# Water Pipe Boyle's Law Mystery.

1. Jul 14, 2007

### CharlesHoward

This is a very curious situation.

I have a large water tank on the top of a hill.
It supplies a gravity fed water system that works fine.
There is an under ground pipe with one end A in the bottom of the tank.
The other end E is about 80 feet below A, and has a faucet.
None of the pipe is above A.
If you open E, then water flows out with good pressure as you expect.
If E is open and the tank empties, air gets into the pipe at A,
and eventually the water stops coming out of E.
All seems normal.
Except ...
If the tank at A is now filled, no water comes out of the open faucet E.

What is the simplest shape of pipe that will cause this?

I am sure the answer involves vapor lock and the fact you cannot siphon water over a 32 foot rise, but thats it.

2. Jul 14, 2007

### Danger

Welcome to PF, Charles.
I have no education, particularly in fluid dynamics (and I've never heard of Boyle's Law), so I'm just going to throw something out. Have you tried refilling the tank with the faucet at 'E' open? Anyhow, Clausius II is probably your best guy to answer this.

3. Jul 14, 2007

### CharlesHoward

Tank was filled with E open. That is the oddity.

4. Jul 14, 2007

### FredGarvin

What size are the tank and the pipe going from A to E?

It's not really vapor lock per se. You are not experiencing evaporation of the source. What I think may be happening is that you have a large air bubble trapped in the pipe after you fill the tank. You open the valve and get no water. Can you hear air flowing out of the valve? I would think you need to purge that line and with a large pipe in relation to the size of the valve, it may take a good amount of time to purge that much air. With a full tank's head pressure on the air bubble, there has to be some decent pressure in that line at that point.

Next time you fill the tank, is there any way to stick a hose down the pipe to prime the pipe before you fill the tank? I'd also open the valve and let it stay open for a good, long time.

Last edited: Jul 14, 2007
5. Jul 14, 2007

### Danger

Sorry that I misunderstood about the state of 'E' upon filling, Charles.
This might sound as if I'm trying to reclaim my comedy badge, but I'm actually quite serious: have you checked the pipe for dead squirrels? If the tank was at all open to the environment, anything that might have fallen in could have gone down the drain when it emptied. I suppose that you have a sump screen of some sort, but it might not be adequate.

6. Jul 14, 2007

### CharlesHoward

I do not know the answer to this problem, but I expect the simplest pipe shape I am interested in that would cause the original behavior at E would not have the cycles.

I really like this puzzle and think it would be a great problem for an advanced physics class.

7. Jul 14, 2007

### CharlesHoward

Danger, the system is not open to the environment other than to air through a very fine mesh screen. This is, I think, a real and hard hydrodynamics problem. Watch postings by FredGarvin.

I negected to mention to him that am morally certain the answer lies on the fact that you cannot siphon water over a 32 foot rise, and that when the system stabilizes after opening E there can be several sections in vacuum and the zero pressure points can be at varying levels.

Last edited: Jul 14, 2007
8. Jul 14, 2007

### Danger

I must be missing something here; I don't see what siphoning has to do with this situation at all. None of your water is going against gravity at any point, is it?

9. Sep 20, 2010

### agentfox

You have an air lock. If your pipe at any point creates a high point (looks like its going over a hill in a profile view) you create a situation where air gets trapped in the high point and does not allow water to flow past that point. You have to install a simple and cheep air/vac valve at the point where the "hill" is. It doesn't matter if all points are below "A" if any point creates a space for air to get trapped.

10. Sep 21, 2010

### maimonides

As long as air comes out of your faucet, you simply need to wait.
32mm diameter gives about .7l/m or 70l/100m.
Now what is your flow rate in normal operation? How fast can you fill a bucket?
1l/sec (10 secs for a 10l bucket) means a minimum delay of 70sec per 100m(or yards) of pipe installed.
Since 10000gallons are swimming pool size, I expect you have at least 5feet water level in your tank. If your assumption "None of the pipe is above A" (and A is the bottom of the tank) is correct, siphoning or airlocks play no role.
If things really stop, you might want to put a pressure gauge at E to see what´s happening.

Edit:
Afterthought: What else is there in your plumbing that might give trouble? Air release valve?

Last edited: Sep 21, 2010
11. Sep 22, 2010

### maimonides

Sorry, I was a bit dogmatic.
I assumed air under water is unstable (it goes up).
If water over air is (meta)stable there is an explanation for the OP and an answer to the "simplest shape" question.

|
|
|
|
| ----____
| |
| |
| |
| |
\ /

| tube; | air filled tube; | water filled tube
Of course the right side need not be vertical, it may be sloped.

Last edited: Sep 22, 2010