ASME B31.3 Allowable Stress (S) & Y-Factor Temperature Derating Guide

Under ASME B31.3 internal-pressure design, required wall thickness depends not only on pressure $P$ and outside diameter $D$, but also on Allowable Stress ($S$) at metal design temperature and the coefficient $Y$ (Y-Factor). As temperature rises, Table A-1 $S$ falls sharply; for ferritic steels above $482^\circ\text{C}$ ($900^\circ\text{F}$), Table 304.1.1 $Y$ also increases. Together, those shifts can move schedule selection by one or more steps.

$$t = \frac{P \cdot D}{2(S \cdot E \cdot W + P \cdot Y)}$$ $$t_m = t + c,\qquad t_{\text{nom,req}} = \frac{t_m}{0.875}$$

Quick Summary (TL;DR)

---

1. Basis of ASME B31.3 Allowable Stress ($S$)

1.1 Elastic-range basic allowable

Basic Allowable Stress $S$ in Table A-1 is the minimum of tensile- and yield-based limits with code safety factors. For ferritic / carbon steel:

$$S = \min\left(\frac{S_t}{3},\ \frac{2}{3}R_y\right)$$

Example — ASTM A106 Gr.B: $S_t = 60\text{ ksi}$, $R_y = 35\text{ ksi}$

$$\frac{S_t}{3} = 20.0\text{ ksi},\quad \frac{2}{3}R_y = 23.3\text{ ksi} \implies S = 20.0\text{ ksi}\ (138\text{ MPa})$$

For austenitic stainless steels, a higher fraction of yield is often permitted:

$$S = \min\left(\frac{S_t}{3},\ \frac{90\%\,R_y}{1.5}\right) = \min\left(\frac{S_t}{3},\ 0.9\cdot\frac{2}{3}R_y\right)$$

That is why TP304 can show ambient $S$ near carbon-steel levels ($\sim 20\text{ ksi}$), while L-Grades (lower $S_t$ / $R_y$) start lower.

1.2 High temperature: transition to Creep & Stress Rupture

At elevated metal temperature, time-dependent Creep and Stress Rupture govern $S$ instead of static yield alone.

Carbon steel also faces Graphitization under long exposure above $427^\circ\text{C}$ (cementite → graphite). Even when Table A-1 still lists an $S$, process practice usually replaces CS with Cr–Mo alloy for continuous hot service.

1.3 Quality Factor $E$ and Weld Strength Reduction $W$

$$t = \frac{P \cdot D}{2(S \cdot E \cdot W + P \cdot Y)}$$

The FieldEngineersKit Pipe Thickness Calculator uses the seamless screening form $S\cdot E + P\cdot Y$ (equivalent to $W=1$) with user-entered $S$ and $Y$.

---

2. Material-Group Derating (CS vs SS vs Low Alloy)

Values below are rounded screening conversions from representative Table A-1 figures ($1\text{ ksi}=6.895\text{ MPa}$). Confirm the project edition and Notes before issue.

2.1 Comparison — Basic Allowable Stress $S$ ($\text{MPa}$)

2.2 Carbon Steel — A106 Gr.B / A53 Gr.B

• Near $371^\circ\text{C}$ ($700^\circ\text{F}$), creep begins to dominate and the $S$ curve steepens.
• At $427^\circ\text{C}$ ($800^\circ\text{F}$), $S \approx 79\text{ MPa}$ ($11.4\text{ ksi}$) — about 43% of ambient $138\text{ MPa}$.
• At $482^\circ\text{C}$, $S \approx 41\text{ MPa}$ ($5.9\text{ ksi}$). For continuous hot service, CS is unfavorable on both creep and Graphitization. Steam / hot-oil headers typically move to P11 or higher.

2.3 Austenitic SS — TP304/316 vs TP304L/316L

Above about $150^\circ\text{C}$, defaulting to “L only” can inflate wall thickness and flange rating without benefit. Dual-certified (304/304L) stock still requires a declared design grade in Table A-1. Prefer Standard or H grades when hot strength governs; prefer L when weld-zone corrosion / sensitization governs.

2.4 Low Alloy — A335 P11 / P22 / P91

• P11 (1¼Cr–½Mo): $\approx 97\text{ MPa}$ at $482^\circ\text{C}$ — more than double CS at the same temperature ($41\text{ MPa}$).
• P22 (2¼Cr–1Mo): Extra high-temperature margin for $500$–$550^\circ\text{C}$ steam / process service.
• P91 (9Cr–1Mo–V): High ambient and elevated $S$ ($\approx 124\text{ MPa}$ at $538^\circ\text{C}$) — core creep-resistant alloy for USC / severe hot piping. Welding, PWHT, and hardness control are mandatory.

---

3. Y-Factor Mechanics & Table 304.1.1 Thresholds

3.1 Physical meaning of $Y$

In Eq. (3a), $Y$ approximates the effect of circumferential stress distribution and plastic strain, including neutral-axis shift in thicker walls.

