Debugging the plumbing

In late October, the plumbing near our water meter started ticking.

Loudly. Loud enough that I could feel the vibration in the pipes, which seemed like an unnecessary commitment to the bit.

I am not a plumber. Everything involved in this story was installed before we moved in. I inherited the pipes, the valves, and whatever decisions brought them together. My understanding of the arrangement was still catching up.

As a developer, I am familiar with taking responsibility for a system I did not build and do not fully understand.

Usually it has fewer opportunities to flood the yard.

The pipe with the strongest opinion

The ticking seemed loudest at a white PVC tee on our side of the meter. I initially called it a “T valve,” which gives you a fairly accurate idea of my qualifications going into this.

It was a fitting. No moving parts inside it. Just a place where pipes joined, apparently determined to make itself the center of the investigation.

The behavior was odd:

• It ticked primarily when nobody was using water.
• I could not see the main meter moving while it ticked.
• Opening almost any faucet, even a little, stopped the noise.
• Stop the water, and the ticking could return.
• The neighboring meter box was making the same sort of ticking noise too.

The faucet trick was especially irritating. I could make the problem disappear by doing something that did not explain the problem.

Apparently plumbing also supports workarounds.

The stationary meter was useful information, but I had only watched it. I had not done a long, precise comparison of readings. “I didn’t see it move” was what I knew. “There cannot be a leak” would have been getting ahead of myself.

That distinction would become something of a theme.

A respectable collection of suspects

There was plenty nearby to look suspicious.

A bronze Hammond check valve sat close to the meter. We thought it might be around 25 years old. Beside it was an old wheel-operated main shutoff that was completely seized. Elsewhere, a larger backflow assembly had corroded shutoffs that reportedly had not been operated in more than ten years.

There were also two downstream submeters and a reverse-osmosis system to consider.

I had started with a noise and was rapidly acquiring a vocabulary.

The check valve became a particularly convincing suspect. It appeared to be a swing check, with a moving disc inside, and Hammond’s own technical guidance describes how an unstable disc can cause noise and vibration. That gave the theory some substance beyond “this thing looks old and is conveniently nearby.”

It still did not establish that our valve was doing it.

Everyone I asked who knew what they were talking about came back with some version of the same answer: a leak on our side. A toilet, a faucet, something quietly letting water through and causing the chatter.

That included a plumber friend I consulted by phone. He had not inspected the installation in person, but the agreement made me think this was a familiar failure mode and a sensible place to start.

I was very ready to believe the people who could correctly identify all the valves. We still had not confirmed a leak.

I left the seized main shutoff and the old backflow shutoffs alone. I was already investigating one problem. Forcing a reluctant old valve seemed like a very efficient way to add another.

The tests that narrowed things down

The two submeters gave me a way to isolate downstream portions of our plumbing using their shutoffs.

I closed one supply. Still ticking.

I closed the other. Still ticking.

With both closed, it continued.

Assuming those shutoffs were sealing, the noise did not need either downstream branch to remain connected. That was useful. It made those parts of the property less likely places to keep looking.

It did not clear the shared piping before those shutoffs, and it did not identify a faulty part.

I was getting better at describing what the tests had actually told me, which was less satisfying than finding the problem but considerably more useful than guessing harder.

Then there was the test at the main meter.

County personnel shut off the incoming supply to our meter. The noise stopped. They turned the water back on. The noise returned.

I was told the problem was on our side and needed to be fixed immediately.

That got my attention.

I had loud vibration in our pipes, several elderly pieces of equipment, and an explanation that seemed to fit the shutoff test. It was easy to start treating the private-side diagnosis as settled and move on to deciding which part was guilty.

The test deserved a closer look.

The clue next door

The ticking next door complicated the explanation that this was all coming from our plumbing. But I kept getting stuck on the timing.

The two services did not always tick together. Ours could go quiet while the neighbor’s kept going, then start again. Having the county shut off our service could quiet our pipes while theirs continued.

A county worker described the properties as being on the same water-system “loop,” or words to that effect. I did not have a map of the underground pipes, but it was another reason to consider something in the shared supply.

I had expected a shared problem to produce matching symptoms. In hindsight, I was giving the mismatched timing too much weight.

The neighbor’s ticking was a clue worth following even if the two meter boxes were unwilling to perform a coordinated demonstration.

