Rainwater system install - Perth

Hi
After getting a quote for $7000 to install a rainwater pipework system I have decided to install it myself and would like some advice or improvements for my current plan.

Details are:
Location - Perth. 1:20 ARI is 140 mm/h which is based on an average rain intensity of 2.33 mm/minute over a 5 minute duration
Roof Area - 420m2 Main roof has a 25 degree pitch. Veranda has a 18 degree pitch
165,000L water tank located 20m from the house

My plan is to connect 16 downpipe outlets to 75mm PVC that connect to a 2 x 100mm PVC ring main. 8 downpipes connected to each ring main.
The ring main will then connect to 2 x 100mm DWV piping buried underground and will be the inlets to the 165,000L water tank.
I also plan to put sediment traps and inspection fittings in the system but am unsure where the best location for these.
I have attached a layout drawing to help


Any help is greatly appreciated
Any help is greatly appreciated.

Very good information.

A problem with wet systems is that the pipe at the head of the system most often builds up sludge due to the flooded pipe having a slow flow rate and this is more pronounced when the end downpipe harvests a small roof area. This will be even more pronounced with your underground wet system having 100mm pipes all the way. I know that there must be a reason why you want to use 75mm downpipes but maybe consider harvesting larger roof areas at the head of the wet system and using a smaller pipe at the head of each run. This is a hard one because special DWV pipes with vortex inducers that agitate and move the bed load are not made.

Unfortunately, the initial use of a smaller pipe at the head of the system which would be better for the health of the system may not be allowed by local regulations which are often not well thought out..

EDIT: Can you check your local regulations re the above as I believe it is allowed otherwise.


Why not merge to a single 150mm pipe that holds twice the volume of a 100mm pipe and which usually costs a bit less than the equivalent two 100mm pipes?

I would also run a minimum 40 or 50mm branch line at the tank to a low restriction inlet fitted about 200mm above the bottom of the tank. This will lower the height of water retained in the wet system from the height of the vertical riser to the height of the water in the tank plus it would also assist the system's discharge rate as well as oxygenate the higher density water at the bottom of the tank.
EDIT: Fitting a low restriction inlet as suggested above cannot be done as there is insufficient head on most downpipes to enable leaf diverters to be fitted to satisfy the requirement for water to pass through < 1mm filter mesh before entering a water tank.

Sediment traps work great and also save water by reducing the amount of water that is flushed. They need to be fitted to a non turbulent section of pipe past the last downpipe or fitting and this distance is usually 4-5 metres. Only suspended fine material will reach the tank and will look like talcum powder.

A Homeone member who installed sediment traps on a 4 tank rainwater harvesting sysyem in QLd posted a 1 year system summary that is worth reading.

viewtopic.php?p=1146540#p1146540

Rainwater is naturally acidic. I would avoid using copper pipe and also put a nylon meshed bag with limestone chips into the tank. The bag will need to be tethered to a buoy for retrieval.

Make sure of the available head Vs required flow rates during heavy rain.

Don't option the tank's top meshed inlet above the outlet that supplies the pump. Even though you will have very little fine debris on the bottom, it is still a good idea to keeep any falling water away from the area that the pump draws its water from.

The best quality water is drawn from 200mm below the water surface. The Homeone member in the thread linked below had his own floating filter inlet made together with a protective collar to prevent it settling on the bottom of the tank. The main discussion starts in the post linked below.

viewtopic.php?p=1146540#p1146540

Should the tank ever fill and you have heavy rain, the overflow will need to be capable of discharging the inflow to prevent overtopping. I can also advise on how to supercharge the overflow pipe's discharge capacity.

I hope that this helps.

Thanks for your reply H2O. Great information that will help me immensely.

The system will have varying gutter heights. Available head of the system is between 300mm and 650mm. Does this impact the system in terms of preferential flow into the ring main?
The gutter pop outlets are 75mm hence the use of 75mm PVC. Would the use of 75mm pipework all the way to the tank help with sludge build up due to increased velocity? If system pressure is constant than velocity increases as pipe diameter decreases. The negative impact however will be reduced flow. Is this correct?

The tank has 2x90mm overflows installed
Thankyou for the links I will check them out.

Are you measuring the available head from the gutter's sole to the tank's top meshed inlet opening or as a minimum to the half way height of the wet system's horizontal pipe above the tank that discharges to the tank's top inlet?

