The Landscaper Who Cut the Line to the Backflow Assembly

Why this matters

An irrigation backflow assembly sits above ground where everyone can see it, and the pipe that feeds it runs a few inches below the bed beside it, where nobody can. A landscape crew renovating that bed is the most likely person in the house's whole life to put a spade through that pipe. This case follows one of those strikes from the cut to the certified retest, including the part that surprised the crew: the cut itself was a small repair, and the expensive part was a valve they closed to stop the water. What an assembly does and why it matters to the drinking water is in What a Backflow Assembly Is Actually Protecting Against; this case is about what happens around one when another trade is digging.

Safety first

  • Water gushing from a cut line near an exterior receptacle, a pool pad, a lighting transformer or a sprinkler controller: keep people off the wet ground near the electrical gear and tell the homeowner to have it checked before it is used.
  • Never improvise a connection on the drinking-water side of a backflow assembly (a hose, a jumper, a bypass to keep the lawn watered). That is exactly the cross-connection the assembly exists to prevent.
  • Closing the main: the homeowner closes their own main shutoff where possible. If the crew does it in an emergency, they write down which valve and how it felt, because that note turned out to be the evidence in this case.

The bed, the riser and the valve

  METER and CURB STOP (street)
     |
     | 1 in service line, about 18 in deep
     |
  ===+=== FOUNDATION WALL ================
     |
  [MAIN SHUTOFF] old gate valve, basement
     |
     +---- TEE -> 3/4 in PVC back out through
     |            the wall, 6 to 8 in deep
     |            under the bed
     |                 \
     |                  X  <- spade cut here
     |                   \
     |                 [BACKFLOW ASSEMBLY]
     |                  above grade -> irrigation
     |
  house plumbing (no pressure reducing valve)

The irrigation branch tees off just inside the house, downstream of the main shutoff, and goes back out through the foundation under the bed. So the house main was the only valve that could stop the cut, and it decided almost everything that followed.

The strike

A three-person crew was replacing plantings and edging in a front bed. At about 10 in the morning, edging along the house side of the bed, the foreman's spade went through a 3/4 in PVC line about 7 in down. Water came up hard.

The foreman did not know the backflow assembly was fed from under the bed, and the crew had not traced it. They looked for a shutoff. The irrigation branch had no isolation valve of its own, and the only valve anyone knew about was the house main in the basement. The homeowner was out, so the foreman went in with the side-door code the homeowner had given them and closed it. The water stopped.

His note in the job record, written at the time: "Shut basement main, old wheel valve, took about 12 turns, last few felt loose. Capped line with 3/4 compression cap. Reopened main about 11:15, many turns, water back on."

Capping was reasonable. A compression cap seals on a wet, cut PVC pipe without solvent, and the crew was not trying to repair the line, only stop the water until the irrigation contractor came. They backfilled loosely over the cap, flagged it, and left a message for the homeowner.

The calls that came that evening

The homeowner got home to two problems. The sprinklers would not run, which they expected from the message. And the shower upstairs was a trickle, the kitchen faucet was weak, and the toilet took several minutes to refill.

They called the landscaper, who called the plumber the shop uses, and the plumber came the next morning.

How the plumber reasoned it

First reading. A pressure gauge on the hose bib by the front door, with nothing running in the house: 62 psi. That is a normal static pressure, and it rules out the obvious idea that the house had lost its water supply.

Second reading. Same gauge, with the tub spout upstairs wide open: 14 psi, and dropping slowly. A healthy service of that size would hold most of its pressure with one fixture running.

That pair is the whole diagnosis in two numbers. Static pressure is what a closed system settles at when nothing flows, and it reaches the full supply pressure even through a tiny opening, because with no flow there is no pressure lost across anything. Flowing pressure is what is left after the water has squeezed through every restriction between the street and the fixture. Full static with collapsed flowing pressure means a restriction in the line, not a shortage of supply.

Where is the restriction? The plumber listed candidates and tested each against the readings.

