ASME B31.1-2024 Leak Test Pressure — FAQ

Question 1

What hydrostatic test pressure does B31.1 require?

Accepted Answer

Not less than 1.5 times the design pressure, at every point in the system — that is ¶137.4.5 of the 2024 edition, and it is the whole of the minimum. Two ceilings sit on top of it: the test pressure may not exceed the maximum allowable test pressure of any component that has not been isolated (the paragraph names vessels, pumps and valves), and it may not push any part of the piping past the ¶102.3.3(b) limit of 90 % of yield strength at test temperature, in hoop or in longitudinal stress. Hold the pressure for at least ten minutes, then you may drop to design pressure and hold for as long as the leak examination of every joint and connection takes. The acceptance criterion is visual: no weeping or leaking, with the sole tolerance being localized instances at pump or valve packing. The phrase at any point matters on tall systems. A gauge at grade reads the static head on top of the pressure at the highest point, so the pressure that must clear 1.5 × P is the one at the top of the riser, and the gauge at the bottom must read that plus the head — and the pipe at the bottom must then clear the 90 %-of-yield cap at the higher pressure. Record both numbers on the test package.

Question 2

Does B31.1 scale the test pressure by S_T/S the way B31.3 does?

Accepted Answer

No. B31.1 ¶137.4.5 is a flat 1.5 × design pressure with no test-temperature stress ratio, and there is no equivalent of the B31.3 ¶345.4.2 Eq. (24) correction. That is the single most common cross-code error on dual-code sites: an engineer who lives in B31.3 reflexively multiplies by ST/S, and a hot power line ends up tested well above what its code of record requires — sometimes above what its flanges will stand. B31.1 handles the hot-line problem from the other end, by capping the test stress at 90 % of yield rather than by scaling the test up to prove hot margin. On the leak test calculator, the B31.1 posture is entered as a hydrostatic factor of 1.5 with the stress-ratio cap set to 1.0, which reduces the ST/S term to unity and leaves the arithmetic exact. Note the corollary for a steam line crossing from the boiler house into the process unit: the two sides of one valve can legitimately carry different test pressures, and the test boundary should be drawn so each is tested to its own book.

Question 3

When is a pneumatic test allowed, and at what pressure?

Accepted Answer

Only when the owner specifies it or permits it as an alternative — ¶137.5.1 — and the Code recommends it only where the system cannot be filled with water or the service cannot tolerate traces of the test medium. Once permitted, ¶137.5.5 brackets the pressure: not less than 1.2 × and not more than 1.5 × the design pressure, again capped by the test rating of any non-isolated component. The procedure is prescribed, not left to the contractor: raise pressure gradually to no more than half the test pressure, then step up in increments of roughly one-tenth of the test pressure until it is reached, hold for at least ten minutes, then reduce to the lesser of design pressure or 100 psig for the examination, which is by soap bubble or an equivalent method at every joint. The test gas must be nonflammable and nontoxic (¶137.5.2), and ¶137.5.4 permits a preliminary test at no more than 25 psig to find gross leaks while the stored energy is still small. Compare the numbers with B31.3: process piping tests pneumatically at 1.1 ×, power piping at 1.2 × minimum with a 1.5 × ceiling — the B31.1 minimum is higher because it never carried the ST/S correction that makes the B31.3 hydrostatic number climb.

Question 4

How long is the hold, and when can I drop to design pressure?

Accepted Answer

Ten minutes at full test pressure for both media, continuously maintained, and only then may the pressure be reduced — to design pressure for a hydrostatic test (¶137.4.5), or to the lesser of design pressure and 100 psig for a pneumatic one (¶137.5.5) — and held for however long the joint-by-joint examination takes. The hold is not the examination; it is the proof, and the examination happens afterward at the lower pressure so that a leak found is not a leak found at 1.5 × P. One exception is built for buried and embedded pipe. Under ¶137.4.6(d), non-boiler-external-piping joints that cannot be seen during the test may be exempted from the visual examination when the owner approves and all of the following hold: every welded joint is 100 % volumetrically examined, the test pressure is held at least one hour for temperature stabilization, pressure and ambient temperature are continuously monitored and recorded into the test record, and any pressure loss not attributable to atmospheric change is traced to its source, repaired and retested. That is a different test with a different record, and it should be planned as one rather than discovered when the trench is already backfilled.

Question 5

Can the pressure test be waived or replaced by examination?

Accepted Answer

Only through the doors the Code opens, and each has a condition attached. The ¶137.8 alternative — visual examination of every weld, 100 % radiography or ultrasonic examination of circumferential, longitudinal and spiral groove welds plus branch connections over NPS 4, and liquid penetrant or magnetic particle examination of every remaining weld including structural attachments — is available only when the owner and designer agree that both hydrostatic and pneumatic testing are impracticable, and ¶137.3.2 lists what impracticable means: a hydrostatic test would damage linings or internal insulation, contaminate a process that cannot tolerate moisture, or risk brittle fracture at the achievable test temperature, and a pneumatic test would present an undue stored-energy hazard that cannot be protected against or the same brittle-fracture risk. Both sides must apply, not one. Separately, ¶137.4.6(c) lets the closure weld between two already-tested systems go untested when owner and designer agree, provided it is visually examined and 100 % radiographed or ultrasonically examined. And ¶137.7 permits an initial-service test where other tests are not practical or leak tightness is demonstrable from the nature of the service — with the flat prohibition of ¶137.7.1 that it is never applicable to boiler external piping.

