Behavior of a water well siphon

It would seem to me that closing a valve at the storage tank will retain water in the 100'+- hose or pipe. Then when you open the valve. provided the high end of the hose is still submerged, the water will flow creating a negative pressure and restart the siphon effect. If there is enough time in between uses to lose enough water to not start an adequate flow I would think adding a valve at the high end will keep the line full. Then to restart it open the upper valve, then the lower. Another thought would be to dig down low enough to have a cistern that has the bottom in the water table and run your hose from there maintaining the water level in the hose.

Quite.

Hmm. Not sure about the valve at the top. If the pipe is unbroken then the siphon will remain until the valve at the bottom is opened. Putting a valve at the top could introduce a source of leaks and the possibility of maloperation that would require the siphon to be re-primed on such occasions.

The siphon will also need re-priming should either end of the pipe become inadvertently uncovered, and this factor needs to be considered in the system design.

Wouldn't it be 115-foot(+) hose or pipe?

How about this--assuming you have or can provide electric power. Use two hoses/pipes, 15-ft(+) and 100-ft. joined at the (now-existing) U bend by a straight section of rigid pipe. The rigid pipe (actually a pipe-teefitting-pipe assy) would have connected, to the skyward-facing tee-fitting opening, an electric, dual-actuated cut-off/cut-on valve to which, at its top, would be fitted a funnel that flared upward. Above the funnel would be suspended the outlet end of another tube (pipe/hose)--a relatively smaller-dia tube--which extended at its distant end into the well water. At the bottom of (or along) the small-dia. tube would be connected (or interconnected) an electric pump. The five or more wires (one wire pair and one wire triplet(+?) from, respectively, the electrically actuated shutoff valve and the "assist" pump would be dressed over and down the cliff along the route of the 100-foot, tank-fill line--possibly be clamped/strapped to it. At the tank (from which you don't want to climb the cliff), each wire circuit (to be) will be connected across a power source, AC &or DC, depending on the top-of-cliff devices; a switch would be incorporated into each circuit. (...a spst switch to control the Assist pump; switching as required to control the cuton/cutoff valve.)

Now, you won't be trying to prevent the water from falling back wellward in the Up linel--and you probably couldn't if you wanted to. Instead you will simply refill that Up line, and any void in the Down line, to reinitiate siphoning, by remote control.

1. With valve at tank closed
2. Switch open the shutoff/shuton valve remotely. Then
3. Switch on Assist pump to raise water 15 feet into the funnel and, thence, into the siphon line...until it overflows the funnel, indicating that the siphon line is refilled, both Up line and Down line. (Watch/listen/detect however you can, or guess when overflow indicates filled siphon line.) (For example, a pan which collected overflow could be drained through a tube which would discharge water drops over the cliff and onto the human operator; thus signalling when siphon line is filled... ...)
4. In quick order, switch the shutoff/shuton valve to closed and quickly open the valve at the tank (or...in reverse order, whichever works best or most easily)
5. ...siphon begins
6. Switch off the Assist pump.

Once that is working satisfactorily, incorporate float actuated valve or whatever automation is desired to make the entire thing work, untended, as a system. Be sure it's done so as to always permit manually activated operation.

Hi Ken

I've gone through the various suggestions, and learned a lot on the way; in the absence of detailed information, it seems the most likely cause is that air is entering through the valve at the bottom when you stem the flow. If that is the case, it is only a problem if the valve is in a section of the pipe where the air can rise to fill a large section of pipe. If you can turn the pipe so that you can mount the valve in a section of pipe (that is near the outlet) where the water is flowing upwards, this should solve the problem.

BTW, if and when you know what the problem really was, I (and doubtless plenty of others) will be very interested to hear.

Regards

Fyz

Do you have a foot valve or non-return in the well?

Does the flow reduce over time?

Do you notice or hear any leaks?

Yes, there is a foot valve in the well to prevent backflow.

When the siphon is run with the lower tank valve wide open, it runs with consistent flow for weeks. However, when I partially close the tank valve to slow the flow, the siphon will fail after 10 or 15 minutes.

I will re-inspect for leaks since it seems obvious that air is entering the intake.

By the way, there are several oil wells in the area and the well water smells heavily of sulphur. Could excessive sulphur gas in the water be causing my problem?

Yes disolved gases (Hydrogen sulphide) could well form a gas bubble. Note check the Ph of your water it could be rather acidic.

Hydrogen sulphide is a killer gas...

If you are to handle/purge reasonable quantity of this gas, extreme care has to be taken.

Slowing the flow from the bottom disrupts the syphon? Three possibilities - sufficient gas in the incoming water (dissolved or not) that collects once the flow slows; a slow leak (given the drop, it could even be that the tube is porous), or air going back up the pipe from the outflow once the water velocity was slow enough to allow this. Intake of free air at the inlet or outlet can be minimised by ensuring the end sections of pipe point upwards, and the top one is quite wide - but I suspect you've probably already done this.

Ken,

You say it seems obvious that air is entering at the intake, does that mean the discharge end is not submerged in the tank? If so submerge the discharge end and let us know if you are still loosing the siphon.

I don't believe 10 to 15 min. is long enough for the disolved gasses to come out of the water and be causing the problem. If there are air leaks in the pipe that are causing the siphon to be brocken you must have a hard time getting the siphon to start. I would be interested in how you start the siphon.

If you shut off the valve suddenly and open it some time later have you lost the siphon?

He says that he is taking water from 15 feet below the surface.......I think that you can ignore the thoughts of air entering the intake at that point....

Andy,

He says the water in the well is 15' below the surface. He does not tell us how deep down the foot valve is in the water.

Correct, I read it wrong.....

A siphon actually works by air pressure. The siphon must always be kept full of water. Small bubbles don't affect its operation much if the water is already flowing, but if there is a bubble at the top bend large enough to break the flow it will stop working. Check your foot valve to see if it's jammed open. If it is, the water inside will drain out. H2S and other gases coming out of solution to form large bubbles within the siohon will also cause it to fail.

From what you are describing, yes, it is possible to maintain the water within with a valve. Just imagine the valves as your own two thumbs when you are using a hose to siphon water out of a fish tank. The important thing is your ball valve or whatever you are using to stop the flow must be at the bottom end of the arrangement. It cannot be at the top bend.

