The short version.

• Corrosion under insulation (CUI) happens when water gets trapped between a pipe's steel wall and the weatherproof insulation wrapped around it — invisible from outside until the insulation is stripped or the pipe fails. Industry-wide, it is a leading cause of hydrocarbon loss-of-containment: one widely cited estimate puts CUI behind roughly 50% of onshore hydrocarbon leaks over the past two decades, and DNV has said it contributed to more than 20% of major oil-and-gas accidents recorded in the EU over 35 years.
• Floating LNG and oil-production units — FSRUs (Floating Storage Regasification Units) and FPSOs (Floating Production Storage and Offloading vessels) — combine three things that each accelerate CUI on their own: constant salt spray and humidity, a hull that never stops flexing (which cracks insulation jacketing over time), and, on FSRUs specifically, cryogenic LNG lines that swing between roughly -162°C and ambient temperature during shutdowns, "breathing" in humid air as they warm.
• Carbon steel and stainless steel fail differently under insulation. Carbon steel loses wall thickness in a way ultrasonic testing can catch before failure, with the highest documented incubation risk between 77-110°C (per API RP 583). Austenitic stainless steel, standard on LNG-service piping, is instead vulnerable to chloride stress corrosion cracking — fine cracking with little visible wall loss, active mainly above roughly 60°C, that can lead to sudden failure without the warning signs a thickness gauge would show.
• A real, named case shows what happens when this is missed: ExxonMobil's Fife Ethylene Plant in Scotland had five separate CUI-caused leaks of flammable gas between February 2018 and September 2019 — one lasting four months — before regulators fined the company and found its insulated-pipework inspection regime "woefully inadequate."
• DNV's DNV-RP-G109, published December 2019, is the industry's standard risk-based framework for managing CUI, scoring four "barriers" — material, coating, water-wetting and design — rather than inspecting insulated pipework blind. No public, named case of a catastrophic CUI failure specifically on an operating FSRU or FPSO was found while researching this piece; the severity claims below are extrapolated from onshore incidents and general offshore-corrosion literature, not a documented floating-asset failure, and that gap is worth being clear about rather than papering over.
• A coatings market has grown around the "coating" barrier in that framework; one Indian example is HPCL's HP TIC/HP CIC insulating coatings and HP NanOKoat, a single-layer nano-material anti-corrosion coating the company introduced in November 2024 for highly saline environments — named here as an industry example, not a verified performance claim.

What CUI actually is, and why it hides

Insulation on process piping does two jobs: it keeps heat in (or cryogenic cold in) for efficiency, and it protects workers from touching a hot or freezing surface. Wrapped around that insulation is a thin metal weather jacket meant to keep rain and spray out. CUI begins when that jacket is breached — through a seam, a fastener hole, a mechanical seal that has degraded, or a crack from vibration — and water gets in without a way to drain or evaporate back out. Insulation is good at holding onto that water rather than shedding it, so it sits directly against the steel pipe wall, out of sight, for as long as the breach goes unnoticed. Corrosion under that trapped water can run at rates commonly cited as up to 20 times faster than the same steel exposed to open, naturally-drying atmospheric conditions, because the insulation blocks the wetting-and-drying cycle that would otherwise slow corrosion down.

The trapped-water mechanism and jacket-breach pathway are standard descriptions of CUI found consistently across engineering and industry sources (Aspen Aerogels, Voliro, and the peer-reviewed review by Alaslani & co-authors in Metals, 2022). The "up to 20 times faster than atmospheric corrosion" figure is likewise a widely repeated industry figure rather than one tied to a single primary measurement; it is presented here as the general order of magnitude reported for CUI, not a precise multiplier for any specific alloy, coating or climate.

Why a floating LNG or oil unit makes this worse

Everything that makes CUI a known problem onshore is present on an FSRU or FPSO, plus three things that are specific to operating permanently at sea. First, the atmosphere itself is more aggressive: constant sea spray and high ambient humidity keep condensation forming on piping exteriors continuously, rather than the intermittent wetting an onshore plant sees with rain. Second, that spray carries airborne chlorides — dissolved salts — which, once past a breached jacket, concentrate on the metal surface and drive the two fastest, most localized failure modes: pitting and cracking, rather than the slower, more even wall-thinning chloride-free water would cause. Third, a floating hull never stops moving. The continuous rolling, pitching and flexing of an FSRU or FPSO in a seaway puts a cyclic mechanical load on rigid insulation jacketing and weather seals that a static onshore refinery structure never experiences, gradually opening the micro-gaps and cracks that let water in.

The marine-environment mechanisms (humidity/condensation, chloride concentration, vessel-motion-driven jacket wear) are described consistently across offshore-corrosion and insulation-industry sources, including EM&I's writing on FLNG/FSRU/FPSO CUI detection and Aspen Aerogels' floating-asset insulation material. These are treated here as an accurate description of the mechanism, not independently re-derived from first principles for this piece.

