Mechanical contractors run and test gas piping constantly — a furnace changeout, a new rooftop unit, a boiler replacement — and then hand a scrap of paper or nothing at all to the inspector. The forms available to fill that gap are the same photocopied sheet the plumbing trade uses, printing "10 psi for 15 minutes" as though it were the national requirement. It is not. It is a local amendment, and it is not the amendment in force everywhere.
IFGC 406.4.1 sets a floor: not less than 1.5 times the proposed maximum working pressure, and never less than 3 psig, irrespective of design pressure. Downstream of the regulator at 7 in. w.c., the multiple works out near 0.38 psig, so the 3 psig floor is the answer. On a 2 psi riser feeding a roof header the multiple gives 3 psig too; on a 5 psig service it gives 7.5 psig and the floor stops mattering. That is why this report asks for working pressure per section rather than printing one number: a job with an MP regulator has two different required test pressures on the same roof, and they are not the same test.
Hold time is where the two model codes actually part company. IFGC 406.4.2 scales with pipe volume — half an hour for each 500 cubic feet or fraction thereof, 10 minutes minimum under 10 cubic feet, capped at 24 hours — while IRC G2417.4.2 governs one- and two-family dwellings with a flat 10 minutes and no volume term. A residential furnace changeout and a commercial rooftop job are genuinely different answers, and a form that prints one duration is wrong on one of them.
The clause that decides whether any of it meant anything is the one nobody prints. IFGC 406.4 limits a mechanical gauge's full scale to 5 times the test pressure, which puts a floor of 20 percent of span on where your reading sits. A 3 psig test needs a gauge reading to 15 psi or less. On the 0-100 psi gauge sitting in most service vans that test lands at 3 percent of the dial, inside the needle's own width, where a real leak produces no visible movement. A "held" reading on that instrument is not evidence. This builder scores the gauge you recorded against the pressure you pulled, prints the finding either way, and fails the report when the instrument could not have shown a drop.
Amendments are the rule, not the exception, so the report does not treat the code as final. Enter the AHJ figure for test pressure or hold time and the verdict is scored against it, with the code minimum and the amendment reference printed alongside — so the inspector reading the PDF can tell at a glance which standard produced the pass.