Somehow I have air in my plumbing?

This just started happening recently.  Primarily one toilet it gets bubbles of air in the water supply which spit and sputter into the tank as it refills.  85% of the problem is at one toilet in the house (by far the highest use toilet).  The remaining spits/sputters happen at the kitchen faucet.  The missus has determined that it seems to happen after using the outside frost proof faucet which is right outside the kitchen.  It's been down below -15F over the past month and I've used that outside faucet lately in extremely cold temps.  The water comes the spigot and some comes out the top of the breather (or whatever it is) of the faucet before the "warm" tap water melts things and all the water comes out the spigot hole.  The toilet is on the other side of the house from the kitchen.  I just took a shower and had a bit of spurting air in there for the first time.  But all the other points of use have been normal.

We have a sandpoint driven well in the basement.  The well is probably 50 years old and the pump and pressure tank are likely 10 years old.  I added some pressure to the tank this past summer using the schrader valve.  I haven't touched the plumbing (other than that pressure tank adjustment) in over a year.

Any ideas where this air could be coming from???

It sound like the outside pipe is draining the pipe inside the house and introducing air into the pipeline causing an airlock.

It does not sound like a venting issue to me because the air bubbles are on the supply side of the water issue, not the drain side of things.

I had this happen to me, and it was a loose fitting on the incoming waterline coming from the submerged pump in the well. The loosened hose clamp allowed air to be introduced, and so every time the pump shut off, water was getting a vent and allowing the water to fall back down the well. Then when the pump came back on, it had to push that air somewhere, so it came out just before the water did.

You have the same problem by the sounds, but it is different in that you do not have a submerged pump. Still, you have air being introduced just ahead of the water being pushed by the pump, and the problem is getting worse.

You do not say whether your system is single pipe or dual, or in other words, shallow or deep, but I suspect shallow well since it is a sand point driven in sand, and thus single pipe. If that is the case your check valve is probably shot. Having that go would do the same thing as my air leak: allow the water to fall back down the well, and when the pump comes back on, there is air ahead of it the water. That is quick, easy and cheap to fix so I would start with that.

Other than that, all I can thing is that you have a bad foot valve, but I am not very familiar with jet pump water systems I admit.

The problem does not sound like it could be the frost proof outside spigot because that is nothing more then a remote valve deep inside the home where it is warm, and the valve body does not freeze. There is no way for air to be introduced into the system because it is under pressure, a failed valve only leaks.

Ditto what Travis said - look into your supply side/check valve. I had same problem on my dug well, turned out to be tiny pin holes in the old copper line that ran from the well to the pump (single feed, no return). When the pump turned on it would suck air into the line (just like a drinking straw with small hole in it). I replaced the line and problem went away.

I was going to say, "Digestion issues?"

From what I can glean from the above, you have a path forward. I hope it yields definitive answers that involve cheap, easy, long-lasting fixes.

-CK

Travis' post is excellent. My thinking also, since the plumbing is constantly under pressure, the air is getting in somewhere early in the system. Has the water table dropped? Is this something you can measure and compare to past measurements? Also, I'm not completely familiar with sand point wells, is the pump submersed down in the well or is it dry, with a line that goes down into the well that feeds the pump?

Thanks team!  I took a photo this morning.  The sand point comes out of the ground and goes to a check valve (brass).  I doubt there's a foot valve at the bottom.  Then to the pump.  The water goes up to the pressure gauge and then to the right to the pressure tank.  Water line goes up just after the pressure gauge.  The kitchen is right above this pump so I can see how air should get there first.  The toilet is a long way and there are many other fixtures that would see the air first due to spots where the line to the toilet drops relative to their supply.  Unless the air bubbles don't rise up into vertical pipe runs and they just follow the flow of the water?

I can definitely see how a tiny leak in the plumbing between the check valve and the water table would allow air to suck into the pipe due to the weight of the water pulling.  I'm thinking if the check valve is bad the water pressure would just push down into the well and it wouldn't involve air at all.  It is old galvanized piping so a leak is definitely possible.  I just hope it's above ground where I can get at it.  I wonder if I could goop up the fittings with something to seal the air leak (if it exists)?

