CUI and Chloride SCC — Why Stainless Fails Fastest Under Wet Insulation

The line that looked fine until the cladding came off

A 4″ 316L line, 80 °C, steam-traced, insulated for personnel protection, twelve years old, in a coastal plant. It is unpainted — it is stainless. At a shutdown a fitter cuts the cladding to fit a new support and finds a wet, grey-brown paste against the pipe and a pinhole weeping at a girth weld. UT over the area reads full wall thickness everywhere.

Guess first: with full wall thickness, what is leaking?

A crack. A branched, hair-fine, transgranular crack network that ran through the 3.0 mm (Sch 10S) wall of austenitic stainless in a couple of seasons without removing a single measurable micron of metal. Chlorides leached out of the mineral wool, concentrated against a hot wall by evaporation, and found the residual tensile stress left by the weld. That is chloride stress-corrosion cracking, and this is its favourite place on a plant.

The belief to kill:

"It's stainless — it doesn't corrode, so it doesn't need painting under the insulation."

Under wet insulation in the 50–150 °C band, austenitic stainless is the more dangerous choice, not the safer one. Carbon steel loses wall you can measure and partly budget for. 300-series stainless loses nothing measurable and then leaks through a crack, with no thickness warning, no useful UT signal, and no corrosion allowance that could ever have helped.

CUI: the mechanism, and the temperature band

Insulation does not keep water out. It is a sponge wrapped in a lid: the cladding stops rain briefly and then stops the water leaving. Water gets in at cladding laps and terminations, at penetrations and supports, from deluge tests, washdown, steam-trace leaks, and — on cold and cycling service — from condensation forming inside the insulation. Once in, it sits against the steel with a continuously replenished oxygen supply, warm, and slowly evaporating, which concentrates every dissolved species it carries.

Three things set the rate, and only the first is on the datasheet:

temperature    →  how fast the reaction runs, and whether the surface stays wet at all
wet–dry cycling →  each dry-out concentrates the salts; each re-wet restarts the cell
what is dissolved →  chlorides, sulphates, and whatever the insulation leaches

The band (carbon steel). CUI is credible from about −12 °C to 175 °C (API RP 583 / NACE SP0198), with the worst rates typically 60–175 °C and a peak around 90–120 °C. Below −12 °C the water is ice. Above about 175 °C a steadily hot surface stays dry and the risk collapses. The trap is the word steadily: a line that cycles, runs intermittently, or is only hot on campaign sweeps through the band every time, and the risk comes straight back.

The band (austenitic stainless, chloride SCC). Susceptibility switches on at roughly 50–60 °C and runs to about 150 °C. It needs chlorides, tensile stress and oxygenated water — and every one of those is present under wet insulation on a hot line.

Drive the slider yourself and watch the band appear — ▶ open the interactive: materials cui scc calc scores a real line and shows the risk peaking in the classic range and dropping away at both ends, then re-flattening the moment you tick "cyclic".

Symbol key — every symbol on this sheet

• CUI — corrosion under insulation: external corrosion of the pipe wall beneath insulation
• CLSCC — chloride stress-corrosion cracking of austenitic stainless
• SCC — stress-corrosion cracking: cracking needing material + environment + tensile stress
• T — metal (skin) temperature, not process temperature · °C
• f(T) — temperature susceptibility factor used by the scorer, 0 → 1 (dimensionless)
• Cl⁻ — chloride ion: the species that breaks the passive film on austenitics · ppm
• σ_res — residual tensile stress, mostly from welding; on its own it is enough for CLSCC · MPa
• PREN — pitting resistance equivalent = %Cr + 3.3 %Mo + 16 %N; higher = more chloride-tolerant
• TSA — thermal spray aluminium: a metallic, sacrificial coating for hot/cyclic CUI service
• CML — condition monitoring location: the fixed spot you strip and re-measure
• PEC — pulsed eddy current: screens wall loss through insulation and cladding
• RBI — risk-based inspection (API RP 580/581), the framework that sets CUI intervals
• C795 / C692 — ASTM specs for low-chloride, inhibited insulation on austenitic stainless
• API RP 583 — the corrosion-under-insulation recommended practice

Chloride SCC — why stainless is the worse choice here

Set the two materials side by side in the same wet insulation, and the difference is not one of degree:

Every row makes stainless worse under insulation, and the last row is the one that does the damage on real plants: the material least able to tolerate the environment is the one conventionally installed without the coating that would have prevented it.

