A tube bundle fails in eighteen months, and the metallurgical report says the alloy was never wrong on paper. Type 316L was rated for the acid, temperature sat inside the published range, and the corrosion allowance was generous. What the specification missed was a trace of ferric ion carried over from an upstream oxidation step, or a shell-side dead leg where chloride concentrated between shutdowns. Exotic alloys are not stronger grades of stainless steel. Each one defeats a specific corrosion mechanism, and each has an environment that destroys it.
TL;DR
-
Alloy selection follows corrosion mechanism, not a general ranking. Titanium and zirconium have nearly opposite failure envelopes.
-
Types 304 and 316 can crack from chloride stress corrosion cracking at under 1 ppm chloride near boiling, so a low chloride analysis alone does not clear them.
-
Nickel governs chloride SCC resistance, chromium handles oxidizing acids, molybdenum handles reducing acids and pitting. Every nickel alloy is a ratio of those three.
-
Zirconium handles sulfuric acid to about 75% at 212°F, then loses resistance abruptly between 77% and 77.5%, and 500 ppm of ferric ion cracks it in hydrochloric service.
-
Fabrication controls decide service life as much as alloy choice. In titanium, iron is treated as an interstitial contaminant rather than shop dirt.
On this page: corrosion mechanisms · the stainless ceiling · nickel alloys · titanium · zirconium · cost and construction · fabrication and inspection · common mistakes · FAQs
The three corrosion mechanisms that drive alloy selection
Three mechanisms account for most exotic alloy exchanger failures, and each answers to a different alloying element.
Chloride stress corrosion cracking needs tensile stress, chloride, and temperature together. It produces branched transgranular cracks with almost no metal loss, so thickness surveys miss it until a tube leaks. Nickel resists it. The International Molybdenum Association states that nickel “increases stress corrosion cracking resistance above about 20%”, which is why most high-performance austenitic grades carry roughly 18–25% nickel where Types 304 and 316 sit near 8–14%. The nickel alloys extend that logic much further, into the 40–70% range.
Pitting and crevice corrosion are localized breakdowns of the passive film, initiated by chloride and accelerated under deposits, gaskets, and tube-to-tubesheet crevices. Chromium, molybdenum, nitrogen, and tungsten resist it, condensed into one figure by the Pitting Resistance Equivalent Number: PREN = %Cr + 3.3(%Mo + 0.5%W) + 16(%N).
General acid attack splits along the distinction that matters most here: oxidizing versus reducing service. Chromium builds the passive film that survives oxidizing acids such as nitric. Molybdenum and nickel resist reducing acids such as hydrochloric. An alloy optimized for one is compromised in the other, and among the commonly specified alloys none handles both extremes at high concentration and temperature. Coverage that broad generally means moving to tantalum, at a cost step most projects will not carry.
PREN ranks similar alloys in chloride service. It says nothing about acid resistance, cracking, or the oxidizing/reducing split, so PREN alone cannot select an alloy for process duty.
The stainless ceiling: where 316L and duplex stop
Most exotic alloy projects begin as an escape from a stainless failure, so the boundary deserves precision.
For Type 316L, both the chloride and the temperature thresholds sit lower than most specifications assume. IMOA reports that under fully immersed, near-neutral conditions “it is rare to see CSCC at temperatures below 50°C (122°F)”, then adds that surfaces which alternately wet and dry, or which evaporate, crack well below that. Near boiling under immersion, the chloride content needed to crack these alloys can be less than 1 ppm. Localized corrosion resistance is equally modest: IMOA states that “Type 316L, with an ASTM G48 CPT of 10°C (50°F), is not suitable for steam condenser tube service in seawater.”
