Regional Storm VS IDF Curves

[Loumenad]

I am performing a hydrologic study between the Rational Method (current method being used), NRCS TR-55 and SBUH (both proposed) and am wondering if IDF curves are based on Regional Storm types. I know that both are used for historical rainfalls, Storm types being rainfall pattern while IDF curves being depth, but have not found documentation explaining if there is a connection.

NRCS use regional rainfall patterns that are generic for the area with just the need for plugging in your storm event depths, but I feel that IDF curves may apply depths similar to storm types given the fact that they are modeled using historical rainfall.

I am trying to determine if this justification is a reasonable contributing factor as to why the Rational Method underestimates volumes of storage for design areas, when compared with TR-55 and SBUH. In addition, I do realize that the rational method was orginally used for conveyance design and peak flow design, not volume runoff calculations, but I am trying to be well versed on the subject to help with analysis of my results.

 

[beej67]

It all comes straight from historical data, neither comes from the other.

The rational method underestimates storage volumes because it assumes a flat hyetograph exactly as long as (but no longer than) the Tc of the storm, and storms don't really happen that way. It's a good assumption to guess the peak discharge (cfs) you're likely to see for a given watershed every X years, but not a good assumption to guess the peak runoff (cf) you're likely to see every X years.

Hope that makes sense?


Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[Loumenad]

I guess the question is then, are IDF suitable for volume calculations or are they only suitable peak flow? The city I am doing this study for uses IDF curves to find the volume of storage required to pass the 50 year storm while releasing at the 5 year pre-developed rate. To do this, the peak flow for the 5 year rate is found using IDF and iteration with intensity and Tc of storm. Then, using the C and A for the improved area in Q=CIAD, the IDF is applied with the 50 year storm and the largest volume required is considered as the storage. SO I and D are based on the curve and simply the pre-rate*D subtracted from the 50 year Q=CIAD gives the storage. Does this process make sense to use, it seems to resemble the Modified-Rational Method.

 

[beej67]

Lets take a step back. When you read off the intensity from an IDF curve, you get a different intensity for any given Tc, right? That's what makes the curve.

Multiply that intensity (in/hr) by the Tc (hrs, not minutes) and you get a rainfall depth, in inches. In theory, that's the same rainfall depth you would have gotten from finding your location on an isohyetal map drawn a storm duration equal to your Tc. (presuming the IDF curve and isohyetal map were both drawn from the same historic data, anyway)

So if you're trying to do the NRCS TR-55 method for a Type II 24 hour storm distribution, you could get your 50 year 24 hour rainfall total from those old nationwide TP-40 maps, or from published 24 hour storm data, or you could get it by finding the value in in/hr off an IDF curve way down at the end where the Tc is 24 hours, and just multiply it by 24 to get your rainfall total in inches.

Did you follow the above bit?

To get any further with your second post, I must admit, I've never seen "D" in the rational equation before. What's that supposed to represent? I sometimes see a Cf, which is a "frequency factor" thrown in for big storms, is that what your "D" is? In particular, I'm not following you here:

SO I and D are based on the curve and simply the pre-rate*D subtracted from the 50 year Q=CIAD gives the storage.

I have seen several different methods for approximating detention volumes using synthetic hydrographs generated by different variations of the rational method. All of these are pretty loose on their assumptions, and none are very accurate, compared to something as robust as the NRCS method. In general, I would say that Q=CIA is for peak flow only, because it's not giving you any information about the storm distribution you're to detain, just the flow rate at the peak of the storm. However, if the municipality has a specific method they want you to use that's rooted in the rational method, and they've had success with it in the past, then don't be shy to use it.



Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[Loumenad]

I see what you did for the IDF curve and finding the 24 hour depth using the 24 hour Tc, but does the IDF necessarily produce the proper rainfall distribution for the region or is it just a rainfall depth divided evenly throughout a 24 hour period?

And the D in Q=CIAD is the duration of the rainfall event, the equivilent to Tc, except its used for the 50 year event to find the maximum volume being stored. SO once you find the pre-developed 5 year flow rate, you would then find the maximum difference in storage by finding the largest difference between the 50 year Q and the pre-rate flow times that duration.

Regardless, I see that by doing this process and talking through this idea, that by doing this necessarily doesn't reflect an accurate rainfall event because if you have 1 in/hr fall for 20 minutes and thats the "peak" that might not neccessarily be a true TYPE 1A storm, or whatever storm you are looking at.

 

[beej67]

And the D in Q=CIAD is the duration of the rainfall event, the equivilent to Tc, except its used for the 50 year event to find the maximum volume being stored.

So your Q in your equation is not flow rate (cfs) but is a volume (cf) and your D is in seconds? If so, that would represent the volume of runoff from a flat 50 year intensity hyetograph, over a period of time equal to the Tc of the basin. Then your delta between your two CIAD's should be the runoff volume difference, which isn't too bad approximation of detention volume, honestly.

I could see that being a very useful design method for small sites, as long as you don't cheat on your Tc. Both Tcs (exist/proposed) need to be the same, and need to be reasonable. Also, as usual, it only works if the regulators buy it.

Have you checked the detention volumes it predicts vs NRCS method?

Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[Loumenad]

Yes, I did some calculations for this for proposed areas of design, and the storage volume required by the rational method is significantly less. As area increases so does the difference. In the case of a 10 acre development my calculations between TR-55 and the rational method had a difference of 20,000 cf.

I have seen a similar presentation of differences like this from a PE in New Jersey in which the idea of the Rational Method predicting close approximations for peak flows but differing considerably for volume calculations which are in line with my results so I think there is commonality there.

Thank you for your feedback, I was really looking to see if IDF curves are more designed for peak flow usage which it appears that may be the only real use they can have, as have a 3 hr event at constant rainfall may be poorly reflecting "real" storm conditions.