Pressure fluctuations causing leak detection triangle to move intermittently?

[Jrangelini]

I am struggling to make sense of the movement of the leak detection triangle on my water meter. I had a leak in the garage (evidenced by moisture coming through the slab), and after various vicissitudes (e.g., having the leak detection company jack-hammer the pipes, which apparently had been roughed in short of the wall, then encased in the slab without sleeving, etc.) the entire area was re-piped, so I decided to run a follow-up leak test overnight. The original leak had tested at 0.5 gallons; now after the repair the meter moved 2.1 gallons overnight!

So I determined that I had a new leak somewhere. I have been running leak tests since then, but getting very strange and inconsistent results. The triangle moves (counterclockwise) with a fitful (start and go) motion, making one full rotation in about 1.5 minutes, when all the lines are open. I then turned off the hot water (handle before the hot water heater), and the triangle appeared to stop, but as I watched it over a 5 minute period, it did move (I'm using a piece of tape to measure movement from side to tip, etc.), then stopped, then after another long period, it moved again. I turned off all the toilet valves to eliminate the flappers as a cause, but the very slow, intermittent movement continued. I then turned off the water at the house to test the service line (from the meter to the house), and even this had an extremely slow movement (10 minutes to move from one side to tip). At this point I thought ALL my plumbing is bad (house was built in 1993), but I wanted to make sure before re-piping the whole thing, so I ran overnight tests again.

The first night, I tested the service line--no meter movement at all.
The second night, I tested the cold water side--no meter movement at all.

So, essentially I am seeing movement in the leak detection triangle during the day that is then NOT confirmed when I test the meter overnight (from 11:00 p.m. or midnight until 7:00 a.m.). Most recently I closed the hot water lines (yes, I have installed valves on basically every line in the house at this point) AFTER the hot water heater, and found the meter moved 0.7 gallons during a 2 hour (daytime) period. So I'm wondering whether the hot water heater, as it heats the water, and then it cools, and re-heats, can essentially be causing pressure fluctuations that would cause water to flow back and forth between my system and the city system (i.e., main).

There is no backflow preventer on the supply line to the house (I have one on the sprinkler line, and I've already replaced this, and currently have the valves shut off to eliminate that and sprinklers as a cause). I don't see any check valve or pressure reducer in the meter box or on my supply line. The water heater is also brand new, and does not have an expansion chamber. I have never seen the triangle move "backwards" (i.e., clockwise), but I wonder if what is happening is a result of pressure differentials between my house and the city water:

- In the daytime, when more people are using water, the city pressure could drop and my water could flow into the system (this would presumably not be registered by the triangle, since I don't believe it can ratchet backwards). When the city pressure builds back up to exceed my pressure, the water would return to my house and move the triangle forward (which incidentally would mean I pay more than once for the water I feed back into the system-!). This could explain why the cold/supply line tests show (SLOW) movement in the daytime but NONE at night, when not as many people would be using water.

- Thermal expansion caused by the hot water heater would increase my pressure and accelerate the rate of displacement to the city supply. This could explain why the triangle moves faster when the hot line is open, and also why the meter moves even when the hot lines were closed AFTER the hot water heater (couldn't be a leak because the cold and supply lines already tested good and I can see everything else!).

If this theorizing is right, then it would never be possible for me to have ZERO triangle movement so long as the hot water heater is on. I'm planning on running more tests (each hot water line with heater both on and off), but I wondered if anyone has experience with this. A simple rule--if your triangle moves, you have a leak--makes sense to me, but this is defying all logic so I'm coming to believe the interplay of pressures actually makes this more complicated. If it is indeed possible for my triangle to move without there being a leak, then I can patch up all the holes in my house and move forward!

 

[Reach4]

If this theorizing is right, then it would never be possible for me to have ZERO triangle movement so long as the hot water heater is on.

You could put in a spring-loaded check valve and a thermal expansion tank. However if you get this indicator movement with the cold supply to the WH turned off, then the WH heating would not seem to be the cause of the indicator movement.

