Why has C22 become the universal alloy in the chemical industry?

In 2008, a petrochemical company commissioned a 100,000-ton-per-year methyl methacrylate (MMA) plant. In the material column for the inner coils of a key heat exchanger, it was written clearly: C22 Hastelloy. The designer’s reason for choosing it was entirely sound—inside the coils flowed a mixture of amide and sulfuric acid at 145–165°C, with both reducing acid and oxidizing conditions. C22 was precisely the material prepared for this kind of “neither one nor the other” service. The tube diameter was φ114 mm, wall thickness 3.02 mm, with ten layers of coils, inside and out. It looked like a carefully considered configuration.

Then it leaked. Not in the place where the corrosive liquid was most aggressive, but in the heat-affected zone on both sides of the weld. During a shutdown for maintenance, the equipment was cut open, metallography was performed, EDS was performed, and the conclusion was one sentence: improper welding had caused high-molybdenum carbides to precipitate at grain boundaries in the heat-affected zone, leaving chromium-depleted regions around them, which were preferentially penetrated in the sulfuric-acid-containing medium.

You paid a high price for C22, and in the end it did not fail in the corrosive liquid—it failed in the welder’s craftsmanship.

The absurdity of this matter lies in the fact that C22 was invented precisely to patch the shortcomings of C276 in the weld heat-affected zone. An alloy sold on the promise of being “more welding-friendly” ultimately exited in an extremely classic welding-failure mode. This is not irony about the material itself; it is the concrete manifestation of the gap between material-selection logic and fabrication reality.

1985: The Birth of a “Patch”

The family tree of the Hastelloy C series was built patch by patch, generation after generation. In the 1960s, C276 arrived, with high molybdenum and high tungsten. It was unrivaled in reducing media such as hydrochloric acid and sulfuric acid, and quickly became the “cure-all” for corrosive chemical service. But C276 had a common weakness shared by nickel-based alloys of that era: after welding, detrimental phases tended to precipitate in the heat-affected zone, chromium was depleted near grain boundaries, and corrosion resistance dropped off a cliff. In addition, its chromium content was only 14.5–16.5%, so in oxidizing media such as nitric acid, chromic acid, ferric chloride, and wet chlorine gas, the stability and self-healing ability of its passivation film were insufficient.

So in 1985, Haynes brought C22 to market. Its positioning was very clear: it was not a fresh start from scratch, but a patch for C276. Chromium content was raised to 20–22.5%, nearly six percentage points higher than C276, specifically to patch the weakness in oxidizing media; carbon was pressed down to ≤0.005%, and silicon to ≤0.005%, with the aim of reducing, from the root of compositional design, the driving force for detrimental phase precipitation in the weld heat-affected zone. PREN—the pitting resistance equivalent number—falls between 65 and 74 depending on the calculation method, placing it in the top tier among nickel-based alloys.

In one sentence: C276 is a powerful but gas-guzzling old off-roader; C22 is a new model on the same platform with upgraded electronic controls and emissions. Its original design intent was to give chemical equipment that might encounter both oxidizing media and welding a choice less likely than C276 to have trouble next to the weld.

“The Transfer Student in Class Who Scores 85 in Every Subject”

If the Hastelloy C series were brought into a classroom, C276 would be like that veteran with uneven grades—able to score 95 in some subjects but hovering around 70 in one or two others. Reducing acids are its home turf; oxidizing media are its weak spot. C22 is a completely different kind of student: it is never the top scorer in any single subject, but its total score is always the most stable.

In reducing media, C22’s molybdenum content is 12.5–14.5% and tungsten content is 2.5–3.5%, with total Mo + W lower than C276 by a noticeable margin. In boiling 10% hydrochloric acid, C276’s corrosion rate is about 30% lower than C22’s. These are the points C22 concedes. But it pulls chromium up to around 22%, strongly making up for it in oxidizing media: in the same 10% sulfuric acid plus 5% nitric acid mixture at 90°C, C22’s annual corrosion rate is 0.05 mm, while C276’s is 0.12 mm—more than double the difference. Ferric chloride, wet chlorine gas, hypochlorites, and oxidizing chlorides containing Fe³⁺ or Cu²⁺—in every subject that tests the “self-healing ability” of the passivation film, C22 stands solidly above 85.

