Siphoning Well Water

This looks like a correct design. Leakproof joints will be necessary to keep from losing the siphoning action. The non-return valve will need to be of a type with low pressure drop.

There are some people here that know the formulas for lift, etc..

I'm not a formula guy so I would probably buy one of these to fill a 55 gal drum at the top of the hill with a hose attachment at the bottom of the drum to get water below. I know that would work. Just a simple hand pump.

You could even put a T on the outlet hose from the drum and switch from straight siphon to using the drum water.

hi ddk,

your idea is great, in fact i also used that method in draining my aquarium tank, however in the case Rtrak, foot valve and priming water port will be needed for every day usage..

in our case our siphon pump will be online from 7am to 11 pm and after that it will be off line...

please refer to my first and second posts for reference.

Mike

I agree: GA. Keep it simple. The whole pipe full of water will weigh about 150Kg, so he should be able to drag it bit by bit to the outlet end.

I agree, I would also add some type of a large area screen on the hose inlet so the hose doesn't get clogged. Simple and cheap.

Your idea sounds workable.

A non return valve will reduce the flow and if not seated properly make priming impossible.

Consider a hand operated diaphragm pump to prime the system.

Firstly many thanks for taking the time to respond, it's much appreciated, sometimes it's like being in a technological vacuum here! I have already got a non-return valve with (basic) filter and the fittings to connect it easily to 32mm tube as I use this size to feed the house via a pump, filter, controller etc. I may have to adjust the spring load although I can blow it open and I'm no Luis Armstrong! I like the idea of the simple lift pump to prime it too, a simple bypass leg with valve would allow siphoning to continue once primed and it'd all be out of site under the Mimosa tree that shades the mine. I had really hoped to be able to install the setup, open the valve and transfer the water, at this time of year I need about 4-5 cu/m (4-5000 litres) per day. My lower mine (feeds the house) holds more than 100,000 litres and replenishes almost instantaneously but I'd have to pump this up about 5 metres and over 100 metres on the level taking about an hour and a half with the house pump maybe my pockets aren't deep enough? Would be nice to use gravity and be more environmentally friendly although they've never heard of water conservation here!

I'm not sure you need a non-return valve. I would put a valve close to the discharge point, and after the valve a small tank, or a upturned bend with free outlet, to make sure the outlet of the pipe stays drowned (to avoid air going back up). Arrange your fill valve (what you called a purge vale in OP) so that when you add water it flows down to the outlet, not into the mine. Fill the small tank to overflow. Close the discharge valve. Fill the pipe via the fill valve, and close the valve. If you then open the discharge valve it should start to syphon, and the air purge out with the water. As the "downward" leg is much longer than the leg in the mine, this should work. If it doesn't you can add a non-return valve.

BTW your 500m elevation is equivalent to about 0.6m water, so won't make much difference. Atmospheric pressure is still about 28 ft, well above your 8 ft lift.

Cheers..........Codey

Non-return valve with (basic) filter and the fittings will only add pressure drop and reduce flow rate. As your differential head is good, you can afford to use it to get required quantity of water.

Flow rate at siphon discharge point is proportional to A(2gh)^1/2 where, A = Area of tube cross section, g = Acceleration due to gravity and h = differential head between mine & field. As all the three parameters A(for 32mm tube), g and h are constants, only frictional losses can be varied by providing valve, filters etc.

Siphon should work well (no pun)

At 500 meters atmospheric pressure is 13.8 psi or close to 32 feet. So a siphon will work on an 8 foot lift. Pay attention to friction loss so you can keep the siphon going. Friction loss will be added to the total head. 32 mm pipe is same as 1.25 inch pipe and relates to about a 40 gpm discharge. Your point of discharge should always be lower than your intake or maximum drawdown. So I would suggest the point of discharge be >32 feet (this is the maximum siphon depth possible at this altitude).

Make sure you put the suction end of the siphon in the well deeper than the 32 feet so you won't be able to siphon below this point and break siphon.

Starting the siphon can be done by either pumping water backwards throug the pipe back to the well. Once the pipe is filled with water and the pump is off, water should flow from the pipe via siphon action.

Another method that may work is to fill the well with water to overflowing with the siphon pipe in the well. This will work as long as the water does not run out the bottom of the well faster than you put it in.

If that fails, you can jig the pipe up and down and have someone put their hand over the discharge end to act like a valve. The hand needs to close the discharge when you are on the upstroke and open on the downstroke. If you make joints in the pipe you will have to ensure no leaks whatsoever. I would suggest a continuous pipe with no connections to the point of discharge.

At your downhill end you will need to make the end of the pipe turn and face "up". If you don't, then there is possibility for air to run back up the inside and kill the prime in the line.