The cryogenic "breathing" effect, and the FPSO's two-front problem

FSRUs add a mechanism that has no real onshore equivalent at the same scale: LNG regasification piping runs at roughly -162°C, and while a line that stays permanently that cold has slowed corrosion kinetics (very little liquid water can exist at that temperature), FSRUs don't run continuously at a fixed cryogenic temperature. Shutdowns, maintenance windows and cyclic operation let those lines warm intermittently toward ambient. As the temperature swings, the insulation effectively "breathes" — drawing in humid, salt-laden air as the line warms, which then condenses into liquid water inside the insulation as the temperature falls again on the next cooling cycle. Each cycle is a fresh opportunity to introduce moisture that a permanently cold or permanently hot line would not get.

FPSOs face a related but distinct problem: they process both very hot streams (crude oil, sometimes cited as accelerating corrosion rates by a similar order of magnitude to the general CUI figure above when combined with insulation) and cooled or cryogenic gas streams, often on the same topside structure. That means a single FPSO's asset-integrity team is managing two materially different CUI temperature profiles — and two different failure mechanisms — across pipework that may be only metres apart, rather than the one dominant profile a single-product onshore plant would have.

The cryogenic-cycling "breathing" mechanism and the FPSO dual-temperature-regime framing are described in industry engineering writing on CUI (StepIn Engineering; general offshore asset-integrity literature); the specific claim that hot-crude processing accelerates corrosion "up to 20 times" is treated here as an application of the general CUI-acceleration figure discussed in Section 1 rather than a distinct, separately-verified statistic for crude oil specifically — no source found for this piece isolated a hot-crude-specific multiplier from the general CUI figure.

Two metals, two very different failure modes

The choice between carbon steel and stainless steel changes not just how fast CUI progresses, but how detectable it is before something fails — which matters enormously on a vessel where stripping insulation for inspection is itself a difficult, deck-space-constrained job.

The carbon-steel 77-110°C incubation-risk figure and the broader 50-175°C NACE SP0198/API RP 583 range are corroborated across multiple independent engineering sources describing those two standards. For stainless steel, the pasted source material this piece was checked against cited a 60-205°C vulnerability range; the primary abstract for the closest matching published study (Fritz & Gerlock, Desalination 135, 2001, fetched directly from ScienceDirect) instead concerns a specialty 6% molybdenum super-austenitic alloy (UNS N08367/AL-6XN, explicitly not standard 304/316-type stainless), which cracked readily at 260°C in 0.02-2.0 wt% chloride solutions and was deemed unlikely to crack "in the range of atmospheric boiling (~100°C)." That confirms the distinction this piece draws: standard austenitic LNG-piping stainless is far more susceptible than this premium alloy, so the lower, more consistently-cited ~60-80°C active-onset range for ordinary stainless is reported here, and the wider 205°C figure is now independently confirmed to describe a different, specialty alloy rather than standard LNG-service stainless.

What it costs when it's missed

CUI's danger is less about any single failure and more about how much damage accumulates before anyone sees it, since the whole point of the mechanism is that it is invisible from outside. The clearest documented example is onshore, not offshore, but shows exactly this pattern: ExxonMobil's Fife Ethylene Plant in Scotland recorded five separate leaks of "extremely flammable" hydrocarbons between February 2018 and September 2019, every one of them attributed to corrosion under insulation, with one leak going undetected and continuing for four months. The UK's Health and Safety Executive concluded the company's arrangements for inspecting insulated pipework were "woefully inadequate," a finding attributed to HSE inspector Lindsey Stein. ExxonMobil pled guilty to breaching the Provision and Use of Work Equipment Regulations 1998 and the Health and Safety at Work etc Act 1974, and was fined £267,000 at Kirkcaldy Sheriff Court — a verdict handed down on 26 August 2026, days before this piece was first published, and coming after the Fife plant itself had already closed in January 2026 after 40 years of operation.

At an industry level, the scale is large enough to be one of the standard justifications cited for risk-based inspection programmes: one commonly repeated figure attributes roughly 50% of onshore hydrocarbon leaks over the past two decades to CUI, and DNV's own published materials, introducing its CUI-focused recommended practice, state that CUI has contributed to more than 20% of major oil-and-gas accidents recorded across the EU over the preceding 35 years. Separately, CUI is often cited as responsible for 40-60% of piping maintenance costs in the oil and gas sector, against a backdrop where corrosion generally costs the oil, gas and petrochemical industry on the order of $1 trillion a year worldwide.

The Fife Ethylene Plant incident (five leaks, Feb 2018-Sept 2019, HSE finding and fine) is drawn from UK press reporting on the HSE investigation and is the one incident in this piece with a specific, named, dated public record behind it. The "~50% of onshore hydrocarbon leaks" and "40-60% of piping maintenance cost" figures, and DNV's own "20%+ of major EU accidents over 35 years" statement, are widely repeated across industry sources (including DNV's own DNV-RP-G109 announcement materials) but are aggregate, mostly onshore-weighted industry figures — no source found for this piece broke these percentages out specifically for floating LNG or oil-production assets, and none should be read as an FSRU- or FPSO-specific statistic.