Get yourself a float operated diaphragm pump and a float valve on the supply line. Once the tank is full, the float valve could close and stop the flow...once the tank level dropped, another float could activate the pump. It would re-prime itself and restart the siphon. It will also act as a check valve in the line to reduce backflow into the well when the system is shut down.

Run the pump on a timer so it only runs long enough to re-prime and start the cycle. The diaphragm pump should let flowing water pass through unrestricted.

I believe you already has an diaphragm pump at the top.

The problem therefore is to get rid of the "air" at the top without running up the hill to fetch a pale of water.

a mechanical prime retainer must be developed.

The first thing that comes into my mind is to have a pressure vessel at the top.

the outlet at the bottom.

inlet about 2/3 up

the top third is for the air.

The priming trips may be less frequent.

If a chemical can be introduced to absorb the gasses it might help.

I didn't see Any mention of a pump...Sorry if I'm mistaken, but if there was already a diaphragm pump in line, there wouldn't be this issue. The reason a dia. pump is recommended is because you don't need to prime them. They will pump air until the water gets primed, then the siphon can take over.

I feel that you have an air leak, check the hose (not with water!) with air pressure and get the hose underwater as you probably have some tiny pinholes, when the siphon is running, they do not matter, but as soon as the siphon stops, they are letting air in(not water out!).

The pinholes that really matter will be on the side down into the water supply and on the side towards the tank, till almost the point where the water level is on the supply side. All lower pin holes will be under water pressure.....if there are any.....

The worst place is between the suction and the highest point of the hose!!!

Buy a new hose and protect it from sunlight!!!!

In short, the answer is No, ...until you open that valve back up.

I think you have to shut it off at the bottom. If you shut it off at the top and leave it open at the bottom, the weight of the bottom 34 feet of water will pull a vacuum, leaving a big bubble of water vapor at the top. I don't think the syphon will start again with that gas in the line.

I guess you need a level sensor at the top and electrically drive a valve at the bottom.

You are right, the water will pull a vacuum.

A vacuum is not a gas, its nothing - just empty.

The vacuum will suck the hose together, till the water is all gone, then it will return to its normal shape again.

Actually, if the hose will withstand atmospheric pressure (15 psi) and not collapse, the "vacuum" will fill with water vapor to whatever the vapor pressure is at the ambient temperature. This bubble of gas will probably keep the syphon from working when he reopens the valve.

Its not likely to be such a hose is it?

I'm thinking it's a pipe and can withstand atmospheric pressure. We've been using hose and pipe interchangeable, and the original poster didn't specify. From his description of the problem - that stopping or restricting the flow breaks the syphon -leads me to believe that it's a pipe and doesn't collapse.

Stopping or restricting the flow results in the pressure in the syphon falling below the vapor pressure of water at the ambient temparature - at which point the water turns to vapor creating a gas bubble, which ruins the syphon.

At normal temperatures* (below 30-deg), water vapour pressure is less than 5% of atmospheric, so air pressure will still support a column** of 28' inside the tube. However, it remains true in principle that, if you close the valve at the top too rapidly, the water's momentum could be sufficient to empty the tube against a vacuum - that depends on whether the diameter of the tube allows the velocity to build sufficiently. This could still happen if the vapour pressure was near zero.
So I'd say that the non-zero water vapour pressure would only make a marginal difference in practice.

Fyz

*Even at 80-degC, atmospheric pressure will still support a column of over 15 feet against water vapour, even at temperatures up to 80-degrees Celsius.
**Rather less if the system is above 2000' elevation.

If the vapor pressure of water is 5% atmospheric at ambient temperature, the weight of about 32 feet of water will balance the difference between atmospheric and the vapor pressure above. Therefore, above 32 feet, the water will vaporize forming a bubble with pressure .05 X 14.7 psi.

If the water is already flowing and is decelerated by restricting the flow at the top, due to inertial forces, less than 32 feet will be necessary. In any case, the syphon won't work with gas in the line and he needs to control the syphon from the bottom.

My suggestion is that he use a float activated microswitch at the top connected to a solenoid operated valve at the bottom.

I'm not disagreeing that inertial effects can cause the column to break and allow the water to run out of the bottom (indeed, if you look carefully, you'll find that was what I was saying - and indeed this was described some time ago in another posting). What I'm saying is that it is merely confusing to bring the vapour pressure into this - so long as the vapour pressure is not large and negative*, the column will break just the same. That means that the vapour pressure is not the crucial factor as you implied in your previous post - though it may just tip the balance between enough water remaining in the pipe to restart the syphon and failure. In any case, a cheaper solution to the issue of inertial effects is to close the valve at the top sufficiently slowly.

*On the basis that under some conditions you could view surface tension as applying "large negative vapour pressure"

You're right, the vapor pressure is small enough to be neglected in the calculations.

Long ago when I was in school and before mercury was considered a hazardous substance, students would make barometers for science fairs. If the valve at the top is closed, what we have here is a giant barometer with water instead of mercury and 34 feet instead of 30 inches. A bubble containing a small amount of water vapor will go from the top of the syphon to 34 feet above the bottom reservoir level.

When you reopen the valve at the top, I would be very surprised if the syphon would work with a 66 foot bubble in the line.

You need to consider the pressure at the top of the bubble. Assuming that the water doe not run out at the bottom, the worst case is that the bubble rises to the highest point in the syphon. That means that the pressure at the top of the syphon is just the vapour pressure of the water. When you open the "top" valve, the atmospheric pressure driving the water will be enough to drive a "head" equal to (atmospheric pressure minus vapour pressure).

On a more practical note, can you suck up water using a partial vacuum in air? How about a using drinking straw. That is no different in principle to the effect of the low vapour pressure.

Fyz

OK, I think you have me convinced. If the pressure is less than about 8.2 psia, the water should flow over the topside. The fact that it breaks indicates that the pressure at the top is higher than this - as mentioned before, probably due to air leak or outgassing of dissolved air.

Phew!!

Fyz

There is a difference, the length therefore the weight of the water plays a heavy role.

Also surface tension in a straw would be a serious effect that would play little or no role in a pipe or hose (except with regard to pin pricks, causing air leaks, that water could not flow thru!)

the vacuum caused with water is only apparent when the water column is greater than 33 feet.....thats a very long straw!!!