The effective head will be 300mm which isn't much at all and I would definitely plumb the two 100mm pipes to a single 150mm DWV pipe to reduce the friction loss over the 20m distance from the house to the tank. When the additional friction losses for the vertical riser plus the elbow at the bottom of the riser are added as equivalent pipe lengths, the (nominal) 100mm and 150mm pipes friction losses over the 20 m distance becomes about 27 m and 28.5 m respectively.

During a minimum 1:20 ARI rain intensity, your roof will yield 420 (sq m) x 2.33 mm/min which is 978 lpm but you need to factor a rain intensity greater than this. Just to put this in perspective, if nominal 100mm and 150mm DWV pipes had a gravity flow of water with a 0.3 m head over their respective distances of 27 m and 28.5 m, the 100mm pipe would flow at about 574lpm and the 150mm pipe would flow at about 1,512 lpm. As can be seen, 574 lpm x 2 which is 1,148 lpm falls short of 1,512 lpm and is only a 17% safety margin over 978 lpm which is the minimum average yield per minute off your roof that qualifies for a bare minimum1:20 ARI classification.

BUT it must also be realised that pipes lose pressure along their flow path and two nominal 100mm DWV pipes or a nominal 150mm DWV pipe will not have 0.3 m head at the head of the 20 m pipe run to the tank as the head will be generated by the downpipes at the house. In other words, the flow rates generated by the examples above will not be achieved by your pipes. The short video below demonstrates pressure drop along a pipeline.

https://www.youtube.com/watch?v=_hSL9_eo4n8

Unfortunately, the flow rate through a connecting 150mm pipe cannot be supplemented by a 50mm low restriction inlet due to there being no leaf diverters fitted to the system. Fitting leaf diverters would decimate the available head.

The above paragraph was edited due to the original advice to also use a 50mm low restriction inlet to supplement the flow rate cannot be done because water must first pass through filter mesh but leaf diverters cannot be fitted due to the limited available.head. There may however be an opportunity to fit a large inlet valve a few hundred mm above the bottom of the tank as the water level in the tank will always be lower than the top of a vertical riser above the tank which gains additional head which might allow leaf diverters to be fitted.

There is also the option of fitting an inlet to the tank's wall at a height higher than the two overflows iand satisfying the need to fit a filter mesh by using the product linked below.

https://rainharvesting.com.au/products/tank-overflows-screens/insect-proof-tank-overflow-screens/mozzie-stoppa-easy-clean/

I would also use a 150mm x 100mm 45 degree junction to seamlessly merge the two 100mm pipes to the 150 mm pipe. A 150mm x 100mm pipe reducer would also be used in one 150mm socket.

You couldn't have 75mm pipes all the way to the tank because they would require a significant head to generate an adequate flow rate only made possible by an unrealistic and unobtainable velocity. .

It is unfortunate that rainwater harvesting regulations are written by novices who take counsel from the unknowing who only reflect on stormwater flow characteristics. Water in a drain pipe flows along the curved bottom with a good turbulent flushing velocity generated with minimal slope but charged (wet) system pipes are flooded and the greater volume of water flows fastest through the core whereas the water on the walls (which includes the bottom of the pipe) is mostly stationary. Two different flow regimes, two different outcomes.

Ideally, the pipe at the head of the system would be smaller than 100mm as it will have better flushing efficiency but flushing efficiency will be an unviable expectation due to the small roof areas.

The discharge capacity should/must exceed the inflow capacity at all times and two 90mm overflow outlets haven't any hope of keeping up with the tank's inflow capability.

A horizontal outlet discharges at different flow rates depending on the depth of water above the outlet's invert.

A 90mm pvc stormwater pipe has an internal diameter of 86.2m but the flow path thrrough the fitting is more like 84mm. With 100mm of water above the invert, each outlet will have a discharge rate of around 225 lpm BUT this is through an unmeshed opening. Water tank overflow outlets are meshed with an open surface area of around 52-55%. Fortunately, the wire mesh used is round which has a less restricting bellmouthing affect than flat surface wire but the flow restriction will still reduce each outlet's discharge rate to around 185 lpm with 100mm of water above the invert.

Torricelli's Theorem is the go to formula for discharge rates through a round hole in a vertical surface.

