Temper embrittlement is not corrosion and it does not thin a wall. It is the material quietly losing toughness while keeping its strength, over years of service, and it is the reason heavy wall hydroprocessing specifications read like chemistry examinations.
What happens
Long exposure in roughly the 650 to 1000 degree F range allows certain impurity elements to diffuse to and segregate at grain boundaries. The boundaries become the weakest path through the material, and fracture switches from ductile to brittle intergranular.
Measured, it shows up as a rising ductile to brittle transition temperature. A vessel that had good toughness at ambient when new can, after years in service, be brittle at ambient. The tensile properties barely move, so nothing in a routine strength check flags it.
The grade most affected in practice is 2.25Cr-1Mo, F22, which is exactly the grade heavy wall hydroprocessing reactors are built from. Hence the attention.
The operating consequence
This is where it stops being a metallurgical curiosity. If toughness has degraded, pressurizing the vessel while it is cold risks brittle fracture, because a brittle material with a flaw and a stress does not yield, it breaks.
The control is a minimum pressurization temperature: an operating rule requiring the metal to be above a defined temperature before the unit is pressurized. It gets revised over the life of the plant as toughness assessments are updated. Operators take it seriously because the failure mode it guards against is catastrophic rather than gradual.
The chemistry controls
Since the mechanism is driven by impurity segregation, the defence is to buy material with fewer impurities. Two empirical indices are used, and both are calculated from the heat chemistry.
J-factor, for base metal
The Watanabe J-factor combines silicon and manganese with phosphorus and tin. Higher silicon and manganese amplify the effect of the phosphorus and tin present, so the index multiplies the pair of sums.
Hydroprocessing specifications cap it. The cap is what forces the mill to a cleaner practice than the ASTM specification alone would require.
X-bar, for weld metal
Weld metal gets its own index, combining phosphorus, antimony, tin and arsenic. It is expressed in parts per million and capped separately, because weld consumables and the welding process introduce their own impurity contributions.
Neither index is part of ASTM A182. A182 specifies the alloy chemistry for F22 and imposes no J-factor limit. So ordinary F22, fully compliant with its ASTM specification, does not satisfy a restricted chemistry requirement. The restriction comes from the project specification, usually drawing on the API RP 934 series, and it must be on the enquiry.
Step cooling
Chemistry indices are empirical correlations, and for critical equipment a correlation is not always enough. Step cooling tests the material directly.
A sample is put through a defined multi-step thermal cycle, holding at successively lower temperatures, which artificially accelerates the segregation that service would cause over decades. Toughness is measured before and after, and the shift in transition temperature is compared against an acceptance criterion.
It demonstrates resistance rather than inferring it. It also takes time, which is part of why restricted chemistry material with step cooling carries the lead time it does.
What this means when you order
The practical point is that restricted chemistry determines which heats of material can be offered at all. It is not a test applied to material after the fact, and it cannot be retrofitted onto a heat that does not meet it.
So the enquiry needs:
- The specification imposing the restriction, with revision
- The J-factor limit for base metal
- The X-bar limit if weld metal is in scope of the supply
- Whether step cooling is required, and the acceptance criterion
- Impact testing requirements and test temperature
- Whether third party witness is required on any of it
Raise it at enquiry, not at order. Restricted chemistry material is a smaller pool, and on some sizes and product forms it is a made-to-order proposition. A requirement that appears after material has been committed is expensive to unwind, and expediting is not usually an option because the constraint is the heat, not the machining.
Where it does not apply
Worth saying, because these restrictions get copied onto specifications where they serve no purpose and simply add cost and lead time.
- F11 does not normally attract them. 1.25Cr-0.5Mo is far less susceptible than 2.25Cr-1Mo.
- Thin section work is less of a concern, because the consequence of brittle behaviour scales with section thickness and stored energy.
- Service outside the embrittling range does not drive the mechanism. It needs sustained time in roughly 650 to 1000 degrees F.
If a specification is imposing J-factor limits on an application that does not need them, that is worth a conversation before the enquiry goes out. We will flag it if we spot it.
See high temperature and hydrogen service for the other mechanisms, and F11 vs F22 for the grade decision.