I was merely trying to explain why the vapour pressure of water was almost irrelevant; so, if I read you aright, you are agreeing with me - what you are saying is that my description is more accurate for the pipe than for the straw.
32' is rather too long unless the hill is at sea level and the water is almost freezing (both possible, of course).
If the content of a section of the pipe is only the water vapour, the effects of surface tension will become very small (the water on the inside of the pipe is at local boiling temperature). This is one of the reasons that an evacuated pipe can leak far more readily than you might expect from pressurised measurements.

Fyz

The problem is that he is stopping/slowing the flow from the bottom, but then the siphon breaks, therefore it can only be due to pin pricks in the area of the pipe/hose subject to less than atmospheric pressure......

Why pin pricks you may ask? Because they would have little effect with full flow, but when partial or stopped, air would bleed into the area at less than atmosheric pressure.

By the way, 33 feet is only an average - a ball park figure, this will change depending upon the height above sea level and the air pressure at the time......

Thanks, you're right - the thread had drifted away from the original problem, and I hadn't noticed. If he is stopping the flow from storage tank at the bottom, only the top 15' in the two branches at the top of the syphon would be below atmospheric pressure. (Of course, if there is air entrapped, it would be the top 15 ft of the water column on the output side, plus the region containing the air). Water leakage from below that would be replaced from the well. I think that restricts the region where he needs to check for leaks.

The 33' would be a maximum, I think, and reduced by any vapour pressure as well as height (and to a lesser extent modified by daily variations in barometric pressure)

Fyz

"The 33' would be a maximum, I think, and reduced by any vapour pressure as well as height (and to a lesser extent modified by daily variations in barometric pressure)"

33ft was just a ball park figure to work with, in practise probably only 30ft or so, and you are probably correct with the water vapour If a vacuum is made, but with the height of the suction side being only 15ft, that will not happen. If the hose had no pinholes, it would just stay there (assuming no dissolved gases).

If you stop the flow from the top, (in the absence of leaks and bubbles entering from the ends) you will get a bubble that stays at low pressure, and allows the air pressure on the well to drive the water. The problem you may have if you stopped the flow suddenly is that the momentum of the water is sufficient to partially empty the bottom of the pipe - maybe sufficient that the pressure drop is no longer sufficient for the water to reach the top of the syphon.

If you stop the flow from the bottom, the pipe can completely fill with water, and the maximum pressure drop relative to atmospheric will be the 15' above the well surface. The pressure at the bottom will be the 85-foot head below the well surface. There is clearly less potential length for air leaks in this case.

Fyz

If the syphon hose is just stuck down into the well 15 ft or so & then run across the top of the ground down hill to the storage tank 100 Ft below & the syphon effect has commenced & you have a flow, as soon as you cut the flow of water with a shutoff valve at the bottom end ,the 15Ft or so stuck down into the well will just run back into the well & empty, put a really good sealing foot valve or one way valve in the pipe end in the well at the top end to stop any run back of water into the well when you shut the flow off .

We had a similar setup to what you described on my parents farm years ago, the poly pipe ran for about 200 metres down hill from a spring fed underground storage tank to a small dam about 30 metres lower down, we just shut the flow off with a brass gate valve to stop the flow, HOWEVER we had to connect a small double piston reciprocating plunger pump that was normally used to pump water from the dam up to the house tanks, a head height of about 30 metres , the inlet hose was coupled to the syphon hose & took a fair while & a jerry rigged priming tank on ther pump to get the flow going in the syphon hose which was 32mm flexible poly pipe, once running no problems, it was only used in dry spells to top the dam which was normally kept filled by a running stream. The uphill property was called "Eversprings" & had never been known to run dry even in the driest of spells. Could do with a few of those in Syney Australia right now. Make sure there are no air leaks in the line!

Brien.

"the 15Ft or so stuck down into the well will just run back into the well & empty". Why on earth should it do that? As the pressure drop is less than 32' something needs to come in to fill the empty space. With the closed valve at the bottom and the tube under the surface, there should be no opportunity - unless there is a leak or the pipe is running with a very large proportion if air.

Put a check valve at each end to maintain water in line. with a valve (open/close) at the tank and one at end of your line along with the check valves should do the trick.

Put a check valve in-line and this should solve your problem.

Standard atmospheric pressure will support a 33 foot column of water. I don't think you could entrain enough air in the water to cause a break in the siphon. Therefore, you must have a leak someplace. Anywhere from inlet to outlet could cause it to lose siphon, depending on the exact location.

Having had a chance to sleep over it I now suspect the following:

The water coming out at the outlet is only flowing at 2/3 of diameter. Air is therefore introduced in the pipe. The air tends to move upwards but is kept back by the flow of water. The moment the flow is closed or reduced at the outlet the bubble of air moves upward until eventually became stuck at the highest point.

To solve the problem 2 things can be done.

1 reduce the flow at the outlet (orifice or second pre adjusted valve)

2 Air release valves close to the outlet.

Neither of these will affect the siphon.

Also note that installing air relieve valves at gradient changes may help.

A detailed sketch will help.

Even in the absence of leaks, and with a proper air-lock at the bottom, I think that while the water is flowing some of the dissolved dissolved gas will tend to come out of the depressurised part. If you run it for long enough, that could eventually stop the syphon (the gas pressure will depend on the origin of the water). So you will end up with a pipe that has a substantial length of low-pressure air with the water running down around it. The gas will form a spring, so that the water at the bottom could continue flowing after you have closed the valve at the top. Even if there is no entrapped gas, stopping the flow at the top could allow the momentum of the water to largely empty the pipe - but in theory you might get around this be turning the flow down very slowly (because of the length of the pipe, that could be slower than we might imagine).

The only way I can see around the issue of collecting gas (on a medium-term basis) is to run the water fast enough that the gases are brought out with the water; then you would need to reduce the flow rate quite gradually if you do it from the top.
Shutting off the flow from the bottom is more predictable - and the pipe will fill up to at least halfway (the pressure at the top cannot be greater than 0.5 atmosphere, or the flow would already have stopped).

