ASME B31.1-2024 Occasional Loads — FAQ (¶104.8.2, ¶102.3.3, ¶101.5)

Question 1

What counts as an occasional load under B31.1?

Accepted Answer

A load that is not always there, evaluated with the sustained loads that are. ¶104.8.2 says the loads described in ¶101.5 — the dynamic effects — may be considered occasional loads if the time limitations of the k factor are met, and ¶101.5 lists them: impact forces from external and internal causes including water and steam hammer from a rapid valve operation (¶101.5.1); wind on exposed piping (¶101.5.2); earthquake on the piping, its supports and restraints (¶101.5.3); vibration, which is handled by arrangement and support rather than by a stress allowance (¶101.5.4); and discharge reactions from pressure and momentum during normal operation and anticipated transients, which is the safety valve case (¶101.5.5). ¶102.3.3(a) adds the temporary supporting of extra weight as another example. Two things people file under occasional that the Code does not: snow and ice are part of the live load under ¶101.6.1, and a pressure or temperature excursion above design is governed by the separate ¶102.2.4 allowance on hoop stress, not by eq. (16). The duration test is the gate — an event that recurs for more than eight hours at a time or more than 800 hours a year has stopped being occasional, and its stress belongs in the sustained check against Sh without any k.

Question 2

What is the allowable stress for the occasional case — where do 1.15 and 1.2 come from?

Accepted Answer

Figure 104.8-1 eq. (16) limits the stress due to pressure, weight, other sustained loads and the occasional load to k·Sh, and ¶104.8.2 defines k by duration: 1.15 for occasional loads acting for no more than 8 hours at any one time and no more than 800 hours a year, 1.2 for no more than 1 hour at a time and no more than 80 hours a year. Both refer back to ¶102.3.3(a), which is the design-criteria statement that occasional loads may exceed the Allowable Stress Table values by those amounts and durations. The percentages are the same ones ¶102.2.4 uses for pressure and temperature excursions, which is deliberate — the Code applies one philosophy of short-term overstress to both — but the two clauses govern different stresses, and an event can invoke both at once. The contrast with B31.3 ¶302.3.6 is the number most engineers carry: process piping allows 1.33 × Sh flat, with an elevated-temperature alternative, and does not grade by hours. On a dual-code site the same wind load therefore produces a different margin on the two sides of the fence, and a line that passes at 1.33 has not been shown to pass at 1.15. Wind, at a few hundred hours a year of design-level exposure, generally lands in the 1.15 tier; a relief valve lift or a design earthquake in the 1.2 tier — but that classification is the designer's, and it should be written down.

Question 3

How is the occasional stress S_o actually assembled?

Accepted Answer

Equation (16) has the same shape as the sustained eq. (15) with the occasional loads added into the force and moment terms. The first term is the axial index Ia times the magnitude of the longitudinal pressure stress Slp plus Fb/Ap, where Fb is the longitudinal force from weight, other sustained loads and the occasional load, compression negative; to that is added the SRSS of Ii·MiB and Io·MoB over Z, the moments being sustained plus occasional; that bracket is squared and added to the square of It·MtB/Z, and the root of the sum is So. The indices are the ¶104.8.1 sustained indices — the greater of 0.75 × the B31J i and 1.00 — not the full expansion SIFs. ¶102.3.2(a)(3) adds a detail that is easy to miss: the pressure P used in Slp for eq. (16) is the pressure coincident with the occasional event being evaluated, which for a relief case is the relieving pressure and for a water-hammer case includes the surge. ¶104.8.4(b) settles the sign question for dynamic results: where the analysis method returns only magnitudes, as an equivalent-static seismic run or a response-spectrum solution does, the most conservative combination of the signed sustained and unsigned occasional forces and moments is used — in practice, adding the occasional magnitude to the sustained magnitude at every point.

Question 4

Do wind and earthquake have to be applied together, and do I apply the ASCE 7 ASD reduction?

Accepted Answer

Not together, and — under B31.1-2024 — not reduced. ¶101.5.2 and ¶101.5.3 were both revised in the 2024 edition and now say the same three things for wind and earthquake respectively: the analysis considerations and loads may be as described in ASCE/SEI 7, except that the Allowable Stress Design reduction factor shall not be applied; where local jurisdictional rules specify wind or seismic loads for piping, those are the minimum design values; and wind need not be considered concurrent with earthquake, nor earthquake with wind. Authoritative local meteorological or seismological data may be used to define or refine the loads, and ¶101.5.3 names ASME B31E as an acceptable alternate method of seismic qualification or as guidance. The ASD sentence is the one that changes numbers. ASCE 7 Chapter 13 component forces and Chapter 26–29 wind pressures are strength-level loads, and the ASCE 7 allowable-stress load combinations apply 0.7 to earthquake and 0.6 to wind before comparing with an allowable; B31.1 says do not take that reduction — apply the full ASCE 7 load and check the result against k·Sh. On this site's ASCE occasional-load card the ASD factors are inputs with B31.3-practice defaults of 0.6 and 0.7; for a B31.1-2024 run, enter 1.0 for both. B31.3 ¶301.5.2 and ¶301.5.3 are silent on the reduction, which leaves it to the owner's design specification there.

Question 5

How are water hammer and steam hammer treated?