Question 6

What stress limits apply to the piping and its supports during the test?

Accepted Answer

The piping limit is ¶102.3.3(b): hoop stress at test pressure may not exceed 90 % of the yield strength at test temperature, and the sum of longitudinal stresses from test pressure plus the live and dead loads present at the time of test — occasional loads excluded — may not exceed the same 90 % of yield. ¶137.1.4 restates that no part of the system may exceed that limit at any time during the test, which is why a hydrostatic test can be the governing case for the wall of a low-pressure, thin-wall line. The supports have their own rules. ¶120.2.1(a) requires rigid supports to be designed for the heavier of the transported fluid and the test fluid, with a carve-out in (b) for large gas, air, exhaust-steam and relief lines where the chance of the line filling with liquid is very remote. ¶120.2.2 says variable and constant supports are sized on operating loads and must not include the test water, but the support must still be capable of carrying the test load unless additional support is provided for the test — which is the code basis for pinning springs and adding temporary supports on vapor lines, and ¶137.2.2 makes those temporary supports a requirement, not a suggestion. ¶121.2(j)(2) allows support material to go to 80 % of room-temperature yield during the hydrostatic test.

Question 7

Which components have to be isolated, and what else does the preparation require?

Accepted Answer

Anything whose test rating is below the piping test pressure, because ¶137.4.5 and ¶137.5.5 both cap the test at the maximum allowable test pressure of every non-isolated component. ¶137.2.4, revised in its entirety in the 2024 edition, says equipment not subject to the test is either disconnected or isolated by a blank or similar means; valves may be used for isolation only if their closure is suitable for the test pressure, the owner is to know the pressure and temperature limits of every valve exposed to the test, and isolated equipment and piping must be vented. Around that sit the practical preparations: vents at every high point in the test position to purge air (¶137.4.2), clean test water chosen to minimize corrosion with the Nonmandatory Appendix IV precautions for stainless steels (¶137.4.3), low-pressure fill lines disconnected or isolated before pressure is applied (¶137.4.4), expansion joints restrained or isolated for the extra pressure thrust (¶137.2.3), joints and welds left exposed for examination unless insulation is agreed in advance with an extended hold (¶137.2.1), and a relief device — recommended set at 1⅓ × test pressure by ¶137.2.6 — so that thermal expansion of a trapped test medium cannot walk the system past its limits. Flanged joints at blinds, and joints connecting subassemblies that were tested separately, need not be retested (¶137.4.6(b)).

Question 8

Is boiler external piping tested any differently?

Accepted Answer

Yes — it is not tested under ¶137.4 at all. ¶137.3.1 sends boiler external piping, as defined by the ¶100.1.2(a) terminal points, to BPVC Section I, PG-99, and requires the test to be conducted in the presence of the Authorized Inspector. That follows from the jurisdictional split B31.1 draws in its scope: Section I holds administrative jurisdiction over boiler external piping — Certification Mark, Data Report, Authorized Inspection — while B31.1 holds the technical responsibility for its design, and the pressure test is part of the administrative side. The practical consequences are a different test pressure basis, a different witness, and a different record; the Certificate holder responsible for the boiler external piping owns the test documentation even where a non-certificate holder installed the bolted or threaded connections. Two smaller BEP-specific rules are easy to miss: ¶137.7.1 forbids the initial-service test for boiler external piping, and the ¶137.4.6(c) closure-weld exemption still requires the volumetric examination. Settle the BEP boundary on the boiler vendor's scope drawing before the test package is written, because the same spool can be on either side of it depending on where the ¶122.1 valves actually sit.

Paragraph	What it sets
¶137.3.1	Boiler external piping is hydrostatically tested to BPVC Section I, PG-99, with the Authorized Inspector present — not to ¶137.4.
¶137.3.2	Nonboiler external piping: hydrostatic test by default; pneumatic, sensitive-leak, or initial-service tests only when the owner specifies them; the ¶137.8 examination-in-lieu route only when both hydrostatic and pneumatic testing are impracticable for the reasons listed.
¶137.4.5	Hydrostatic test pressure at any point not less than 1.5 × design pressure, capped by non-isolated component test ratings and by ¶102.3.3(b); 10-minute hold, then reduce to design pressure for the leak examination.
¶137.5.5	Pneumatic test pressure not less than 1.2 × nor more than 1.5 × design pressure; stepped pressurization; 10-minute hold; examine at the lesser of design pressure or 100 psig.
¶137.5.4	Optional preliminary pneumatic test at no more than 25 psig to find gross leaks first.
¶102.3.3(b)	During test: hoop stress ≤ 90 % of yield at test temperature; longitudinal stress from test pressure plus live and dead loads ≤ 90 % of yield.
¶137.2.6	A relief device set at 1⅓ × test pressure is recommended against thermal expansion of the test medium.
¶137.4.6, ¶137.7, ¶137.8	Closure welds, initial-service tests and the examination-in-lieu alternative — the three ways a joint or system avoids a full pressure test.