CUI Monitoring: Moisture Sensors, Thermography and Inspection Ports

Short answer: CUI monitoring means watching for the conditions that cause corrosion under insulation, mainly water in the insulation, rather than measuring metal loss directly. The main tools are permanently installed moisture sensors, periodic infrared thermography or neutron backscatter surveys to find wet insulation, and sealed inspection ports that allow quick checks at fixed points. A monitoring alarm does not tell you how much wall has been lost; it tells you where to send NDE such as profile radiography, guided wave screening or UT after insulation removal.

That distinction, between monitoring that finds the cause and inspection that measures the damage, is the key to using CUI monitoring well. This guide explains how each monitoring approach works and where it falls short, how codes such as API 510 and API RP 583 frame CUI, how inspection ports should be designed and maintained so they do not become a leak path themselves, and how to build a trigger-to-NDE workflow so that a wet-insulation signal turns into a wall-thickness answer quickly.

Monitoring versus inspection: why the difference matters

Corrosion under insulation needs three things: a susceptible metal, a temperature in the corrosive range, and water held against the surface. Insulation hides all three from view. Traditional CUI programmes deal with that by inspection: remove insulation at selected windows, look and measure, or use NDE that can see through insulation. Inspection answers the question "how much metal is left?" but only at the places and times you look.

Monitoring answers a different question: "is water getting in, and where?" Water ingress usually comes before corrosion, sometimes years before significant wall loss. If you can detect wet insulation early and fix the cladding or sealing, you may stop CUI before it starts, and you know which locations deserve inspection first. The trade-off is that no moisture sensor measures metal loss. A dry reading today does not prove there is no historic damage, and a wet reading does not prove there is corrosion.

Used together, monitoring improves where and when you inspect; it does not replace inspection under the owner's code-based programme. API 510 (2022 edition) notes in its discussion of intervals that, in some circumstances, permanently mounted sensors that can be continuously monitored may be more appropriate than frequent manual inspections. That is support for sensors as part of the programme, not a substitute for the code's inspection requirements.

Where CUI happens: the temperature window and the usual suspects

Monitoring is only worth installing where CUI is credible. API 510 (11th edition, 2022) says CUI inspection shall be considered for externally insulated vessels, and those in intermittent service, operating within these ranges:

The code also warns that nozzles and supports can move in and out of the CUI range even when the vessel body operates outside it, and that this should be considered in susceptibility and documented in the inspection plan. Intermittent service matters too: equipment that cycles through the range, or sits cold and wet during shutdowns, can be worse than equipment that runs steadily in it.

API 510 lists the most susceptible areas on vessels as above insulation or stiffening rings, nozzles and manways, other penetrations such as ladder clips and supports, damaged insulation where water can enter, failed caulking, top and bottom heads, and other water traps, while cautioning that CUI locations can be very unpredictable. Piping has its equivalents: low points, dead legs, supports, branch connections, valve and flange insulation boxes, and vertical runs where water collects at the bottom of each section. Those are the locations where monitoring and inspection ports earn their keep. API RP 583, the dedicated recommended practice for corrosion under insulation and fireproofing, covers susceptibility, NDE options and risk assessment in more depth; see our summary of API RP 583.

Permanently installed moisture sensors

The fastest-moving part of CUI monitoring is installed sensors that report moisture continuously, often wirelessly. A useful public benchmark is the Belgian InnovateCUI project run by Sirris, which tested five sensor principles side by side for 18 months on a 50-metre replica of 3-inch piping in the port of Antwerp, insulated mainly with mineral wool and aluminium cladding, with controlled water injection points. The principles tested were:

• Air humidity spot sensors placed within the insulation, measuring relative humidity and temperature at a point.
• Electromagnetic guided wave sensors running in the annulus between pipe and cladding.
• Capacitance or resistance sensors measuring electrical properties across the insulation, which change when it is wet.
• Optical fibres on the outside of the cladding measuring temperature along the line, where wet insulation shows as a thermal anomaly.
• Sensor wires on the pipe surface using electromagnetic guided radar to locate water along their length.

All five detected water infiltration. The project's most useful practical finding was that they did not always agree: a given day could be classified as wet by one sensor and dry by another, even though each was working correctly within its own measurement principle. The lesson for owners is to understand what each sensor actually measures (humidity at a point, an electrical property across a volume, temperature on the cladding, or water along a line) before deciding how to set alarm thresholds and what an alarm means. A second phase of the project began in 2026 to test under more realistic industrial conditions.

Two practical questions decide whether sensors work in a plant. First, installation: most sensors that sit inside the insulation are fitted when insulation is replaced, so they suit new builds, turnarounds and re-insulation projects better than retrofits. Second, every penetration through the cladding for a cable or transmitter is a potential water path. Sealing details matter as much as the sensor itself.