There are a few methods whereby the discharge rate can be considerably improved but not to the extent where a flimsy 90m pvc stormwater pipe will discharge at around 500 lpm which would be crazy to do anyway. You could however make two 100mm outlets acceptable for purpose with a few tricks that I can impart.

Because overflow pipes never have less than 70-75% air unless there is a restriction in the pipework, you can use one overflow pipe to service the two outlets.

How much mitigation will there be above the overflow inverts?

Your knowledge on this subject is amazing. Thanks for sharing H2O
Your calculations sure do embed what I thought were small changes impact greatly on system efficiency. I will go away and read this a few more times and come up with a plan that I hope you can review in the not to distant future.


Thanks

Despite the clever merging of the two overflow pipes to a single vertical pipe, the last photo posted also shows several common installation and design flaws. Who wants to have a go at picking what they are?

Hi H2O
Been offline for a few weeks. I have asked the builders plumber for advise and he is telling my system design is way oversized and complete overkill. His idea is 75mm pvc dowpipes into a single 90mm pvc ringmain then into 100mm DWV pipe for the underground connection to the tank. His reasoning is higher velocity and turbulence with less losses on first flush. So much conflicting info I am unsure now.
Can I please ask, If you were to install the system for my home above what would your recommendations be

I have just done an approximation of your roof area. Is the 420 sq m you stated the 'plan' area (including veranda roof) or was the roof catchment area factored to allow for wind driven rain on the roof slope? My approximation of the plan area comes up short of 420 sq m. The roof plan area is used to calculate the amount of water draining off the roof during varying rain intensities.

Can you confirm the roof plan areas (as seen on the plan from above) thanks?

The builders plan stated 420m2 includes all verandas. I checked the builders Calcs and came up with 413m2.
There was no factor for wind driven rain.

Thanks

Are you able to upload another house/drainage plan that is cropped to the downpipes (so it is larger) and also number the downpipes so I can explain a few things much easier? Start at the bottom left and number those DPs 1-4 and the DPs further down on the left 5-8.

Also and most probably just for interest, can you also show the approximate maximum head of each downpipe?

Your plan has areas of concern that I want to point out and also offer suggestions on design improvements.

The plumber's idea of using two 100mm DWV pipes over the final 20 metres to the tank is not good as I have already explained in post 4 re flow rate Vs available head.

Hope this can help

Appreciate your assistance mate

The plumber only wants to use a single 100mm DWV pipe to the tank
Downpipes
1-4 / 15 &1 6 approx 400mm head
5-12 approx 650mm head
13 & 14 approx 300mm head

Unfortunately, this is yet another sad indictment on the woeful lack of knowledge of rainwater harvesting, storage and hydrostatics in the plumbing industry due to plumbers not being trained in rainwater harvesting best practice or any practice for that matter.

Thanks for posting the diagram so quickly.

I'll go over the new diagram now and reply in the morning. It has been a long day.

I'll just go through a few best practice parameters that the system should to be designed around so you have an understanding of my suggested alterations.

PIPE SIZES:

Subsurface stormwater pipes and charged (wet system) pipes have different flow and flushing characteristics. It is best to use small pipes at the head of a wet system to promote higher velocities along that initial run provided that the carrier pipe is not smaller than the downpipe but this should not apply if leaf diverters are used. Note that I can find no regulatory reference re the necessity for a charged system carrier pipe to be as large or larger than the downpipe

I have once again looked through Parts 2 (BCA) and 3 (PCA) of the National Construction Code (NCC) plus AS/NZS 3500.1 and AS/NZS 3500.3 for regulations re minimum pipe sizes for rainwater harvesting charged systems but the only reference found was in AS/NZS 3500.3 Section 7 Surface and subsoil drainage systems - Installation which states under the heading of Connections to pits and arresters, Minimum diameter for single dwellings in rural areas and residential buildings on urban allotmentswith areas less than 1,000 sq m, the minimum pipe size shall be DN90.

I also spoke with a knowledgeable plumber at the VBA who said that subsurface rainwater charged system sizing calculations are not covered under the plumbing regulations and so they don't check for this. This contradicted what I was told several years ago by another VBA plumber who said that the minimum size was 90mm but I believe the more recent advice. Unfortunately, a much needed separate category for charged rainwater harvesting systems is totally neglected as is the training for plumbers.

EDIT: (Clarification).
The regulations cover roof drainage. surface drainage, sub surface drainage, stormwater drainage and pumped systems.

