 000 m3.)
But regarding the reservoir routing... I'm quite lost.

What should i do to get a good estimate of the peak flow at the downstream end of the lake (where there is a small dam)?

Thanks for any clue regarding the method i should use for the reservoir routing!

*

http://files.engineering.com/getfile.aspx?folder=62b5643d-0251-4f9c-a7f9-50

Thank SMIAH
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Eng-Tips Forums is Member Supported. Click Here to donate.
Helpful Member!Helpful Member!psmart (Civil)
28 Jan 09 13:11
1) You should be able to use a standard reservoir routing, such as the Storage-Indication method, which is part of all H&H programs.

2) The dam would normally be modeled as a weir outlet on the pond.

3) You will need to have some stage-storage information above the weir crest (such as contours) in order to do an accurate routing.

4) If there is a pre-storm outflow from the lake, you will need to incorporate this as a base flow and/or starting elevation. The exact mechanism varies depending on the software you use.

You also need to allow for the direct precipitation on the surface of the lake. This is generally done by including the surface "runoff" using a curve number of 98+


Peter Smart
HydroCAD Software

http://www.hydrocad.net

Thank psmart
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

SMIAH (Civil/Environme)
28 Jan 09 13:32
Thanks Peter.
This is precious informations



Thank SMIAH
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

SMIAH (Civil/Environme)
28 Jan 09 13:40
One thing i'm not familiar with is the lag time for the lake?

I would tend to use a very short time.
Correct?


Thank SMIAH
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Helpful Member!Helpful Member!psmart (Civil)
28 Jan 09 13:49
If you wanted to calcualte the Tc all the way to the dam, you could include the lag time for the lake. But since you're going to perform a separate resevoir routing for the lake, the Tc path for the contributing watershed would generally end at the water's edge.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

Thank psmart
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Drew08 (Civil/Environme)
28 Jan 09 16:53
You should also use a CN=100 for the reservoir area so S=0 and P=Q, i.e. no infiltration, storage, or inital abstraction losses.

Thank Drew08
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Helpful Member!Helpful Member!psmart (Civil)
28 Jan 09 17:32
Drew08 - Since there are still some losses (such as evaporation), a lower CN of 99 or 98 can be appropriate for direct rainfall on the water surface. In most cases, it won't make much difference to the final results. If in doubt, try a range of values and compare the results.

I beleive there was a recent thread on this topic...


Peter Smart
HydroCAD Software

http://www.hydrocad.net

Thank psmart
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

msquared48 (Structural)
28 Jan 09 22:03
I would assume the worst case here with no available extra storage in the lake and, depending on the length of the flow through of the lake, tend to set the time through the lake to zero, neglecting any upstream effects.

Mike McCann
MMC Engineering

Thank msquared48
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Helpful Member!Helpful Member!psmart (Civil)
28 Jan 09 22:10
Depending on the exact storage and discharge characteristics, the lake could provide significant attenuation of the peak inflow (runoff). To estimate this effect you need to perform a reservoir routing.

Peter Smart
HydroCAD Software

http://www.hydrocad.net

Thank psmart
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

msquared48 (Structural)
28 Jan 09 22:27
Yes, but then you are assuming that when the design storm hits, the reservoir will be drawn down to some degree. Unfortunately, most major storms happen when the reservoirs tend to be higher, if not full.

If the structure is in the FERC network of dams, there could be enough time to draw the reservoir down prior to the storm, but that could be a luxury here. Even the FERC and Corps models are not perfect as the recent storms in the Northwest here proved recently. I would be conservative and design as if there is no storage. There could be a backwater effect, and I might yield on that issue, but not the available storage.

Mike McCann
MMC Engineering

Thank msquared48
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Helpful Member!Helpful Member!psmart (Civil)
29 Jan 09 1:36
I was not assuming any draw-down prior to the storm. In fact, I was assuming that the initial water surface was already overtopping the dam, in order to produce a pre-storm discharge, as you would commonly see for a natural lake.