All else looks good. 8' lift should be OK, tube diameter seems OK, fall to other side seems OK. Flow rate will be whatever it is. You will not run "sprays" on the pressure, but there will be good flow.

Suggest also to use ball valve or cock instead of normal tap. The water pressure may not be enough to break the washer away from the seat if it were overtightened.

Good luck with the practical stuff.

once flow is established ad ddk and me, even if the end is kept not facing up, no water will enter in. Water will enter in and stop the flow, only if the suction end of the hose in the well comes out of water even for a second.

The answer #4 has a good idea. Get a hose. (or plastic pipe). Put a good valve on one end, on/off is ok, with the valve open, coil the pipe under the water in the mine, letting the water fill the pipe. When the pipe - or hose - is full, close the valve, and pull the full pipe down the hill to your watering place, leaving the open end fastened underneath the water level in the mine. The hose will work better with no other valves, tees, or other restrictions except the shutoff valve: when the valve is wide open, the flow is probably fast enough to wash the remaining air out the bottom, if there is any. Having no leak at all is the best way to ensure good function in a siphon pipe.

hi,

If your set - up looks like the picture below, then you got it. it will surely work

Mike.

take a look at

WWW.pipeflowcalculations.com

I am not a pipe expert, but playing with the equations would suggest it will work and achieve the flow you need. Total drop is about 12 feet and that will give a static pressure of about 6 PSI.

My gut feel though is you may need to eliminate all the bends and restrictions possible. I would just use a ball valve at the bottom to shut it off. If flow is too low a larger 50mm pipe will have a large impact.

Wikipedia - siphon has formulas for flow, using Bernoulli's equations.

the velocity of the outlet flow would be v=√2gh

in your case g = 9.8, h ~ 3.7 so v = 8.5 m/s

The area of the pipe is .00096 m^2 so you will get about .5 cu m / min or about 700 cu m per day. (that will be a high estimate since fittings and pipe roughness was not included.)

If I try to throw in the pipe calculations I get something like 115 cu m /day, but I am definitely out of my depth here.

Best guess -- you have a winner, since you only need 5 cu m per day!

hi GW:

the link that you given is great for designing purposes, however in our case during building of our siphon set - up where the only available spare is the hose and one gate valve, we used all alternatives to make it work and honestly speaking it worked well with out any paper works..

for the check valve, we just use a trunk of bamboo and an old slipper..

for the discharge valve we just use a tree branch that fits perfectly on the hose..

for the priming container, a used 5 gallon pail is used and the only available gate valve..

sounds funny but it worked very good and almost free...

by the way, we used this set up to maintain the water level inside our mining tunnel.

I have no doubt the siphon worked.

On an engineering forum I find a distinct lack of "engineering" with the use of formula and applied science to model and predict actual performance.

A side question would be for 5 cu m / day what is the minimum hose size required?

It could even be much smaller at lower cost. But no one has done the math or modelling to show this!

The question was about design principle, not empirical data. You can assume that the OP has in fact surmised that an average flow rate from a 32 mm pipe from 7AM to 11PM is sufficient. You could ask him why he thinks that. That would be a useful comment. Saying you are surprised at the lack of applied science is not really useful. Help him get a good answer! The fact that so many people are fixated on this "hose" tells me that not too many people have pulled a hundred feet of hose up and down a steep slope at 7AM each day.

OP never asked for flow rates, so why should we find any formula and use it. Engineering forum doesn't mean that we always use some formula and make thread LOOK more intelligent.

What was needed was very simple.. bring the water from one point to some other point at lower elevation. Simplicity is the beauty.

Not so...see post #5

at this time of year I need about 4-5 cu/m (4-5000 litres) per day.

Ok, thanks.

So OP realized his needs afterwards. But basic question was very simple and simple answers were expected.

An ancient age solution, doubt whether it will work in present days of contaimination.

hi,

Sometimes ancient age solution works great than the present day engineering... also, ancient engineering is our base...

Siphon pump works very well, I built one and already tested in every day application.

Mike

My estimate of flow is 4.6 m³/h.

I read the OP as meaning the pipe outlet is 20ft lower than the TWL in the mine, so Δh is 20 ft. Also read it as 60ft pipe length from top of bank to outlet, so allowed 75ft total plus 4 bends. Assumed 32mm OD pipe (metric) so used 29mm bore.

Cheers........Codey

OK, now you follow my thread!

I used a lower "head" (8 up, 20 down diff of 12 ft)

I used a 35mm bore that is 1.45x the area of 29mm bore.

We both come up with a plausible result.

Now we need the "Myth Busters" to prove it!

With your figures (still 75ft long) I make it ~ 5.8 m³/h.

Cheers.........Codey

Do you have your formula?

We are in the same order of magnitude. Without pipe friction I have a much larger flow.