ROOF DRAINAGE:
This calculates the maximum roof area that can drain to varying gutter cross sectional areas and downpipe sizes. A 1:20 year regional storm event that is based on an average rain intensity over a 5 minute duration is used as the base for eaves gutter calculations and a 1:100 year rain intensity is used as the base for box gutter calculations. Roof drainage regulations only cover from the roof to the downpipe.

SURFACE DRAINAGE:
Surface water drains via grates set in paths etc and via surface gutters. This water must divert to a silt pit before connecting to a sub surface stormwater pipe.

SUB SURFACE DRAINAGE:
Agricultural (Ag) pipes must also divert to a silt pit before connecting to a sub surface stormwater pipe.

STORMWATER PIPE:
These are laid wth a minimum gradient that varies with the pipe size and collects the roof, surface and subsurface water. Because pervious and paved surface areas vary as well as the drained subsurface areas, calculations must be made to design the stormwater pipe size to cope. Stormwater pipe systems operate with gravity flow and should not retain water between rain periods. This also provides regular flushing every time it rains.

RAINWATER HARVESTING CHARGED SYSTEMS:
These are not covered under stormwater drainage because:

• A subsurface charged system cannot accept surface and subsurface water nor should this water be diverted to a tank in any case.

• Storm water and water in a charged pipe have different flow characteristics. For example, a gravity drained stormwater pipe at the head of the system will drain with an agitated velocity with a small amount of flow whereas a full flow charged pipe has a much slower flow velocity for the same amount of inflow. Using large pipes at the head of a charged system builds up sludge with accompanying bacterial activity that depletes the waterr of oxygen when water is retained between rain events. This is the biggest problem with charged systems.
• Stormwater pipe sizing regulations cause serious sediment build up in charged rainwater harvesting pipes.
• 90mm pvc stormwater pipe is not suitable for use in a charged system.
• Additional regulations need to be applied to the best practice use of flexible couplings on vertical risers and overflow pipes.

The system you have designed has crud accumulation areas at:
DP1. 12 m to the nearest tee.
DP5. + 6 m to the tee.
DP6. + 6 m to the elbow.
DP13 6 m to the nearest tee.
DP16. + 6 m to the nearest tee.

75mm pvc stormwater pipe has an internal diameter of 71.8mm (4.05 L/metre) and walls 1.6mm thick whereas the 80mm DWV pipe has an internal diameter of 76.2mm (4.56 L/metre) and walls 2.9mm thick. I don't know of an adaptor that will connect 75mm stormwater pipe to 80mm DWV pipe. Either of these pipes would be better to use than the thin walled (1.9mm) 90mm pvc stormwater pipe that holds 5.8 L/m. I also haven't checked whether a 80mm DWV pipe could be sealed to a 75mm pop.

MAJOR TURBULENCE AREAS (Pressure loss).

These are the areas where the carrier pipes 'balance', no doubt to compensate for wind driven rain causing a rain shadow on one side of the house roof. The balance fitting served by DP 13 is particularly worrisome as DP 13 is the last bastion for a pressure boost during heavy rain.

PRESSURE LOSSES THROUGH PIPES.

I linked a video in an earlier post that demonstrated dynamic pressure losses along a pipe but downpipes diverting to a charged system largely offset these losses by mimicking water towers boosting a carrier pipe's pressure by releasing pressure via a water column that is positive to the pressure in the pipe. One day I will make a video that demonstrates this.

DWV 'tees' are called junctions and the branch has a small bias towards the direction of flow through the run. This makes a DWV junction more efficient than a stormwater tee but 45 degree DWV wye junctions used to connect to the carrier pipe are much more efficient than DWV junctions.



The problem is obviously with DPs 13 & 14.

During heavy rain, if the greater amount of water in the system cannot all be delivered to the tank, DPs 13 & 14 become the next points of least resistance for the excess with obvious consequences. One alternative is to have both of these downpipes drain to a separate tank but first look at my alterations to the original plan.

RED: 75mm stormwater or (preferably) 80mm DWV.
PURPLE: 100mm DWV
BLUE: 150mm DWV.

DPs 2 & 3 provide recharge and flushing turbulence to the long pipe served by DP 1.

The single 100mm pipe connecting to a single 150 mm pipe after the first 8 DPs accomodates flow vagaries caused by wind driven rain.