Most critically, it has no weak subject. What chemical plants fear most is never “one particular medium is especially aggressive,” but rather service conditions flipping back and forth between oxidizing and reducing, or the medium containing both types of corrosion factors at the same time. C276’s 95 in the reducing side is reassuring, but once process fluctuations push oxidizing conditions up, its margin for error narrows. C22’s selection logic is: 85 in every subject, the most stable total score; when you do not know which way the next batch of material will lean, betting on it is least likely to fail. Thus its status as the “universal spare tire” took shape. Not because it is strongest in every category, but because it is “good enough and unlikely to make a big mistake” in the greatest number of service conditions. The material-selection decisions of chemical industry folks are often not made in the laboratory, but among schedule, budget, medium analysis reports, and “what the previous unit used.” When a material’s reputation is “an upgraded version of C276,” “stronger in oxidizing media,” and “better corrosion resistance in the weld zone,” it is almost inevitable that it will be played as the card of “try C22 first for any service condition.”

But “No Precipitation by Design” Does Not Equal “No Trouble in the Field”

C22’s metallurgical design does indeed move in the direction of “causing less trouble in the weld heat-affected zone.” Carbon and silicon are pressed to extremely low levels, with the aim of thermodynamically reducing the nucleation opportunities for detrimental precipitates such as μ phase, P phase, and carbides in the sensitization range of 650°C to 1100°C. On paper, this is a good move.

The problem is that welding is never a process executed “as on paper.”

Welding is local metallurgy. The weld pool melts and solidifies within seconds; the heat-affected zone is dragged through a series of temperature ranges, and the cooling rate depends on plate thickness, ambient temperature, bead arrangement, interpass temperature, heat input, shielding gas, and even the welder’s torch-travel habit that day. C22’s compositional design can lower the “precipitation tendency,” but it cannot lock every on-site thermal cycle to an ideal curve. When heat input is excessive, interpass temperature is out of control, and high-temperature dwell time is prolonged, the heat-affected zone will still enter that temperature window where it should not linger, and let what should precipitate precipitate.

The failure of that coil in the MMA plant followed exactly this path. Improper welding caused high-molybdenum carbides to precipitate at grain boundaries in the heat-affected zone between the base metal and the weld. Molybdenum was pulled away by the carbides, creating molybdenum-depleted microregions in the surrounding matrix. In the oxidizing sulfuric-acid-containing medium, these local regions lost enough support from corrosion-resistant elements, and grain boundaries became preferential dissolution paths. Corrosion proceeded along the grain boundaries inward and eventually penetrated the tube wall. There is another sentence in the paper that every person involved in equipment should read more than once: solution treatment and rapid cooling can eliminate the precipitation of high-molybdenum carbides at grain boundaries, thereby avoiding this type of leakage. In other words, the problem is not that “C22 cannot be welded,” but that “after welding, what should have been done was not done, or the welding process itself did not leave the material room not to fail.”

Another example, the failure of a spiral heat exchanger tube, had a similar outcome but added a layer of field realism. Obvious local thinning appeared in the heat-affected zone on both sides of the weld. The analysis conclusion was likewise that an unreasonable welding procedure led to intergranular corrosion in the heat-affected zone, while erosion by the solution accelerated the damage. Here, “erosion” is not a supporting role. Once intergranular corrosion tears open tiny selective channels in the heat-affected zone, the flowing medium carries away corrosion products, exposes fresh metal surfaces, and turns what might have been a self-limiting intergranular attack into continuous thinning. Welding opened a breach at the grain boundaries; the flow field is responsible for tearing that breach wider and wider.

Why Did It Fail Precisely in “the Area It Was Most Proud Of”?

Because the four words “welding-friendly,” in purchase orders and technical agreements, are often translated into “you can weld it any way you like,” while in metallurgy their true meaning is “more forgiving than C276, but forgiveness has a limit.”

In the C276 era, welders and process engineers had muscle memory for “nickel-based alloys will have trouble in the heat-affected zone after welding.” By the time C22 arrived, material salespeople would say “weld-zone corrosion resistance is better than C276,” and design institutes would write in specifications that “welding performance is good.” As a result, the field easily developed a dangerous complacency: since it welds better than the old material, following the pace of stainless steel should be no big problem.