Duplex grades buy real margin, but not temperature. IMOA tabulates the maximum ASME design temperature for 2205 and 2507 at 600°F (315°C), with European TüV practice more conservative at 480°F (250°C). The controlling mechanism is alpha prime formation in the ferrite phase, which IMOA places in the 575–980°F (300–525°C) band.
| Grade | UNS | Mo (%) | PREN | Note |
|---|---|---|---|---|
| 316L | S31603 | 2.0–3.0 | 24–26 | G48 CPT near 50°F (10°C) |
| 2205 | S31803 | 2.5–3.5 | 33–35 | Older, wider nitrogen range |
| 2205 | S32205 | 3.0–3.5 | 35–36 | Current grade; ASME design limit 600°F (315°C) |
| 2507 | S32750 | 3.0–5.0 | 40–43 | Super duplex; same 600°F (315°C) ASME limit |
| 254 SMO | S31254 | 6.0–6.5 | 41–43 | 6-Mo superaustenitic |
| AL-6XN | N08367 | 6.0–7.0 | 43–45 | 6-Mo superaustenitic |
Compositions and PREN values per IMOA fabrication guidelines.
A procurement trap sits inside that table. Both S31803 and S32205 are sold as “2205” under EN designation 1.4462, but only S32205 carries the tightened nitrogen range and the higher PREN.
The 6% molybdenum superaustenitics are the usual step above duplex, and their advantage is structural as well as chemical. IMOA gives the ASME allowable stress for N08367 at 800°F (427°C) as 18 ksi against 9.6 ksi for Type 316 — nearly double, which translates into thinner walls for the same duty.
Nickel alloys: matching Hastelloy and Inconel to the acid
Nickel alloys read as three-element recipes. Nickel suppresses chloride SCC, chromium handles oxidizing media, molybdenum handles reducing acids and pitting. Copper, where present, is there for sulfuric acid.
Hastelloy C-276 (N10276) balances 16% chromium against 16% molybdenum with 4% tungsten, which is why it tolerates both acid types better than single-purpose alloys. Haynes publishes a critical pitting temperature of 302°F (150°C) and a critical crevice temperature of 131°F (55°C) in acidified 6% ferric chloride per ASTM G48. Its limit is hot oxidizing acid, and the published nitric acid data shows how sharply temperature dominates concentration: at 200°F (93°C), 70% nitric gives 2.62 mm/y, while at boiling even 60% nitric reaches 18.42 mm/y. Nitric service is a temperature question before it is a concentration question.
Hastelloy B-3 (N10675) is the specialist, at roughly 28.5% molybdenum with only 1.5% chromium, built for pure hydrochloric, hydrobromic, and sulfuric acids and other non-oxidizing acids. With essentially no chromium it has no oxidizing-acid capability, and oxidizing-ion contamination is the recognized restriction on B-type alloys: a B-3 bundle in hydrochloric service is at risk if an upstream unit puts ferric or cupric ion into the stream. Confirm the tolerance for a specific stream with the mill, because the margin is narrow.
Monel 400 (N04400) holds a niche nothing else fills: Special Metals states it “offers exceptional resistance to hydrofluoric acid in all concentrations up to the boiling point.” Its weakness is stagnation. The same bulletin reports low corrosion rates in flowing seawater but crevice and pitting attack under stagnant conditions, which makes shell-side dead zones and idle-period layup a design concern rather than an operating detail.
| Alloy | UNS | Key composition | Selected for | Avoid |
|---|---|---|---|---|
| C-276 | N10276 | 16Cr–16Mo–4W | Mixed oxidizing/reducing acids, chlorides | Hot nitric — 18.42 mm/y in 60% boiling |
| C-22 | N06022 | 22Cr–13Mo–3W | Oxidizing acids, mixed service | 70–90% H₂SO₄ at and above 150°F |
| B-3 | N10675 | 28.5Mo–1.5Cr | Pure HCl, HBr, H₂SO₄, non-oxidizing acids | Oxidizing media; ferric/cupric ions |
| 625 | N06625 | 21Cr–9Mo–4(Nb+Ta) | Cryogenic to 1800°F (982°C); seawater | — |
| 825 | N08825 | 21Cr–3Mo–2Cu–Ti | Sulfuric and phosphoric acid | Above 1000°F (540°C) |
| 600 | N06600 | 15Cr, no Mo | High-temperature oxidation, caustic, high-purity water | Chloride pitting service |
| 400 | N04400 | 30Cu | Hydrofluoric acid to boiling | Stagnant seawater; oxidizing media |
Compositions and temperature limits per Haynes International and Special Metals technical bulletins. Special Metals attributes 825’s acid resistance to nickel acting in conjunction with molybdenum and copper.
