ASME B31.1 Safety Valve Reaction Force Calculator (Appendix II, Open Discharge)

Computes the reaction force F₁ at the discharge elbow of a steam safety valve with open discharge, per ASME B31.1 Nonmandatory Appendix II ¶II-2.3.1.1, from the exit velocity and pressure of ¶II-2.2.1, and the design load F₁ × DLF of ¶II-3.5.1.3. B31.1-2026; these paragraphs read the same 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.

When a steam safety valve lifts, the jet leaving the discharge elbow pushes back on the valve with a force that arrives in a few hundredths of a second. That force bends the inlet nozzle and the header under it, and it is the load the stanchion or bracket at the elbow is there to take. B31.1 Nonmandatory Appendix II gives a hand method for it. This calculator runs that method for an open discharge installation, the usual power-plant arrangement where the elbow fires up into a vent pipe it is not attached to. It works out the velocity and static pressure at the elbow exit from the stagnation enthalpy, turns them into the steady reaction force F₁, and applies the dynamic load factor you enter. The steam constants a and b of Table II-2.2.1-1 are inputs: type them from your copy of the Appendix, or pick the steam condition on the card to autofill them (Ultimate). The method and the worked example below are free to read; the live card is part of PiperPSV, a Pro tool.

Method

Conditions at the discharge elbow exit, Point 1, B31.1 Nonmandatory Appendix II ¶II-2.2.1(a):

V₁ = √( 2·g_c·J·(h_o − a) / (2b − 1) )

P₁ = (ṁ / A₁)·((b − 1) / b)·√( 2·(h_o − a)·J / (g_c·(2b − 1)) )

Reaction force at Point 1, ¶II-2.3.1.1, and the design load:

F₁ = ṁ·V₁ / g_c + (P₁ − P_a)·A₁

ṁ = 1.11 × stamped relieving capacity

F_design = F₁ × DLF

The force has two parts. The first is momentum: mass flow times exit velocity. The second is pressure: the flow leaves the elbow choked, so the static pressure at the exit plane is above atmospheric and acts over the exit area. The Appendix writes the mass flow as the capacity stamped on the valve times 1.11, with ṁ in lbm/sec, and the calculator does that conversion from lb/hr. h_o is the stagnation enthalpy at the valve inlet, from steam tables at the relieving condition. a and b are the two constants of Table II-2.2.1-1, selected by steam condition and by the range the exit pressure falls in, so check the computed P₁ against the range of the row you used. The constants g_c = 32.2 lbm-ft/lbf-sec² and J = 778.16 ft-lbf/Btu are as printed in the Appendix. US customary units throughout: Btu/lbm, in², psia, lbf. The dynamic load factor covers the suddenness of the load and is your input.

Limitations — what this calculator is not

• Open discharge only, within the geometry Appendix II sets for it in ¶II-1.2: valve inlet length L_o no more than 4·D_o and elbow offset m no more than 6·D_o of the discharge pipe. Outside those limits the designer has to judge whether the method still applies.
• Closed discharge systems are not covered. ¶II-2.2.2, ¶II-2.3.2 and ¶II-3.5.2 describe transient pressures and unbalanced forces that may need a time-history analysis.
• Steam only. The exit-state equations and their constants are written for steam; ¶II-2.2.1 itself says several methods exist and makes the designer responsible for a conservative result. The table's superheated row carries a note on its upper pressure limit, above which an alternate method is to be used for verification.
• The reaction at the elbow is all that is computed. The vent pipe forces F₂ and F₃, the blowback check of ¶II-2.3.1.2 and the elbow's maximum operating pressure P_1a from Figure II-2.2.1-2 are separate steps of the Appendix.
• A force, not a stress check. Turning F₁ × DLF into moments at the branch and header (¶II-3.5.1.1) and checking them against ¶II-4.1 belongs to the stress analysis.
• The dynamic load factor is typed. The calculator warns below 1.1, the floor in ¶II-3.5.1.3, and does not compute the installation period or read Figure II-3.5.1.3-2.
• One valve at a time. ¶II-3.5.1.2 asks for the combinations of valves that can blow together on a common header.
• Appendix II is nonmandatory. The mandatory requirement is ¶101.5.5: the piping shall withstand the reaction forces. An owner specification or a valve manufacturer's reaction force may govern in place of this method.