But C22’s welding window is not wide. Public process literature repeatedly emphasizes the same thing: low heat input, fast welding, short arc, no weaving or slight weaving, and strict control of interpass temperature. Some specifications cap interpass temperature below 90°C; others set upper limits of 100°C or 150°C. But they all point in exactly the same direction—the heat-affected zone must not stay too long in the high-temperature range. Heat input is recommended to be controlled in the range of 100–200 J/cm, and welding current should be 20–30% lower than for stainless steel of the same thickness. The groove before welding should be machined and cleaned with acetone. After welding, under certain severe service conditions, solution treatment is required, with a temperature range of about 1120°C to 1170°C, followed by rapid cooling after holding, to rehomogenize the segregated microstructure and redissolve precipitates that should not remain at grain boundaries. None of these requirements is difficult when written on paper; but each one, when implemented in the workshop, requires people, machine, material, method, and environment to all be in place. Do the welders have specific qualifications for nickel-based alloys? Was interpass temperature actually measured with a temperature crayon, or does “it feels not hot” count? Were the heat input parameters specifically qualified for C22, or copied from a C276 procedure card and just had the alloy name changed? Was post-weld solution treatment written into the manufacturing outline, or omitted under the vague wording “generally no heat treatment is required”?

C22’s “welding friendliness” is relative to C276, not relative to “any welder, any set of parameters, any cooling rate.” When one link in process discipline slackens, the heat-affected zone will remind you with grain-boundary precipitation and chromium-depleted microregions: the material’s design margin is not there for the field to make mistakes with.

 A Practical Checklist for Chemical Industry Folks: Write the Weld Bead into the Contract as Part of Material Performance

The material-selection decision for C22 usually ends at “the material is specified as C22.” But the two cases above show that the material grade only locks in the potential of the base metal; the weld and heat-affected zone are the “lower-limit-determining zone” of the equipment’s actual service performance. When purchasing C22 equipment, the following items should enjoy equal contractual status with the “material grade.”

First, welder qualification must be specific to nickel-based alloys and the welding method. Not “has a welder’s certificate,” but “has a qualified item for GTAW of C22 or the same family of nickel-based alloys,” and the qualification must cover the actual plate thickness and joint type. Second, the welding procedure parameters should be made as a technical annex to be locked down: heat input range, interpass temperature upper limit, bead layout, shielding gas, and cleaning procedure. Open-ended wording such as “per manufacturer recommendation” is not acceptable. Third, post-weld treatment must clearly state “do” or “do not do.” If solution treatment is to be done, list the temperature, holding time, cooling method, and inspection method item by item; if not, there must be a written basis based on the service medium and welding procedure qualification, rather than a sentence saying “usually not required.” Fourth, corrosion inspection of the heat-affected zone cannot rely only on the base metal mill certificate. The intergranular corrosion tendency and metallographic structure of the weld and heat-affected zone should be part of the factory documentation.

These requirements may seem trivial, but they correspond to a very simple fact: C22’s high score in oxidizing media belongs to this equipment only on the premise that the heat-affected zone on both sides of the weld has not been artificially “ruined.” No matter how beautiful the base metal’s PREN is, once the heat-affected zone becomes chromium-depleted due to the welding thermal cycle, the corrosive medium will not check the material certificate. It only recognizes how much chromium and molybdenum remain at the grain boundary in front of it.

C22 is a very honest alloy. It was born in the 1980s, designed to patch C276’s weaknesses, and then over decades in chemical plants it was played as a universal card: “try C22 first.” Most of the time it caught the ball, because its total score is indeed stable. But it never promised “no trouble no matter how you weld it.” It raised welding friendliness above C276, pressed carbon and silicon to extremely low levels, and made the passivation film in oxidizing media more robust—and then it left the last line of defense to process discipline.

When that line of defense cannot be held, the failure report does not say “C22 corrosion resistance insufficient”; it says “improper welding.” An alloy whose mission is to be a patch ultimately needs another layer of patch to protect the area it is most proud of.

C22’s corrosion-resistance ceiling is not given by the material; it is given by the weld bead.

➤