Rainwater catchment in tropics (New build)

Hi Leon.

I should have the time to answer this tonight, I couldn't log in for 2 days but all is fine now.

Thanks for all the vital information, it certainly makes it easier to answer your queries.

I see that you are near Batangas, did the earth move for you too?

Your posted information shows that you have a very good understanding of best practice. Best practice isn't involved nor should it be expensive, in fact it is frequently cheaper.

Your 1:20 ARI is severe but the Phils get typhoons every year and their size, intensity and frequency is increasing. The waters around the Phils are warmer now but not spectacularly so like many other regions are but with the atmosphere now holding about 7% more moisture and twin and triple typhoons no longer rare in your zone, the Phils will face many challenges in the coming years.

A few notes.

A 1:20 storm's figures are based on an average intensity over 5 minutes, that intensity is not transposed to an expected hourly figure.

The roof multiplier factors wind driven rain but the total roof area does not collect more water than a flat roof would. Depending on wind direction during severe winds, a lot of the water draining from the upper roof will probably never reach the lower roof and a lot of water falling on the lower roof would also be blown off. The real test will come with heavy rain when the wind has dropped.

I can see in the house photo that you have a large gutter. If your roof drainage installation is like here in Australia, the 77mm hole in the gutter is for the gutter 'pop' which is also referred to by various other names. Your 90mm downpipes will attach to the pop, usually by pop rivets.

In Australia, 90mm round downpipes use to be rated at 3.5 litres per second (lps) but manufacturers now usually quote 4.2 lps although I have also seen claims of 4.5 lps. Given that a pop drains the gutter and not the downpipe, we should look at AS3500.1 Section 8 which quotes a vertical 75mm pipe with a 50mm head as 2.63 lps and 3.22 lps with a 75mm head when draining a storage tank. That same pipe/pop will however drain more water from a sloped gutter because of the flow turbulence within a restricted space. Having a good gutter slope is important because strong winds can impact on the gutter's flow plus the gutter's high points determine the areas of roof that drain to each downpipe.

Abutting walls also need to be factored for additional wind driven rain.

A (nominal) 150mm DWV pipe holds about 18 litres per metre, this equates to a flow rate of 1,080 litres per minute at a velocity of 1 metre per second. I would be looking at either using a calming inlet or preferably diverting into a settling tank via an inlet fitted a couple of hundred mm above the bottom of the tank at the inlet's lowest point. This would also give you more head but if either alternative is used, then you really need mosquito proof leaf diverters fitted to the downpipes but finding a typhoon proof leaf diverter could be a real challenge.

Large tanks have some mitigation above the overflow but you would still need a massive overflow capacity and the best way to achieve this would be to install a vertical bell mouth on top of a 150mm diameter elbow extension coming off the overflow pipe. A bell mouth can generate a syphon but that won't happen in your situation but a degree of bubble flow which is the third stage of syphonic purge would happen during extreme inflows. Having a larger bell mouth and bubble flow substantially increases the overflow pipe's flow capacity but bubble flow will also ensure that full syphonage with its associated negative atmospheric pressures within the pipe are not realised.

 Like  ⋅  Add a comment  ⋅  Pin to Ideaboard  ⋅  Share



 Like  ⋅  Add a comment  ⋅  Pin to Ideaboard  ⋅  Share

Re pumping water to a second tank, I would have either a fossil fuelled or a solar powered pump on hand rather than rely on grid electrical power. If massive rain was forecast, I would fill the second tank prior to the event and also mitigate the settling tank. I would not consider a 50k settling tank and would have 100k as a minimum. You also need to ensure that the high velocity overflow did not cause problems at the discharge point.

The two calculators below should be of benefit to you. PVC pipe has a roughness coefficient of 150 and you must also allow for friction losses through fittings (elbows etc) by adding an equivalent pipe length for each fitting.

http://www.calctool.org/CALC/eng/civil/hazen-williams_g

http://www.nationalpump.com.au/calculators/friction-loss-calculator/

A sediment trap is of benefit if you have a lot of dust blown onto the roof and it will take 3-4 metres for the majority of suspended sediment to settle as bed load. You would have to reduce the fitting to 100mm and then to 40mm or 50mm. If you have a sediment trap and also a low restriction inlet, you need to close the inlet valve and use the head provided by the downpipes. With your large roof areas and massive annual rainfall, minimising the volume of discharge from the sediment trap would obviously not be an issue.

