ASME B31.1 Hydrostatic Test Pressure Calculator (¶137.4.5, 90% of Yield)

Returns the ASME B31.1 hydrostatic test pressure — a flat 1.5 × design pressure under ¶137.4.5, with no test-temperature stress ratio — the pneumatic range of ¶137.5.5, and the hoop stress that test puts in the wall against the 90%-of-yield limit of ¶102.3.3(c) (¶102.3.3(b) in B31.1-2024).

Method last updated 2026-11-11 (calculation changelog) · fixture-verified on every build — most recently 2026-11-12.

Built and fixture-verified by Matthew Norris, P.E. — active P.E. licensure in Arizona, California, Kansas, Missouri, North Carolina, Texas.

Power piping is tested to a rule that fits on one line: hydrostatic test pressure not less than 1.5 times design pressure. There is no S_T/S term to work out, which surprises engineers who came up through B31.3. What B31.1 has in its place is a ceiling. During the test no part of the piping may see a hoop stress, or a longitudinal stress from test pressure plus live and dead loads, above 90% of yield strength at test temperature, and the test may not exceed the test rating of any component left in the boundary. This calculator gives the hydrostatic minimum, the pneumatic range, and the hoop half of the stress limit for the pipe you describe, with the test pressure at which the wall would reach the limit. Use it to write the test pressure on a test package, and to find out before the pump is connected whether a thin, low-pressure line has room for it. The factors and the yield strength are typed inputs.

Method

Required test pressures, B31.1 ¶137.4.5 and ¶137.5.5 (identical wording in B31.1-2026 and B31.1-2024):

P_T,hydro = 1.5·P

1.2·P ≤ P_T,pneumatic ≤ 1.5·P

Stress limit during test, ¶102.3.3(c) in B31.1-2026 (¶102.3.3(b) in B31.1-2024), hoop direction:

t_test = t_nominal·(1 − mill tolerance)

σ_hoop = P_T·( D_o − 2·y·t_test ) / ( 2·t_test ) ≤ 0.9·S_y

The hydrostatic and pneumatic factors are entered, not built in, so the result states the basis it was run on. The stress check needs a word of explanation, because the Code gives the limit and leaves the stress formula to the engineer. The calculator uses the ¶104.1.2 straight-pipe equation solved for stress, a conservative basis with two deliberate choices. The wall is the least the mill may have delivered, nominal × (1 − under-tolerance), because that is the wall the test has to be safe on. The additional thickness A is not deducted: on the day of the test the corrosion allowance is still metal. The Code gives the 90% limit and no stress formula, so this form, on the minimum wall with no allowance deducted, is the stated and deliberately conservative basis of the check. Setting y = 0 turns the expression into the outside-diameter form P·D/2t, which is the conservative bound if your procedure calls for it. S_y is the yield strength at test temperature, from BPVC Section II, Part D, Table Y-1 for the material; the calculator has no default for it and reports pressures only until one is entered. It also returns the pressure at which the hoop stress reaches the limit, which is the number to compare with a test pressure raised for static head.

Limitations — what this calculator is not

• Hoop stress only. ¶102.3.3 puts the same 90%-of-yield limit on the sum of longitudinal stresses from test pressure and the live and dead loads present at test time. That needs the weight of the water-filled line and its support arrangement, and it is not evaluated here.
• Straight pipe wall only. ¶137.4.5 also caps the test at the maximum allowable test pressure of every non-isolated component — vessels, pumps, valves, expansion joints, instruments. The calculator checks none of them.
• The pressure is the pressure at the point being checked. ¶137.4.5 sets the minimum at any point in the system, so a gauge at the bottom of a tall system reads the static head on top; add the head yourself before comparing with the pressure at the limit.
• Boiler external piping is not tested to ¶137.4. ¶137.3.1 sends it to BPVC Section I, PG-99, with the Authorized Inspector present.
• The Code sets the 90%-of-yield limit but gives no stress formula. This calculator uses the pressure-design hoop form on the mill-minimum wall with no allowance deducted, which is conservative: a wall that passes here has margin against a form that credits the allowance or the nominal wall. A project procedure that specifies a different form governs.
• S_y must be the yield strength at the metal temperature during the test. A typed value is used as entered. The brittle-fracture consideration ¶137.1.2 raises for tests at low metal temperature is outside this calculation.
• Whether a pneumatic test, an initial service test or the ¶137.8 examination route is permitted at all is decided by ¶137.3.2 and the owner, not by this arithmetic.

Worked example — fixture-verified

Cold reheat line from the high-pressure turbine exhaust to the reheater: NPS 24 Sch 40 seamless carbon steel pipe (24 in OD, 0.688 in nominal wall), design 650 psig at 700°F, hydrostatic test with water at ambient temperature. ASSUMED inputs for this example: yield strength Sy = 35,000 psi at test temperature, y = 0.4, 12.5% mill under-tolerance. Read Sy for your own material from BPVC Section II, Part D, Table Y-1.

