Please explain curious behavior with my rainwater barrel

[DannySmythe]

I have a barrel that collects rainwater from the gutters on my house. At the bottom of the barrel is a manual 'faucet' with a round handle for turning the flow on and off. On that spigot, I have installed a "Y" fitting.
One side of the "Y" goes to the irrigation hose which I turn on when the garden gets dry and there is water in the barrel. This irrigation hose goes down a slope and is always below the bottom of the rainbarrel.
The other side of the "Y" goes to the bottom of a clear 1/2" plastic tube that is mounted alongside the rainbarrel so I can see how much water is in the rainbarrel without opening it. The top of the tube is open.
Now comes the curious thing. With the rainbarrel full and the plastic tube showing full, I turn on the 'Irrigation' side of the "Y". The water in the plastic tube immediately drains all water and stays empty while the irrigation is on. When I turn off the irrigation, the plastic tube fills back up.
So I have 2 questions.
1. How do I get around this? I really want to know the water level as I am irrigating so I can use a certain amount and then turn off irrigation. I might want to use 10% every day until empty.
2. How can this be explained?

 

[DEvens]

The explanation is that when you open the valve, the pressure at the bottom of the clear tube changes. This likely has two components.

First, for the water to flow there has to be a pressure drop. The place the water is moving to has to be lower pressure than the place it comes from. The water is flowing out of the barrel so the pressure at the valve has to be lower than the pressure just inside the barrel. As well the pressure has to get steadily lower until the water exits your irrigation system. Otherwise, water would not flow in that direction. So opening the valve means the pressure has to drop all along the path of water flow.

Second, Bernoulli has something to say about it.

http://en.wikipedia.org/wiki/Bernoulli's_principle

 

[DannySmythe]

Thanks for the reply.
I'm not convinced of the explanation though.

I don't see how the 'distributed pressure drop' explains this behavior. For the tube to drain completely means that the pressure at the 'Y' is 0 (relative). How can this be? This means I could start cutting the tube down to the Y and still have no water. Then I could look into the Y and see the water flowing without any leaking out! There is something else going on here.

I haven't done any calculations, but it doesn't seem like the Bernoulli effect is significant enough. Maybe so though...

 

[Travis_King]

Likely if you put a valve there and leave it only partially open you will not see this issue.

The water is acting as a siphon. All the pressure has to be at the wye is less than atmospheric. If the atmospheric pressure is greater than the pressure at the wye plus the pressure drop in the tube and fittings, then you will see the water drain. The siphon effect is very strong. In piping systems were siphonage occurs, water can "suck" a vertical column of water up to ~30 ft depending on altitude, piping geometry and material, and atmospheric conditions. So it's quite easy to imagine that it could "suck" the water from the vertical column that's already at a higher elevation.
http://en.wikipedia.org/wiki/Siphon

Additionally, what is likely occurring is something like a jet pump or ejector, where the higher velocity liquid from the tank creates a sort-of venturi effect, causing a motive force for the liquid in the vertical column.
http://en.wikipedia.org/wiki/Injector

 

[OmCheeto]

Thanks for the reply.
I'm not convinced of the explanation though.

I don't see how the 'distributed pressure drop' explains this behavior. For the tube to drain completely means that the pressure at the 'Y' is 0 (relative). How can this be? This means I could start cutting the tube down to the Y and still have no water. Then I could look into the Y and see the water flowing without any leaking out! There is something else going on here.

I haven't done any calculations, but it doesn't seem like the Bernoulli effect is significant enough. Maybe so though...

My guess is combination of both "entrainment" and perhaps a slight "Bernoulli effect".
Wiki's first line on this, is jibberish to me; "Entrainment is the transport of fluid across an interface between two bodies of fluid by a shear induced turbulent flux."
But that's probably my fault.

In the navy, we had pumps called "eductors", which used a water stream as "the pump". The reason I think entrainment is predominant in your situation, can be shown by experiment:

Take a water hose, connected to your water main, put it in your rain barrel, turn the water on, full blast, holding the exit of the hose underwater, and point it out of your rain barrel. You will notice that your rain barrel will start to empty, implying that the water from the rain barrel, is somehow latching onto the flow of water out of the hose.

It's been about 35 years since I've studied this, but the following website seems to confirm what I remember about it:

http://web.clark.edu/ggrey/PHYStext/07Matter.html#Viscosity

Nonideal fluids have viscosity. This is a property which, when large in value, might be thought of as stickiness. An object moving through a viscous fluid tends to drag a surface layer of fluid with it.

Likewise, fluid flowing through a pipe will move more slowly because it adheres to the pipe surface. The boundary layer adhering to the pipe slows down the next layer in and so on, giving a velocity profile which changes from zero at the wall to a maximum speed in the center of the pipe.

Likewise, flowing water, due to this "stickiness", will drag the non-flowing water along with it.

 

[DannySmythe]

Likely if you put a valve there and leave it only partially open you will not see this issue.

The water is acting as a siphon. All the pressure has to be at the wye is less than atmospheric. If the atmospheric pressure is greater than the pressure at the wye plus the pressure drop in the tube and fittings, then you will see the water drain. The siphon effect is very strong. In piping systems were siphonage occurs, water can "suck" a vertical column of water up to ~30 ft depending on altitude, piping geometry and material, and atmospheric conditions. So it's quite easy to imagine that it could "suck" the water from the vertical column that's already at a higher elevation.
http://en.wikipedia.org/wiki/Siphon

Additionally, what is likely occurring is something like a jet pump or ejector, where the higher velocity liquid from the tank creates a sort-of venturi effect, causing a motive force for the liquid in the vertical column.
http://en.wikipedia.org/wiki/Injector

I don't have a siphon setup. There is no inverted U shape flow.
But the ejector idea seems like the right track.