Bed load is mostly stationary and it clumps in small 'colonies', only breaking off at the front once the 'colony' becomes over populated, the water becomes turbulent or the flow rate increases. For fine detritus, the critical flow rate is about 0.6 metres per second and higher for heavier materials. Because a pipe's laminar flow is almost stationary at the boundary layers, fine particles actually have more resistance to flow than do chunkier materials that protrude into the faster flowing streamlines closer to the core.

The video below is worth watching, note how the turbulence created by the elbow allows the larger heavier particles to travel up the vertical pipe but it is disappointing that the video didn't also show fine particle clumping.

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

Having a settling tank's pump's draw outlet a minimum of 300mm above the bottom of the tank at the outlet's lowest point will supply very good quality water to the pump. A floating suction outlet hose that takes water from a few hundred mm below the water surface allows you to fit the valve a lot lower and will also provide the best quality water.

http://www.crystalclearwater.com.au/waterboy/

 Like  ⋅  Add a comment  ⋅  Pin to Ideaboard  ⋅  Share

The ARIs are a measure of rain intensity and you are correct with your figures apart from the wrong application of the multiplier when determining the total amount of water harvested. The 29.84mm is the total over a 5 minute duration.

If you lived in Adelaide where the 1:20 ARI is 120mm/hr (an average of 2mm per minute over a 5 minute duration), you could experience a 1:20 ARI yet only have maybe 11mm of rain for the hour. It does create confusion when the ARI is expressed as an hourly figure which is why I try not to forget to explain that the figure is based on an average over a 5 minute duration.

I recommend that people do all hydraulic calculations as per the 1:20 ARI and then add a safety margin but that is my judgement and recommendation, nothing official.

Some peripheral information.....

The 1:20 ARI is used here in Australia for eaves gutter and downpipe roof drainage compliance, there are no similar regulations for harvesting rainwater to rainwater tanks but the authorities often erroneously and unknowledgeably state what must be done. As an example, I am in Victoria and the Victorian Building Authority (VBA) has a technical solution sheet that states that the tank's overflow pipe must be the same size as the inflow pipe. The diagram was inherited from the former disgraced Victorian Plumbing Industry Commission.

http://www.vba.vic.gov.au/__data/assets ... -Tanks.pdf

A charged wet system pipe could be connected to several downpipes and the inflow capacity during heavy rain could easily exceed a tank's overflow capacity...which often happens. During the Millennium Drought, the Brisbane City Council (BCC) mandated that rainwater tanks must harvest 50% of the roof area or 100 sq m, whichever was the greater. Brisbane's 1:20 ARI is 240/250mm per hour yet there was no information given about tank overflow capacities and when the drought broke, over-topping tanks became a big problem.

Water discharge flow rates through a tank's horizontal overflow pipe is subject to Torricelli's theorem and the calculations should have immediately raised concerns about inadequate overflow capacities yet most tank overflow outlets also have mesh that has not much more than 50% open area that restricts the flow. The BCC gave no advice about the hydraulic benefits of removing the mesh and installing an external flap valve or any other recommendations...nothing!

If a water tank is connected back to the stormwater via the overflow, the water tank is effectively part of the stormwater system but this often escapes the attention of many plumbers and other industry 'professionals'.

Tanks of 100kL size are very expensive here. First quote I got was about AUD $35k.

My builder has suggested (he'd like his guys to remain in work so eager to do some more jobs for me) that they build it using concrete hollow block and add a non-toxic coating to prevent leaking. It'd be a square shape of about 7m x 7m x 2m. I'm happy for them to do this but wondering if anyone can point me to some good info about building tanks with concrete, especially from a civil perspective. I want to be sure that what they build will hold that amount of water with no problems...

My main worry is the tank won't be very earthquake proof, but considering the amount that a proper tank is going to cost me, I'm willing to take the risk. I'll situate it at the lowest point on the property so if it does happen to burst I think the house and garage will be safe.

Round tanks are inherently stronger than square tanks. You may need to talk to an engineer if you are considering that option.

They build the bigger round tanks on site here and cost nowhere near the amount you have been quoted. You tube should have examples.

22,500 litre tanks are popular here and are made en masse. Do they have a rainwater harvesting industry and manufacturers in the Phils?

What is FRP?