Infrared thermography surveys for wet insulation

Infrared thermography is the most common periodic survey for wet insulation. Wet insulation conducts heat much better than dry insulation, so on a hot line the cladding over a wet patch is usually warmer than the surrounding cladding, and on a cold line it is usually colder. A thermographer scanning from the ground, a vehicle or a drone can survey long runs quickly and mark anomalies for follow-up. API RP 583 includes thermography among the screening methods for wet insulation.

Thermography has well-known limits that should be built into the survey plan:

• It needs a temperature difference. Lines close to ambient temperature give little contrast; surveys are usually planned for when the difference between process and ambient is greatest.
• Weather and sun interfere. Solar loading, wind and recent rain change cladding temperatures; many surveys are done at night or under overcast conditions, with the conditions recorded.
• Cladding surface affects readings. Shiny aluminium or stainless cladding has low emissivity and reflects its surroundings, which can mask or mimic anomalies. Painted or weathered cladding usually images better.
• It finds water, not corrosion. An anomaly is a reason to look, not evidence of wall loss. Equally, a location that was wet for years and has since dried may show nothing even though damage is present.

The output that matters is a marked-up list of anomaly locations, tied to line numbers and isometrics, with the survey conditions recorded so the next survey is comparable. Our article on infrared thermography in NDT covers the technique more broadly.

Neutron backscatter and other wet-insulation screening

Neutron backscatter detects hydrogen, which in insulation usually means water. A source emits fast neutrons; those that collide with hydrogen nuclei slow down and are counted by a detector. Higher counts indicate more hydrogen and therefore probably more moisture. API RP 583 includes neutron backscatter among the techniques it describes for finding wet insulation. It works regardless of process temperature, which makes it a useful complement to thermography on lines close to ambient, but it is a point-by-point technique, slower than an IR scan, and like any radioactive-source method it requires appropriate licensing and handling. Some insulation types and hydrogen-bearing coatings or process fluids can raise background readings.

Visual surveys remain the cheapest monitoring of all: stains, rust weeping from cladding seams, damaged jackets, failed caulking, missing end caps and vegetation growth at the base of insulated lines. API 510 expects external inspections to note areas where insulation coverings and penetrations may be letting moisture in on vessels susceptible to CUI. A structured walkdown checklist, repeated on a fixed cycle and linked to the same line and CML numbering as the inspection programme, catches a large share of ingress points before any instrument does.

Inspection ports and plugs: quick access without becoming a leak path

Inspection ports (also called inspection plugs, CML plugs or insulation windows) are removable openings in the insulation and jacketing at fixed points. They let inspectors take UT readings, look at the surface, or check for water without stripping insulation. They are popular because they make repeat thickness readings cheap and repeatable at the same spot.

They also introduce a weakness: every port is a break in the weatherproofing. API 510 (2022) states that on insulated vessels, CML sealing systems such as CML plugs should be kept in place so that insulation covers stay sealed and moisture does not enter and cause CUI. Older editions of API 570 made the same point for piping: locations where insulation plugs have been removed for thickness readings need particular attention, plugs should be promptly replaced and sealed, and commercially available removable plugs can be used to mark inspection points for future reference. Check the current API 570 edition for its present wording.

Good practice for ports:

• Use purpose-made plugs or covers with a positive seal, installed so water sheds away from the opening (avoid the top of horizontal lines where possible, unless the CML must be there).
• Record each port against its CML number so the reading history stays tied to a physical location.
• Make re-sealing part of the inspection task, with a check that every plug opened has been closed and sealed before the crew leaves.
• Look inside the port each time: wet insulation, rust staining or scale at the opening is itself a monitoring result.
• Remember ports only show what is under them. CUI is often concentrated at water traps that are not at a port; ports complement, not replace, screening and targeted insulation removal.

From alarm to answer: the trigger-to-NDE workflow

Monitoring only pays off if a signal leads quickly to a decision. A simple workflow that many owners adopt:

1. Signal. A sensor alarm, IR anomaly, neutron backscatter high count or walkdown finding is logged against the line or equipment number and location.
2. Triage. The inspector considers material, operating temperature against the CUI range, consequence of a leak, insulation type, coating age and CUI history to set priority.
3. Screen through insulation where possible. Profile radiography can show wall profile and wet insulation through the jacket on smaller-diameter piping; guided wave testing can screen longer lengths of piping from a few insulation windows; pulsed eddy current is another option some owners use for average wall thickness through insulation.
4. Prove up. Remove insulation at the indicated area, inspect visually, and measure with UT thickness, UT scanning or corrosion mapping. For austenitic and duplex stainless steels, the concern is chloride cracking, so surface examination (for example, penetrant testing) replaces thickness measurement as the key check.
5. Evaluate and record. Measured thickness goes into the owner's evaluation: corrosion rate, remaining life and, where needed, an engineering assessment commissioned by the owner. The ingress point is fixed and the monitoring record updated.