Bonus:  For those who don't know what a sand point is, it's a simple driven well.  You get a "point" from the hardware store which is a pointy piece of heavy metal with a fine mesh cylindrical screen above it.  That screws onto your pipe and you drive it down into the sand.  Keep adding pipe and couplings until you hit water.  Then you suck on the top of it with a pump to get water out.

Mike Haas wrote:

I can definitely see how a tiny leak in the plumbing between the check valve and the water table would allow air to suck into the pipe due to the weight of the water pulling.  I'm thinking if the check valve is bad the water pressure would just push down into the well and it wouldn't involve air at all.

I can, with almost certainty, guarantee the air is coming in somewhere on the suction side from the pump back down to the well. Everything on the other side of the pump should be under constant pressure supplied by the pressure tank, and air will not infiltrate on that side of the plumbing.

It is old galvanized piping so a leak is definitely possible.  I just hope it's above ground where I can get at it.  I wonder if I could goop up the fittings with something to seal the air leak (if it exists)?

Yes you can, and it involves unscrewing the fitting, applying pipe thread compound (there are many flavors to choose from), and sewing them back together. There is also a tape version that is wrapped on male pipe threads instead of goop.

So...  Is there pressure on the input side of the pump, between it and the brass check valve?  I'm not actually sure but I'd think so (once the pump turns off and is idle).  So my hunch is that the leak is before the check valve.  

Thanks James, I'm sure if I disassembled everything I could use pipe dope and seal it back up successfully.  I'm a bit worried about taking it apart and then not being able to reprime the line to the pump.  Would the pump be able to draw water up through the empty well pipe?  

I was trying to ask if I could use a more redneck solution like pasting something on the fittings to seal up the tiny air leak.

Mike Haasl wrote:So...  Is there pressure on the input side of the pump, between it and the brass check valve?  I'm not actually sure but I'd think so (once the pump turns off and is idle).  So my hunch is that the leak is before the check valve.  

It depends. A pump may have a built in check valve right at the output side. I looked closer at your picture and all I can see on top of the pump is a reducing bell, and nothing that looks like a check valve. And upon closer viewing, have you inspected that union? It appears to my eye that it is slightly tweaked, with the left side level, and the right side heading slightly uphill to the elbow. If it's not dripping, then no air is coming in at the union. I mention the union because I've had difficulty with them sealing properly if they're not aligned just about perfectly.

If we assume there is pressure within the pump going back to the check valve, and there are no drips, then indeed there is a leak between the check valve and the point or the water level in the ground has dropped below the top of the screen on the point, and is pulling air in there.

Mike Haasl wrote: I'm a bit worried about taking it apart and then not being able to reprime the line to the pump.  Would the pump be able to draw water up through the empty well pipe?  

It depends on the kind of pump. Some water pumps can do this and pump air, and a good example is a septic pumping truck. Those guys throw an empty like 5 inch diameter hose down into a septic tank and turn on the pump, and it can pull all that air out of the hose.

I was trying to ask if I could use a more redneck solution like pasting something on the fittings to seal up the tiny air leak.

It's certainly worth a shot. I'm guessing if air is being pulled in when the pump is on, then water is leaking out when the pump is off. Getting the leak dry so some sort of goop to bond and stay in place may be challenging. What comes to mind for something that might work is that jb-weld epoxy stuff, and abrading the metal surface with sandpaper or something will help it bond to the metal as long as it can stay dry while it cures. Keep in mind though that if something like this does work, it is really a bandaid and I think it is a matter of time before if fails and starts leaking again.

You have one of two problems:

1. An air leak between the check valve and the water in the well
2. Sand particles in the check valve partially keeping it open

You can tell how bad the air leak is by how much air is present between when the pump comes on. If you run the pump for a few minutes, then shut it off, wait a minute, and then turn it back on...if there are no air bubbles it is a teeny, tiny leak, like you will not be able to see the pin hole leak. But if you do that same thing and there is a lot of air, then you will know it is significant. But if you do that same thing, but wait a half hour before you run the pump again, and there is quite a bit of air...well you can hopefully see the amount of air, by the amount of minutes that lapse between the pump coming on, will indicate how big of an air leak you have. In can not be stated enough, IT WILL BE TINY...