Worked example — fixture-verified

Hot reheat safety valve on the reheater outlet lead, set at 750 psig with steam at 1,000°F, stamped 400,000 lb/hr, discharging through an NPS 10 Sch 40 elbow (10.02 in ID, 78.85 in² flow area) into a separate vent pipe. ASSUMED inputs for this example: ho = 1,512 Btu/lbm, DLF = 2.0, and round stand-in steam constants a = 820 Btu/lbm and b = 4.3. The stand-ins are not the Table II-2.2.1-1 values: read a and b from your own copy for the steam condition and exit-pressure range, or use the picker on the card.

Step by step

1. Mass flow, ¶II-2.3.1.1: ṁ = 1.11 × 400,000 / 3,600 = 123.33 lbm/s.
2. Enthalpy term: h_o − a = 1,512 − 820 = 692 Btu/lbm; 2b − 1 = 7.6.
3. Exit velocity: V₁ = √(2 × 32.2 × 778.16 × 692 / 7.6) = √4,562,966 = 2,136.1 ft/s.
4. Exit pressure: P₁ = (123.33 / 78.85) × (3.3 / 4.3) × √(2 × 692 × 778.16 / (32.2 × 7.6)) = 1.5642 × 0.7674 × 66.339 = 79.63 psia. That sits inside the pressure range of the table row the stand-ins represent.
5. Momentum term: ṁ·V₁ / g_c = 123.33 × 2,136.1 / 32.2 = 8,181.8 lbf.
6. Pressure term: (P₁ − P_a)·A₁ = (79.63 − 14.7) × 78.85 = 5,120.0 lbf.
7. Steady reaction: F₁ = 8,181.8 + 5,120.0 = 13,302 lbf.
8. Design load: F₁ × DLF = 13,302 × 2.0 = 26,603 lbf, acting on the elbow opposite the jet.

Thirteen thousand pounds of steady thrust from one valve, and the pressure term is 38% of it: leave that term out and the force is badly understated. The dynamic load factor of 2.0 is the ceiling for a single ramp load on a one-degree-of-freedom installation, used here because the installation period was not calculated. Working ¶II-3.5.1.3 through with the valve's opening time usually gives something lower, never below 1.1. Either way, the next step is the moment at the branch connection, the header stress under it, and the support that carries the load at the elbow.

Sources & citations

• ASME B31.1-2026, Power Piping, Nonmandatory Appendix II — ¶II-2.2.1(a), pressure P1 and velocity V1 at the discharge elbow outlet; Figure II-2.2.1-1; Table II-2.2.1-1, values of a and b by steam condition (typed from your copy, or autofilled by the picker on the card and cited there). These paragraphs and the table read the same in B31.1-2024.
• ASME B31.1-2026 Appendix II ¶II-2.3.1.1 — reaction force F1 at the discharge elbow, with mass flow taken as the relieving capacity stamped on the valve × 1.11, and a dynamic load factor to be applied.
• ASME B31.1-2026 Appendix II ¶II-3.5.1.1 and ¶II-3.5.1.3 — reaction moments and the branch-connection floor DLF × F1 × Lo; dynamic load factor from the installation period and Figure II-3.5.1.3-2, never taken less than 1.1.
• ASME B31.1-2026 Appendix II ¶II-1.2 (open and closed discharge definitions and the Lo and m limits), ¶II-2.3.1.2 (vent pipe forces and blowback), ¶II-4.1 (loading criteria); ¶101.5.5 (discharge reactions) and ¶107.8.4 (reference to the Appendix) in the body of the Code.

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

Why is the mass flow 1.11 times the stamped capacity?

Because the Appendix says so: ¶II-2.3.1.1 defines the mass flow in the reaction equation as the relieving capacity stamped on the valve multiplied by 1.11. The Appendix does not explain the factor, and the usual engineering reading is that it undoes the capacity derating built into certification. A safety valve's stamped capacity is a certified figure set below what the valve passed on the test stand, so that the nameplate is a capacity you can rely on for overpressure protection. For reaction force the risk runs the other way: you want the most steam the valve can pass, not the least. Multiplying by 1.11, which is one over 0.9, brings the nameplate back up toward the flow the valve can really deliver. Two practical points follow. Use the stamped capacity from the valve nameplate or data sheet, not the required relieving load from the overpressure study, which is smaller. And if the manufacturer publishes a reaction force for the valve, compare it with this result; the owner's specification should say which one governs.