Step by step

1. Hydrostatic test pressure, ¶137.4.5: P_T = 1.5 × 650 = 975 psig. No S_T/S term enters, however hot the design temperature.
2. Pneumatic alternative, ¶137.5.5: not less than 1.2 × 650 = 780 psig and not more than 1.5 × 650 = 975 psig.
3. Wall the test has to be safe on: t_test = 0.688 × (1 − 0.125) = 0.602 in. The corrosion allowance is still in place and is not deducted.
4. Hoop stress at the hydrostatic test: D_o − 2·y·t_test = 24 − 2 × 0.4 × 0.602 = 23.5184 in; σ_hoop = 975 × 23.5184 / (2 × 0.602) = 22,930.4 / 1.204 = 19,045 psi.
5. Limit, ¶102.3.3(c): 0.9 × 35,000 = 31,500 psi. Ratio 19,045 / 31,500 = 0.605, so the hoop check passes.
6. Pressure at which the wall reaches the limit: 31,500 × 1.204 / 23.5184 = 37,926 / 23.5184 = 1,612.6 psig.

The pipe has plenty of room: the gauge could read 1,612 psig before the mill-minimum wall reached 90% of yield, so the 26 psi of static head at the bottom of a 60 ft riser changes nothing. The pipe wall is not what limits this test. A line this size is hung on springs sized for steam, and filled with water it carries roughly double its operating weight, so the test package needs the travel stops in and the longitudinal stress checked. And every valve, expansion joint and instrument left inside the test boundary has to be good for 975 psig or be isolated.

Sources & citations

• ASME B31.1-2026, Power Piping — ¶137.4.5, required hydrostatic test pressure: not less than 1.5 times design pressure at any point, not over the maximum allowable test pressure of any non-isolated component nor the test-stress limit; 10-minute hold, then reduction to design pressure for the leak examination. Same wording in B31.1-2024.
• ASME B31.1-2026 ¶102.3.3(c), During Test — hoop stress, and the sum of longitudinal stresses from test pressure and live and dead loads at test, each limited to 90% of yield strength (0.2% offset) at test temperature. This is ¶102.3.3(b) in B31.1-2024; ¶137.1.4 and ¶137.4.5 of both editions print the cross-reference as 102.3.3(b).
• ASME B31.1-2026 ¶137.5.5 — pneumatic test pressure not less than 1.2 nor more than 1.5 times design pressure. ¶137.3.1 — boiler external piping tested to BPVC Section I, PG-99. ¶137.3.2 — hydrostatic test by default for nonboiler external piping. ¶137.2.6 — relief device at 1⅓ times test pressure recommended against thermal expansion of the test medium.
• ASME B31.1-2026 ¶104.1.2(a), eq. (3) (eq. (7) in B31.1-2024) — the straight-pipe relation the hoop stress form is taken from. Yield strength at test temperature: ASME BPVC Section II, Part D, Table Y-1 — typed from your own copy.

Per the source & citation policy, allowable-stress and factor table values are inputs — typed from your governing edition, or autofilled on paid tiers from the embedded datasets enumerated there, always cited by standard, edition, table and line. Where a page does reproduce specific ASME data (the B16.5 ratings, the quick-reference tables), it states the source table and conditions inline.

FAQ

Is the B31.1 hydrostatic test really just 1.5 times design pressure?

For the minimum, yes. ¶137.4.5 requires a hydrostatic test pressure of not less than 1.5 times the design pressure at any point in the piping system, and the paragraph reads the same in B31.1-2026 and B31.1-2024. There is no multiplier for the ratio of cold to hot allowable stress. That is the difference from B31.3, where ¶345.4.2 scales the test up by S_T/S when the design temperature is above the test temperature, and it is the commonest mistake on sites that run both codes: a main steam line tested to a process-piping habit ends up far above what its own code asks for, and sometimes above what its valves will take. B31.1 comes at the hot-line problem from the other side, with a ceiling: the test may not exceed the test rating of any non-isolated component, and it may not stress any part of the piping past 90% of yield at test temperature. Hold the pressure ten minutes, drop to design pressure, and examine every joint. The B31.1 leak test FAQ walks through the rest of ¶137.

Is the 90% limit in ¶102.3.3(b) or ¶102.3.3(c)?

Both, depending on the edition, and the 2026 text makes it easy to trip over. In B31.1-2024, ¶102.3.3 has two items: (a) occasional loads during operation and (b) the limit during test. B31.1-2026 inserted a new item (b) covering occasional loads in the creep regime, and the test limit moved down to (c) with its words unchanged. The cross-references did not follow: ¶137.1.4 and ¶137.4.5 of the 2026 edition still send the reader to 102.3.3(b). Read literally, that would make the creep-regime occasional allowable the limit on a pressure test, which makes no sense for a test run with water at ambient temperature. The sensible reading, and the one this page uses, is that the test limit is the paragraph headed During Test, wherever the edition puts it. On a calculation, cite it both ways — ¶102.3.3(c) in B31.1-2026, ¶102.3.3(b) in B31.1-2024 — and name the heading, so a reviewer holding either edition lands on the right sentence.

