Stormwater time of concentration

[RjMelbourne]

Colleges

I have been asked to calculate the discharge of a 25 ha rural catchment and compare it to what the local authority has on its data base. Based on my calculations trunk drainage will be designed and my client will be able to claim reimbursements from the authority. I have done so how ever I am not happy with the answers I have computed.

Below is what I have done and a description of my catchment area. I would appreciate any feed back on if you feel I have gone the write way or if my model needs a calibration.

CATCHMENT
Area:- 25ha
Average Slope 2.0%
Length 750m
Large field with grass up to 30cm high

My area has no contributing area above so I have assumed that the top of my field will sheet a maximum length of 200m (as allowed in local authority manual) to calculate my tc for sheet flow I have used Friends equation of tc = (107.n.L0.333)/S0.2 . For this I used a Horton's retardant factor of 0.045. This gave me a sheet flow time of around 19.5 minutes

I have then assumed after the 200m of sheet flow my discharge will become stream flow and used the Kinematic wave equation. tc = 6.94.(L.n)0.6 / (I0.4.S0.3) with a retardant factor of 0.2 as detailed in the Australian Rainfall and Runoff guide 1987

How ever as your aware this formula requires the input of a Intensity in mm/hr. To calculate my I0.4 to use within the kinematic wave formula I did the following

I = I0.4
I0.4 = 6.94 (L.n)0.6 / S0.3
there for I = time (min) x I (mm/hr) 0.4

Based on the intensity table for my area gave me a tc of 90 min and Intensity of 36.6mm/hr to use within my formula, with this information when I calculate my kinetic wave equation this gave me a tc of around 83.5 min.

I have then added the 19.5 min from my previous sheet flow calculation and come the the conclusion that the tc for my catchment is 103min. I have then gone ahead and proceeded to calculate the run off using the rational method of , flow = C.I.A / 360

For my intensity of my rational calc I have used 105 minutes as this is a standard ARI tc and is the closest to my calculated tc of 103 min. (I should noted the next standard ARI tc is 90 minutes)

With all of the above my figures calculated that the run off of the catchment was 1.7m3/sec. This seems very low to me. I understand that the catchment has a flow length of 750m and based on the tc of 105 minutes the storm intensity is very low, Plus a catchment of the characteristic will has a large pervious area and will absorbed allot of water but for a 100 year storm I feel discharge is very low.
Based on the above does anyone have any comments or can recommend any calibrations

 

[francesca]

Unless the field has been graded, you are highly unlikely to get sheet flow for the full 200m. The TR-55 method allows a maximum sheet flow of 300' (less than 100m), and even that is difficult to achieve on an ungraded surface. Bumps, stones, even larger plants will quickly turn sheet flow into shallow concentrated flow.

In my time doing drainage calcs for a year and a half, I never got a time of concentration greater than half an hour for a basin of around 60 acres (25 ha). The largest component of Tc was always the sheet flow component; once the flow starts to concentrate, it rapidly picks up speed. Granted my basins did not have an average slope of 2%, but many did in the channel section, even more reinforcing that your sheet flow component should make up at least half of your Tc.

Finally, I see you have rounded 103 minutes up to 105 minutes, which is wrong. To be more conservative you should've rounded down.

 

[blueoak]

For the TR-55 I normally only assume 50' of sheet. It is rare that I come across an area with strong enough soils.
I would have trouble finding a watershed less than a few square miles with a Tc that long (lots of hectares).

The following web page has a couple methods. I use Kirpich as a guess.

http://www.lmnoeng.com/Hydrology/TimeConc.htm

As a rule of thumb I divide the length of the watershed (straightline) and I guess 4 fps for small sheds for 100yr and 3 fps for 50 yr and less and then up size the pipe. But then again I like to work harder on initial loses and rainfall pattern.

I really dislike rational for areas this big, just because I like to see the loss rather than wrapping it in a C factor. It can show how small upstream storage can help your drainage.

Do you have any regression equations for the region? These are nice as a reference. They are almost always lower than hydrologic analysis, because it is easier to justify being conservative than flooding people.

Oh, my guess is 10 minutes. (The smallest my rationla allows.)

 

[francesca]

The upper limit on drainage basin area for the rational method is around 81 ha/200 ac. I would shy away from it getting close to that size basin, but theoretically you should be fine for 25 ha/60 ac, as long as you've done your homework on a good composite Coefficient. I would never use the rational method to calculate anything but a ballpark guess on a basin of this size, though. With cheap software, the TR-55/SCS methods are quick and easy to use and give a more robust answer.

Depending on the rainfall pattern for the region, I would hesitate to slap on losses; if it has rained for several days prior to the storm in question, the ground could be saturated and if it's winter rainfall, evaporation could be minimal. From a liability standpoint I would rather highlight were it is possible that people could get flooded, rather than be overly generous with my losses. After all, the SCS Curve Numbers or Rational Method Coeficients include losses in them.

 

[env21tech]

There is no relation between rainfall intensity (what is happening in the sky) and Tc(what is happening on the ground). The raincloud will act the same whether it is raining on a parking lot or on a heavily wooded forest.
The water generation capability of rainclouds is available in IDF curves (Intensity = average intensity for the duration, rainfall depth = duration x average intensity).
For an area 25 ha in size the 5-minute raincloud will be more than large enough to deluge the site, so use the average intensity associated with the 5-minute duration.
Use the Rational Method to estimate runoff.

 

[LCruiser]

env21tech - I think you mistyped. I don't think you actually mean that:
"There is no relation between rainfall intensity (what is happening in the sky) and Tc(what is happening on the ground)."

Realistically anyway...

 

[cvg]

if you compute approximate flow velocity of the stream flow you are getting about .3 feet/sec which is extremely low - and this is reflected in your very high 83 minutes. I think you need to revisit this part also. You have nearly two hours to concentrate flow on 60 acres which is unrealistic. Using a regression equation is a good idea to get a gut check for if your calculations are in the ball park.

 

[env21tech]

Lcruiser -- you are correct. I should have typed " there is no relation between rainfall intensity (what is happening in the sky) and Tc (surface water flow patterns).

 

[blueoak]

env21tech,
I think I disagree with your point,
It all depends on your storm of concern. I use different surface response (Tc) for different types and sizes of storms. On a large watershed my best guess for a PMP lag was significantly shorter than a <100yr I was calibrating with.
You really do have to look at the storms that were used to derive the lag relations.

 

[env21tech]

hello blue oak -- my thinking is that we should not use Tc to predict how hard it is going to rain. This prediction should be based on IDF curves.