Free Backflow Test Report Template for HVAC Contractors

They differ, which is why the form changes with the type. A reduced pressure assembly takes a reading on check valve one, a tight-or-leaking result on check valve two, and the relief valve opening point. A double check takes a numeric reading on each of its two check valves. A pressure or spill-resistant vacuum breaker takes the air inlet opening point and a check valve reading. An atmospheric vacuum breaker takes none at all — it has no test cocks and is not a testable assembly.

No. On a reduced pressure assembly the second check is recorded as closed tight or leaking, not as a differential pressure. That is how the USC procedure and AWWA M14 both record it. The builder offers it as a tight-or-leaked result rather than a number box, so there is nothing to type a misleading value into.

No — but something related does still apply, and the two get conflated constantly. The Ninth Edition said the first check "should be 3.0 psid greater than the relief valve opening point". The Tenth Edition replaced that soft margin with a hard two-part criterion: the reading must be above the relief valve opening point AND at least 5.0 psid. So the 3.0 psid margin is gone, while the requirement to beat the relief valve is not. The 3.0 figure survives only in Chapter 10, where it governs how a model gets onto the USC approved list — a tester never applies it with a gauge.

Because of the second half of the criterion. If the relief valve on that assembly opened at 5.4 psid, then a first check holding 5.2 psid is below the opening point, and the Tenth Edition requires it to be above. The builder prints both numbers and names which condition was missed, rather than showing a bare fail against a threshold you appear to have cleared. Worth saying plainly: an assembly failing this may still be preventing backflow. USC is explicit that assemblies which fail the field test can still be protecting the supply — this is a test verdict, not a safety verdict.

Relief valve first, then the second check, then the first check. That is the USC Tenth Edition order and it is not arbitrary — the Foundation warns that testing the check valves first can affect the relief valve opening point reading. Several published procedures, including some state and manufacturer cards, run the checks first. The form asks for the readings in USC order so it does not quietly encourage the wrong sequence.

You can change the reduced pressure first check minimum, and every export prints the criteria that were actually applied so nobody reading the report has to assume. Be aware of what that setting does and does not cover: it changes the threshold, not the shape of the rule. The builder always applies the USC Tenth Edition form of the criterion — at least the minimum AND above the relief valve opening point.

No, and the differences are larger than a single number. New York and Wisconsin set a flat 5.0 psid with no link to the relief valve opening point at all, so a check reading 5.2 against a relief valve opening at 5.4 passes there and fails under USC. Massachusetts adopts the USC Ninth Edition, which still requires the first check to sit 3.0 psid above the relief opening — stricter than the Tenth. Los Angeles County has historically used a tightness-based criterion with no numeric floor, though as of July 2026 that applies only to internal protection assemblies; meter protection there now requires ABPA or ASSE certification, and ABPA mandates the USC Tenth Edition. This builder implements the USC Tenth Edition rule. If your purveyor uses one of the other forms, check the printed criteria against theirs before you submit.

The builder flags it. If the calibration date you entered is more than twelve months before the test date, the report says so on screen and in a note on every export. It does not block you — the readings are still yours to record — but most purveyors will not accept readings from a gauge outside its calibration interval, and finding that out from the rejection letter is worse than finding it out here.

As both tests. Enter the initial readings, describe what you repaired, then enter the retest readings. The retest is what decides the assembly, so a rebuilt check that now holds reads as passed — but the failing initial reading stays on the report, because a repair with no before value is not a repair anybody can audit.

Yes. A submission listing three passes and one fail is not a pass, and rolling it up as one would be the most dangerous rounding this document could make. The summary shows the counts separately so you can see exactly where the failure is.

No, and every export says so. A test report is the certified tester’s record of what the gauge showed. Acceptance of the assembly rests with the water purveyor or the authority having jurisdiction. The form references them rather than impersonating them.