Backflow preventer testing, start to filed report

A backflow prevention assembly is a mechanical device sitting between a public water main and something that could contaminate it — a boiler, an irrigation zone, a fire loop, a process line. It works by holding pressure differentials in the right direction. Nothing about it is visible from outside: a check valve fouled by a piece of pipe scale looks identical to a check valve doing its job.

Testing is the only way to know. A tester connects a calibrated differential pressure gauge to the assembly's test cocks, isolates it, and measures the pressure difference each internal component actually holds, in PSID. The test is pass or fail against numeric criteria, and the numbers are the evidence — which is why a report with the readings blanked out and only a “passed” box ticked is worth very little.

Only assemblies are testable. Devices such as atmospheric vacuum breakers and hose bibb vacuum breakers have no test cocks and no shutoffs, so they are inspected and replaced rather than tested. If there are no test cocks, there is no differential test to run.

Annual testing is the industry standard of care rather than a single federal rule. The USC Foundation for Cross-Connection Control and Hydraulic Research recommends annual field testing of all backflow prevention assemblies in its Manual of Cross-Connection Control, Tenth Edition, and notes that AWWA Research Foundation report 90928F found 94% of water suppliers require field tests annually or more frequently. The recommendation comes out of the Foundation's approval program, where an assembly must survive a one-year field evaluation before it is listed.

States adopt that into enforceable language. Washington's plumbing code, for instance, requires the owner to have the assembly tested by a state-certified backflow assembly tester at installation, repair or relocation, and at least annually thereafter unless the authority having jurisdiction requires it more often. Most programs read the same way.

Testing is restricted work. It takes a certified backflow prevention assembly tester and a calibrated differential pressure gauge, and most purveyors additionally require the tester to be registered with them before a report is accepted. The credential side — who issues it, the prerequisites, renewal, and whether it also covers repair — is covered in full on the backflow certification guide.

• Certification, current and in scope. Number and expiration go on the report. Some jurisdictions certify testing and repair separately.
• A calibrated gauge. Make, model, serial and calibration date go on the report. Readings taken past the calibration date are commonly treated as invalid, which means retesting every assembly touched since.
• Local registration. A purveyor can reject reports from a tester who is not on its approved list, even when the credential itself is valid.
• Fittings, bleed-off arrangement and a way to catch discharge. An RP relief valve will dump water; a PVB body has to be drained for the air inlet to open fully.

The assembly type decides the reading set. Getting this wrong is the most common reason a report is incomplete: an RP report with no relief valve opening point, or a PVB report carrying two check readings that the assembly does not have.

The RP first-check figure changed, and it matters

The USC Foundation's published comparison of the Ninth and Tenth Edition field test procedures replaced the older wording — first check reading should be 3.0 PSID greater than the relief valve opening point — with a direct requirement: the first check reading must be above the relief valve opening point and at or above 5.0 PSID. The stated reason is to provide a required minimum value for the first check.

So the 3.0 PSID margin between check #1 and the relief opening is no longer the criterion in the current procedure, and BackflowPass treats a thin margin as an advisory warning by default, not a failure — while still surfacing the number, because some jurisdictions and trainers continue to enforce it. It is a per-device setting; the calculator on the pass/fail calculator page runs the same rule engine so you can see both behaviours.

Per-assembly walkthroughs, field notes and worked examples live on the template pages: RPZ test report, DCVA test sheet and PVB inspection form.

Follow the procedure your jurisdiction accepts, step for step — the sequence is not cosmetic. Two of the Tenth Edition changes exist purely to keep water moving through the assembly and the test kit so an RP relief valve does not discharge during setup, and another exists to keep the gauge at the same elevation as the assembly so check readings are not skewed by head. The shape of the visit, though, is the same everywhere:

1. 1

   Identify the assembly and the service

   Type, make, model, serial, size, location, hazard and service type — domestic, irrigation, fire protection, boiler or process. This is the identity the purveyor matches the report to; a serial that does not match its record is a common rejection.

2. 2

   Notify, then isolate

   Water stops for everything downstream. Tell occupants, and for fire protection coordinate with the monitoring company and the authority having jurisdiction before touching a valve.

3. 3

   Flush the test cocks and the kit

   Clearing debris out of the cocks and air out of the hoses is what makes the reading real. Grit in a test cock produces a fictional failure.