Fyz

I assume you have checked for air leaks in your pipes and there is no sharp bend where the pipe comes out of the well where you can get cavitation. Also since you are using a siphon you are up in the mountain where you don't ahve electrisity.

Is the discharge end of the pipe submerged in the storage tank, this would prevent air from entering entering from the discharge end.

In my well the pump fills the preasure tank and the pump shuts off. The preasure in the tank causes back preasure on the foot valve to keep it closed. You just have the 15ft of water holding the foot valve closed.

How do you start the siphon? Perhaps you could simplify it with a water pump to put water in a preasure tank down near your storage tank and a venturi valve to reintroduce the water in to the line and create a suction to restart the siphon.

Here is a link to information on water pumps.

http://journeytoforever.org/at_waterpump.html

PS. let us know how you solve your problem.

FINAL DEMAND FOR INFORMATION IN TERMS OF REGGULATION XXX.PLEASE.NOW

Sketch with information please.

Gradients.

Priming method (do you carry a pale of water up the hill like Jack and Jill?)

Outlet arrangements.

etc.

Your problem can be solved!

If you already found your solution switch the light OR say good by.

Geez Hendrik, I need some time to investigate. Hold your horses.....

Thanks Hendrik, and all for your thoughts and well founded suggestions. I will carefully investigate for leaks and double check the foot valve to make sure it is holding. And, I promise to report back on my findings and my solution if I find it. Otherwise, I will post a diagram and mucho detail. No doubt this brain trust can solve it!

Los Olivos, Ken

Is this sketch depict the situation?

If so, there should be nothing wrong with the set up.

A NRE at upstream and Block Valve at down-stream should be just enough.

But: I am yet see true NRV over it's Life-cycle.

And Gate valve would permit air thro' gland on partial closure (in full open 'back-seat' might prevent leak). Diaphragm valve might help.

If rubber hoses are used, possible innumerous tiny cracks might together permit huge air.

Tiny leaks are good enough to break the siphon. If flange joints are there without raised face but with full gasket, could be problem source.

Very cool sketch, Yesyen. Yes, that's it exactly.

By the way, what software did you use for the sketch? I like it.

Los Olivos, Ken

The original description appeared to imply that the water level was 15ft below the ground surface. If that is what was meant, then the 15' should be from the surface of the water to the top of the pipe - not simply from the pipe inlet.

Yes, I agree.

Is it right now?

Los Olivos, Ken: Thank, I use Flash, a wonderful product from Macromedia. Such simple sketches are very minimal compared to its capabilities.

Looks good to me - the hill looks awful steep, but I doubt that is material to the issues

Just joking, I forgot the

Perhaps more expensive--perhaps less expensive--but definitely quick-and-dirty...

Why not just a suction pump at the bottom, to "reprime" siphoning? Could be powered by your vehicle; or by autostart from a generator next to the tank. Once siponing starts, pump shuts off. Easy setup, virtually fail safe, few or no moving parts to fail within the siphon, no maintenance except refilling fuel tank.

In order to analyze the problem the parts of the system is separated.

Foot valve

If close to the ground sand or rubble may be sucked in which may prevent the flap from sealing or settle at the bottom with no flow creating a plug.

If the valve is close to the surface a vortex may form and air will be introduced through the in-take.

Siphon

Up to about 30' below the highest point can be considered the siphon.

Gravity line

The rest of the system must be considered as such.

Because of the slope available in this section the flow rate will exceed the capacity of the siphon unless restricted at the outlet.

Turbulent flow can be expected with voids being created and collapsing.

Cavitation is possible.

Pipe

The type of pipe may also play a part. Example the inside lining on a rubber hose may separate from the rest and create an obstruction.

Fittings

Pressure fittings may actually allow air to enter when a vacuum is applied. (lip seal or even o ring)

Please also check

Are bubbles of air coming out at the outlet?

Can you hear the flow?

Smooth or erratic flow?

Vortex at the top?

sand or particles at the bottom?

A lot of homework?

A systematic analysis...

But it was told that the siphon has no problem while in full flow. And looses priming when restricted or closed for a while.

If so do you doubt: vortex, high rate of flow, cavitation?

I too strongly believe air is getting into system when siphon is not in operation.

But we went to the extend of even including a pump, vacuum pump like costlier options, I feel like suggesting the following set-up.

1. A 'Tank – TK2', having adequate capacity (more than the pipe line hold up) is connected as shown thro' a NRV.

2. 'Vent Vessel - VV', 'Ejector – EJ', 'Solenoid Valves - SV1, SV2' are the other components.

Operation.

· During normal siphoning, the tank TK2 collects/ensures a level of water equal to reservoir.

· Siphon stops by closing the block valve – V1,

· When required, SV1 and SV2 are to be opened remotely. Through SV2 water fills and air vents through SV1, until the level equalizes in the pipe and TK2. Ensuring this, SV1 and SV2 are to be closed.

· TK2 is big enough and should have still good quaintly of water.

· Now V2 is to be opened. This acts as motive jet for ejector EJ. This should suck the air/vapor from the pipe.

· Ensuring the pipe is filled with water, V2 is to be closed and V1 is to be opened to establish a fresh flow.

The hardware requirements are tank, valves and pipeline. Piping and fitting need not be big in size.

Does this sound practical and feasible?

It might be possible for a small quantity of air to pass through the system running at full blast. With the valve being closed this air is trapped.

Closing the valve at the bottom Will reduce the vacuum therefore less chance for air leaks at that stage.

Rubble sucked in at the foot vale may be kept suspended at full flow.

A 20' vent pipe installed 10' to 15' below the surface of the water (top end 5' to 10'above) might do the trick. When the valve is closed at the bottom the siphon will keep on delivering until the vent pipe is filled up to the water level. any air in the system will escape. any water leaks will result in a drop in level and activation of the siphon.

Your priming system may work but also consider a bigger tank at the top with an inline venturi right next to it to fill the tank.

I still think all the possible causes should be checked and discarded or cured.

"Closing the valve at the bottom Will reduce the vacuum therefore less chance for air leaks at that stage."

But, the originator 'Guest" says in his mail 5 as, "When the siphon is run with the lower tank valve wide open, it runs with consistent flow for weeks. However, when I partially close the tank valve to slow the flow, the siphon will fail after 10 or 15 minutes."

