Time of Concentration / Peak Runoff 1

[Rookie2]

I'm confused, please help. Why is it that as time of concentration for a watershed goes up, peak flow goes down?
And does it make sense to split a relatively small watershed(<100ac)into subwatersheds with different time of concentrations and combine hydrographs? Does this assume that a storm with different rainfall intensities is occurring at the same time? TIA

 

[RWF7437]

1. As time of concentration goes up, the Rational method assumes the storm duration also goes up. Because of the shape of Intensity-Duration curves, the intensity goes down as duration increases. I-D-F curves are derived statistically from precipitation records.

2. It seldom, if ever, makes sense to divide a drainage basin into pieces which have a time of concentration of less than 5 or ten minutes. My own opinion is that ten minutes is a reasonable minimum.

3. It never makes sense to use the Rational method for large drainage basins. FEMA limits its use to basins smaller than 1 square mile ( 640 acres). Many local reviewing agencies limit it even more. Where I live, Salem, Oregon, the City limits it to 100 acres or about 0.16 sq. mi.

4. The Rational method was developed in 1873 in Rochester NY and is based on only a few, small urban drainge basins studied then. It hasn't been shown to be accurate much beyond those original conditions. For larger basins you are probably better off to use stream gage records or programs such as NFF the National Flood Frequency program available free on the net. Do a web search for it. You'll find lots of other useful information along the way.

Good luck

 

[Rookie2]

Thanks so much for responding to my post. I am using the SCS TR55 method, but every method I know of (I generally work in small watersheds(<500ac), as time of concentration goes up, peak flow goes down. I understand that it is because that's what the statistics say, but if you have a five minute storm intensity over a watershed with a time of concentration of 45 minutes, does that peak from the five minute storm not travel to the point of interest? I know my question may not be clear; my specific problem may help.

We are developing a site at the downsteam end of a 100 acre watershed. The watershed consists of three well definable sub watersheds. 3/ac residential, 8/ac residential and pasture. All with a time of concentration of 30 to 45 min.
The site we are developing is at the end of this watershed with a tc of 5 min. I can show by combining these hydrographs that we are not increasing the peak flow at all because by the time the peak of the larger watershed has reached our site, our site peak flow has already left. Is this reasonable? And in the process of thinking through this I stopped to think, why does peak flow go down as time of concentration goes up. I know that's what the formula says, but i need more.

Thanks again

 

[RWF7437]

BE CAREFULL !

Your post suggests that you're making a number of unsubstantiated assuptions. Whether or not your development will increase the peak flow depends on the the size, shape, pattern and direction of travel of the &quot;storm cell&quot; which is causing the runoff. We don't know any of those things, nor can we predict them with any accuracy. The prudent thing to do is assume the worst combination of these things which will almost certainly increase the runoff downstream from your development.

Routing only one assumed combination of hydrographs does not answer the question either since they are based on a single type of storm, not all possible storms.

There is more art than science to hydrology so try to find an artful solution. Remember too, some reviewing agency will look at your design and they are very unlikely to accept an assertion that development will not increase runoff.

Good luck

 

[Rookie2]

That's why I'm asking for advice. I don't know what the conservative approach is. If we route the flow from our site through a pond then the tc goes up and actually increases the peak flow for the larger watershed. Our options are decrease the runoff for the 5 min 10 year storm and increase the runoff for the 45 min 10 year storm; Or increase the runoff for the 5 min 10 year storm and decrease the runoff for the 45 min 10 year storm. We can provide detention to do either. help? any answers concerning the tc / runoff relationship would still be appreciated.

 

[RWF7437]

If you're building a pond and outlet control structure using TR-55 you can't possibly be increasing the runoff either upstream or downstream ! Remember, the pond has no effect on the flows coming in to it and can only protect lands downstream from it. Juggling numbers cannot change the physical and climatological facts.

A pond designed for only one small set of conditions is not very useful. You use TR-55 to get an initial pond size (volume) and outlet control size. Then, you check it against a range of possible storm conditions. Even if the design standards you are required to use don't require this level of checking, the prudent engineer will want to do this. Without such a check, you don't know how good, or bad, your design is. When will it fail? How will it fail ? What is the overflow control ? Are you satisfied that the pond will meet not only the mandated design criteria but will it function safely over the range of possible storm events ?