• $Y = 0$: Thin-membrane extreme — no plastic redistribution (cast / ductile iron).
• $Y = 0.4$: Default for ductile steel at ambient / moderate temperature.
• $Y \to 0.5$: Approaches an idealized thin-wall plastic limit state.
• $Y > 0.5$ (typically $0.7$): Accounts for stress redistribution under thick-wall creep → larger $P\cdot Y$ term → slightly smaller calculated $t$.

In practice, the $S$ drop increases $t$ far more than any $Y$ increase can reduce it.

3.2 Table 304.1.1 — $Y$ values

Interpolate between tabulated temperatures. Example — ferritic at $485^\circ\text{C}$ ($905^\circ\text{F}$):

$$Y = 0.4 + \frac{905 - 900}{950 - 900}\,(0.5 - 0.4) = 0.41$$

Austenitic steels keep $Y=0.4$ through $1000^\circ\text{F}$, so their $Y$ rise lags ferritic / alloy steels.

---

4. Worked Example — $200^\circ\text{C}$ vs $485^\circ\text{C}$ Derating

Common design data

NPS 8 commercial walls (B36.10M): Sch 40 $= 8.18\text{ mm}$, Sch 80 $= 12.70\text{ mm}$, Sch 160 $= 18.23\text{ mm}$.

Case A — $200^\circ\text{C}$, ASTM A106 Gr.B

• Table A-1 interpolation: $S \approx 19.91\text{ ksi} = 137.3\text{ MPa}$
• Ferritic $Y = 0.4$ ($\le 482^\circ\text{C}$)

$$t = \frac{4.0 \times 219.1}{2\bigl(137.3 \times 1 \times 1 + 4.0 \times 0.4\bigr)} = \frac{876.4}{2(137.3 + 1.6)} = \frac{876.4}{277.8} = 3.16\text{ mm}$$ $$t_m = 3.16 + 1.5 = 4.66\text{ mm},\qquad t_{\text{nom,req}} = \frac{4.66}{0.875} = 5.32\text{ mm}$$

Select Sch 40 ($8.18\text{ mm} \ge 5.32\text{ mm}$) with comfortable margin.

Case B — $485^\circ\text{C}$, A106 Gr.B vs A335 P11

$485^\circ\text{C} = 905^\circ\text{F}$ → $Y = 0.41$ (interpolated).

A106 Gr.B:

$$t = \frac{876.4}{2(39.4 + 4.0\times 0.41)} = \frac{876.4}{2(39.4 + 1.64)} = \frac{876.4}{82.08} = 10.69\text{ mm}$$

P11:

$$t = \frac{876.4}{2(96.3 + 1.64)} = \frac{876.4}{195.88} = 4.48\text{ mm}$$

Interpretation

1. Raising temperature from $200$ to $485^\circ\text{C}$ drops A106 $S$ from $137$ to $39\text{ MPa}$ and drives $t_{\text{nom,req}}$ from $5.3$ to $13.9\text{ mm}$ — Sch 40 → Sch 160.
2. $Y$ only moves $0.40\to0.41$; it does not reverse the schedule jump.
3. At the same $485^\circ\text{C}$, P11 still passes on Sch 40 — material upgrade cuts schedule, weight, and often flange class together.
4. Leaving A106 at $485^\circ\text{C}$ retains Graphitization and creep-rupture risk. A listed $S$ does not make continuous CS service acceptable.

---

5. Field Checklist & Calculator

High-temperature line-spec checklist

Interactive calculator

---

FAQ

Q1. Why does schedule jump so hard at high temperature?

Because $t \propto 1/(S\cdot E\cdot W + P\cdot Y)$. When $S$ falls by half or more, $t$ nearly doubles. The modest rise in $Y$ only slightly offsets that. Case B (A106) is the textbook illustration.

Q2. Is L-Grade stainless forbidden on hot lines?

Not forbidden — often inefficient. L-Grades help corrosion / sensitization but carry lower Table A-1 $S$, so the same $P$ needs a thicker wall and possibly a higher flange class. When hot strength governs, specify TP304/316 or H as the design grade.

Q3. If $Y=0.7$, can I safely thin the wall?

$Y$ is only a stress-distribution coefficient in Eq. (3a). It does not replace creep-rupture, oxidation, or Graphitization limits. Even when $Y$ rises and $t$ dips slightly, low $S$ and material degradation still control design.

Q4. When is the calculator’s $S=138\text{ MPa}$ valid?

It matches A106 Gr.B ambient / low–moderate temperature Table A-1 binding ($20.0\text{ ksi}$). Near $200^\circ\text{C}$ the value is still close, but above roughly $400^\circ\text{C}$ you must enter the temperature-specific $S$. Using ambient $S$ on a hot line produces non-conservative (dangerously thin) walls.