What dynamic load factor should I use?

Anything from 1.1 to 2.0, and the Appendix tells you how to earn a value below 2.0. ¶II-3.5.1.3 treats a rigidly supported installation as a one-degree-of-freedom system loaded by a single ramp from zero to the steady force. For that case the factor lies between one and two, depending on how the valve's opening time compares with the natural period of the valve standing on its inlet pipe. The procedure has three steps: calculate the installation period from the weight of the valve and its attachments, the height of their center of gravity above the run pipe, and the stiffness of the inlet pipe; divide the valve opening time by that period; and read the factor from Figure II-3.5.1.3-2. A fast valve on a flexible, heavy installation reads near 2.0. A stiff, short installation reads much lower. The paragraph sets a floor: the factor shall never be taken less than 1.1, and a less conservative value has to come from calculation or test. With no period calculation in hand, 2.0 is the defensible number.

Which steam constants a and b do I use?

The row of Table II-2.2.1-1 that matches the steam condition at the exit and the range your exit pressure falls in. The table has three rows — wet steam, saturated steam and superheated steam — and the second and third are each stated for a range of P₁, the static pressure at the elbow exit. That makes the selection slightly circular, because P₁ is one of the things the constants are used to calculate. In practice you pick a row, compute P₁, and confirm it lands inside the range of the row you picked; if it does not, switch rows and run it again. Note that the selection follows the exit pressure, which is far below the set pressure, so a valve relieving superheated steam does not automatically call for the superheated row. The Appendix's own sample problem in ¶II-7 is worth reading for exactly this point. This page prints no table values: read them from your copy, or pick the steam condition on the card, which fills both constants and cites the table.

Where does the reaction force go?

Into everything between the elbow and the nearest restraint. With no support at the elbow, F₁ × DLF acts at the elbow exit and bends the valve body, the inlet nozzle and the header or run pipe under it. ¶II-3.5.1.1 gives the simple treatment: the moment at any point is the force times the moment arm times a suitable dynamic load factor, and the moment used at the branch connection below the valve is never less than DLF × F₁ × L_o. Those moments go into the loading criteria of ¶II-4.1 together with pressure and weight, with the reaction treated as an occasional load against k·S_h; the occasional loads FAQ covers the k factor. Where the moment is more than the nozzle and header will take, the answer is a support at the elbow that carries the thrust to steel, which then has to be designed for the same force. The relief-valve piping supports guide covers that load path, and the safety valve reaction FAQ covers the rest of Appendix II.

Does this method work for a closed discharge system?

No, and the Appendix is direct about why. In an open discharge installation the valve fires into the atmosphere or into a vent pipe it is not coupled to, the flow reaches a steady state almost at once, and a single force at the elbow describes the load well. A closed discharge system carries the effluent away through pipe connected directly to the valve. When the valve opens, a pressure wave runs down that pipe and can steepen into a shock before it reaches the end, and until steady flow is established every straight run between two elbows sees a momentary unbalanced force. ¶II-2.2.2 recommends a design pressure for the closed discharge pipe of at least twice the steady-state operating pressure for that reason, and ¶II-3.5.2 says these systems do not lend themselves to simplified analysis and that a time-history analysis may be required. The steady-state pressures can still be estimated with the ¶II-2.2.1 method, but the support loads cannot. Treat a closed system as a transient problem from the start.

Related calculators & tools

• ASME B31.1 Power Piping Wall Thickness Calculator — Pressure design of the header and inlet branch the valve sits on
• Equipment Nozzle Load Check Calculator (API 610 / NEMA SM-23) — Compare the reaction with a drum or header nozzle allowable
• Water Hammer Calculator — Joukowsky Surge Pressure and Wave Speed — The other fast momentum load in the same occasional case
• AISC 360 Steel Member Check Calculator (ASD) for Pipe-Support Steel — Size the stanchion that carries the thrust at the elbow
• B31.1 PSV reaction forces — FAQ — The rest of Appendix II: branch moment floor, closed discharge, stress criteria
• B31.1 occasional loads — FAQ — The k·Sh limit the reaction stress is checked against
• PSV relief-valve piping supports guide — Getting the reaction into steel: stanchions, stops and stiffness
• Load cases for B31 stress analysis — Building the relief-thrust occasional case in the stress model
• PiperPSV — relief valve sizing (Pro) — Size the valve first: API 520 orifice area, with this card alongside