 

[Sponsor]

 

[Jrangelini]

I do get some movement with the cold supply to the WH turned off, but it is significantly less than with it turned on. So my theory is that this is a compound problem--some due to drops in city pressure (drawing water out from my supply, even when the hot water is off) and even more due to my thermal expansion (which increases my pressure, so would increase the water flowing to the city over just the cold water amount).

 

[Jadnashua]

Since water doesn't really compress and only changes volume with temperature variations, changing the city's water pressure would not normally cause water to actually flow back and forth. Now, I suppose that supply hoses on valves and maybe the washing machine might balloon a little and store some energy that could push water back to the city, but that volume would be very small, and only happen once each pressure variance.

 

[Themp]

So the leak detection company made it worst with the re-pipe. .5 gallons to 2.1 based on your statement. What does this company say about this? Should they not be on the hook to fix all leaks.

 

[Jrangelini]

I didn't really pursue it, because the original leak was definitely addressed--the whole section beneath both the cold and hot manifolds was repiped (and insurance paid for this--not the actual "repair" per se, but the estimate on the cost of fixing the drywall, etc., was enough to cover my cost). The company did send a leak detection guy out free of charge to try (in vain) to locate the second leak. Bottom line is plumbing companies here just want to do whole-house repipes, and there's not much technology (for example, no one does video inspection of supply lines) or interest in finding and repairing leaks (i.e., minimizing the cost). I am still questioning whether I even have a second leak, and it's not just thermal expansion from the hot water heater causing movement back and forth between my system and the city system (causing the micrometer to move as the contraction from cooling draws water in).

 

[Themp]

You need to know what the city pressure is and then use a pressure gauge that has a tattle needle to register your highest pressure during a time period that shows a leak. If the pressure exceeds the city pressure then you may be right on this. And can you call the city and ask if your meter has a backflow check valve in it? Most cities do include a backflow valve in the meter.

 

[Reach4]

Yes. Pressure gauge.

You don't have a thermal expansion tank, but your new pex pipe expands and contracts a little bit. If your triangle is super sensitive, maybe that could be the difference between the old galvanized and the new. Outside pressure changes could cause tiny flows.

 

[Jrangelini]

I don't really have any pex pipe in the systems at all--it is all copper, except for one 12 foot run I had re-piped with CPVC (thinking that was the leaking pipe before I realized there is a lot more to all this...). The hot water heater is connected to copper with some flexible lines having a metal exterior.

I did buy a pressure gauge, it reads 85 psi at the hose bib above my house shut-off. I was thinking I might be able to see pressure decreasing (if there is a leak) or fluctuating (if the supply varies), but the needle seems remarkably steady. It also measured 85 psi at both the cold and hot outlets to my washing machine. That seems kind of high, so now I'm questioning whether I should install a pressure reducer? Rheem's technical person says the water coming into the heater should be in the range of 60-70 psi and I'm well above that.

I also spoke with a technical person at the water company, and he says my meter (a positive displacement type) is capable of measuring backflow (from my house to the city supply) and would "deduct" any water returned. If this is true (I'm skeptical, never having seen it move backwards), then over time the pressure differentials would tend to even out. This could conceivably explain why I observe some (minimal) movement on the cold and supply during the day, but then these test at zero overnight.

I'm still completely stumped as to this scenario--I turn off the hot water heater at the circuit and run out all the hot water, filling it completely with cold water (to eliminate any thermal expansion effects). I leave the supply to the heater on, but close the hot water lines after the heater (having installed cut-off valves on these lines). No water is being used, the water is contained just to the cold side (which tested zero) and the turned-off heater (cut-off from the hot side), yet the micrometer moves! Where is this water going??? I just measured .2 gallon over a one-hour period. When I shut-off the supply to the heater, there is zero movement over a one-hour period--demonstrating the integrity of the cold side. So how does the water heater (cold, unpowered) "absorb" water? I'll be testing overnight to see if there is some limit.