&quot;Local&quot; detention ponds are not very useful, efficient or rationally designed in most cases and may represent a liability for the people who have to maintain them. Regional ponds and conveyance systems make much better economic and engineering sense, in my view.

End of lecture.

Good luck anyway,

Russ

 

[Rookie2]

Thanks again, but you're getting the wrong idea. I'm not trying to manipulate the numbers to give us the most &quot;economical&quot; solution. I just really don't know the answer. We have a channel coming through our site (not a stream) on which we could provide detention for the larger watershed (maybe a 48&quot; culvert) or we could pass the channel through our site with a culvert and provide a pond to retain the site only (8&quot; orifice). You know what, nevermind, we're gonna do the culvert and retain the 2 and 10 from our site. I think that's conservative. Thanks, and have great weekend.

PS - still looking for answers on the tc / runoff relationship

 

[ptmoss]

The simple answer is just that you have the same relative volume of water.

short tc -- the rate at which the water leaves the basin is very high.

long tc -- there is a longer period of time that the water is leaving the basin, hence lower peak.

As the tc is the time for the most hydraulically distant part of the watershed to reach the basin outlet point then with a longer tc the lower parts of the basin are all contributing water as that most hydraulically distant point winds its way down.

Try four or five different tc's, say 5, 10, 30 minutes, etc., and while your peak rate will vary substantially your volume will probably stay within 5%.

Detention can have high unintended consequences if the entire receiving water (stream) is not carefully considered. Can you use small infiltration/bioretention areas throughout the development instead of one detention area?

 

[Rookie2]

Thanks ptmoss for the tc/runoff explanation. That's the direction I was headed in (no, really). I actually was doing that calculation. Not much of an option on the detention pond. As I said before we are developing about 10 acres at the end of the watershed on which about 300' of the channel draining 100 acres passes. We either let the channel pass through and retain the small site to pre-developed conditions or pass the small site through and retain the larger watershed to predeviloped conditions. Or am I missing something completely?

 

[ptmoss]

To answer your predicament you really need to evaluate what's downstream from you. What does this channel empty into? What are you actually needing to design for? I think one of the things I read into Russ's comments, and I agree with, is that you may have a regulatory requirement to detain to a pre-development peak (or some other parameter), and you can fairly easily do that, but by doing so you might be creating a whole other set of problems. And these would be problems that you're not mandated to check by the reviewing authority you have to satisfy.

If you have a lake right below you it might not make much difference (detention-wise). But if there is another development right below you then you need to evaluate a larger range of events than the reviewer might require to see if your detention is having some other adverse effect.

There's a school of thought that says staged detention with some form of water quality treatment included is the way to go. Another puts higher stock in the water quality. Then there's the client (developer). They just want you to meet whatever the minimum requirements are that the regulating authority is looking for. And there are plenty of engineers that will do that and feel they are meeting their obligations just fine.

Am I starting to stray off track here?

 

[Rookie2]

One more question about tc / runoff relationship.

Sorry, i've been known to ask stupid questions and make things more difficult than they should be but.. tcmoss spoke about the runoff volume being relatively the same
for different tc's and I understand this....here is the way I'm thinking, tell me where I'm wrong (I'm thinking in terms of SCS method)..

In terms of a 10 yr Storm Hydrograph:
1) A ten year 24hr storm distribution (rainfall intensiy) looks the same no matter what the time of concentration is. (rainfall not runoff)
2) The 5 minute peak runoff is based on the average rainfall that occurs in the 5 minutes centered on the peak ordinate of the 10 year hydograph.
3) The 45 minute peak runoff is based on the average rainfall that occurs in the 45 minutes centered on the peak ordinate of the 10 year hydograph.

So, for a watershed with a time of concentration of 45 minutes the 5 minute peak intensity occurs. Is this not the peak runoff? The runoff from the 5 minute peak will travel 45 minutes and impact the point of interest, right? But the way I think the peak is calculated for the 45 minute storm is by using the average rainfall that occurs in the 45 minutes centered on the peak ordinate of the storm hydrograph.Does this make any sense? I think I've confused myself; Be gentle....The only way to learn is to ask questions.

 

[Rookie2]

Nevermind, I've sorted the whole thing out. The results I was getting made me question, and want to be sure of my understanding of hydrology. I think this illustrates why detention ponds might not be all they're cracked up to be.
By detaining the peak from our small site, we actually increase the peak from the larger watershed. By increasing the time of concentration from our site (routing through pond) our site peaks at a time closer to the peak of the larger watershed, therefore increasing the total peak. We are going to retain runoff from our site to a point where neither the site runoff or the runoff from the larger watershed is increased. Who knows what this does to the peak of the recieving stream. I know that this one site will not effect the peak greatly, but what about when the entire watershed is developed with detention ponds?

