Main Water Feed - 20psi Pressure Drop When one Fixture is Open

[stoopalini]

I'm looking for assistance in trying to diagnose a 20psi pressure drop whenever a single fixture is opened. This can be a sink, an outdoor spicket, or a flushing toilet, etc ...

The house is in Texas, was built in 1998, is slab on ground, has a 3/4" PVC feed from the meter box to the house. The house then has 3/4" copper main lines with 1/2" copper branches to each fixture.

About 10 years ago, I had a water softener system installed in a small shed behind the house. To plumb it, we intercepted the water main line at the front, and created a loop. I installed three 3/4" ball valves at the intercept point, so I could bypass the softener loop if there was even an issue with it.

I made a diagram to try and help explain the issue, and what I've tried to resolve it. Points C, D and E are outdoor water spickets. C and D are on the "send side" of the softener loop. I had these installed because I wanted a couple of outdoor spickets that were not soft water. Points A and B are just points where the city took pressure measurements.

and here's a closer look at the softener bypass:

Last week, the PRV went bad and was leaking horribly. I had a plumber come out and replace it with a new Watts LF N55B M1 brass unit. I asked him to set the pressure to 70psi, which he said he did. After he left, we noticed the pressure wasn't as good as before the repair, so I bought a gauge and checked it. It was at 60psi. So I adjusted the PRV to set it to 70, and then checked with a sink flowing, and the pressure dropped to around 50psi. I then increased the PRV to 80psi, and checked again, and the pressure is dropping to 60psi when a single sink is open.

So I called the plumber and explained the issue, and he recommended I call the city. They came out to troubleshoot and took some pressure readings. At point A on my diagram, the pressure was 84/78 (84 static, 78 running). At point B, the pressure was 84/65. So they decided to replace the meter.

After the meter replacement, I still had the 20 PSI drop at the house, and the city said it must be something on my end.

To rule out the softener loop, I bypassed it by closing valves 1 and 2, and opening valve 3, to send water direct to the house. Then I took measurements at point E, and still see a 20psi drop with one sink flowing.

To rule out the house plumbing, I opened valve 1, and closed valves 2 and 3, then checked with a gauge on point C, with the spicket at point D flowing water, and I still saw the 20 psi drop.

So I called the plumber back, thinking the PRV may be bad. He pulled the PRV, the check valve, and the cutoff valve, and we inspected everything. They all seem fine, with no buildup or corrosion. So I asked him to replace the PRV anyway, which he did with a Wilkins 70XL. After the 2nd new PRV, nothing changed and I'm still experiencing a 20 psi drop when a single fixture is opened.

This makes me think there's either an issue with the city supply, or with the pipe between the meter and the 3 valve bypass, or possibly a system design issue.

The city supply is showing a 6psi drop, measure at point A. Would this translate to a 20psi drop due to distance of pipe to where I am taking measurements? The city tells me the main feed to my house is a 3/4" copper line, with a single 90 bend, then 100' of 3/4 copper pipe to a 4" main.

I've also read that a PRV needs a delta of 10psi between the incoming pressure and the outgoing pressure to function properly. So I backed the PRV down to 50psi and tested. At 50psi, the drop is no longer 20psi, but is around 10psi.

We really need a min of 60psi due to a recent master bath remodel, where we installed multiple shower heads. So to get even close to 60psi when the shower heads are in use, I've set the PRV to 80psi; but I would prefer to be able to set the PRV to something lower, like 70, if I can get the system to maintain that pressure when a fixture is in use.

I assume I didn't notice the pressure issue before this because the original PRV most likely failed a while ago, and was probably allowing full pressure to pass. I did have 3 hot water heater tank failures (internal tank ruptures) in the past 5 years. I've since replaced one tank with a tankless, and added an expansion tank to the other. These failures may be unrelated, or could be due to water pressure spikes from the city and a failed PRV not regulating the pressure.

Any help would be greatly appreciated!

 

[Reach4]

You don't mention a thermal expansion tank. You need one with your setup.

I understand that you can not take pressure readings as A and B. Only the city can. The next time the plumber does work at the PRV, I suggest you have him install a yard spigot near there, if it would be an OK place. That would let you check the pressure there.

I am not a pro. Do you have any city water towers within a few miles of your house? f there are water towers in your system, expect the water to peak in pressure around 4 AM. But water towers tend to keep an upper limit on pressure. If so, if it were mine, I would consider bypassing the PRV, as long as the pressure stays under 90. I I understand that is not the official word.