I would be interested how the pipe equations are throttling the system down. I am not familiar with those equations. I do see the type of pipe can make a very large difference, from poly, to copper, to galvanized, etc.

I use the Fanning equation Δh = 4*f*L/D*(V²/2*g) where is the friction factor. That's the difficult bit as f depends on the Reynolds and the pipe relative roughness (= absolute roughness/diameter D). The formula for f cannot be solved directly, but it's easy enough to get a value using iteration on XL. The formula's available on the web.

I do it on Mathcad and just found by trial the flow needed to give 20ft Δh.

In this case f ~ 0.009 for 29mm dia, 4.6m³/h. Used absolute roughness 0.2mm.

Cheers..........Codey

Ok, now if we are talking of calculations,

Your calculations are correct, but you have not given any consideration to the frictional losses. (though you have stated about roughness)

v=√2gh is very basic formula, without consideration of frictional losses. Frictional losses will depend on the number of bends, pipe inside surface finish and the velocity itself.

Here in this case, only two bends are sufficient, + pipe losses + ball valve losses.

My guess is flow will reduce by a factor of 2, as you calculated

I have been plugging through more of the literature. Poly Pipes has some good info.

It looks like for the modest volume the OP requires a 12 mm (1/2 inch) to 16mm (3/4 inch) would have been sufficient.

Any comments?

I have tried a set up very similar this with some success, but the siphon did break on a regular basis, this we put down to the water degassing at the gas in the top of the siphon. The flow was intermittent, and never allowed to flow at full speed to flush the accumulated gas out of the system. A smaller pipe may have helped, but the solution we settled on was a solar panel/pump tank and float switch. Its been running 5 yeas now and on the second pump.

J

To prevent gas from entering your pipe,simply put a u-shaped bend in the pipe at the bottom suction end.The gas will not flow downward into the "U".No other parts required.

Have you noticed the dates of this thread?

Day late,dollar shy.DUH!

I Didn't notice,just replied.I hope the problem is resolved by now!

Virtually the same design (pipe sizing, total length, the drop were about the same) worked well for years at a tea plantation in SouthIndia (elevation nearly 1000 m above MSL).

Bioramani

With a total positive head of 12 feet, you will have a static pressure at the bottom of the hose of aprox. 5 psi.

As long as the hose remains full, you will have a constant flow of water available.The volume is uncertain due to many variables that require more information.

Use a wide-open gate valve(WOG) to control the flow.

To increase flow rate,increase hose size.Remember, if you double your hose diameter, you will get 4 times the flow.

Good luck.

It appears this is a real apparatus, not a science fair project, so dragging a hose back and forth every day would be a lot of work. I would build the line from 32mm PVC, with no air leaks, from the bottom of the well, to the discharge point, which has a ball valve allowing you to easily stop flow with an airtight seal. I would in fact recommend a couple of centimeters of turn up on the pipe. It will slightly slow the flow, and reduce air cavitation. At the highest point of elevation, cut a tee into the pipe, and extend 100 mm of 32mm pipe vertically to a ball valve. To operate the system, close the drain valve at the bottom, open the ball valve at the high elevation, and funnel in enough water to fill the down hill sections. Close the ball valve at the high point. Open the ball valve at the bottom. You will need a simple flapper valve at the intake to the well. (operating as a check valve, or foot valve)

No body said that OP has to drag the hose back and forth everyday. So your point is invalid.

Ok, putting PVC pipe has its advantages... simple hose may have pinches inbetween, obstructing flow. That will not be there with PVC. But PVC will need priming from top.

My earlier solution is just simplest one, and cheapest one.

Now with many replies, it has become unnecessarily costly.

This also might work

I like the drawing, if you ever visit the 'selsufficientish' website have a look at my post entitled 'WATER MINE (not yours?)'

http://www.selfsufficientish.com/forum/viewtopic.php?f=38&t=18095

Not sure if I'm allowed to link to other sites, apologies if not.

Next task is to try and get the guys in the merchants to understand my awful Portuguese. I usually do some drawings which they seem to like and understand. I went in the stores with the guy one day and there was a collection of my 'sketches' on the notice board!

The picture looks like it would work great but I would add a couple things.

As was mentioned earlier, an upturn at the bottom or perhaps a trough.

A second 'Priming Water Control Valve'. Have one just above the T and one just below the storage tank. Open the one at the top and let the connecting tube fill with water. Close it and then open the one at the bottom and let the water in the connecting tube and the air in the siphon tube exchange places. Close it. Repeat until the siphon starts.

A diverter valve at the bottom of the siphon to connect the outflow to a ram-pump. Use it to fill the priming tank for the next day. Once full switch the diverter valve back to full flow.