The mechanics are simple and unforgiving. CLSCC needs three things simultaneously — a susceptible microstructure (austenite), a tensile stress, and chlorides in an oxygenated aqueous film above about 50 °C. Remove any one and it stops. But the tensile stress does not have to come from pressure: the residual stress left by a girth weld is at or near yield, which is why the cracks start in the heat-affected zone of welds, at cold-worked bends, and at attachment fillet welds, on lines that are barely pressurised.

Peel it apart in 3D — ▶ open the interactive: materials cui scc 3d strips the cladding and insulation off an insulated line, animates the water tracking in at a stub penetration and wicking into the shoe crevice, and switches the damage between carbon-steel wastage and the branched crack network on an austenitic weld.

The real barrier is the coating — and the details around it

Rank the layers by what actually keeps the steel dry:

1. The coating on the steel. This is the barrier. Everything else is weather protection.
2. Vapour barrier / sealed jacket — essential on cold and cycling service, where the driving force pushes moisture inwards.
3. Insulation type — whether the material holds water against the steel and what it leaches.
4. Cladding — a rain shield only. Once water is past it, the cladding keeps it in.

Coating choice follows temperature and cycling, not habit: modified epoxy / epoxy phenolic and novolac systems to roughly 200 °C; inert multipolymeric matrix or silicone for higher; and thermal spray aluminium (TSA) for hot, cyclic or hard-to-reach service, where it is the best-performing option by a wide margin — it survives from cryogenic to about 595 °C, protects carbon steel sacrificially even where damaged, and, importantly here, mitigates CLSCC on austenitic stainless by holding the surface cathodic.

Insulation choice. Cellular glass is closed-cell and simply does not absorb water; mineral wool to ASTM C795 is low-chloride and inhibited; aerogel blankets are thin and low-chloride. Standard calcium silicate is the worst common choice on a CUI-critical line — it holds water like a brick and leaches. On austenitic lines, C795 is not optional.

The details are where it leaks in. Cladding laps oriented to shed water downwards; banding and sealant at every termination; weep holes at low points so what gets in can get out; penetrations for tappings, vents, drains and supports caulked and flashed; shoes and stiffener rings detailed so the insulation is not crushed or interrupted; insulation stopped clear of flanges rather than jammed against them; removable panels at CMLs so inspection does not depend on destroying the system.

Susceptible locations — the CUI walk-down list. Supports, shoes and clamps through the insulation · nozzles, tappings, vents, drains and instrument penetrations · dead legs and low points · steam-traced lines (heat plus a leak source) · lines under damaged, dented or badly-lapped cladding · personnel-protection insulation on hot lines (installed for touch safety, usually unpainted, rarely inspected) · intermittent and cyclic service · below-ambient lines with a breached vapour barrier · anything within splash of a cooling tower or in coastal air · the 6 o'clock position everywhere.

Worked example, pitfalls, and the mitigation ladder

The line from the opening. 316L, 80 °C, standard mineral wool, no coating, 10 years old, with a support through the insulation, an instrument penetration, and steam tracing. Scored:

temperature band   f(80 °C) = 1.00  → 25.0 / 25     (CLSCC band is 50–150 °C)
coating condition  bare steel       → 30.0 / 30     ← dominant driver
insulation         20 × (0.5×0.80 water + 0.5×0.70 chloride) = 15.0 / 20
location detail    3 items × 3      →  9.0 / 15
age                10 × (10/20)     →  5.0 / 10
                                      ---------
                                       84  →  VERY HIGH

Two actions, in the order the tool ranks them: recoat with TSA (coating 30 → 1) and re-insulate to ASTM C795 (insulation 15 → 9.5) take the same line to 50 — Medium. Nothing about the process changed. The dominant driver was never the fluid; it was the absent coating.