Titanium: oxidizing chlorides, and where it stops
Titanium’s record in chloride cooling water is hard to argue with. ATI reports that millions of feet of welded titanium tubing are in power plant condenser service with no reported corrosion failures on the cooling water side, and that its resistance extends across seawater, salt brines, and hot chloride salt solutions.
Grade 2 (R50400) is the commercially pure workhorse for shell-and-tube duty. Grade 7 (R52400) adds 0.12–0.25% palladium for reducing-acid and crevice resistance at roughly five times the price of Grade 2. Grade 12 (R53400) adds small nickel and molybdenum for higher strength and better crevice resistance.
Crevice corrosion sets the service envelope. ATI publishes the commercially pure and palladium-alloyed limits as a temperature ladder, and the Materials Technology Institute adds the Grade 12 figures:
-
Unalloyed Grades 1–4 typically do not suffer crevice corrosion below 176°F (80°C) at any pH
-
Palladium-alloyed Grades 7, 11, 16, and 17 typically do not suffer crevice corrosion below 480°F (250°C) above pH 1
-
Grade 12 can be used to 464°F (240°C) above pH 2, and holds to about pH 2 in hydrochloric service against pH 0.1 for Grade 2
Three limits disqualify titanium outright, and all three appear routinely in chemical plants.
Fluoride attacks titanium by dissolving the protective oxide. The widely repeated 5 ppm limit deserves scrutiny; a fabricator’s technical author states in print that he could not locate its primary source. What is documented is the pH dependence — above roughly pH 3.2 most fluoride exists as free F⁻ rather than undissociated HF, and free fluoride destabilizes the oxide film at tens of ppm. Hot boil-outs are a recognized damage scenario.
Hydrogen absorption normally occurs, per ATI, when temperature exceeds 176°F (80°C) and the titanium is galvanically coupled to an active metal or under impressed current, or when pH falls below 3 or rises above 12. Galvanic coupling in a mixed-metal exchanger is a metallurgical decision, not a corrosion-allowance question.
Anhydrous chlorine is the third. ATI lists titanium as incompatible with fluorides, strong reducing acids, very strong caustics, and anhydrous chlorine. Dry chlorine can ignite titanium, and the metal depends on a minimum water content to stay passive. That moisture limit comes from chlorine industry standards rather than the metal supplier, and it must be confirmed against the producer’s specification for any chlorine duty.
Zirconium: reducing acids, and the oxidizing-ion trap
Zirconium covers almost exactly the ground titanium cannot. In pure hydrochloric acid it resists attack at all concentrations well above the boiling point, usable above 480°F (250°C) in 10% HCl and to roughly 266°F (130°C) at 37%. In seawater to 392°F (200°C) ATI reports little or no corrosion with no pitting or crevice attack, noting that zirconium “will not be affected as the titanium is by crevice corrosion” in boiling seawater.
Sulfuric acid is where the data becomes decision-grade, and where a rounded rule of thumb is dangerous. From ATI’s corrosion tables for Zr 702:
| Condition | Zr 702 rate |
|---|---|
| 0–75% H₂SO₄ at 68°F (20°C) | Under 1 mpy |
| 75% at 212°F (100°C) | Under 5 mpy |
| 76% at 212°F (100°C) | Under 10 mpy |
| 77% at 212°F (100°C) | Under 20 mpy |
| 77.5% at 212°F (100°C) | Over 50 mpy |
| 80% at 86°F (30°C) | 20–50 mpy |
| 62% boiling | Under 5 mpy |
| 75% boiling | 10–20 mpy |
Two things in that table matter more than the individual numbers. The rate rises steeply through the 75–77% band and then breaks between 77% and 77.5%, so a unit specified at 75% has roughly two percentage points of margin, not five. And temperature moves the whole curve: 75% acid runs under 5 mpy at 212°F but 10–20 mpy at boiling. Grade choice between Zr 702 and Zr 705 should be checked against the specific concentration and temperature rather than inferred from strength, because their relative performance changes across the range.