Which wall and which formula does the hoop stress check use?

The Code leaves both to you; ¶102.3.3 states the limit and stops. This calculator makes two choices and prints them so they can be reviewed. The wall is nominal × (1 − mill under-tolerance), the least wall that can legitimately be in the line, without deducting the corrosion allowance, because none of it has corroded yet on test day. For a line that has been in service, enter the measured minimum wall with a tolerance of zero. The formula is the ¶104.1.2 straight-pipe equation solved for stress, σ = P·(D_o − 2·y·t) / 2t, which is the same relation that sized the wall. If your test procedure calls for the plain outside-diameter form P·D/2t, enter y = 0 and you have it; it runs a few percent higher. The choice matters only when the result is close to the limit. When it is, that is the moment to stop refining the hoop formula and look at the longitudinal stress with the line full of water, because that half of the same limit is the one that usually governs first.

When does the 90% of yield limit actually control the test?

When the as-built wall is already working hard at design pressure relative to its cold yield strength. The test raises the hoop stress to 1.5 times its design-pressure value, so the limit binds once the hoop stress at design pressure in the delivered wall exceeds 0.6 of yield at test temperature. Most carbon and low-alloy steel lines sit well below that, as the worked example does at a ratio of 0.605 of the limit. The cases to watch are the ones where the allowable stress is a large fraction of a low yield strength, annealed austenitic stainless steel being the usual example, combined with a wall ordered with little to spare. Static head adds to it: the bottom of a tall riser sees the test pressure plus the water column. And thin, large-diameter, low-pressure lines reach the limit in the longitudinal direction first, because a water-filled span carries bending the steam-filled line never did. Enter your own numbers and read the pressure at the limit; that figure tells you how much room the test has.

Can I test pneumatically, and to what pressure?

Only as an alternative the owner specifies. ¶137.3.2 makes the hydrostatic test the default for nonboiler external piping and lists the other routes — pneumatic, mass-spectrometer and halide, initial service — as options when the owner specifies them. Once a pneumatic test is on the table, ¶137.5.5 brackets the pressure: not less than 1.2 and not more than 1.5 times design pressure, and never above the test rating of a non-isolated component. In the worked example that is 780 to 975 psig. The procedure is stepped: raise the pressure gradually to no more than half the test pressure, then in increments of about a tenth until the test pressure is reached, hold ten minutes, and reduce to the lesser of design pressure or 100 psig for the soap-bubble examination of every joint. The stored energy in a compressed gas is the reason for every one of those steps. The hydrostatic vs pneumatic guide sets the two methods side by side, with the B31.3 rules alongside.

Does this apply to boiler external piping?

No. ¶137.3.1 takes boiler external piping out of ¶137.4 entirely: it is hydrostatically tested in accordance with ASME BPVC Section I, PG-99, and the test is conducted in the presence of the Authorized Inspector. That follows the split B31.1 draws for this piping in ¶100.1.2, where the Code supplies the technical rules and Section I supplies the administrative ones. The practical upshot is a different test basis, a different witness and a different record for the piping between the boiler and its terminal valves, even though it may be the same pipe specification as the line downstream. Draw the test boundary at the same valve that marks the jurisdictional boundary, so that neither side is tested to the other's rules by accident, and confirm that boundary on the boiler manufacturer's scope drawing before the test package is written. The numbers on this page apply to the nonboiler external piping on the far side of that valve, which is where most of a plant's B31.1 piping sits.

Related calculators & tools

• ASME B31.1 Power Piping Wall Thickness Calculator — Sets the design pressure and the wall the test is checked on
• ASME B31.1 Maximum Allowable Pressure Calculator (Power Piping) — The design-pressure rating of the wall, next to its test limit
• Hydrostatic / Pneumatic Leak Test Pressure Calculator (ASME B31.3) — The B31.3 test with its ST/S ratio, for the other side of a code break
• Pipe Support Spacing Calculator — Maximum Deflection-Limited Span — Water-filled spans: the weight the longitudinal test stress comes from
• Variable Spring Hanger Load Variability Calculator — Springs sized for steam have to be pinned for the water-filled test
• B31.1 leak test pressure — FAQ — The rest of ¶137: holds, isolation, the alternatives to a pressure test
• Hydrostatic vs pneumatic leak testing — Hydrostatic against pneumatic, with the stored-energy argument
• Load cases for B31 stress analysis — The hydrotest case in the stress model: water weight, pinned springs