Hey everybody,

I'm a journeyman plumber in the state of Montana, and I have a real head scratcher on my hands.

I installed a well pump and pressure tank at one of my customer's houses a few days ago. I have been working on systems like this for ten years, so not my first rodeo. Everything that could go wrong went wrong in the process. The new pressure tank had a problem from the factory and had a leak, so I replaced it. But the pump itself is baffling me.

The customer had an old Sears pump, 1/2 horse shallow well with 115v power supply connected to a 30/50 pressure switch. The reason for replacing these things is the pump is 20 yrs old, the pressure tank 12 yrs old and the pressure tank's bladder had failed. The customer figured why not replace the whole thing. I replaced the pressure tank and set the psi in the tank to 28 lbs, considering the 30/50 pressure switch. I then replaced the Sears pump with a Meyers 1/2 horse that was factory preset to 115v. The pump ran exactly one time. When I bled the pressure back down to get the air out, the pump would kick on and hum like the impellar was stuck, and then kick out on thermal overload. Whatever, I figured, faulty pump from the factory. Weirder things have happened. I take the pump back, but this time I buy a higher end A.O. Smith pump. Once again, 1/2 horse 115v. Both times they come with a new pressure switch, both of them being 30/50.

I install this one, and it works. Bleed the air out, it turns on at 30, off at 50. No leaks, no drips, no errors. A day later, the customer calls me. The pump won't shut off. I run out to his house and find that the pump will only build about 38 lbs. of pressure in the system. Now I'm blown away. Why would it work for a day and then not be able to build the pressure? I put a gauge on the pump as it's running and it's producing 30 lbs. of pressure as it is running. I attempt to bleed any air out of the system, there isn't any. So I figure, maybe that pump is faulty too. It happens to be Thanksgiving, so no parts supply houses open, at least not for a reasonable amount of $. So I take his old pump, which was working fine, and install it to get him through the weekend till I can get a new pump. I install it and it works fine. This time, however, I set the cut out pressure a little lower, to 45 lbs., figuring that should get him through the weekend. Now he calls me this morning, a day later, and once again, the pump system will only produce about 33 lbs. of pressure and won't shut off.

I have NO IDEA why the system will work for a day and then this problem shows up. Here are some things you probably want to know.

Well tank: 33 gallon Pentair tank set to 28 lbs.
Pump: 1/2 hp A.O. Smith, 1" Driscoll line in to pump, 3/4" copper line out piped into tank tee on pressure tank and then to house water.
I replaced the power cord as well. Could it be caused by not enough amperage to pump? Normally that causes heat and thermal overload issue
There is a check valve in place before the pump. I pulled the check valve and did a pressure test on the line. It holds and doesn't bleed off.
I haven't pulled the foot valve, basically because customer has no idea where it is, and there's 6" of snow on the ground. But the pressure test should have showed this being the problem if it was, right?
I don't know well depth or distance to well, but the pump systems are the same, horse for horse, volt for volt. So I have a hard time believing this would be a factor.

This is the first time in ten years that I have had this sort of problem. I hope someone out there might know something I don't!!

 

Here is an idea. I've been doing some research. Could this problem be caused by pump cavitation? I did reconfigure the layout a little bit, and the suction pump is shorter then 10x the diameter of the pipe before the first elbow. I heard that can cause cavitation. I've never had that problem before, but it sounds like a problem that is a perfect storm like problem. I did have to put two new barbed 90's in to make it work, but I shortened the line by about 6 ft., so I would think that any extra resistance caused by those 90's would be negated by the shorter pipe. Any thoughts on this?

 

Do you think it may be drawing air from some place?

 

It certainly doesn't exhibit the signs of sucking air. The pressure test on the suction line held, the pump isn't losing it's prime, and when I open the valves by the pump I don't get any air. BTW, when I install a pump system I always add a ball valve right off the top of the pump so that I can prime the pump if need be. That is the one I'm opening to see if it has any air in it. So I am literally right at the pump with no air there, as well as in the lines as well.

The more I read on this cavitation thing, the more it seems like that might be what's going on. It apparently is a hard to diagnose, intermittent problem that happens in a perfect storm like setting. The reason I think that is when I hook a hose to the top of the pump and let the water run while the rest of the system is valved off I get good flow directly from the pump. I put a pressure gauge on the hose, and the pump while running is producing 30 lbs. of static pressure, which should be more than enough to fill an expansion tank up to a 45 lb cutoff. What I think is happening, however, is that the suction side of the pump, due to this cavitation, can't overcome the static pressure on the discharge side of the pump and so the pump "hydro-locks" for lack of a better term. Basically, it hits a pressure at which it equilibriates, and then just runs forever. The reason I say this is because on the first pump I switched out, I tinkered with it for awhile, and sometimes it would run to about 38 lbs. of pressure and sit, others it would run to about 32 lbs. of pressure and sit, and so on. Where it decided to sit and run without building more pressure seemed to change, so it sounds like it wouldn't be a pump issue.