 

[Sponsor]

 

[stoopalini]

Thanks for the input.

I do have an expansion tank in the system, it is attached to the hot water tank, and was installed last year when that tank was replaced. Do you mean I need an expansion tank somewhere on the main line coming in to the house?

Not sure what you mean by asking the plumber to install a yard? Do you mean a yard gauge, at the PRV?

There is a water tower about a mile up the road. I had thought about installing a PRV bypass, so I could bypass it for now, and then re-engage the PRV in the future by operating the valves, in case the city increases the line pressure. There is local construction going on, and I'm thinking when the new houses are finished, the city may increase the line presssure. For this reason, I'm hesitant to remove the PRV altogether.

 

[Dj2]

You could have a leak in the PVC pipe from the meter to the house. Shut off the water at the house, check and observe the meter for more than 10 minutes: is it moving at all?

Also, 3/4" main supply is too small, I would replace it with a 1" copper L.

 

[stoopalini]

We did a leak test, and the meter stayed constant. So I don't think there's a leak anywhere.

I thought the 3/4" main feed seemed small as well. Even the city workers were confused by it. About 50' from the meter box, on the city side, there a hole where you can see down to the pipe and the saw marks from when they dug the channel. So it's obvious it's a 3/4" main feed. They're supposed to come fill it in, but maybe I should tell them to replace it instead?

Is there a way to determine if having that small of a main feed is causing my issue?

 

[Jadnashua]

There are on-line calculators that will tell you the pressure drop with a particular size pipe, a specific flow rate, and a specific distance. What you are describing is a flow-related problem. Are you sure each of the valves are fully open? A 20# drop with a simple faucet opened indicates a significant restriction somewhere in the line.

If you bypass the softener, does it improve?

Since this happened with the replacement of the PRV...it's possible that it is defective.

 

[stoopalini]

Positive the valves are 100% open. Bypassing the softener results in the same 20# drop.

Does the 6psi drop at point A indicate a city problem? To test point A, they removed the meter and connected an elbow with a gauge and a spicket at the end. They checked static pressure with the spicket closed, and it read 84psi. Then they opened the spicket on the test pipe, and the pressure dropped to 78psi.

 

[stoopalini]

Oh, and the PRV was replaced again. So I've had 2 new PRVs this week. Check out my posts above for the details.

 

[Jadnashua]

Other than frictional losses during water flow, the only other constant is a change in pressure due to elevation changes. Friction can come from the line being too small for the desired flow, too many fittings or changes of direction, or too long of a run. A kink or crimp might have occurred in the supply line somewhere. Water pressure will change about 0.43#/foot elevation change...it drops as you go up, and increases as it goes down, so if the house is built on a hill, the pressure will be lower as you go up, and higher if you're downhill from the supply.

 

[stoopalini]

Great info, thanks! The house is on a level plot, with no significant changes in elevation. The more info I get, the more it appears the main branch from the city may be too long, or there an obstruction in their line, or there's a valve on their end somewhere which may not be open 100%.

I'll give the city a call again after the holidays and see what they say.

 

[Reach4]

The more info I get, the more it appears the main branch from the city may be too long, or there an obstruction in their line, or there's a valve on their end somewhere which may not be open 100%.

As far as you know, the pressure at point B (meter output) stays 65 PSI or more, right? How could you conclude that the problem is before the meter?

Are you in the warm part of Texas or the cold part?

How about moving the PRV to the input to the bypass? And add a gauge to the input and output of the PRV.

 

[stoopalini]

I was basing that assumption on the fact the main (i.e.: point A) experiences a 6psi drop with a single spicket flowing, and that point B experiences a 19/20psi drop with a single spicket. These tests were done while the main was disconnected from the house, with a test elbow, a gauge, and a spicket attached.

I should have had the city run the pressure test again after they replaced the meter, but unfortunately I didn't.

I do think it's a good idea to install two pressure gauges though, one before and one after the PRV. I think I'll look at doing this next.

 

[stoopalini]

Oh, and I'm in central Texas. Just outside Austin. It's typically warm, but had a cold snap recently. Back up to the 70's now though.

 

[Themp]

So I called the plumber and explained the issue, and he recommended I call the city. They came out to troubleshoot and took some pressure readings. At point A on my diagram, the pressure was 84/78 (84 static, 78 running). At point B, the pressure was 84/65. So they decided to replace the meter.

