High Temperature and Hydrogen Service

Chrome-moly grade selection is a damage mechanism question. The flange rating is almost never the binding constraint. What binds is creep, hydrogen attack, sulfidation, and on heavy wall hydroprocessing equipment, temper embrittlement.

1. Creep

Above roughly 800 degrees F, steel under sustained load deforms slowly and permanently. Not at yield, at stresses it would carry indefinitely when cold. Given enough time it eventually ruptures. That is creep, and it is time dependent, which makes it unlike every other strength limit in ordinary design.

Chromium and molybdenum in solution, and stable carbides, slow the process. In Grade 91 the vanadium and niobium additions form fine carbonitrides that pin dislocations and give a further large improvement.

Two design consequences follow:

• The allowable stress becomes time based. In the creep range the design code allowables are derived from creep rupture data for a design life, not from yield.
• The design code takes over from the flange rating table. Once you are in the creep range, the governing number comes from the code, and the ASME B16.5 rating table is no longer the whole story.

2. High temperature hydrogen attack

This is the mechanism that dictates grade selection across refinery hydroprocessing, and it is the one people most often underestimate because there is nothing to see until the material is already degraded.

What happens

At elevated temperature and hydrogen partial pressure, molecular hydrogen dissociates at the steel surface and atomic hydrogen diffuses into the metal. Inside, it reacts with carbides to form methane. Methane molecules are far too large to diffuse back out, so they accumulate at grain boundaries under enormous local pressure.

The result is internal fissuring, blistering and decarburization. The material loses strength and toughness, and the damage is distributed rather than local.

It is not repairable. Affected material gets replaced. There is no heat treatment that undoes it.

The Nelson curves

API 941 gives the operating limits, as curves of temperature against hydrogen partial pressure, one per material. Read your operating point against the curve for the grade you intend to use. Inside the curve, with margin, the material is considered resistant.

Adding chromium and molybdenum moves the curve up. The chromium forms stable carbides that hydrogen cannot readily strip, which is precisely why F11 and F22 are the hydroprocessing grades.

Do not confuse this with hydrogen embrittlement, which is a near-ambient loss of ductility from dissolved atomic hydrogen and is largely reversible by baking it out. Different mechanism, different temperature range, different remedy.

3. Temper embrittlement

This one is not corrosion and not a wall thinning mechanism. It is a slow loss of toughness in the material itself, and it matters most on the heavy wall 2.25Cr-1Mo used for hydroprocessing reactors and their connections.

Long exposure in roughly the 650 to 1000 degree F range allows tramp elements to segregate to grain boundaries. The material keeps its strength and loses its toughness, which shows up as a rising ductile to brittle transition temperature. A vessel that was tough when new can become brittle at ambient after years in service.

How it is controlled

By restricting the elements that drive it, at the point of ordering the material:

• The Watanabe J-factor on base metal, an empirical index computed from silicon, manganese, phosphorus and tin. Hydroprocessing specifications cap it.
• The X-bar on weld metal, computed from phosphorus, antimony, tin and arsenic. Also capped.
• A step cooling test, which artificially ages a sample through a defined thermal cycle and then measures the toughness shift, to demonstrate resistance rather than infer it from chemistry alone.
• API RP 934 series guidance for heavy wall hydroprocessing reactor construction, which is where these requirements are usually drawn from.

The operating consequence of temper embrittlement is the minimum pressurization temperature rule. Once toughness has degraded, pressurizing a cold vessel risks brittle fracture, so operations keep the metal above a defined temperature during pressurization.

4. Sulfidation

Sulfur compounds in hot hydrocarbon streams react with the steel surface to form iron sulfide scale. The scale is not protective in the way a passive film is, so the wall thins steadily and predictably. Rates rise sharply with temperature.

Chromium is the element that slows it, because chromium-bearing scales are far more protective than iron sulfide. This is the main reason F5 and F9 exist in refinery service, and it is often the reason a crude or vacuum unit specifies 5Cr where the temperature alone would not demand it.

Rate estimation comes from established curve sets:

• Modified McConomy curves for high temperature sulfidic corrosion without hydrogen.
• Couper-Gorman curves where hydrogen is present, as in hydroprocessing.

Two related mechanisms often travel with sulfidation and change the answer:

• Naphthenic acid corrosion, driven by the acid content of the crude and by velocity. It attacks in a way chromium alone does not fix, and the usual answer is an austenitic stainless with molybdenum rather than a chrome-moly.
• Silicon content in carbon steel, which affects sulfidation rate and has caused failures where low silicon components sat alongside normal ones in the same service. It is a carbon steel issue rather than a chrome-moly one, but it turns up in the same units.

Putting it together

A single service usually involves several of these at once. A hydroprocessing reactor connection sees temperature, hydrogen and sulfur simultaneously, and if it is heavy wall it carries temper embrittlement restrictions on top. The grade has to satisfy all of them, so the binding constraint is whichever is most demanding.

What that means for an enquiry:

1. Design temperature, and the design life if the specification sets one
2. Hydrogen partial pressure, not just "hydrogen service"
3. Sulfur environment, and whether naphthenic acid is a factor
4. Any restricted chemistry requirement, with the specification imposing it
5. Hardness limits, and whether NACE MR0175 or ISO 15156 applies
6. Impact testing requirement and test temperature

With that we can confirm the grade before you buy it. Plenty of chrome-moly gets ordered a grade below what the Nelson curve actually allows, and the flange is not the place to find that out.

See chrome-moly grades for the material reference, and preheat and post weld heat treatment for the fabrication side.

Common questions