The part I had misunderstood

The following day, the water department returned and repaired a malfunctioning regulator on its side of the water system.

Our ticking and vibration stopped immediately.

Our check valve, seized shutoff, backflow assembly, and white PVC tee were all still there. None had been repaired or replaced to make the noise go away.

That was a much more convincing result than the collection of theories I had assembled around them.

It also changed how I understood the earlier shutoff test.

Turning off our incoming water had stopped the noise. I had accepted that as evidence that the faulty equipment must be on our side.

But closing that connection could also stop an upstream disturbance from reaching our plumbing. The observation fit more than one explanation.

The shutoff test showed that the noise depended on the incoming supply. It did not, by itself, establish which side contained the fault.

Once the county repaired its regulator and our untouched plumbing went quiet, that gap in the original reasoning became much easier to see.

This was familiar troubleshooting territory. Disconnect something, watch the symptom disappear, and feel briefly certain you have located the bug. Then discover you removed a condition the bug needed in order to show up.

I just usually get to learn that lesson somewhere with a chair.

What I can actually say happened

The county repaired a regulator on its side, and our noise disappeared immediately afterward. That is the strongest evidence in this story.

I do not know the regulator’s exact model, what failed inside it, or precisely what the repair involved. I also do not know the exact path the vibration took, why that PVC tee was such an obvious place to hear it, or whether our check valve was responding to something coming from upstream.

There is published research on pressure-regulating valves interacting with water networks and producing pressure oscillations. That provides some background for how a problem like this could happen. It does not document what happened inside the county’s equipment in this case.

Opening a faucet reliably quieted our ticking. I can report that observation without pretending I measured the mechanism behind it.

I also cannot confirm whether the neighbor’s noise disappeared after the repair. Our service went quiet, and that is the outcome I can speak to.

There is still a seized valve

The seized main shutoff has not been repaired yet. It is now right at the top of the homeowner’s list of Stressful Projects I Would Very Much Like to Have Already Finished.

Troubleshooting has a way of doing that. You go looking for one problem and come back with a collection of awkward things you really should address. They were probably there yesterday, but now you have met them properly.

The valve still needs attention, and the rest of the plumbing deserves a closer look too. Fixing the noise did not give the whole installation a fresh service history.

At least I could separate those maintenance jobs from the ticking investigation.

That distinction matters. Otherwise, any old part discovered during an investigation can quietly become part of the diagnosis, whether the evidence points to it or not.

Other people’s noisy pipes

Some of the homeowner discussions we started with are worth keeping here. These are people’s accounts of their own problems and repairs; the symptoms and outcomes vary.

• Thrumming pipes with no apparent water use: a particularly close match, including a vibrating meter and noise that stopped when a cold faucet opened. The homeowner said plumbers had pointed to a toilet, but a later follow-up reported that replacing the pressure-reducing valve finally silenced the pipes.
• Humming despite a stationary meter: this homeowner also questioned whether the source was outside the house, but reported that replacing their own pressure-reducing valve fixed it. In a more recent main-line vibration discussion, the poster also reported a regulator replacement resolving the noise, without establishing who owned that regulator.
• An issue at another house affecting several homes: one contributor said replacing their own regulator did nothing. After months working with the city, they reported tracing the disturbance to another house. The original question involved noise during flow, so it was a different pattern from ours, but the shared-system possibility was interesting.
• A check-valve replacement that bought three quiet days: then the vibration returned. That post does not establish the eventual cause, but it was a useful reminder that replacing a suspicious part and getting some quiet are not necessarily the end of the investigation.

Those stories gave me possibilities to consider. The county’s repair, followed immediately by our pipes going quiet, was the result that settled this particular investigation for me.

What I am keeping from this

The next time I troubleshoot something unfamiliar, I want to be more deliberate about writing down the observation before attaching an explanation to it.

The sound was loudest at our tee. That was an observation.

The source must be right there. That was an assumption.

Closing the incoming supply stopped it. Another observation.

Therefore the faulty part belonged to us. Another assumption.

The neighbor’s similar symptoms, the downstream shutoffs making no difference, and the county repair finally silencing everything on our service were all pieces that mattered. None required me to become a plumber to pay attention to them.

I still do not fully understand that plumbing installation. I understand it a little better, and I have a much better appreciation for how easily a useful test can acquire a conclusion it did not earn.

The pipes are quiet now.

I am trying to let them enjoy it before I turn this into another project.