The same tool, run the other way. Preset "200 °C hot oil, cal-sil": well above the band, sound immersion-grade coating, one penetration, 8 years → score 10, Low. Now tick cyclic / intermittent service and nothing else: the temperature factor floors at 0.85 and the score goes to 49, Medium — a five-fold jump for a line whose nameplate temperature never changed. A hot line is only safe while it stays hot.

Common pitfalls

• Assuming stainless needs no coating under insulation. It needs it more than carbon steel does.
• Specifying "mineral wool" on 300-series line without ASTM C795 — that is specifying the chloride.
• Judging risk from process temperature. CUI is set by the metal skin temperature, and by whether it cycles.
• Treating cladding as protection. It is a rain hat; once water is inside it becomes the lid.
• Insulating hot lines for personnel protection and never painting them — the single most common CUI population on any plant, and the easiest to delete with a standoff guard.
• Looking for CUI with spot UT. The damage is localised and hidden; use profile radiography, pulsed eddy current, or strip-back at chosen CMLs.
• Adding corrosion allowance "because of CUI". CLSCC is a crack; CA buys nothing.
• Re-using old cladding and old sealant after a strip-back, then recording the line as inspected.
• Leaving hydrotest or deluge water in insulation over a shutdown.

The mitigation ladder (top of the list first, because it removes the mechanism rather than slowing it):

1. Do not insulate it. If the insulation exists only for personnel protection, a standoff guard or cage does the same job with no annulus to hold water.
2. Coat the steel properly for the actual temperature and cycling — immersion-grade system, or TSA for hot/cyclic and for austenitics where CLSCC governs.
3. Choose insulation that does not hold water or leach chloride — cellular glass, or mineral wool to ASTM C795 — and maintain a continuous vapour barrier on cold service.
4. Detail the water out: laps, terminations, sealant, weep holes, protected shoes.
5. Upgrade the material where CLSCC risk cannot be engineered away — duplex 2205 is far more resistant (not immune), and higher-nickel alloys such as 825 or 6Mo more so again.
6. Inspect for what you cannot see, on a risk-based interval (API RP 583 / 581), targeting the walk-down list above rather than the accessible straight runs.

Outcome

• CUI is a wet-insulation problem, not an insulation-material problem: the cladding keeps water in, and the annulus concentrates whatever the water carries.
• Carbon-steel CUI band ≈ −12 to 175 °C, worst 60–175 °C, peak near 90–120 °C. Cycling or intermittent service defeats the upper limit — the metal keeps re-entering the band.
• Chloride SCC of austenitic 300-series needs chlorides + tensile stress (weld residual stress is enough) + water above roughly 50–60 °C → branched transgranular cracking, no wall loss, through-wall in weeks. Stainless is the more vulnerable material here, not the safer.
• The coating is the barrier; cladding is a rain shield and insulation is a sponge. Choose the coating for temperature and cycling, and use ASTM C795 low-chloride insulation on austenitics.
• Most CUI lives at supports, penetrations, dead legs, low points, steam-traced lines, damaged cladding and personnel-protection insulation. Inspect those, not the easy straight runs.
• Interactive: ▶ open the interactive: materials cui scc calc · 3D: ▶ open the interactive: materials cui scc 3d.

Open items

• Add a cold-service branch to the scorer (vapour-barrier integrity, condensation cycles, ice-plug effects) — the current curve is tuned to hot and cycling service.
• Cross-link to a CUI inspection-technique comparison: profile RT vs pulsed eddy current vs guided wave vs IR moisture scanning, with detection limits and cost per metre.
• Worked TSA versus epoxy-phenolic life-cycle cost for a 5 km insulated network, including the strip-and-recoat campaigns each system implies.

Educational teaching material — simplified and illustrative. Not for engineering design use.
© EPiC Circle · provided by PipeAxis