The failure mode that catches specifiers is oxidizing-ion contamination. As little as 50 ppm of ferric ion triggers pitting in 10% HCl at 212°F (100°C), and at 500 ppm ferric, stress corrosion cracking occurs at every temperature tested. Above 20% HCl the attack mode shifts to intergranular.
| Service | Titanium | Zirconium |
|---|---|---|
| Seawater, brackish water | Excellent; crevice risk above 176°F (80°C) for CP grades | Excellent to 392°F (200°C); no crevice attack |
| Hydrochloric acid | Poor — strong reducing acid | Excellent if free of oxidizing ions |
| Sulfuric acid | Poor | Good to ~75% at 212°F (100°C); breaks between 77% and 77.5% |
| Nitric acid | Good | Under 1 mpy to 98%, but SCC noted |
| Wet chlorine | Good — requires moisture | Fails, above 50 mpy |
| Anhydrous chlorine | Fails — ignition risk | Fails |
| Ferric / cupric chloride | Good | Fails, above 50 mpy |
| Hydrofluoric acid, fluorides | Fails | Fails |
Per ATI corrosion data for Zircadyne 702/705 and corrosion-resistant titanium alloys.
Wet chlorine and ferric chloride are the cleanest demonstration that these metals are not interchangeable. Titanium handles both; zirconium fails in both.
Cost, thickness, and construction
Allowable stress often beats price per pound. Ti Grade 12 carries an ASME Section VIII Division 1 allowable of 20,000 psi at 100°F, declining to 12,300 psi at 600°F. Zr 702 runs 15.7 ksi from -20 to 100°F, falling to 5.2 ksi at 700°F, while Zr 705 runs 22.9 ksi to 11.3 ksi over the same range. Those values set wall thickness, and thickness sets cost. The Materials Technology Institute published a case in which Grade 12 at $15.07/lb against Grade 2 at $13.34/lb — a 15% premium per pound — produced a vessel costing 21% less in material, because the higher allowable permitted thinner sections. Comparing alloys on price per pound inverts that result. Titanium in appropriate service is also commonly specified with no corrosion allowance, which removes sacrificial metal from the wall, though that is a design convention to be agreed with the fabricator rather than a code entitlement.
Thermal conductivity is less of a penalty than engineers expect. ATI publishes Zr 702 at 22 W/m·K and describes zirconium’s conductivity as more than 30% higher than that of stainless steel alloys. Commercially pure titanium sits in the same general band as austenitic stainless rather than well below it. The meaningful gap is against carbon steel, near 54 W/m·K, and in fouling-limited or film-limited service that difference is usually secondary to the controlling film coefficients. Where a design is genuinely conduction-limited, the tube wall calculation should be run on the actual grade’s published conductivity rather than on a family assumption.
Clad construction limits exotic alloy to the wetted surface. Clad plate in these metals is covered by ASTM B898 for titanium, zirconium, tantalum, and niobium, and by ASTM A263, A264, and A265 for stainless and nickel cladding, the latter two requiring shear and bond strength testing. One published fabricator guideline puts the crossover where a solid design would need cladder thickness above roughly 0.625 in. for reactive metals and high-nickel alloys, or about 1 in. for austenitic and duplex stainless. A second constraint belongs in the specification. In weld-repair development work on titanium-clad carbon steel, positive material identification showed contamination becoming problematic once remaining clad thickness fell below 0.048 in., and the study concluded that welding on clad thinner than 0.078 in. is highly risky. Specify clad thickness with future repair in mind, not only initial corrosion allowance.
Fabrication and inspection controls that decide service life
For reactive metals, fabrication practice is part of the metallurgy rather than an overlay on it.
Iron is an interstitial contaminant, not shop dirt. AWS G2.4/G2.4M:2021, the guide for fusion welding titanium, contains a clause titled “Oxygen, Nitrogen, Carbon, and Iron in Titanium,” grouping iron with the interstitials that embrittle the metal. ASTM B614 states that cleaned zirconium should show “no evidence of paint, oil, grease, glass, graphite, lubricant, scale, abrasive, iron, or other forms of contamination,” and ASTM B600 covers the equivalent for titanium. Both are a practice and a guide rather than mandatory specifications, so they take effect when the purchase documents invoke them. IMOA is blunt about the consequence for stainless: fabricating stainless and carbon steel in the same shop is particularly prone to iron contamination from carbon steel grinding dust. Dedicated tooling, stainless-only brushes, segregated storage, and pickling are the controls.