Your minimum qualifier for a 1:20 ARI is an average of 2.33 mm/minute over a 5 minute duration. 420 sq m x 2.33 mm/min = 978.6 mm/min but because we need a safety margin and using the equivalent of a 225mm head where DP13 connects to the 150mm pipe (this is also assisted by DPs 10, 11 and 12) and calculate flow by using a total friction loss of 28 m of equivalent pipe length for the pipe run to the top of the tank, we get a flow rate of 1374 L/min. NOTE: The 225mm head example does not equate to a 225mm head at DP13 as it will have have friction losses. Using 45 degree junctions with DPs 13 and 14 at the 150mm pipe intersect is advised.

On the other hand, a single 100mm (104mm ID) DWV pipe with a (fanciful) hydraulic head of 300mm at a point where it will have an equivalent friction loss pipe length of 28 m to the top of the tank will provide a flow rate of 562 lpm.....of course a 300mm head at that point will not be achievable.....but if the plumber was to conjure up a 1 metre head, he would get 1,077 lpm. I would not consider 1,077 lpm to be a suitable safety margin plus having a velocity of 2.113 metres/sec is not wise.

I would also fit the sediment trap about 4 metres past the last junction. These work great but I use an invert taper instead of the pipe reducer now. The inspection opening (IO) is just in case something gets lodged in the entrance to the small flush pipe. The smaller pipe increases the flush velocity and wastes less water. All off grid properties should use these simple traps.

Thankyou so much H20
Your time and expertise is greatly appreciated. I will pass all this info to the plumber and report back

You know more than the plumber does about this subject.

Are you trenching and fitting the pipes or is the plumber doing it?

If you use mosquito proof leaf diverters on downpipes 5-12, you could also use 65mm or even 50mm DWV pipe for the initial pipe lengths as no large objects will reach the subsurface pipe.

The 65mm pipe has an internal diameter of 63.6mm that provides a volume of 3.2 litres per metre which would require an unobtainable 77 lpm to achieve a velocity of 0.4 metres per second but its use would be better than the 80mm pipe.

The 50mm pipe has an internal diameter of 51.5mm that provides a volume of 2 litres per metre and so a velocity of 0.4 metres per second is attained at 48 lpm. This equates to a rain intensity of 1.8 mm/min for your average size roof area.

Standard gutter flow along the bottom of a stormwater pipe requires a design minimum flushing velocity of 0.7 metres per second but larger heavier objects fall down open downpipes not fitted with mosquito proof filter mesh. The video linked below may interest you as it demonstrates different types of debris transferred along a flooded (charged) pipe line. Unfortunately it doesn't use dye to verify laminar flow nor use fine sand spread along the pipe but it is a good video nevertheless.

https://www.youtube.com/watch?v=c1xX90ZfBj4

The standard mosquito proof leaf diverters use 955 micron mesh but I have a as yet unreleased compact leaf diverter with intended 500 micron filtration. I have previously mentioned this in the past on this forum and while a 2016 release was the intended goal, the filter screen which includes surface effects can't be made by injection molding. Right now I am waiting for new product developments in SLA 3D printing that include a suitable UV stabilised resin and while the technology is advancing at a rapid rate and there have been many new resins released lately, I am still waiting and commercialisation is unlikely to happen soon unfortunately.

The only current leaf diverter that I recommend is the ICON Leaf and Debris Controller. Being larger than others, it loses about 70mm more head when compared to the market leader but this would not be an issue with DPs 5-12.

The trench is dug and I plan to do the pipe installation myself. The roof has stainless gutter guard installed with access hatches for cleaning and maintenance.

Id did have a thought of also installing some mosquito mesh cut to size and install it inside the gutter over the top of each DP outlet, essentially giving me a dual barrier for debris.
Have you seen this done before?

Access hatches are essential yet commonly overlooked in the design. Well done.

No problem then if you use subsurface pipe smaller than the downpipe for all initial runs and then upsize just before the second DP on the run.

I don't recommend it. It isn't needed for your set up and is only asking for trouble. It must also be remembered that most gutter debris is flushed from the gutter during heavy rain.

It reminds me of the early leaf diverters that had a course outer mesh (about 8mm apertures) and a small inner mosquito proof mesh. They constantly blocked and overflowed, particularly during heavy rain,,,a bad idea proven!

Thankyou H2O you are a gem