Why "just fix it" is harder at sea

An onshore refinery that finds CUI on one processing train can usually isolate that train and keep the rest of the plant running while insulation is stripped, the pipe is tested, and it's re-jacketed. An FPSO or FSRU rarely has that luxury: topside space is far more constrained, a major leak or a precautionary shutdown for inspection often affects the whole vessel rather than one isolable unit, and specialist inspection and maintenance crews have limited deck space and berth space to work from once they're mobilized offshore at all. For an FSRU specifically, insulation that has taken on salt water also stops doing its main job efficiently, which shows up as increased LNG boil-off and lower regasification efficiency well before any leak actually occurs — a hidden running cost on top of the structural risk. When a shutdown does happen, the economics are correspondingly worse than onshore: a full-vessel shutdown on a producing FPSO or an operating FSRU is commonly described as costing millions of dollars a day in lost production, a figure this piece has not independently verified against a specific vessel's numbers but which is consistent with the general economics of single-point offshore production assets with no redundant train to fall back on.

How the industry is meant to manage it

The standard industry answer, since December 2019, is DNV's DNV-RP-G109, "Risk based management of corrosion under insulation" — developed with oil-and-gas industry partners to replace blind, calendar-based insulation-stripping inspection with a risk score built from four factors: the material underneath, the coating protecting it, how likely water is to actually reach the steel (the "water-wetting" barrier), and design features that either help or hinder drainage and inspection access. The intent is to focus expensive, difficult offshore inspection effort on the pipe sections genuinely most at risk — hot or cryogenic-cycling lines, stainless sections in the CSCC-prone temperature band, areas with known jacketing damage — rather than spreading a fixed inspection budget evenly across a vessel where most insulated pipework is not actually at meaningful risk.

DNV's own 2020 announcement of the recommended practice, fetched directly, confirms this in DNV's own words: “We published this methodology as Recommended Practice RP-G109 in December 2019,” assessing “four (4) CUI barriers: material, coating, water wetting and design.” The practice has since been revised; DNV's current listing shows a June 2025 edition, so the framework in active use today is a later revision of the one first published in December 2019.

The "coating" barrier in that framework is itself a product market, and one Indian example illustrates the kind of response it has produced. Hindustan Petroleum Corporation Limited (HPCL) markets HP TIC and HP CIC — thermal- and cold-insulating coatings pitched specifically at managing temperature extremes and preventing corrosion under insulation — alongside HP NanOKoat, a single-layer nano-material coating HPCL introduced in November 2024 aimed at structures in highly saline environments, and a general-purpose HP Primer base coat. These are named here as an example of a coatings-market response to the problem this piece describes, not a recommendation or an endorsement of any specific product's performance.

HPCL's own site, fetched directly, confirms HP-CIC (a cold insulating coating usable to -20°C, thermal conductivity 0.07 W/m·K, rated to a 1,000-hour ASTM B117 salt-fog test, pitched as a full PIR/PUF replacement) and lists HP NanOKoat alongside HP Primer under its Protective Coatings product line. HP TIC specifically and the "up to ten years" HP NanOKoat durability figure sometimes cited elsewhere were not locatable on HPCL's own site even on a direct fetch, so those two specifics remain reported as claims this piece could not independently verify rather than confirmed facts. HPCL also markets outboard-engine marine oils and rust-preventive products (HP-MARIGEN, HP Milcy Marina, HP RUSTOP); those protect a different asset class (engines, stored parts) from a different corrosion mechanism than the insulated-piping CUI this piece covers, so they are omitted here as off-topic rather than folded in.

What doesn't follow from any of this

The case for taking CUI seriously on floating LNG and oil assets is real, but it rests on combining a well-documented general mechanism (CUI itself, onshore incident data, published temperature standards) with the logical expectation that a marine, cyclically-loaded, thermally-cycled hull makes it worse — not on a specific, publicly documented FSRU or FPSO catastrophic failure. No such named case was found while researching this piece, and that absence is reported as such rather than as evidence the risk is overstated; floating-asset incident data of this kind is often held privately by operators and insurers rather than published. Similarly, the 60-205°C stainless-steel vulnerability range volunteered in the source material behind this piece does not match what the corrosion-science literature checked here actually reports, and this piece reports the narrower, better-supported ~60-80°C active-onset range instead rather than repeating the wider figure uncritically. Readers evaluating a specific vessel should treat every general industry percentage in this piece (the 50% and 20% hydrocarbon-leak figures, the 40-60% maintenance-cost figure) as industry-wide and mostly onshore-weighted context, not a substitute for that vessel's own risk-based inspection data under a framework like DNV-RP-G109.

Sources and caveats