 

[DannySmythe]

My guess is combination of both "entrainment" and perhaps a slight "Bernoulli effect".
Wiki's first line on this, is jibberish to me; "Entrainment is the transport of fluid across an interface between two bodies of fluid by a shear induced turbulent flux."
But that's probably my fault.

In the navy, we had pumps called "eductors", which used a water stream as "the pump". The reason I think entrainment is predominant in your situation, can be shown by experiment:

Take a water hose, connected to your water main, put it in your rain barrel, turn the water on, full blast, holding the exit of the hose underwater, and point it out of your rain barrel. You will notice that your rain barrel will start to empty, implying that the water from the rain barrel, is somehow latching onto the flow of water out of the hose.

It's been about 35 years since I've studied this, but the following website seems to confirm what I remember about it:
Likewise, flowing water, due to this "stickiness", will drag the non-flowing water along with it.

This makes perfect sense to me (not necessarily meaning it's correct).
I really like your example of the hose in the rainbarrel.
If this is the correct explanation, then I should be able to poke a hole in the barrel slightly above the bottom and attach my clear tube there. Since there is no rapid water movement, the level in the tube should be the same as the level in the barrel.
I'll give this a try and post the results.
Thanks to everyone!

 

[Travis_King]

It's one of the two effects. I suspect that it has more to do with the venturi effect as well, however the fittings in that arrangement may result in a partial vacuum when the fluid is allowed to flow, causing the fluid in the column to fill the space (and thus resulting in another partial vacuum behind it). True it is not a true siphon, but I meant that it may be the result of partial vacuum due to the gravitational potential of the water in the hose, rather than the result of the low pressure zone created by the transition of the liquid energy from gravitational potential to kinetic energy (velocity) which is what the venturi effect describes. Though again, I suspect it's the latter.

And you are correct. Whichever phenomenon is responsible for draining the view tube, if you de-couple the view tube from the hose outlet, you will not see that issue.

 

[OmCheeto]

...
1. How do I get around this? I really want to know the water level as I am irrigating so I can use a certain amount and then turn off irrigation. I might want to use 10% every day until empty.
...

I can think of 4 options, two involve removing the "level gauge" from the existing "Y" connector.
1. Install a second faucet for the level gauge, away from the draining faucet. (my preference).
2. Install a double u-tube level gauge(might get an air bubble at the top, rendering it useless)
3. Install a throttle valve on your existing level gauge. (Fun experiment!)
4. Install a siphon to fill the level gauge (least cost, least amount of work)

Doh!

This makes perfect sense to me (not necessarily meaning it's correct).

I hope I didn't imply somewhere, that I knew what I was talking about.

I'll give this a try and post the results.
Thanks to everyone!

Yay! Science!

 

[DannySmythe]

I did some quick experiments with the rainbarrel which is currently full.
1. Both sides of the Y are wide open. By raising and lowering the end of the irrigation hose and always keeping the rest of the hose below the Y, the level gauge follows the level of the hose end.
2. Both sides of the Y are wide open. By raising and lowering the middle of the irrigation hose and always keeping the end of the hose 3' below the Y, the level gauge follows the level of the hose middle.
3. The level gauge side of the Y is wide open and the hose follows about 45 degree slope downhill with the end of the hose about 5' below the Y, the level gauge shows empty. By gradually closing the Irrigation side of the Y, the level gauge slowly fills back up to 'full' when the valve is finally off.
I'm not sure what this proves to be honest.
When the barrel is empty (in about 2 weeks) I will install a new tap near the bottom of the barrel on the other side from the faucet. I'm hoping this arrangement will give me an accurate level regardless of the irrigation flow rate.
I'll keep you posted.

 

[OmCheeto]

I did some quick experiments with the rainbarrel which is currently full.
1. Both sides of the Y are wide open. By raising and lowering the end of the irrigation hose and always keeping the rest of the hose below the Y, the level gauge follows the level of the hose end.
2. Both sides of the Y are wide open. By raising and lowering the middle of the irrigation hose and always keeping the end of the hose 3' below the Y, the level gauge follows the level of the hose middle.
3. The level gauge side of the Y is wide open and the hose follows about 45 degree slope downhill with the end of the hose about 5' below the Y, the level gauge shows empty. By gradually closing the Irrigation side of the Y, the level gauge slowly fills back up to 'full' when the valve is finally off.
I'm not sure what this proves to be honest.

From the data you've collected so far, I'd say you've proven nothing!
But that's ok. I've done lots of experiments that proved absolutely nothing.

#2 makes no sense whatsoever to me. And that's alway a good thing. It tells me that you somehow changed the system, other than simply raising and lowering the middle of the hose.

When the barrel is empty (in about 2 weeks) I will install a new tap near the bottom of the barrel on the other side from the faucet. I'm hoping this arrangement will give me an accurate level regardless of the irrigation flow rate.
I'll keep you posted.

It should.

 

[Baluncore]

There is another important effect here that I think, after a quick scan of the thread, has been missed. That is the “vena contracta” or “pinch effect” in the pipe where water flows from the tank, into the faucet and “Y” fitting. The water must change from slow radial flow to fast linear flow as it enters the faucet, momentum and viscosity pinch the flow.
See: http://en.wikipedia.org/wiki/Vena_contracta
In effect that is a restriction orifice that drops much of the head pressure of the water in the tank. The column of water in the hose pipe draws water from the sight tube because of the obstruction at the vena contracta.

I expect you get air entrained in the water from the hose when the sight tube is drawn low by high hose flow.