I know this is an oddball, more rare problem, and I hope there is someone out there who has run into this before...

 

I'm not a pro, just a DIYer that has screwed things up before. Maybe you can learn from my mistakes.

You could be overpumping the well and running it dry. That would cause cavitation. The old pump may have been so weak that the well had no trouble keeping up.

Or there might be a leak on the intake side. That could cause cavitation, too. If a pump is cavitating, the body might heat up and the pump will sound "funny" like there is gravel inside it. If it heats up too much, even once, it will ruin the threads on the PVC and that will cause a leak. Only good way to fix that is to change the fitting. Tape won't do it (take my word for it.)

If you have a T and a valve ahead of the tank, (and it sounds like you do) you can let the pump run free and then you will see pops and bursts of air and water if it is present. Let it run for two or three minutes to see if you are overpumping. More time if the guy will be irrigating since all the water in the tank will be exhausted and the only water he will get is from the pump if he leaves the hose on for 30 minutes.

You can test the intake joints for leaks by using the "shaving cream test." I've described it here in detail before, so search here or on the 'net. I like barbasol. Intake leaks are really hard to find, so take your time. A really tiny leak may let air build up in the pump.

Also, particularly on the intake side, when it comes to gluing PVC, I have found that it takes a full 24 hours before testing to get good results. I know what the can says, but every time I have gotten impatient and fired things up after, say 18 hours, I have had to glue it again because it leaked. Threaded PVC is also notorious for leaking, so try to minimize the use of that.

Make sure that the intake rises from the well to the pump. I pretend it works like a reverse drain. If the level pipe or hose rises above the pump, or it is even dead level, an air bubble can form in the intake pipe and stop the flow. So the bottom of the body of the pump should be located above the output of the well, which should be 18". I put my pump on concrete blocks about 2' up. This bubble may not happen right away.

I'm hitting all the high spots. Hope it helps. Have you read through our little sticky at the top? It might give you some ideas.

 

I appreciate your feedback.

There is no PVC in the line anywhere. I completely agree with you about PVC, even CPVC. The can actually states on the glue that "before a pressure test 24 hours must pass," It's one of the reasons I don't use PVC on pressurized water lines, even sprinkler lines. Not to mention the cold winters and the cold water wreak havoc on the tinsel strength of the PVC.

The gravel sound you mentioned I did hear upon firing up the second pump. I thought it was just a little debris from re-priming the pump, but like I said, I did have to do some reconfiguring to install the new system. The customer had an old Sears set up in the house where the pump mounted to the top of the pressure tank. They don't really install that set up very often anymore. I know they exist, but they are on their way out the door, so I installed a system where the pressure tank and the pump are separate units. By doing so, I shortened the intake pipe by about 6 ft. to get the pump closer to the floor, and added another 90. The piece of pipe between the 90 and the pump is only about 6 in. long, which isn't 10x the diameter of the 1 in" pipe. That's why I'm thinking I have some cavitation going on. The funny thing is that I installed the old pump again, and now it's doing the same thing. So something besides bad pumps has to be going on. I wonder if it's cavitating enough to damage the pump to where eventually it won't build to cutoff pressure anymore, or if under a perfect storm it just "hydro-locks."

The well does rise to the pump as well. Like I said, the weird thing here is that a brand new system will work for pretty much exactly a day, and then it won't build pressure anymore. Then when I immediately turn around and install a new pump, it works again, for 1 day. Could that pump cavitate, heat up to where something inside the pump breaks, and then not work anymore?

 

I'm not a big fan of more than one 90 between the well and the pump. When I am talking about a rise, I'm talking about cocking things a little so the pipe rises without an elbow. Think reverse "drain" on the intake side, not "pressure".

You want to make it as easy as possible for the pump to pull since pumps are much better at pushing than they are at pulling.

Maybe you can take and post some pictures? I dunno how to do that, but others have instructions. It could be something obvious that we could spot.

 

Do you have access to the well so you can measure the level? The recharge of the well might not be keeping up with the pump. The well may have a large enough volume to keep up with the pump for a day, then need a day to recharge. With the water level below the intake for the pump you would get cavitation. Just something to think about.