4. 4

   Run the required tests in order

   Check differentials, relief valve opening point on an RP, air inlet opening on a vacuum breaker, then shutoff tightness. Keep the kit at the assembly's elevation.

5. 5

   Record every reading as taken

   Write the measured value, including the ones that fail. Do not round toward a pass, and do not leave a required element blank — a missing reading is an incomplete report, not a pass.

6. 6

   Repair and retest if anything failed

   Clean or rebuild, then run the full reading set again and record both sets. The initial failure stays on the record; the retest is what shows compliance.

7. 7

   Return the service, then file

   Reopen slowly, check for discharge and leaks, confirm downstream service, then complete the report with your certification and gauge details, signatures and dates, and file it by the purveyor's deadline.

The BackflowPass generator follows exactly that order on a phone: it shows only the readings your selected assembly needs, grades each one as you type, computes and flags the RP margin, opens repair and retest fields the moment something fails, holds your certification and gauge details, takes both signatures with timestamps, and outputs a one-page PDF. It works with no signal, so a basement or a vault is not a problem.

Failures are mundane and repetitive. Knowing the pattern shortens the visit and tells you which repair kit to carry.

Use manufacturer repair parts. The USC Foundation's approval listing is tied to the assembly as evaluated, so non-original parts can affect an assembly's approved status — see the USC cross-connection reference for how approved-assembly lists work.

The test is only half the job. Reviewers reject reports for missing data far more often than for bad readings. Where a purveyor mandates its own form, that form governs — but the data set below is what reviewers consistently expect.

• Facility and service identity — name, address, contact, purveyor, meter or account number, and where on the site the assembly is.
• Assembly identity — type, manufacturer, model, serial, size, line pressure, service type and hazard classification.
• Every required reading as measured — including failing values, plus the derived figures such as the RP margin between check #1 and the relief opening.
• Failure reasons, repairs, and the full retest set — a bare “repaired” note with no retest readings is not a compliant record.
• Tester and gauge credentials — certification number and expiration; gauge make, model, serial and calibration date.
• Signatures with dates — the tester, and the facility representative where the form asks for one.
• Test date and test type — initial, annual, post-repair or survey — and the final pass/fail state.

Keep your copy. When the same assembly comes up next year, prior-record data is what turns a twenty-minute paperwork job into a two-minute one — BackflowPass keeps searchable history and starts the next annual test from the previous record with the readings blank, so the old report is never overwritten.

A notice saying your backflow test is due is a compliance deadline on the property owner, not on the plumber. Practical version:

1. 1

   Find out how many testable assemblies you have

   Purveyor records list them by serial and location, and the notice usually names them. Buildings routinely have more than one — domestic, irrigation and fire lines are separate services.

2. 2

   Hire a certified tester registered with your purveyor

   Ask for the certification number and expiration, and confirm the purveyor accepts reports from that tester. Get the gauge calibration date too.

3. 3

   Plan for the water being off

   The test isolates the service. Schedule around tenants, kitchens, medical equipment and irrigation controllers, and coordinate fire-line testing with your monitoring company.

4. 4

   Budget for the possibility of repair

   A failed assembly must be repaired or replaced and retested. Rebuild kits are inexpensive relative to replacement; a large or fire-line assembly replacement is a different order of cost and often needs a permit.

5. 5

   Confirm the report was filed, and keep a copy

   Filing by the deadline is what closes the obligation. A passing test that never reached the purveyor still reads as non-compliant, and unresolved non-compliance can escalate to service interruption.

Costs are set by local labour rates, assembly size and access, so ask two testers in your area rather than trusting a national figure — including this page's, which is why it does not quote one.

• Your water purveyor's cross-connection control office — the operative authority for frequency, accepted form, submission method, deadlines, tester registration and correction windows.
• Your state drinking water or plumbing program — who may test, which certifications are recognised, and the code language adopting the annual requirement.
• USC Foundation for Cross-Connection Control — the Manual of Cross-Connection Control, Tenth Edition (field test procedures in Chapter 9), the annual-testing recommendation, and the List of Approved Backflow Prevention Assemblies.
• US EPA Cross-Connection Control Manual — background on hazards, assembly selection and program structure.
• The assembly manufacturer's service literature — repair kit part numbers, torque and rebuild procedure for the exact model in front of you.

Before production use, verify thresholds and report requirements against the current procedure your purveyor accepts. Rules change between editions — the RP first-check change above is the clearest example.