Why A106 Grade B Can't Do What P22 or P91 Does at High Temperature
The substitution comes up in enough different forms that it’s worth addressing directly. A project needs alloy steel pipe for a high-temperature service — a process heater outlet line, a main steam header, a reformer feed line. The alloy steel specified is backordered, or the price difference is significant, or someone on the team notes that A106 Grade B is listed as carbon steel and the temperature looks manageable. The question becomes: can we just use A106?
The answer, for P22 and P91 at the temperatures and pressures these grades are typically specified for, is no — and the reason involves a mechanism that carbon steel simply doesn’t have.
What Creep Is and Why It Changes Everything
At temperatures below roughly 370°C (700°F), steel behaves elastically and plastically in the way structural steel behaves at ambient temperature. Load it to below yield and it springs back. Load it beyond yield and it deforms permanently, but the deformation stops when the load stops. The material responds to stress, not to time.
Above about 370°C, something different starts to happen. Steel under sustained stress begins to deform slowly and continuously over time, even at stress levels well below the yield strength. This is creep — time-dependent plastic deformation under steady load at elevated temperature. It’s not a dramatic event. It happens slowly, measured in millimeters of deformation over years of service. But it’s irreversible, it accumulates, and if it accumulates enough, the pipe wall ruptures.
The temperature where creep becomes significant relative to service life is not a fixed number — it depends on the material. For plain carbon steel (A106 Grade B), creep becomes practically significant at around 370°C. For the chromium-molybdenum alloy steels in A335, the alloying elements — chromium, molybdenum, and in P91 also vanadium and niobium — stabilize the microstructure and dramatically slow the creep rate. P11 (1.25% Cr, 0.5% Mo) extends reliable service into the 450–550°C range. P22 (2.25% Cr, 1% Mo) goes higher still. P91 (9% Cr, 1% Mo) with its additional precipitation-hardening elements can be used reliably at 600°C and above.
The Practical Consequence at the Design Temperature
The design pressure of a pipe is determined by dividing the allowable stress for the material — at the operating temperature — by a formula that accounts for wall thickness, outside diameter, and applicable safety factor. ASME B31.3 and B31.1 both tabulate allowable stresses for listed materials at temperature.
For A106 Grade B, the allowable stress tables in ASME B31.3 drop sharply above about 370°C, and the standard restricts A106 Grade B to a maximum temperature of 427°C. Beyond that, it’s outside the code’s coverage entirely.
For P22 per ASTM A335, the allowable stress at 500°C is still a useful value — high enough to result in manageable wall thicknesses for moderate-pressure systems. P91 retains significant allowable stress at 600°C, which is why it’s used in ultra-supercritical steam systems operating at conditions that would make any other material the designer might consider either impractical (too thick) or non-code-compliant.
At 480°C and 10 MPa — conditions that appear in process heaters and steam superheaters fairly regularly — A106 Grade B is simply not in the code tables. It has no listed allowable stress at that temperature. The ASTM A335 pipe grades were developed precisely to fill the range that carbon steel can’t cover.
What Actually Happens When Carbon Steel Is Used at High Temperature
A106 Grade B pipe placed in service at 480°C under sustained pressure doesn’t fail immediately. The initial behavior looks acceptable — the pipe holds pressure, the system operates. Creep begins accumulating.
Over months and years, the pipe wall thins and deforms slowly — not uniformly, but at stress concentration points: at weld heat-affected zones, at supports where secondary stresses add to the primary pressure stress, at areas of local thinning from corrosion. The deformation is not visible during routine inspection because it happens slowly and the geometry change is gradual. What triggers attention is when a wall thickness measurement drops below the minimum required, or when a slow-developing crack at a creep-damaged zone progresses to leakage during a transient condition — startup, shutdown, or a pressure spike.
Creep ruptures in process plant piping are not rare events. They appear regularly in API 579 fitness-for-service assessments, in ASME post-incident analyses, and in refinery and power plant maintenance records. The common factor in cases involving substituted carbon steel for alloy steel is that the failure mode was predictable from first principles — the material didn’t have the creep resistance for the service — but the long incubation period meant the substitution looked successful for years before the consequences appeared.
The Weld Factor
One additional consideration for any comparison between A106 and A335 grades: welding requirements differ significantly.
A106 Grade B has a relatively low carbon equivalent and can be welded with minimal preheat in many configurations. Field welding is straightforward for experienced welders.
P22 requires preheat and post-weld heat treatment (PWHT). P91 requires strict control of preheat, interpass temperature, and PWHT temperature, and it’s sensitive to deviations from the specified PWHT range in ways that can significantly reduce the long-term creep strength of the weld. The welding procedure specification and welder qualification requirements for P91 are considerably more demanding than for carbon steel, and the inspection requirements — including hardness testing after PWHT — reflect the importance of getting the heat treatment right.
This doesn’t make P91 a difficult material to work with, but it does mean that substituting A106 into a system originally designed for P91 and then having to convert back — or introducing a field repair with A106 weld metal into a P91 system — creates complications that are easier to avoid by specifying correctly from the start.
Where the Substitution Question Actually Has a Valid Answer
There is a version of this question where the answer is different: for lower-temperature services where A106 Grade B is within its allowable stress limits, carbon steel is the correct choice and alloy steel would be over-specification. A steam utility line at 200°C and 1.5 MPa doesn’t need P11. A condensate return line doesn’t need P22.
The substitution question only becomes a problem when the temperature pushes above the range where A106 Grade B has listed allowable stresses in the applicable code, or when the creep life calculation for the system doesn’t close with carbon steel properties. Those are engineering calculations, not procurement judgments, and they need to be made with the correct temperature and pressure data for the actual service.
When the engineering says the service requires alloy steel, the alloy steel needs to be procured. The lead time or price premium for P22 or P91 is real, but it’s bounded. The cost of a creep failure in a high-pressure, high-temperature line is not.