 

[RWF7437]

Please BE CAREFUL!

From what you've told us so far,you have demonstrated that for one type of storm, and for one storm duration ( 24 hours ) and for one case of antecedent moisture conditions, that it is possible for the peak flow to be increased. But you have NOT shown that this is true for every possible storm. You haven't even analyzed other storm intensities, durations, frequencies, or sets of assumptions about development patterns.

Better luck,

Russ

 

[Rookie2]

Russ, I have analyzed for the 2, 10, and 50 year 24 hr type II storm distribution. I think regardless of amc and duration of storm the relationship of the time to peak for the subwatersheds will remain the same, Am I wrong in saying that the 24 hr storm duration is the conservative approach in determining the peak flow? Yes there may be a storm distribution that produces different results, but I can't analyze for the unknown. All I can do is analyze for the worse possible case for the typical storm for this area, and I think I have done that. If you disagree please let me know, and be specific. The general criticisms and wishing me better luck is not very helpful.
Thanks

 

[RWF7437]

Partly in answer to your question, we don't know what the critical storm duration is. It may be 24 hours or it may not. TR-55 uses 24 hours only because there was lots of 24 hour rainfall data available in 1972 073 when the method was developed. So I cannot say whether in your case 24 hours is either conservative or not.

Also, TR-55 is not the only method available. There are others including the Snyder Unit Hydrograph, Clark Unit Hydrograph, Santa Barbara Unit Hydrograph and simulation programs such as HEC-HMS, SMADA, SWMM and others. All of these help us to look at the situation from many points of view and use estimates of peak flow calculated by other means such as Regional Regression Equations as does NFF ( mentioned in an earlier post).

Finally we should ask what level of protection we are really providing. The &quot;100 year storm&quot; has a 1% chance of ocurring in any one year. Over the life of a 30 year mortgage, the &quot;100 year storm&quot; has a 24% chance of occuring; 1 in 4. Over 100 years the &quot;100 year storm&quot; has a 67% chance of ocurring. In other words, it more likely than not that it will occur one or more times.

Since you find luck un helpful, I'll send you no more.

 

[Rookie2]

Russ, I didn't mean to offend. Thanks for the input, it has been very helpful. I'll take another look at it.

 

[johnhan76]

Rookie and Russ,

I am going to attempt to jump in. Keep in mind that some storms move from the downstream up and some storms move from the upstream down. In your case the entire watershed would probably receive the storm at the same time since it is only 100 acres. I would just keep this in mind for future jobs because sometimes detention will not increase the peak downstream if the storm moves in one direction but will increase the downstream peak if it moves in the other.

Russ, i was under the impression that the 24 hour storm has all of the other storms (6 hour, 12 hour, etc.) nested in it. I may be wrong.

 

[Rookie2]

Hadn't thought of that, thanks

 

[RWF7437]

johan76,

I don't believe the 24 hour storm has all the other, shorter duration, storms &quot;nested&quot; in it. Nor can it possibly represent longer duration events. It is an artificial storm derived from statistical analysis of rainfall records back in the 1970s. Many states have recently updated those record adding 40 more years of data.

Also, it is very unlikely that weather accross the US can adequately be represented by the four SCS storm types. There are many microclimates, especially in the mountainous northwest, which experience different weather patterns.

Whenever possible, I try to route actual, historical storms through detention ponds and look at the results. (Rainfall data is usually available to do this these days ) Did the pond fill ? Was it overtopped ? Might it have failed in some way during such an actual storm ?

There is a lot of software now on the market which will take much of the work out of this so there is little reason NOT to do it.

Even so, flow and volume estimates are rarely more accurate than plus or minus 30% and we can only say t6hat these are the probable flows and volume.

 

[RWF7437]

And some useful information on the &quot;shape&quot; factor for the SCS hydrograph:

Basin Characteristics Attenuation Factor

Rural, flat 150
Rural, gently sloping 200
Rural, rolling hills 300
Mixed urban and rural 400
Mixed urban-rural, gently sloping 484
Urban, steeply sloping 575
Rational Formula 645