Rainwater harvesting pipe size calc 2

[swazimatt]

I am looking at collecting rainwater off an industrial roof building. The existing downpipes are at irregular intervals and connect to the below ground stormwater network.
i want to disconnect this from the SW system and connect these downpipes to a below ground pipe to the tanks
gutter is 8m above ground, and outlet into tanks is not more than 3m above ground - 5m static head. Total building length is about 100m and then another 40m to the tanks

I am not sure how to calculate total losses in the pipe to make sure i have enough head to get the total calculated runoff (approx 50l/s) to the tanks
Do i start with the furthest downpipe, size it for the runoff from it's roof catchment and then calculate head loss to the point the next downpipe connects to the main pipe and calculate head losses for the next section with the increased flow and carry on like that?

It would be quite straight forward if there was only one downpipe but i am getting confused with multiple supply points

I think i have got to the point where i have overthought this and lost site of the simple solution. I have thought of modelling this with epanet with a tank at the top of each downpipe but that would only work with equal flows to each downpipe and my catchments are different

Any advice welcome

 

[BigH]

Make sure your state and/or city permits rainwater harvesting . . . some do not.

 

[SlideRuleEra]

Do i start with the furthest downpipe, size it for the runoff from it's roof catchment and then calculate head loss to the point the next downpipe connects to the main pipe and calculate head losses for the next section with the increased flow and carry on like that?

Yes.

To prevent pipe that is oversized (say, used at full capacity only in a 50 year rainfall event), what flow rate should be used is the real question.

 

[TerryScan]

I can't make sense of your post:

On one hand you say "connect these downpipes to a below ground pipe to the tanks". Then you say that the outlet into the tanks is 3m above ground.

Do you plan to pump the water into the tanks?

 

[cvg]

you need to start at the downstream end and work upstream...
also suggest that you should not undersize this pipe. any backup could end up causing problems on the roof. make sure you either make the pipe large enough or provide an overflow to prevent backup if the pipe is clogged (or too small)

 

[swazimatt]

On one hand you say "connect these downpipes to a below ground pipe to the tanks". Then you say that the outlet into the tanks is 3m above ground.

Do you plan to pump the water into the tanks?

No it will be a reverse siphon

Make sure your state and/or city permits rainwater harvesting . . . some do not.

I am in New Zealand, rainwater harvesting is encouraged to some extent, in this case we are adding additional impervious area so am using the rainwater harvesting as detention and reuse to avoid increasing peak runoff from the site

you need to start at the downstream end and work upstream...

would i just use the hazen-williams formula with known Q and size the pipe?

 

[cvg]

HW or mannings, either could be used for this application.
definitely allow for minor losses, air, debris etc

expect that an inverted siphon will collect sediment and debris and will eventually clog and fail

 

[LittleInch]

I don't know how EPANET works, but don't understand why your "feed" tanks would need to be the same flow? It sounds like you know what the flows are in each down pipe?

But you could just size the last 40m with all the flow and either leave it at the same size or reduce diameter as you work backwards. You can work backwards from each connection node reducing flow as you go, but remember the head at the node is the same for the header and the down pipe.

you could also do this by a series of calcs in an excel sheet.

The key will be to figure out your worst case thunderstorm and ensure that the furthest downpipe doesn't result in more than 5m of head above the tank inlet level.

I would also allow for some flushing connections on your header from one end to the other as long term low flow (summer) could lead to build up of dirt, leaves, dead birds etc

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[cvg]

EPANET is typically used for pressurized water networks, probably not an ideal application. however EPA SWMM would work well for this. but unless this is a very large and complicated system, the spreadsheet method should be more than adequate

 

[swazimatt]

I don't know how EPANET works, but don't understand why your "feed" tanks would need to be the same flow? It sounds like you know what the flows are in each down pipe?

Epanet works backwards in that you say what you want to flow out the end and it then works out how much would be flowing from each tank, but i actually need to calculate if my head and pipe size allows for my total flow into the tank under 5m head (minus friction and minor losses)


Thanks, I will give it a crack and see how it goes

 

[LittleInch]

What's the pipe size now and what's the end point?

Just make it 50% more diameter and you won't be wrong so long as the existing system works OK.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[stanleyshum1997]

If the storage tank is above ground, you can only collect the roof runoff. Make sure there is enough slope for the runoff and the pipe. The usual slope of 1" to 12" is not enough. Make clean-out every 100'. Anchor blocks are required in entry corners. That is where the pipe is bent. Always oversize your pipe and storage tanks. A tank is not a good idea because it only has a limited retention capacity. Overflow drain is needed for the tank. Beware of vertical drops in a pipe when there is a connection. water from vertical drop gravity might damage the connection. Clamps are needed for the pipe. Aware of roof ponding and the pipe opening can drain faster than the water accumulates. Use closed pipe flow formulae to size the pipe. Size it bigger than requires. Use energy grade line and heading lose due to pipe friction to find the flow rate. Do not forget to consult the local plumbing codes.

 

[NOLAscience]

How will the client use the water that you collect? How quickly? During the rainy season, water will not be needed so much for watering landscape, if that is the plan. During a drought, the tank will empty quickly. Last year I hardly ran the drip irrigation system in my garden, and this year, I can't keep certain plants alive regardless of how much I am watering.

 

[swazimatt]

The tanks are both for stormwater attenuation and reuse (orifice outlet set a certain height above the base of the tank, irrigation storage below the orifice). We are increasing the hardstand areas and are trying to avoid increasing the peak flows from the site (this would result in having to increase pipe sizes in the downstream network. The reuse will be for irrigation of the landscaped areas

During the rainy season, water will not be needed so much for watering landscape, if that is the plan. During a drought, the tank will empty quickly

This is true, and something i questioned as the climate here is wet winters and dry summers. I researched it as it is now a requirement for new buildings in Auckland (reuse should also be used in the house) and while it is true and does not result in a major water solution for the individual, if multiple houses do this it puts less pressure on the dams during the wet season and allows them to fill up fully. this then results in them being able to supply water over the dry period. I believe they are currently not able to reach 100% full before the summer period.
For this development it is just a minor cost saving at the start of summer, when the tank is empty they will switch over to mains supply - they had to install the pipes and tanks for the attenuation and increasing the tank size for a bit of storage was not a huge increase

yep, thanks

Use closed pipe flow formulae to size the pipe. Size it bigger than requires. Use energy grade line and heading lose due to pipe friction to find the flow rate.

That is pretty much how i did it, ended up well below hgl