Now the leak HAS to be between the check valve and the water line in the well. IT HAS TO BE. It cannot be lower than the water in the well, because there is no air there. And it cannot be past the check valve because that is what the check valve does; hold pressure. So your issue resides between those two points.

The best way to check, is to get everyone out of the house, turn everything off, and het the house as quiet as you can. Then get the pump to turn on, and then shut it off.

NOW LISTEN

If you have an air leak, you will hear it whistle. This whistle is the water dropping down the well, and creating a vacuum through the pin hole. It will be small, so it will be slight, but it should have a whistle.

If you do not hear the whistle, it is most likely a piece of sand in the check valve not allowing it to close. In that case you will not hear a whistle because it is just not doing its job.

I am betting almost 90%, that your problem is actually in the check valve. They are super cheap, I think $10, and easy to replace. Yours is not too bad because it is located between a Union and the elbow going down the well.

Here is your check valve...

By the way, when listening for a whistle, you can use a stethoscope, even a Fisher Price Stethoscope will work. An ear held to the handle of a screwdriver held against the pipe will work.

Ok, I'll listen for the whistle  Based on all the conjecture, I'm pretty sure it's a pin hole leak below the check valve.  

If there was a little piece of sand in the check valve, the pressure tank would be slowly pushing water back down into the well.  If that was happening, there wouldn't be air in the line and the pump would be coming on randomly (I think).

Mike Haasl wrote:Ok, I'll listen for the whistle  Based on all the conjecture, I'm pretty sure it's a pin hole leak below the check valve.  

If there was a little piece of sand in the check valve, the pressure tank would be slowly pushing water back down into the well.  If that was happening, there wouldn't be air in the line and the pump would be coming on randomly (I think).

If the check valve is closing, but not sealing, then you could still build pressure and have the pump shut off. Now, I think what Travis said is the MOST likely. But a grain of sand in the check valve, plus a toilet fill valve that needs sufficient pressure to seal properly is also possible. If enough water drained through the check valve to reduce the pressure, AND the toilet fill valve lost its seal at that time, that would suck in air at the toilet.

Great troubleshooting tips that folks have shared!

I see you've got a pressure gauge on there.  If all your faucets are closed, how quickly (if the power to the pump is turned off) does the pressure drop?  If it is holding fine, then you aren't losing water back down your well pipe through the check valve.

I'm pretty sure I'd hear the pump come on if it did at night.  And that isn't happening.  So I'm nearly sure it isn't leaking back past the check valve.

The missus dug up a stethoscope and I listened around all the joints (round speaker thing removed and trying four spots around each joint with the end of the rubber hose).  I didn't hear any noise anywhere....

I came up with another possibility.  Might some air be leaking through the pressure tank membrane into the water supply?  If so, is there a way to identify that or troubleshoot if that's happening?

Mike Haasl wrote:I came up with another possibility.  Might some air be leaking through the pressure tank membrane into the water supply?  If so, is there a way to identify that or troubleshoot if that's happening?

It could be, but you would know it pretty quickly because the pump would be cycling a lot. I also would think you would be showing a loss of pressure in the system since you would only be getting pressure when the pump was on. Your gauge is either stuck in the picture, or the pump was just run because it is showing pressure. But if you really wanted to tell for sure, you could put a tire gauge on the valve stem of the cold water tank and see how many pounds of pressure it has.

I sooner think you have a problem in the suction side of your pump from the water level in the well, to the check valve, or a bad check valve itself.

You could have a problem down your well too. This is still the same problem described in the sentence above. It depends on how your plumbing is put together. Most places use plastic pipe since it comes in long lengths, but in my grandmothers house, her well was old and so it had lengths of 10 foot pipes coupled together. If yours is like that, then you might have a pin hole in one of the couplings. If the water in your well is 12 feet down, and there is a coupling at the 10 foot level, and has a pin hole, you would have the problem you are having. I would have thought listening with a stethoscope down the well would have picked up a whistle, but again it will be ever slight, and hard to hear.

In some ways this is the more likely place you will have a pin hole. That is because if you have a coupling at the 10 foot depth, and the water in your well changes from 9 feet to 15 feet in depth, that means the coupling is subjected to cycles of water and then air. It also means the rising and falling water will scour away the zinc coating if your pipes are galvanized. Keep in mind, this has been down that hole for fifty years, that is a long time.