  • The utility side. A problem at the meter or the street would hit the neighbors too, and the gauge at the hose bib, which is downstream of the main, showed full static. It also could not explain why the house was fine until 11 the day before. Set aside.
  • The cap or the cut. The cut branch is capped and dead-ended; a capped branch carries no flow, and a branch off the tee is not in the path from the main to the fixtures. A leaking cap would show as static pressure that will not hold, and static held at 62. Ruled out.
  • Debris pushed into the house by the cut. Dirt drawn into the open line can clog aerators and cartridges. But it clogs fixtures one at a time, and the tub spout, which has no aerator, was as weak as the kitchen. A restriction that affects every fixture equally is upstream of all of them. Not this.
  • The one thing that changed. The main shutoff had been closed and reopened, and the foreman's note said the last turns "felt loose". An old gate valve has a wedge-shaped gate on a threaded stem. When the connection between them wears or breaks, the stem turns but the gate stays down in the flow path, and the valve feels like it opened.

This house had no pressure reducing valve. In one that did, the same two readings would also fit a reducing valve stuck nearly shut, and the plumber would have read pressure on both sides of it before opening anything, because a restriction is found by where the pressure drops, not by what changed most recently.

The confirmation. The plumber closed the curb stop at the meter (the utility-side valve, which the plumber was permitted to operate in that jurisdiction), opened the gate valve's bonnet, and found the gate sitting loose in the body, separated from the stem. It had probably been worn for years and was finished by being operated for the first time in a decade.

The repair. The plumber replaced it with a full-port ball valve, reopened the curb stop, and measured again: 62 psi static, 48 psi with the tub wide open. The 14 psi reading was the gate.

The irrigation side, and why it needed a certified test

The irrigation contractor repaired the cut with a slip coupling and new pipe the same week. Before reconnecting, they broke the union at the assembly's inlet and flushed the line, because a cut line fills with soil when the water is shut off and the pipe drains back, and that soil goes straight into the assembly when the water returns.

Then a certified backflow tester tested the assembly. Most water suppliers' cross-connection programs require a test by a certified tester when an assembly is installed, repaired or relocated, and on an annual schedule; disconnecting and reconnecting its inlet put it in that category here. The details sit with the local water supplier and the state: Texas, for example, licenses backflow testers and landscape irrigators separately, and some states require a licensed plumber for any work on the potable side of the assembly. Check yours. This is not legal advice.

The first test failed: the check valve did not hold its required pressure difference. The tester opened it and found grit on the seat, cleaned it, retested, and it passed. The test report went to the water supplier, and a copy to the homeowner. Had the crew's cut not been followed by a flush and a test, the assembly would have sat in the yard looking normal with a check valve that could not stop a backflow.

What it cost and who owed what

Three trades came out because of one spade cut: the plumber for the main valve, the irrigation contractor for the line, and the tester for the assembly. The landscaper paid for the line repair and the test as its own damage, because the crew cut a line in a bed they were working without tracing what fed the assembly beside it. The main valve was a harder conversation. The valve was worn before the crew touched it, and it would have failed the next time anyone used it. The shop and the homeowner split it, and both wrote down that they had agreed to.

What the crew changed in its walk-through

The shop rewrote its bed walk-through around three facts this case taught them.

  1. An above-ground assembly means a buried supply line nearby. Every backflow assembly in or near a work area gets its supply traced to the house or to a tee before any edging or digging, using the clues in How to Find the Private Lines a Utility Locate Will Not Mark. Within 2 ft of that line, digging is by hand.
  2. Know the shutoff before you need it. The walk-through now asks the homeowner where the irrigation shutoff is, if there is one, and where the house main is, and whether either has been turned recently. If the only shutoff is an old house main, that is written down before work starts.
  3. The person closing a valve writes down how it felt. The foreman's "last few felt loose" is what took the plumber straight to the valve. The shop now asks for that note every time.

References

  • Your water supplier's cross-connection control program and your state's backflow tester licensing rules
  • See related: What a Backflow Assembly Is Actually Protecting Against, Backflow Prevention Testing Reference
  • See related: Irrigation Systems for the Outdoor Trades That Dig Around Them
  • See related: How to Hand Off a Problem You Found in Another Trade's System