This statement of his is confusing and contradicting the logics.

I agreer that it do not sound logical.

Coming to think of it my argument for a vortex holds for air leaks as well.

The flow and velocity will expel the small quantities of leaked air.

Closing the valve a fraction will create a positive pressure (less negative) above the valve and will keep the air trapped until enough accumulates to loose prime.

It will not be noticed right away because of the length of the pipe.

What do you think of a second valve at the bottom that can be preset to keep the outlet flow equal to the capacity of the siphon?

What do you think of the idea of the vent pipe?

An air valve at that position may also help.

Detecting air leaks on a 100' m pipe within 10 to 15 min may be impossible. Can you recommend any method?

"What do you think of the idea of the vent pipe? An air valve at that position may also help."

A vent pipe with 'an air valve' anywhere (above the source water level), while siphon is on or off will only admit air in to system.

"Detecting air leaks on a 100' m pipe within 10 to 15 min may be impossible. Can you recommend any method?"

Yes, Detecting air leaks within 10 to 15 min may be impossible. But we can 'hydro test' duly isolating the line from service and taking care of the test pressure being comfortably less than what pipe can handle (more so if flexible hose is involved). This should revel pinholes or joint leaks either visibly or by not retaining the test pressure under watch.

By the way, is our 'guest-originator' is still around? How do we to understand a guest is a visitor or the originator?

...it does not contradict the logic if for example he has multiple pin prick holes in the part of the pipe (the first and second 15 foot aproximately) that is slightly under a vacuum.

Pin prick holes would allow air in but not let water out. When the flow is high, the air would be swept along with the water, when the flow is low, the air will collect at the highest point and as some point break the siphon effect.

I feel that at least the first 30 feet of hose or pipe should be replaced or at least checked for air leaks....using air pressure inside the hose/pipe and either a soapy solution or a bathtub filled with water!

The pin prick solution fits perfectly as detailed, but another solution might also, if there is more info that we have not yet been given......its just that the owner of the Blog seems to have lost interest in us all......

It must be the same feeling a doctor get when loosing a patient.

Finding an air leak is difficult and performing all the proposed test are hard work.

Somebody must invent a CR4 transporter whereby interested persons can be beamed to a site for instant first hand information. The inventor should be rewarded with 50 perpetual motion credits. (that is 50 posts before any poster may use the word stupid.)

When I saw this discussion first, I thought siphon is a simplest physics principle. But going through the discussions, I wonder there are many things we have not thought of and need to reaffirm. Each of us is exploring and explaining in our/his own way. I could be saying the same thing as the other but differently. I think we will consolidate the understanding.

Towards this, I have a diagram here and ask you gentlemen to suggest the right answers.

We will first discuss about the pressure existing inside the pipe at indicated points. For convince we will keep the unit for pressure as water column. And may ignore the velocity component for the time being. I have shown the pressure at point P2 equal –15' (?) water column in the manometer.

What are the pressures at various points:

P1 = ?

P2 = ?

P3 = ?

P4 = ?

From P1 to P3 you have equilibrium, nothing happens so to say. The water weights on each side are equal.

But between P3 & P4 you have an imbalance, that is a weight of water only on one side.

The height of the siphon being only 15 feet (water will vacuum around 32 feet or maybe less), therefore the extra weight of water between P3 and P4 will allow that water to fall out of the hose at P4, lowering the pressure between P1 & P3 and therefore sucking in new water at P1......your siphon starts working.

Starting with the only thing we actually know: the pressure at P4 is atmospheric (A). Other than that, we would need to make assumptions:

Assumptions1: No air in pipe, pipe diameter uniform, negligible viscosity:
P3 = P1 = A-5'.
P2 = A-20'

Assumptions2: No air in pipe, flow limited by viscosity:
P3 ~ (>) A-5'
P1 ~ (<) A
A-20' < P2 < A-15'
(As drawn, P2 is nearer A-15)

Assumptions3: No air on pipe, variable pipe diameter, negligible viscosity:
A-5' < P3 < A
A-5' < P1 < A
A-20' < P2 < A-15'

Add air in pipe at your discretion - or take your pick.

Fyz

I am fairly certain that we are being overly complicated here, the water must flow over a 15 ft hump, that works he says, when the flow is not restricted at the bottom.

The water drops about 85ft, so he should get a good pressure at the bottom tank, provided the flow is not restricted anywhere - in fact he says that it works fine until he turns off the tap at the bottom tank!

I personally cannot believe (but I could be wrong!) that there could be enough gas dissolved in the water to break the siphon.....but let us not completely ignore that....

So the only other alternative that I see is an air leak.....into the hose/pipe somewhere between P1 and P3, most likely on or near P2, where the vacuum will be at its maximum, but anywhere will do it, where the pressure is lower than the atmospheric pressure!

Theoretically, provided he has a sealed system and there are no vast amounts of dissolved gases in the water, he should be able to turn on and off the tap at the bottom and not need to start the siphon each time!

Where am I going wrong Guys?

Hi Andy - Although I don't fully agree with you, your contribution has done much to clarify my thoughts.

Unless there is something unexpected about this shape of the pipe, I think it should be possible to restart the syphon - even if the entire section above P1-P3 is full of air. If you restart gradually (so that we are looking at the hydrostatic limits), the syphon should restart - provided that the bottom 15' of height is full of water once the entrapped air has expanded to about twice its initial volume (less if the expansion is adiabatic - but that's probably not slow enough).

Assume that the pipe is of uniform cross section, and the top section between P1 and P3 (above the surface of the well) is 45' long, that means that there will be a good enough "vacuum" to restart the flow when the water in the lower section has moved 45' along the pipe. Even it the remaining section of pipe is vertical, that leaves a "pulling" drop of at least 40' (=100'-15'-45'), which is more than sufficient to restart the flow.
(We also have to take into account that "the syphon breaks" if he slows the flow too much (without deliberately stopping it), and momentum is tending to keep the flow going under these conditions)

So we need one of the following:
i) The pipe contains enough air during the flow so that not only the top section is filled with air, but also a significant amount of the lower section. If that is so, either the air is entering between P1 and P2 (so that the average density of the upwards-flowing fluid is reduced), or the continuation of the flow is reliant on the air being significantly compressed when the water is running fast.
ii) The leak is fast enough that it can fill the pipe once the flow goes below some level that is needed to pull the air with the water. (It might also leak quicker if the leak is somewhere that the vacuum increases as the flow slows)
iii) The pipe leaks water (but that wouldn't explain the problem with slow flow, and should be easy to find)
iv) The air leak is actually in the valve that is near the bottom. Constraining the flow will cause a pressure drop in the valve, that can suck air in.
v) There is a long length of pipe in the top region (P1-P3), or its cross section is much larger than that in the remainder of the pipe. (But this would not really explain the flow stopping when the flow rate is reduced but not stopped)

I'm likely missing something here; but in the absence of additional information, I'd say that the problem was most likely to be in the valve. The simplest solution is to ensure that the valve is in a short section near the end where the water is flowing upwards.