You just need a way to get the 8 foot rising pipe filled with water, then let it flow down to the field. I would try to devise a hand pump to lift the water to the ground, then throw a manual valve to start it flowing to the field. The siphon's vacuum would take over to keep the flow going.

Actually you need to pump it up the 8 feet then 8 feet 1 inch down the other side or it will just siphon back into the mine.

Unless you meant to pump it all the way the ground at the bottom then open a manual valve at the bottom. That would work even better.

The members have said it all really. I would add, it's a walk in the park.

Take a normal hose pipe, a small bucket, and a short rope, to the well.
Tie the rope to the bucket and collect water from the well as required.
Have the two open ends of the hose pipe near to you, and fill the pipe
at one end with water. (from the bucket) When the pipe is full, i.e.both
ends are over flowing, throw one end into the water of the well, and the
other end down the slope. Leave to run dry! n.b....

You may have to carry the top end of the pipe down the slope a little, to get
below the 8ft drop of the well, and while doing so, you should cover the end
of the hose with your thumb to prevent the water running back into the well.
It also helps to sink the well end of the hose by a small weight, a stone or such.

I have emptied whole lakes like this, containing 100,000's of gallons of water,
where contractors have quoted large sums of money, for pumps, etc.

A walk in the park. (ps.. if you want to stop the flow, clamp the down hill end.)

jt.

Don't you just love simplicity!

I got the bits & pieces yesterday and dug a channel across my (granite) road. I laid the 32mm (od) pipe on the ground on a route that could be dug-in later (20 years ago I'd have dug it in straight away!).

I installed a ball valve at the bottom end with another piece of pipe about 500mm long on the end. At the highest point I installed the 'T' facing upwards with a piece of pipe about 1m long, secured vertically to a timber stake with Zip ties, this terminated in a blank. At the well/mine end I installed the non-return valve, on one end of the pipe, I fixed the valve to a weight to hold it under the water (itself having a rope attached for removal), the other end was installed in the 'T'. With the exit valve closed I primed the vertical pipe using a funnel and a 30 litre water container (30 litre = 30 Kg I'm 52 not 22). I secured the filling point with the blank and trotted off to the exit end.

I opened the valve and there was a gurgling, then the flow started, nice clear water, then some dirty water, then it stopped! I walked back to the well/mine I decided that there was too much pressure loss across the NRV. I removed the NRV and re-secured the open end of the pipe to the weight, I then put another isolating valve on the other end of the pipe before the 'T'. I closed both valves and re-primed the system. I then opened the valve next to the 'T' and trotted off to the discharge end again. This time it was much better, clean water, air, dirty water, clean water, it ran for about an hour before - the mine was dry!

I need to fine tune things a bit, the pipe isn't very flexible so I think I'll terminate it with a loop and valve on the retaining wall. Maybe I'll build a trough from which I can run hosepipes as needed. I've started to clean out the well/mine, I pulled one piece of wood out that was 1.5m below water level so it looks like I've got another job on now, a pity really as the mosquitos really like the well/mine

Sincere thanks to all contributors, your input has been invaluable, you can practically guarantee that someone somewhere else in the world has had a similar experience, thanks for sharing them with me.

Good job, congrats

outstanding job. I might suggest that if you are emptying the well, slowing the flow could improve the volume delivered between service (the flow stops because the well is dry) just reduce 32 mm orifice to 25mm at the delivery spigot. You may be able to do it by throttling the ball valve, but that opening will be pretty turbulent.

I think you have not read the post 5 carefully, My lower mine (feeds the house) holds more than 100,000 litres and replenishes almost instantaneously.

in post 43, OP describes setup, working, and says "after 1 hour mine was dry".

Sorry about the confusion guys. The mine in this post is the 'upper' mine. This mine hasn't been used for years but the locals say there's plenty of water in it. I'm in the process of digging it down at the moment. The mine that feeds the house etc is lower down the site, this is the one with the large volume that replenishes instantaneously. Unfortunately, the level at which I need large volumes of water is above the level of this mine, meaning that I would need to pump the water. Back to the spade and bucket for me now!

so the sketch, the many, many descriptions of this system, are not correct. @##@@!

The sketch and discussions all relate to the upper mine and are absolutely correct, in fact the sketch couldn't be more accurate if the contributor had visited the site! I hesitated to mention the lower mine in case of confusion!

I am a little confused between a mine and a well.

A mine to me is an excavation to remove minerals, with a shaft and tunnels, or an open pit excavation. Usually of large volume.

A well is a drilled hole, relatively small diameter possibly going down hundreds of feet to an aquifer, or seldom used now, a larger diameter relatively shallow hole that originally were hand dug that collects surface water or in the surface water table. Comparatively small volume to a mine.

What is the volume of your "mine".

When a mine is abandoned they often fill with water.

It will work.