Shielding requirements differ categorically from stainless work. AWS G2.4 specifies shielding gas purity and dew point, and documents trailing shields, backup shields, and argon-purged chambers as standard practice. The process coverage itself is telling: G2.4 addresses GTAW, GMAW, PAW, and laser welding for titanium, while AWS G2.5/G2.5M:2025 addresses only GTAW and PAW for zirconium. On austenitic and duplex stainless, a nitrogen backing purge is accepted practice and small carbon dioxide additions stabilize the GMAW arc. Neither gas belongs near titanium or zirconium.
Weld color is a controlled acceptance criterion. Both AWS G2.4 and G2.5 include a table and figure covering weld surface color, running from bright silver through straw and blue to gray and white flaky powder as contamination increases. The accept/reject boundary is set by the governing code or purchase specification, so it belongs in the inquiry documents rather than on the shop floor. One code detail is often missed: AWS D1.9, the structural welding code for titanium, specifically excludes titanium pressure vessels and fluid-carrying pipe lines from its scope.
Qualification and post-weld treatment. Welding procedures are qualified by base metal grouping under ASME Section IX Table QW/QB-422, and titanium and zirconium each occupy their own P-number ranges separate from the stainless and nickel groups. Confirm the applicable group against the current table when writing the specification, because it determines which qualifications a shop must already hold. ATI specifies a stress relief for zirconium at 1050°F ±50°F for half an hour to an hour, and states that Zr 705 requires stress relief within 14 days after welding. Haynes recommends re-annealing C-276 once outer fiber elongation exceeds about 7%, because cold work can affect stress corrosion cracking resistance.
Inspection methods and their limits. Eddy current testing is the applicable tube examination, and ASTM E426 covers titanium, stainless, and similar low-conductivity alloys including nickel alloys — low electrical conductivity is the shared challenge the standard is written around. ASTM B338, the titanium condenser and heat exchanger tube specification, builds electromagnetic, ultrasonic, hydrostatic, and pneumatic testing in rather than leaving them to the purchaser. Where cross-stream leakage is unacceptable, helium mass spectrometer leak detection detects leakage down to 1 × 10⁻⁸ Pa·m³/s under ASTM E498, and a vented double tubesheet makes any tube-joint leak externally detectable. Penetrants used on titanium and stainless must be controlled for halogen and sulfur content.
One hazard specific to zirconium. At intermediate corrosion rates, zirconium grains can become trapped in the surface oxide, creating a pyrophoric fire hazard. ATI advises steam or hot-air treatment before the equipment is exposed to air. That is a turnaround procedure, and it rarely appears in operating manuals written around stainless equipment.
Common mistakes in exotic alloy specification
-
Ranking alloys instead of matching mechanisms. There is no “better” between titanium and zirconium; one handles oxidizing chlorides and fails in reducing acid, the other reverses it.
-
Leaving trace contaminants out of the design basis. Ferric and cupric ions disqualify both Hastelloy B-3 and zirconium; fluoride disqualifies titanium and zirconium. These species are often absent from the data sheet and present in the stream.
-
Specifying nominal concentration without upset conditions. Zirconium’s sulfuric acid resistance breaks between 77% and 77.5%, so a unit designed at 75% has roughly two percentage points of excursion margin.
-
Treating a low chloride analysis as clearance for 316L. Concentration under deposits, at wet/dry interfaces, and beneath insulation causes cracking well below the bulk threshold.
-
Comparing alloys on price per pound. Allowable stress determines thickness, and thickness determines cost.
-
Assuming a shop that welds stainless can weld titanium. Segregation, shielding, purity control, and weld color acceptance form a separate discipline with its own AWS guides.