Myself, if I really wanted to get rid of the problem, I would tear apart and replace everything from the reducer going to the well, all the way back to the union, including the check valve of course. That is cheap, easy and quick to do, and is 90% likely to take care of your problem. What is the worse that can happen; you spend $20 in parts, and have an hour of wrenching? If you replace what I have circled in green, put pipe-dope on the threads, you will have eliminated 12 potential pin hole leaks alone.

Or you can just wait, and if the problem gets bigger, it will be easier to find.

I have blown this picture up, and really tried to see in better detail, but I cannot say for sure, but there almost looks like the two circles are witness marks...or in this case, a drip. I cannot quite tell, but if this was my pump, I would be looking and listening really hard at this spot (the inlet of the check valve).

Thanks Travis!  I'll check the pressure tank pressure.  I can't remember what I pumped it up to.  I think I turned off the well and let the water pressure drop to nothing and then pumped the tank up to 28 or 32 or something like that.  It had been at 18 so at some point or somehow it had lost that pressure.

That witness mark in the upper circle is just a drip of the pipe dope from the original installation.  I can't tell what the lower mark is since I've pulled off the cover.  Nothing onerous looking on the cover though.

Once the pressure tank has drained, I'll check the pressure and report back.

Do you know if this pump can suck the water back up the well (reprime itself) once I've torn apart those fittings and replaced them?  I'd hate to tear apart the plumbing and not be able to get it working again (angry wife, uncomfortable life)...

I am slightly concerned that the leak is at a fitting underground.  But then again, how much air is down there?  The well casing is galvanized pipe so I'm guessing they drove it in 5 or 10' lengths.  The last length is just sticking out of the slab so the highest buried joint is at least 4.5" underground.  It's under the house so it's probably pretty dry soil down there.  And definitely sandy.

If this is still ongoing ... would you humour me and see what happens when you hold a lit match near the bubbles? Maybe a lit match held by metal tongs, with all flammable objects kept away?

My immediate thought for a gas in well water in the rural US is methane from fracking. I'm probably wrong (I don't even know if you have fracking nearby), and probably an alarmist, but it's a 2 min test, and something you really do want to know before you start tearing stuff apart or welding.

Edit - look up "flame test for gas" but if it's methane, it should be a fireball, oxygen, should make the flame brighter, CO2 and other non flammable gases should extinguish the flame - I've only ever heard of methane, hydrogen sulphide (which you would probably smell),  and carbon dioxide as gases in water underground.

Good idea Catie!  I don't think there's any fracking within about 200 miles of me, and it's a shallow well so I doubt that's the problem.  But I'll give it a shot.

I do have new information though!  I let the pressure tank empty (pump turned off) and a lot of air came out of the kitchen faucet as the pressure petered out.  I tested the pressure tank pressure and it was at zero.  Hmm.  I turned the pump back on and it pressurized the water line (and the schraeder valve on the pressure tank).  So I'm thinking it might be the pressure tank???

Here in the South (and probably elsewhere) the water system pressure tanks often did not have bladders in them.  Instead the air that served as a cushion above the water in the tank and which provides for the compressibility to pressurize the system was introduced by a small weep hole placed  in the well pipe going down into the well.  With each pump cycle some water would leak out into the well which then allowed the introduction of a little air into the line which in turn would get pumped up into the pressure tank.  The reason this was necessary is because without a bladder, the air in a pressure tank slowly gets absorbed or mixed with the water and disappears from the tank...  The result of this is a "waterlogged" tank and a pump that cycles too frequently.  One could replace the lost air by periodically pumping some in.  The advent of bladder tanks prevented the mixing of the air and water which in turn made recharging the air unnecessary.  It made for a simpler system,.. as in less to go wrong...   Nonetheless the bladders do break and leak.  I still have a non bladder system and like it,...  An interesting thing about it is that each time the pump starts a bit of air is ejected from the tank by a  float valve designed to keep the tank from getting too much air in it.  This is a long reply to this post about air in your system but it illustrates how the air is getting into your lines...  You have a leak in the line going up from the well to your check valve.  It is introducing air into your water.  In my system it is on purpose,...  in your case it is accidental.  You need to pull that line and patch the leak...  Unfortunately if it is a driven sand point line that may be impossible...  A simpler thing might be to put a non-bladder pressure tank in there that would allow for the valve that would bleed off the excess air.