Over to you...

Fyz

Really top quality Physicist, we all thank you for your time and input.

I am not quite so certain with regard to:-

"iv) The air leak is actually in the valve that is near the bottom. Constraining the flow will cause a pressure drop in the valve, that can suck air in."

I believe the water is at its highest pressure here.....more likely water out than air in!!

We can assume that the water leaves the pipe at atmospheric pressure. So, if the water is flowing downwards from the valve, the side of the valve nearest the outlet will be below atmospheric pressure. Constricting the flow further reduces the pressure (Bernoulli's principle). Consider the limiting case where the valve constriction is so short that we can ignore local viscous losses - the constriction works because the potential energy driving the flow can only support a limited velocity - at which point the pressure is equal on both sides of the valve. If you think of it as a venturi, you can easily see why air will be drawn in; the difficulty is why it would leave the valve on the inlet side of the pipe. My suspicion is that, as you turn the valve down, the situation goes through a highly turbulent phase, and this mixes the air and water so that some is expelled on either side of the valve - part of the valve body might even be acting as a vortex pump; whether this is possible depends on the design of the valve.

Obviously, I have no idea whether the valve that Ken is using can do this - but in the end the reason for proposing this as the solution is that the other options appear even more remote (assuming of course that Ken has been able to check for gross leaks, and that there is not a vast volume of pipe above the level of the well surface)

Regards

Fyz

We try working out together and find if we are somewhere going wrong.

Physicist's explanation sounds good.

For simplicity we consider the assumptions1: No air in pipe, pipe diameter uniform, negligible viscosity:
P3 = P1 = A-5'.
P2 = A-20'

Considering the gauge pressure only (not considering the atmospheric pressure), can we rewrite the above as:

P3 = P1 = -5'

P2 = -20'

If yes, we will now close the flow at P4 and analyze what P1, P2, P3 and P4 would be?

We will now close the flow at P4 and analyze what P1, P2, P3 and P4 would be?

Please note that there is no foot-valve (as to resemble a true siphon) at the inlet.

If there is no air entrapped:
P1=A
P2=A-15'
P3=A
P4=A+5'

That didn't take long - but unfortunately it didn't provide me with much illumination either

Fyz

Fine, let us now analyze when downstream is 100' for the following 3 conditions.

Condition1: Valve full close, what are P1, P2, P3 and P4.

Condition2: Valve full open, what are P1, P2, P3 and P4.

Condition3: Close the valve to a stage such that P4 start building up a pressure than atmospheric. What are P1, P2, P3 and P4?

That's the real question.

Without additional data, the answer is.... undecidable - except P4 (on downside of valve - always atmospheric), and P1, P3 (valve open) always less than atmospheric. If we assume the water is flowing (or just that the up-pipe is full) and make the pipe diameters equal at P1, P2, and P3, we can add P1>P2+15'>P3

Fyz

Hi,

did you mean?:-

If there is no air trapped and the tap at the bottom is closed.

P1=A
P2=A-15'
P3=A
P4=A+85'

Remember, although when the tap at the bottom is closed, both P1 & P3 are at atmospheric pressure, but when the tap is opened, there is less pressure than the atmosphere as water will then be "sucked up" by the weight of water falling out at P4!! So P3 (and of course P2!) will also be at less than atmospheric pressure too.....

After P3, the3 pressure will at some point equal atmosphere and under that it will progressively increase to some point on the way to P4, then the pressure will again decrease again, probably pulling a rubber (plastic?) hose together a bit as the water tries to get out as quickly as possible!

So the above formula is a bit misleading at best......

I hope that my comments are helpful to you.

Surely that is condition 4 - not mentioned by yesyen.

Fyz

In your reply 62 you've indicated (for the downstream length = 20'):

Assumptions1: No air in pipe, pipe diameter uniform:
P3 = P1 = A-5'.
P2 = A-20'

On the same grounds, for the downstream length = 100', a mathematical extension tents to give:

P3 = P1 = A-85'

P2 = A - 100'

But Physics is different from Mathematics.

That is why, you have said, "the answer is.... undecidable".

Andy, "when the tap is opened, there is less pressure than the atmosphere as water will then be "sucked up" by the weight of water falling out at P4!! So P3 (and of course P2!) will also be at less than atmospheric pressure too.....".

While we wonder the exact pressure around P2 is 'A – 100', for sure it is low, low pressure, conducive for water to boil at the prevailing ambient temperature.

If so, how siphon happening when valve is fully opened:

'The free fall of huge water in the long downstream column doesn't allow the forming vapors to collect and break the flow.'

But what could happen when reducing the flow?

Restricting the flow, reduces the velocity and increases the retention time to that extend the vapors objectionably occupies the pipe and break the siphon.

Then, what is the simple remedy?

Shift the outlet valve location to an elevation such that P2 will not get an undesirable pressure condition of vapor forming under any circumstances. Say at 20' ~ 25' from the top?

I will try and answer your points for you, but I have the impression that you have never made a long (high?) siphon!! If true, you should experiment a little, then you will better understand the simple physics we are talking about...

While we wonder the exact pressure around P2 is 'A – 100', for sure it is low, low pressure, conducive for water to boil at the prevailing ambient temperature.

Not true, to get a vacuum, you need to have a water column higher than aproximately 32 feet. For example, one that is open (but under water) at the bottom and closed at the top. Use a rigid tube for this experiment.

If so, how siphon happening when valve is fully opened:

Sadly you show here what I mentioned before, no practical experience of a siphon.....