Specify the environment, not the alloy
The most useful thing an engineer brings to an exotic alloy inquiry is not a material preference. It is a complete environmental description: full ionic composition including trace oxidizing species and fluoride, pH range across startup and upset, maximum skin temperature rather than bulk temperature, and an honest account of stagnant volumes and shutdown conditions. Every alloy here has a documented environment that destroys it, and in most failures the destroying species was present in the process and absent from the data sheet.
Harris Thermal fabricates shell-and-tube heat exchangers and ASME Section VIII pressure vessels in high-alloy construction, with in-house thermal and mechanical design and fabrication, examination, and hydrostatic testing under one roof. Where a bundle needs replacement and the original drawings are gone, positive material identification and reverse engineering establish grade and geometry before material is ordered. The TEMA classifications guide covers what each class controls in a fabricated unit, and the chemical processing equipment page covers separation and reaction service applications.
FAQs about exotic alloy heat exchangers
When should a specification move from 316L to a duplex or 6% molybdenum grade?
When chloride can concentrate at the metal surface, or when temperature approaches 122°F (50°C) in chloride service. Type 316L has an ASTM G48 critical pitting temperature near 50°F (10°C), and near boiling it can crack at under 1 ppm chloride. The 6% molybdenum superaustenitics carry PREN values of 41–45 against 24–26 for 316L.
What is the practical difference between Hastelloy C-276 and C-22?
C-276 runs 16% chromium against 16% molybdenum; C-22 runs 22% chromium against 13% molybdenum. Higher chromium improves oxidizing-acid performance, higher molybdenum favors reducing acids. Haynes publishes elevated C-22 rates in 70–90% sulfuric at and above 150°F. For C-276 in nitric acid, the published rate is 2.62 mm/y at 70% and 200°F (93°C), rising to 18.42 mm/y in 60% acid at boiling. Temperature dominates concentration in that service.
Can titanium be used in hydrochloric or sulfuric acid service?
Not as a general solution — ATI lists titanium as incompatible with strong reducing acids. Grade 12 extends usable range to about pH 2 in hydrochloric service against pH 0.1 for Grade 2, but zirconium is the alloy selected for hydrochloric duty at meaningful concentration and temperature, provided the stream carries no oxidizing ions.
Why does ferric ion contamination matter so much for zirconium?
Zirconium’s passive film is stable in reducing conditions and breaks down under oxidizing attack. As little as 50 ppm of ferric ion triggers pitting in 10% hydrochloric acid at 212°F (100°C), and at 500 ppm ferric, stress corrosion cracking occurs at every temperature tested. Upstream equipment corrosion is a common source, so the design basis should reflect achievable stream purity. The same restriction applies to Hastelloy B-3.
What is the upper temperature limit for duplex stainless in a pressure vessel?
IMOA tabulates the maximum ASME design temperature for 2205 and 2507 at 600°F (315°C), with European TüV practice at 480°F (250°C). The limit exists because alpha prime forms in the ferrite phase, which IMOA places between 575°F and 980°F (300–525°C). That is separate from sigma phase, which forms roughly between 1300°F and 1830°F (700–1000°C) and is a welding and annealing concern rather than a service one. Confirm the governing figure against ASME Section II Part D for the specific product form.
When is clad construction preferable to solid exotic alloy?
One published fabricator guideline puts the crossover where a solid design would need cladder thickness above roughly 0.625 in. for reactive metals and high-nickel alloys, or about 1 in. for austenitic and duplex stainless — a rule of thumb rather than a code rule, and worth pricing both ways on a specific job. Clad plate for titanium and zirconium falls under ASTM B898; nickel and stainless clad under ASTM A265 and A264. Specify thickness with future weld repair in mind, since development work on titanium-clad carbon steel found repair welding highly risky below about 0.078 in. remaining clad thickness.
What should a fabricator be asked before an exotic alloy exchanger is awarded?
Whether reactive-metal work runs in a segregated area with dedicated tooling; how shielding gas purity and dew point are controlled and verified; which weld color acceptance criteria apply and to what standard; how procedures are qualified under ASME Section IX for the applicable P-number group; and how post-weld thermal treatment is handled for zirconium. AWS G2.4 and G2.5 govern titanium and zirconium fabrication, and neither overlaps ordinary stainless practice.