With the old meter, you were seeing 78-65=13 psi drop across the meter(running). I am going to assume the new meter did not change this, otherwise you would not still have your 20 psi drop because the new meter would have provided more output pressure at point B. Thus, you are only really looking for a 7 psi drop after the meter. I think you are going to have to set the PRV to 80 or remove it to get what you need for the shower. You mentioned a new expansion tank with the new water heater, has its pressure been set correctly to what the PRV is set to, so that you are not getting pressure spikes when no water is running and the water heater is heating? I only mention the expansion tank setting because people set the PRV but forget about the expansion tank and then it fails prematurely because the bladder is stretched more.

Finally, I am confused on this statement: "I've since replaced one tank with a tankless, and added an expansion tank to the other." You have a tankless and a water heater? And before the tankless you had more than one water heater?

 

[Reach4]

Let me explain the idea of moving the PRV farther toward the bypass. If you know electronics, a PRV is like a regulator. It should be able to accept a drop in the line, and regulate it to be more consistent at its output.

Are these replacement PRVs the same model? What model are they using?

 

[stoopalini]

The test on the old meter output was 84psi static, and 65psi flowing. The test on the main (i.e. In front of the meter but behind the city cutoff valve) was 84/78.

I had a plumber install the expansion tank on the water heater tank when it was replaced, but am not sure what he set it to, or how he did it. I don't see a gauge on the expansion tank either, should there be one?

And ya, this house had 2 hot water heater tanks. One for the master suite and kitchen, and the other for the rest of the house. And they're in the attic! That just seems like a really bad place for hot water heaters.

After having 2 tank failures which feed the master/kitchen (1st one was the original tank from 1998, and the 2nd was a replacement which failed within the warranty period), I replaced it with a tankless. Then when the tank which feeds the rest of the house failed, I had it replaced with a new tank and an expansion tank.

 

[stoopalini]

I understand electronics, and get what you're saying about moving the PRV. It makes sense to me. But if the pressure drop is happening pre-PRV, is still see a drop, right? I'm sure I could get my plumber to move it though. There's already a meter box at that location for the bypass valves anyway.

The PRVs are not the same brand. I put the details in the 1st post as to the models, but the 1st one was a Watts LF N55B M1 and the second one (now in service) is a Wilkins 70XL.

I really appreciate the help guys!

 

[Reach4]

The reason I asked about the cold is that it is good if you don't have a hard time keeping pressure gauges from freezing.

But if the pressure drop is happening pre-PRV, is still see a drop, right?

Maybe not much. If you have a 5 volt regulator, it doesn't much matter if the 12 volt input sags to 7. If the input sags to 5.5, it depends on the regulator. Not a complete analogy, but how low the input can go and maintain regulation varies with the regulator. PRVs are not as sharp acting as most electronic regulators.

If you replaced your 3/4 inch PVC in the yard with 1 or 1.25 poly, the pipe would drop less pressure.

The drops will be bigger with higher flow. Your typical flows are usually under 10 GPM. So PRV loss should not be a lot of what you are losing. If you wanted to minimized the unwanted drop in a PRV, you could use a 1 inch. There would be some extra plumbing adapters.

When you do a pressure loss test, it is good to know the flow rate. A good way to get that is to fill a bucket of known volume, time the fill, and do some math.



From http://content.zurn.com/web_documents/pdfs/specsheets/REG-70XL.pdf :

 

[Themp]

I thought on this more and you have a 6 psi drop from the city when you turn on your faucet in the house. So, the static 84 psi is not really a factor here since the city cannot maintain 84 psi. So, this coupled with the 13 psi drop across the meter leaves you with a 19 psi drop. So, the problem lies with the the city and nothing in your house. I am assuming here that the C, D, and E psi measurements are based on 84-20=64 psi. If you saw 78-20=58psi at C,D, and E, then you are back to just finding the 7 psi in your house.

 

[stoopalini]

That was my thinking too. C, D and E measurements show 80psi static (because that's what the PRV is set for) and 60psi with a sink open.

I probably don't quite understand fully how a PRV works, but I do understand it's a spring/diaphragm setup. So if the static incoming pressure is 84psi, and I set the PRV to 70 (static pressure), the spring is then tensioned accordingly based on these two pressures. Then if the incoming pressure drops, say to 78, the spring tension wouldn't be correct to maintain 70psi anymore; right?

I understand a PRV is to protect against presssure spikes, or increases, but not regulate against a pressure drop. I didn't think they were able to maintain the set pressure if the incoming static pressure that was used to set the spring tension, dropped.

Am I misunderstanding the working dynamics here?