'The free fall of huge water in the long downstream column doesn't allow the forming vapors to collect and break the flow.'

I have no idea what you meant here, unless you are talking about the pin holes I mentioned before, being too small to have any effect when the bottom valve is fully open......and the water is flowing rapidly.

But what could happen when reducing the flow?

Restricting the flow, reduces the velocity and increases the retention time to that extend the vapors objectionably occupies the pipe and break the siphon.

Then, what is the simple remedy?

Shift the outlet valve location to an elevation such that P2 will not get an undesirable pressure condition of vapor forming under any circumstances. Say at 20' ~ 25' from the top?

But you do not mention if the pipe still extends the full 100 feet or not in your idea! If not, the speed of the flow thru the siphon will be severely reduced due to the fact that you have lessened the weight of water on the down side from P3 towards P4.

If the pipe is still there, but not under water at the bottom end, it will simply empty when the valve is closed.......the siphon will probably not restart, we are back where we were at the very begining!

If it is under water, a huge vacuum will form between the valve and the water due to the fact that the column is higher than the 32 feet I mentioned....as long as everything is sealed VERY WELL, then water will boil in the vacuum formed, vapour will fill the vacuum and before long, the whole of the column of water will drop out of the pipe.....

e.g. the valve you propose will cause as many problems for the user as the other ideas from others for valves in various places but not at P4!!!

Let me state here once and for all time, the ONLY place for the valve is at P4, nowhere else.

Also please go and play with some water, a 20 foot length of hose and your bath. Fill it with water and experiment with a siphon. Go downstairs and see how much pressure you get from just upstairs to downstairs! Relate that to an 85 foot drop!!!! You will quickly understand the principles concerned.

Have fun, experiment and learn. Hands on is always better than just reading!!

"If it is under water, a huge vacuum will form between the valve and the water due to the fact that the column is higher than the 32 feet I mentioned....as long as everything is sealed VERY WELL, then water will boil in the vacuum formed, vapour will fill the vacuum and before long, the whole of the column of water will drop out of the pipe....."

If the water could boil instantaneously, and the water flow was not at all limited by viscosity, restricted the minimum length of column at which this could happen would be 64' (because of momentum). But the bottom of the pipe would still be the tank, and the pipe now has water vapour at low pressure. So the water will refill the pipe to a height of up to 32 feet (or whatever height you would get with a water barometer). That would still allow the flow to restart - in the absence of leaks. In any case, it wouldn't matter if the bottom pipe is empty - so long as the top one has no leaks.

I think yesyen's idea is to keep the flow relatively slow, so the pressure in the top section never falls below A-20'. It means that the surfaces of the pipe can always be covered with liquid water, which tends to act as a seal for tiny pinholes. Unfortunately, if there is any entrapped gas, a slow flow is precisely what is required for the gas to float to the top and break the syphon. So I agree that the best place for the valve has to be near the bottom. In case the valve sucks in air as you close it, it would be a wise precaution to place it in a short section of thepipe where the water is rising again

20' does not, of course, answer the new questions that appear as you exceed 32'.

As regards physics, mathematics is the formal representation of physical principles. If you omit one of the principles, the maths and reality naturally part company.

In this case, P2 cannot be A-100'. If you include the required physical-mathematical constraint P2>0, the maths shows this perfectly.

The reason the situation is undecidable is that we don't know enough about the pipe and the viscosity - which give additional mathematical constraints. If we know that the pipe is uniform and always at the same angle and remains full of water, we can write some simple equations - but there's still a missing variable - the pressure on the pipe-side of the valve. However, we know that these constraints do not apply to the actual situation - because we have been told the syphon fails. So modelling this set of constraints will in no way help to answer the original question.

Fyz

I still put my money on pin prick air leaks (or bigger, he has not mentioned if he sees water leaks at any time) between P1 & P3. It fits perfectly the description of the problem as described and it fits in with the hardware as described.

Its a shame that our Blogger is not reporting!!!

Is this on the basis that the leaks simply become greater than the flow when the flow rate is reduced?

Fyz

......in proportion yes, with flow reduced (or stopped), the water cannot sweep all before it, the air gathers at the highest point - P2 and eventually breaks the siphon.

In all that has been spoken of in this blog, this is the only scenario which fits the details we have been given!

For example, people have been adding valves at P1, and at other various positions, but they do not actually exist!!! (according to our Blogger in command!) Therefore these imaginary valves should not be used in calculating the problem!!

I somehow have the impression that a lot of people do not have a clue as to how a siphon actually functions.......! Or have never made one with say garden hose, where how fat (or thin) the hose gets will tell you aproximately the relative pressure (or lack of relative to atmospheric pressure, at that point, as the siphon functions.....just by looking at the hose!

Actually, this Blog is now getting boring with our "Blogger in Chief" basically ignoring us and our efforts!!

Mostly agreed, except that blogger said no detectable leaks - and I think that a venturi in the valve at the bottom is just the sort of thing that would provide an increasing leakage rate as you stem the flow, and is unlikely to be detected.

Fyz

I agree air leaks is the no one contender but if not other causes must be investigated.

He declared somewhere that there are no leaks but air leaks are hard to spot.

The ±85' drop below your point p3 will introduce a lot of suction, turbulent flow etc and may be the cause of air leaks.

My one suggestion is to install an additional valve at the bottom that should be preset to have a zero pressure at p3 and a positive pressure in the 85' section.

another suggestion is to install a vent pipe at p3 this will have the same effect.

I prefer a footvalve at the top. The flap will be open when in operation at a cost of an additional few inches of friction. The only task of the footvalve is to prevent it from accidental loss of prime when not operating for a few days.

Did slong ever mentioned the diameter of the hose?

maybe it is there, Friction and losses must be included in calculations.

A vent at P3?? Surely hat is guaranteed to allow air to enter under the action of whatever vacuum is providing the excess "suction" that gives the flow some speed.

A footvalve at the top? - when and how is this to be closed - remember that the syphon "breaks" before the valve at the bottom is fully closed.

I think that everyone agrees that the problem has to be air or gas entering the syphon. The basic question is what mechanism would allow a rapid flow but cause failure as the flow is restricted from the bottom.