I've seen mentioned a few times that the tank inlet should be at least 300mm below the bottom of the leaf eater/ rainhead. If I calculate friction losses for about 70m of pipe I get 600mm. Do I then add that to the aforementioned 300mm? Or is the aforementioned 300mm just a rough figure people throw around to take into account friction losses on a simple setup?

Thanks SaveH2O

Unfortunately the builder already has the 90mm DWV downpipes on site, but I might see if I can get him to change just a few to 100mm (the ones that will get the highest flow rate) as well as increasing the size of the "pop".

"reinforces that you have a good understanding of proper design" - most of this comes from reading your posts here and on the ATA!! I feel like I'm back at school ! It took me a few weeks of procrastinating before I fully dived into the rainwater design, was one of my last big worries with the house build. Now that I've sunk my teeth into it I feel a lot more confident about doing it.

Thank you heaps for your assistance - it's great to get so much professional advice from someone on a forum, it is very much appreciated.

The standards were updated in 2015 but the relevant information is the same.

A tank's mitigation capacity is critical when there is a statistical probability of the inflow exceeding the overflow at times when the tank is full but the mitigation compartment of course increases more slowly than the inflow.

The other thing to remember is what I mentioned in an earlier post about preparing for a severe rain event by filling the second tank and emptying the larger settling tank. If there was a drain valve on the bottom that ran to a suitable discharge point, then this could be left open and the inflow would also give the tank a good clean. Once the severe rain intensity eased off, the drain could be closed and the recharge would be pristine water. The drain outlet may need to be only 40-50mm maximum.

Another consideration is the mosquito proof mesh that should be on the overflow outlet and I assume that you will also have this. BTW, mosquitoes do go up pipes.

Mosquito proof metal mesh has a very low % of open area...barely 50% and this restricts the flow. The actual flow restriction is complicated because round wire effectively creates micro bell mouths but the restriction (flow rate reduction) is significant regardless.

Using an external flap valve in place of mosquito proof mesh is very highly recommended for extreme overflow situations. I haven't seen the (nominal) 150mm flap valve that I have linked below but our stormwater and DWV pipes both have outside diameters of 160mm, the difference being that stormwater pipe contains a higher % of regrind in the core and the two wall thicknesses are 3.2mm and 4.2mm respectively.

https://plumbingsales.com.au/pvc-flap-valve-150mm.html

There is a trend now here in Australia for some tanks to be manufactured with very high overflow outlets that provide negligible mitigation to increase the tank's capacity as a sales advantage but if the public was aware of the shortcomings of doing this, would they buy them? Unfortunately, the public is treated like mushrooms...kept in the dark and fed bs. Who would buy a tank with a side 90mm meshed overflow outlet if they wanted to harvest more than one downpipe and were told that the overflow's discharge capacity was <120 litres per minute?

A note here about vertical drain pipes. When you pull the plug out of a basin or bath etc, you will see an annular vortex. This is not air escaping from the pipe, it is air being drawn into a pipe, the reason why downpipes never contain less than 3/4 air unless blocked or flow restricted such as a charged (head restricted) system. The vortex is also the reason why water swirls down a round downpipe.

Increasing the orifice diameter (adding a bell mouth) increases the pipe's capacity because it floods (adds a higher % of water to) the smaller pipe. If you had a 150mm pipe with a 250mm bell mouth, you will have a massive increase in overflow capacity. Combined with a reasonable mitigation capacity that you would most probably (and unintentionally) have anyway and not forgetting that mitigation also increases the head, you will not have any worries. A 150mm to 250mm diameter increase is a much lesser % increase than the bell mouth in the photo I posted and there is no reason you couldn't go bigger than 250mm.

You might be able to modify the end of the pipe to a bell mouth shape by applying a heat gun to the end of the pipe and shaping the softened uPVC into shape. I have never done that to this size pipe but it should work, the trick is to keep the heat source moving, don't keep it in the one spot. There are some You Tube videos showing it being done to smaller uPVC pipes.

Re question 2, you really don't want to create even a short duration syphonic effect down a vertical pipe exiting at the bottom of the tank. A standing pipe would also be hard to brace. Just keep it simple and remember that it is insurance.

Both.

The bell mouth could potentially provide enough water and turbulence in your system to generate an occasional syphonic pulse down a longer vertical pipe. If so, the sudden acceleration and the resultant vacuum break would be undesirable due to 'kick' transferring via the elbow at the bottom of the pipe to the tank’s outlet unless the pipe was somehow braced. I have never conducted tests of this scenario but I wouldn’t chance it.