Accepted Answer

As impact loads under ¶101.5.1, which requires impact forces from all external and internal conditions to be considered and names the pressure wave from a sudden change in fluid momentum — water or steam hammer from a rapidly opened or closed valve — as one example, warning that other causes exist. B31.1 gives no method for computing the load and no separate allowable; the stress produced enters eq. (16) as an occasional load if its duration fits the k limits, and a design-basis valve closure or a turbine trip fits the 1.2 tier comfortably. The method is the designer's, and the honest hierarchy runs from the Joukowsky screening estimate of the surge pressure, through an unbalanced-force calculation on each straight run between elbows with a dynamic load factor, to a time-history hydraulic analysis feeding a dynamic pipe-stress run — the last being the only one that captures the timing of the force on each leg. Two cautions specific to power piping: steam hammer from condensate slugs in warm-up and from rapid stop-valve closure on reheat and extraction lines has broken supports that were never designed for an axial load, and ¶101.5.4 with ¶120.1(c) require vibration and resonance to be addressed by arrangement, dampers and restraints, which is a support-design problem rather than a stress-allowance one.

Question 6

Is the safety valve reaction an occasional load?

Accepted Answer

Yes — it is the ¶101.5.5 discharge reaction, and on steam headers it is usually the governing occasional case. The force is computed by the Nonmandatory Appendix II method for an open discharge, momentum plus pressure at the elbow exit with the stamped capacity increased by 1.11 and a dynamic load factor never less than 1.1, and the resulting moment at the branch connection under the valve may not be taken less than DLF · F1 · Lo. That moment goes into eq. (16) with the sustained-plus-occasional indices, checked against k·Sh, and Appendix II-4.1 restates the same three checks for every critical point of the installation. The duration tier is the designer's call: a valve that lifts rarely sits in the 1.2 tier, and a valve that cycles through every transient may not. Multiple-valve headers must be checked for each valve and for the combinations that can occur, and closed-discharge systems are a different problem entirely, with transient pressure waves that the simplified method does not capture. The separate B31.1 PSV reaction force FAQ on this site walks through the appendix; the point here is only that the reaction is an eq. (16) load, and the header and its supports — not just the branch — must carry it.

Question 7

How does the 15 % / 20 % pressure–temperature allowance differ from the k factor?

Accepted Answer

They share the hours but govern different stresses. ¶102.2.4 says a piping system may be considered safe for occasional short operating periods at higher than design pressure or temperature if the computed circumferential pressure stress does not exceed the maximum allowable stress for the coincident temperature by more than 15 % for events of no more than 8 hours at a time and 800 hours a year, or 20 % for no more than 1 hour and 80 hours a year — subject to any tighter limit from component standards or manufacturers' ratings. That is a hoop-stress allowance for a pressure excursion, and it is what lets a relieving pressure above design pass the wall-thickness check. The k factor of ¶104.8.2 is a longitudinal-stress allowance for an added load, applied to the whole of eq. (16). A relief event can invoke both — the relieving pressure through ¶102.2.4 and the reaction force through k — and each is checked against its own limit. What neither one does is extend a flange rating, a valve rating or any other component standard's pressure–temperature table; ¶102.2.4 says so in its opening sentence, and the component rating usually binds before the pipe wall does. The B31.3 counterparts are ¶302.2.4 for the excursion and ¶302.3.6 for the load, and the numbers there are different — do not carry either set across the code boundary.

Question 8

What about the hydrostatic test — is that an occasional case?

Accepted Answer

No, it has its own limit, and the occasional loads are excluded from it. ¶102.3.3(b) says that during a ¶137 pressure test the hoop stress shall not exceed 90 % of the yield strength at test temperature, and the sum of the longitudinal stresses from test pressure plus the live and dead loads present at the time of test — excluding occasional loads — shall not exceed the same 90 % of yield. ¶137.1.4 repeats that no part of the system may exceed that limit at any time during the test. So the test case is checked against yield, not against k·Sh, and wind or earthquake are not combined with it; B31.3 ¶302.3.6(b) says the same thing in its own words. The test case still needs to be run, because on a vapor line filled with water it is the heaviest weight case the pipe ever sees: ¶137.2.2 requires temporary supports where the test liquid weight demands them, ¶120.2.2 requires springs to be capable of the test load or supplemented, and ¶121.2(j)(2) lets the support material reach 80 % of room-temperature yield during the test. The load case set for a power line therefore has the hydrostatic condition beside the occasional ones, with a different allowable and a different set of supports active.

Paragraph	What it sets
¶104.8.2	Sustained plus occasional stress per eq. (16) not to exceed k·S_h; k = 1.15 for ≤ 8 hr at a time and ≤ 800 hr/yr, 1.2 for ≤ 1 hr and ≤ 80 hr/yr; ¶101.5 loads may be treated as occasional when those limits are met.
¶102.3.3(a)	The occasional-load allowance during operation, by the amounts and durations of ¶104.8.2.
¶101.5.1 – ¶101.5.5	Dynamic effects: impact including water and steam hammer; wind and earthquake per ASCE/SEI 7 without the ASD reduction factor, jurisdictional minimums, B31E as alternate; vibration; discharge reactions.
¶104.8.4(b)	Where a dynamic analysis returns magnitudes without sign, use the most conservative combination of signed and unsigned forces and moments.
¶102.2.4	Pressure–temperature variation allowance: hoop stress may exceed the allowable by 15 % (≤ 8 hr, ≤ 800 hr/yr) or 20 % (≤ 1 hr, ≤ 80 hr/yr) — a different rule from k.
¶102.3.3(b)	Test condition: hoop and longitudinal stresses limited to 90 % of yield at test temperature, occasional loads excluded.
¶120.1(b), ¶121.2(j)	Supports designed for the concurrently acting loads including wind and earthquake; 20 % allowable increase for short-time overloading.