The vent is supposed to be at point p4 post #73 (I hope I got the P3 from another sketch)

P4 can however be dropped down to ±30' below the top. (experience cause me to limit to below max.)

The siphon is a complete separate unit and will work.

In this setup any leaks below p4 will not matter.

Air released out of the water may kill the siphon at any time and therefore disregarded as it is stated that the siphon only stops when the valve is closed.

The rapid flow in the bottom part of the pipe do not contribute to the siphon and this is proven because the siphon do not loose prime at full flow the siphon part is definitely without leaks.

Air sucked in will be discharged through the outlet (because of the velocity) or will be trapped in the bottom piece of pipe. Even a bubble of air in the pipe below p4 will be at less than atmospheric pressure and the siphon will not stop.

If the valve is closed when a bubble of air is trapped in that section the air may be able to move right up to the top and prime will be lost.

A preset valve will cause a positive pressure in the bottom part and zero pressure at P4. No chance of air entering the system.

Your intended location makes good sense - but I'm not certain it takes everything in the original question into account. The critical feature was that the syphon when breaks when the valve is partially closed, and water is still flowing - albeit at a reduced rate. That does allow the possibility that air (from whatever source) is swept out when the water is flowing fast, but is allowed to accumulate when the water slows. The only alternative that I can see is that the air can enters from the valve or below the valve when the flow is slower. As the velocity through the valve itself must actually increase as the valve is closed, this is unlikely unless there is turbulent mixing in the valve - this is more than possible if the valve was not designed to quench the flow with a head of 85'.

Your proposal to place the valve 30' below P2** and provide a vent immediately below the valve will avoid this excess pressure at the valve - but I'd still prefer to place the valve in a section of the pipe where the flow is upwards. The residual potential problem is leaks - but I think these are less likely to be problematic if there is never sufficient height of water to break the column (I'm assuming that surface tension will help with seals). However, it's not a solution of the problem is larger leaks or entrapped gas. But the height of the column below the valve makes that appear unlikely to me - and I'm not sure that there is a passive solution in that case....

Personally, I still favour placing the valve at the lowest point and in an upwards-facing section of pipe - preferably vented at the top of that section as you propose. So far as I can see that would require massive air leaks to fail - or a slow leak of water as well.

**I'd go for 25' myself, in case the water gets warm on a sunny day and the vapour pressure rises.

Fyz

I hate to say this, but I personally could make no sense at all from what you have written.......!!

Obviously English is not your prime language, I speak German fluently, if you can write better in German, I will translate it into understandable English for you.....

Sady my French & Portuguese have degraded too much to be able to make the offer in either of those languages.......

Hi Andy

It's highly cryptic - but I found it fine in the context of Hendrik's previous posts - and as an answer to questions. Obviously, English is not my first language either...

Functional description of proposal:
Refer to yesyen's diagrams
Resite P4, where the valve is, to 30' or less below P3 (vertical component)
Continue the pipe down to the tank*
Provide a pipe to vent the down-pipe that joins it just below the new valve position P4
*My note: you can now call the tank P5 if you like

Advantage: prevents the pressure in the syphon ever dropping below the vapour pressure of the water

Hope that helps

Fyz

Correction: apologies - I typed P3 when I meant P2 - i.e. the relevant line should read

Resite P4, where the valve is, to 30' or less below P2 (vertical component)

Fyz

Does this depict the current discussion?

It is getting there. The exceptions are that:
a) The idea of the vent pipe is to prevent the weight of water below P4 creating sufficient draw for the water to "boil" **. So it is counterproductive to place a valve in the vent pipe; and
b) The vent pipe itself should go an appreciable distance above P4, so there is no possibility that any of the water will be lost through the vent. It will be safe if it goes above P1, but that is probably further than is necessary.

As an additional safety factor, I would be inclined to bend the main flow so that it is heading upwards through P4. The main flow would then be bent back down again. I think the vent pipe is best placed at the highest point following the valve (i.e. where the pipe bends down again). This is the point that should be at most 30' below P2

** It seems to me that there are two reasons for wishing to avoid the water "boiling".
First, that absorbed/dissolved gases come out very rapidly when the water boils (possibly even somewhat before it, so 30' may prove to be too much if that is the cause of the problem).
Second, that surface tension will seal small leaks, which will prevent them being detected with pressure tests. Such leaks should mostly stay sealed so long as there is water in contact - even at significantly reduced pressure.

Regards

Fyz

P.S. It seems that a good picture is indeed worth 10^4 words...

Thanks Physicist.

The problem is with moving the valve is the fact that all the water under the valve will leave the pipe, even if it is under water in the bottom tank, the drop (70 feet?) is so grat that atmospheric pressure cannot support that weight (anything over 32 odd feet!)

It will also not fix the problems seen as they are due to either minute airleaks in the first 30 odd feet of pipe, or to too much gas dissolved in the water....

I will be signing off from this blog anyway as we have discussed it to hell and beyond and the owner is long gone......

The objective is to avoid the pressure at the top falling too close to the vapour pressure of water. The reasons this looks hopeful are that it minimises two effects that appear when the ambient pressure is equal to or below the saturated vapour pressure:

The originator said there were no leaks. I take that to mean that pressure testing didn't show any. Micro-leaks can remain sealed for long periods by water - but the water will evaporate rapidly (boil) if the local pressure goes below the vapour pressure, which will open these leaks (and they mightn't be as "micro" as all that).

Similarly, as the pressure falls towards the saturated vapour pressure, bubbles of absorbed gases desorb and then expand rapidly. With increased size comes increased buoyancy that allows the gas to separate rapidly from the stream. Again, limiting the "pull" of the column so that the pressure is always at least twice the vapour pressure will dramatically reduce this problem.

Obviously, in the absence of further information, we cannot guarantee that this will work - but I think it has more chance than anything that we had seen previously.

Fyz

My German is even worse (Can understand technical material)

My intentions were:

1 The vent pipe must extend to above the maximum water level at P1

2 No valve is needed at P4

motivation

Slong may not want to climb up the hill to close the valve.

When the valve is closed at P5 the siphon will keep on flowing until the water level in the vent reaches the same level as P1.

3 It may even be better to install a second valve at P5. should be preset to limit the flow, ensuring a positive pressure below P4.