Even when the initial WSE is above the spillway, the storm will increase the WSE even further, and this increase in storage will cause attenuation of the peak flow.

This effect can be simulated with a reservoir routing, based on the dam's stage-discharge relationship, and the stage-storage relationship as the water exceeds the spillway. A narrow spillway will "hold back" more water and produce more storage and attenuation than a wide spillway.



Peter Smart
HydroCAD Software

http://www.hydrocad.net

Thank psmart
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

msquared48 (Structural)
29 Jan 09 2:05
OK.. I agree - backwater analysis as I said.

Mike McCann
MMC Engineering

Thank msquared48
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

SMIAH (Civil/Environme)
29 Jan 09 8:35
This conversation is interesting.
I am assuming that the reservoir is full (i.e at the top of the crest of the small dam at the downstream end).

But combining a 100-year/6-hour SCS II storm + a full reservoir might be too conservative?

We don't have any legislation in here (Canada *sigh*) regarding those issues.




Thank SMIAH
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Helpful Member!cvg (Civil/Environme)
29 Jan 09 10:15
100-year storm plus full reservoir starting condition is the typical procedure here. Level pool routing is often used to route through the reservoir. For large reservoirs or long dams, a two dimensional routing method may be used which may do a better job of estimating water surface elevations than level pool methods. Also, if this is a larger reservoir or dam, then the spillway generally must also be large enough to pass the PMF.

Thank cvg
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

Helpful Member!Helpful Member!psmart (Civil)
29 Jan 09 10:17
A backawater analysis is usually not required to evaluate the peak-flow reduction in this type of scenario. As long as the outlet control (dam) is relatively small in relation to the lake, it will create a zero-velocity level-pool, which can be modeled using a reservoir routing, such as the storage-indication method.

On the other hand, if the level-pool assumption is not met, then a backawater analysis would be required. This would be the case with a dam across a river, but that's not how I interpreteed the question at hand.

"But combining a 100-year/6-hour SCS II storm + a full reservoir might be too conservative?"

If the lake is normally filled to the spillway, this does not seem overly conservative. But it's really a matter of the standards you are trying to meet, and the downstream consequences of any increase.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

Thank psmart
for this valuable post!


Inappropriate post?
If so, Red Flag it!


Check out the FAQ
area for this forum!

SMIAH (Civil/Environme)
29 Jan 09 10:26
Thanks!

Thank SMIAH
for this valuable post!



SMIAH (Civil/Environme)
29 Jan 09 13:15

Taken from the FERC:

http://www.ferc.gov/industries/hydropower/safety/guidelines/eng-guide.asp

 

[RWF7437]

For the few (two?) who might care.

Based on the original post, and using the criteria in the FERC document provided by SMIAH, I get the following:

1% AEP/ 24 hour storm Type IA rainfall ( 4.7 inches) AMC 2:

Peak Flow = 377 cfs

PMP/ 24 hour storm Type IA rainfall ( 13.0 inches) AMC 3:

Peak Flow = 2091 cfs


PMP/ 24 hour storm Emergency Spillway Type rainfall ( 13.0 inches) AMC 3:

Peak Flow = 2355 cfs ( About 13 % higher than Type IA, immediately above )

PMP/ 6 hour storm Emergency Spillway Type rainfall ( 5.7 inches) AMC 3:

Peak Flow = 2246 cfs

PMP/ 72 hour storm Emergency Spillway Type rainfall ( 17.5 inches) AMC 3:

Peak Flow = 1133 cfs

All flows calculated using HydroCad 8.0.

 

[DamEng1]

I got lost in all the responses, but the first thing you should do is to check with the state dam safety office and determine what criteria they require for the design of dams. Also, the temporal distribution of the rainfall can have a large effect on the resulting hydrograph. The emergency spillway distribution with a 40+" rainfall is essentially like designing for a Type II rainfall of around 30".