Mike, do you have a water softener after the pressure tank? That often will have a pump that generates pressure in the desulfuring tank with another check valve that can be the problem. Ask me how I know...

Mike Haasl wrote:
I do have new information though!  I let the pressure tank empty (pump turned off) and a lot of air came out of the kitchen faucet as the pressure petered out.  I tested the pressure tank pressure and it was at zero.  Hmm.  I turned the pump back on and it pressurized the water line (and the schraeder valve on the pressure tank).  So I'm thinking it might be the pressure tank???

With this new information, I do believe this is the culprit, especially since you mentioned in the original post that air in your lines this just started happening recently. David's post above is excellent, and I won't repeat anything. I think the bladder in your pressure tank has failed, and it's time for a new one. This to me seems the best place to start, but I think it's also possible there could still be a leak between the check valve and the point. How old is the pressure tank?

I agree...if you have no pressure in the tank, then the bladder has a leak in it. I would start by replacing that first. That is a really quick and easy job and they only cost $50 or so.

I would have thought your pump would have been cycling a lot, because that is what they do when they are waterlogged, but I admit it is hard to diagnosis a problem from afar because you cannot see and hear things.

Thanks team, I'd much rather change out the pressure tank (less to go wrong) and see how that works.  I'll do that today.  I don't know how old it is, it came with the house so it's at least 6 years old.  I'm not sure why the pump isn't coming on and off a lot, maybe there's still 1psi in the bladder that it's compressing.

Since I let all the air out of the system, there hasn't been as much burbling out into the toilet tank.

I don't have a water softener, it goes straight from the photo above into the cold water network.

Good to know David, I didn't know about that kind of tank.  I am pretty sure this one has a bladder.  

Ok, it seems like it was the tank.  I took it out and after draining all the water out, it still had a quart or two sloshing around in it.  I'm guessing it was on the wrong side of the bladder.

In removing it, I noticed that the galvanized piping was looking pretty blocked.  I ended up replacing everything after the pressure gauge.  You could barely see a spot of light when looking through the 15" long assembly.  That may explain why we don't seem to have tremendous flow at the taps...

Put it all back together and only had one leak :)   Now I'm just giving it time to see if any other leaks develop before putting my tools away.  I'll probably wait a couple days before claiming victory on the air issue.

Travis Johnson wrote:I agree...if you have no pressure in the tank, then the bladder has a leak in it. I would start by replacing that first. That is a really quick and easy job and they only cost $50 or so.

I wish.  $149.00 plus tax...

That style of pump has to be primed if the drop pipe is empty. I just keep 2 5g buckets full and use a funnel to fill the pump and pipe. I put a t on top instead of a elbow, pull plug fill drop pipe. Not sure on the sand point, but most well systems need a check valve at the bottom of the drop pipe. This could be built into the Sandpoint, mine hase a foot valve.

We have 2 55g barrels in the basement, which we use to store water. Pump problems, power out, we have enough water for our house for 2 weeks. This is also the water we use it we lose prime in our drop pipe. Usually takes about 5g, but if you don't have it might not get the pump working.

We also have some nice small ponds in the front yard. 1500g of water to flush the toilet. Plan ahead and the disaster is just a change of plans.

Thanks
Brian
3HR

Thanks Brian, the thing I'm confused about is how you actually get water into the line from the check valve down to the water table.  If I had a tee with a plug, as soon as I remove the plug, any water in the line will drop to the water table.  If I pour in 500 gallons of water, it will still just drop to the water table before I can get the plug back in. Unless I had a ball valve above the tee and I could deluge that pipe with water and shut the valve to hopefully catch it all in the line.

Unless of course my sand point has a foot valve.  But I don't know if it does and I suspect it doesn't.  Otherwise why would they have also put in a check valve?  Maybe for insurance but I'm not sure enough to just willy nilly dig into the plumbing below that check valve.

Getting a few bubbles out at the toilet still but I'm hoping that's just from exhausting air from the work I did yesterday.

I think I can officially claim success (knock on wood).  No more air in the lines and the pressure is better than ever (due to replacing the partially blocked pipes).  Thanks everyone for your ideas and helping me figure this out!