MIDUSS - remove volume from flow?

[cht13er]

Good day!
I'm trying to model rainwater harvesting in a development; we'd like to have the first 20mm of rainfall that hits the roofs be captured and reused. Since we're assuming 50% of the lots are roofs (employment uses), this is a major component of minor storm events (e.g. the 2-year storm is 36mm).

I can't model this as a diversion, because diversions deal in flow rates - this 20mm over the roof area is a volume.

Does anyone have an idea of how I can model this correctly? Thanks in advance!

Chris from Ontario Canada

 

[dicksewerrat]

What are the reuse uses? watering plants? washing cars? flushing toilets? What will it be stored in? barrels? How will the 21st mm of water know to run off into the ditch and not to storage? I would rather want to store the later rainfall, after the roofs were washed clean from contaminants.

Richard A. Cornelius, P.E.

http://WWW.amlinereast.com

 

[cvg]

could this be modeled as initial abstraction?

 

[cht13er]

dicksewerrat: reuses will be greywater, irrigation, etc. stored in cisterns - alternatively, the sites can choose to send this water to underground infiltration.

The first 20mm that lands on the (flat) roof will be captured; the 21st mm that lands will find that the cisterns or infiltration galleries are full and will be either sent to grade or else diverted to the storm sewers (as long as the infiltration galleries are lower than the storm sewers this should work).

You raise an interesting point about dirty water from the roofs .. I'll look into that. Typically though we assume that roofwater is 'clean' and can be infiltrated immediately.

-----

cvg: you are a legend among men. I'll simply modify my Ia to 11.1mm (50% of my site is roof, 40% is parking lot, so 20mm over the roofs is equivalent to 11.1mm over all the impervious surfaces) and the runoff hydrograph is modified just how I want it to be. Thanks!!

 

[cvg]

primary contaminants on roofs are leaves, dirt and bird poop. I have a flat roof on my house and you would be amazed at what you can find on there. A few leaves and pine needles block the scuppers, dirt and water build up and form a layer and pretty soon you have a rooftop garden - which by the way is a good use for the water.

 

[cht13er]

For a light industrial/employment development of 65 ha (160 acres), the impact of trees and wildlife should be minimal. Thanks for the follow-up!

 

[psmart]

Model the cistern as a retention pond sized to overflow after 20mm of rainfall. That is, after all, the physical situation you are trying to simulate.

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[cht13er]

psmart: below are three hydrographs. The first shows the effect of Ia=20mm, the next two show the typical runoff and then the effect of a pond sized to hold the first 20mm of volume and then discharge uncontrolled after that volume. As you see, not only does it hold the first 20mm, but it also attenuates the peak flow (we didn't intend to do that!) and then after discharging a whole bunch also attenuates flows and volumes after the peak.

http://files.engineering.com/getfile.aspx?folder=05512a18-da78-4dea-84f6-74a7185fd739&file=Ia.JPG

http://files.engineering.com/getfil...e-8caf-4bd9-a0c0-dc4a104f8dbd&file=Runoff.JPG

http://files.engineering.com/getfile.aspx?folder=aeb2ba6f-5220-4a52-917c-d5bf70251ced&file=Pond.JPG

Chris

 

[beej67]

FREEZE.

What's your intent with this modeling? There's two things you could be modeling here -

1) Water Balance - where you show that you'll always be able to flush your toilets or water your plants even in the dry season, because of how much water you've stored up from the wet season, or

2) Detention - where you model what a rare event storm (e.g. 25 year or something) does when routed through your system.

For Water Balance, you typically pick monthly rainfalls, abstract those to volumes, subtract your monthly usage, and calculate a running volume of water in your cistern. Usually you pick rainfall numbers that show a dryer season than usual. This sort of thing is typically just done in Excel for a development application.

For Detention, you usually don't take credit for volume in the cistern, because to do so wouldn't be conservative. What happens if there's a slow drizzle for a week before your 25 year storm? What happens if the 25 year storm hits during Christmas and everyone's been on vacation, so nobody's flushed the toilets in a week? If the cisterns are full before the storm starts, then the volume you may have assumed for detention isn't there, which can cause downstream flooding. This is especially true for irrigation cisterns, because you don't irrigate during the wet season, and the wet season is when you get your peak storms.

If you're going to use any dead storage in your cistern for detention volume, talk the assumption over with your reviewer, explain your assumption to your client, and document those conversations, at a minimum. Whether or not you're technically allowed to do so will vary by municipal and state code. If you do take volumetric credit for them, at least do a "cisterns full" run of your stormwater detention model to check for flooding conditions, for your own edification.



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

http://www.campbellcivil.com

 

[cht13er]

Great advice, beej67; I will run the model without the reduction to see the effects. Following that, I will talk to our client and the reviewer to ensure that it's an acceptable approach - it might be that I end up assuming the cisterns are full (as you assume) or that half of them aren't working .. etc.

 

[psmart]

Regarding the retention pond alternative, the amount of peak reduction will depend on how you model the overflow. But I believe this is an accurate representation of the cistern/pond. Even if it is full at the beginning of the event, there will be some additional storage as the WSE elevation rises above the outlet, and this storage will cause some reduction in the peak.

But I certainly agree with others that it's problematic to assume that the cisterns will be empty at the start of the event. A conservative approach would assume they're full and that all the runoff will appear downstream. Talk to your reviewer.

Peter Smart
HydroCAD Software

http://www.hydrocad.net