For more on through-insulation methods, see CUI detection without removing the insulation and the CUI detection guide.

Designing a CUI monitoring programme

A workable programme usually combines several layers rather than relying on one technology:

• Everywhere in the CUI range: structured visual walkdowns on a fixed cycle, plus inspection ports at CMLs, maintained and resealed.
• Long runs and large areas: periodic IR thermography (and neutron backscatter where temperature contrast is poor) to find wet insulation.
• High-consequence or historically wet locations: permanently installed moisture sensors at known water traps, fitted during re-insulation.
• Every signal: a defined route to NDE prove-up and a feedback loop that updates the monitoring plan when damage is or is not found.

Before buying sensors, decide what an alarm will trigger, who owns the data, how sensor locations map to the CML and isometric numbering, and how false alarms will be handled. Mitigation also belongs in the plan: thermal spray aluminium and good coatings under insulation reduce the consequence of ingress (see TSA for CUI mitigation), and insulation type matters because some materials absorb and hold more water than others.

Regulatory overlay in the US and Canada

For US facilities covered by OSHA's Process Safety Management standard (29 CFR 1910.119), paragraph (j) requires inspection and testing of process equipment following recognized and generally accepted good engineering practices, at frequencies consistent with those practices and manufacturers' recommendations, with documentation. CUI programmes are typically built on API 510, API 570 and API RP 583 as those practices. Monitoring data can support and target the programme, but the inspection and documentation obligations remain. EPA's Risk Management Program (40 CFR 68) has parallel mechanical integrity requirements for Program 3 processes. In Canada, provincial pressure equipment regulators such as ABSA in Alberta and TSSA in Ontario oversee owners' integrity management programmes; confirm how monitoring is recognised in your approved programme with the jurisdiction.

Common mistakes with CUI monitoring

• Treating "no moisture alarm" as "no CUI" on equipment that has been insulated and wet for years.
• Installing sensors without deciding what an alarm triggers, so alarms are acknowledged and ignored.
• Comparing sensor types as if they measured the same thing.
• Running IR surveys in poor conditions or on low-emissivity cladding without recording conditions.
• Leaving inspection plugs open or poorly sealed, creating the ingress the programme is meant to prevent.
• Forgetting nozzles and supports that pass through the CUI range when the main body does not.
• Measuring wall thickness on stainless steel and missing chloride cracking.

How Atlantis supports this

Atlantis NDT does not supply CUI sensors or set owners' CUI risk rankings. We perform the NDE that turns monitoring signals into wall-thickness answers: profile radiography through insulation, guided wave screening of piping runs, UT thickness surveys at CMLs and inspection ports, corrosion mapping after insulation removal, and visual and penetrant examination, by ASNT-certified technicians under ASNT Level III oversight, reporting to your API-certified inspector. See our CUI inspection services. Quotes within 24 hours: request CUI follow-up NDE.

Frequently asked questions

What is CUI monitoring?

Continuous or periodic checks for the conditions that cause corrosion under insulation, chiefly water in the insulation, using installed moisture sensors, IR thermography, neutron backscatter, walkdowns and inspection ports. It targets inspection; it does not measure wall loss.

How do you monitor corrosion under insulation?

Combine visual walkdowns and sealed inspection ports at CMLs, periodic IR or neutron backscatter surveys for wet insulation, and installed moisture sensors at high-risk water traps, then send every signal to NDE prove-up.

Do corrosion under insulation monitoring sensors detect corrosion?

Generally no. Most detect moisture, humidity or temperature anomalies. Wall loss must be confirmed with NDE such as profile radiography, guided wave screening, or UT after insulation removal.

Which CUI moisture sensor is best?

It depends on the application. In the InnovateCUI field trial all five sensor principles detected water, but they sometimes disagreed on whether a given day was wet. Choose based on what each measures, coverage, insulation type and installation opportunity.

Can infrared thermography find CUI?

It finds wet insulation, which is the precursor to CUI, by showing temperature anomalies on the cladding. It needs a temperature difference between process and ambient, and it does not measure corrosion.

What temperature range is CUI a concern?

API 510 (2022) says CUI inspection should be considered for carbon and low-alloy steel between 10°F and 350°F, austenitic stainless between 140°F and 350°F, and duplex between 280°F and 350°F, including intermittent service.

Do insulation inspection plugs cause CUI?

They can if left open or poorly sealed. API 510 says CML sealing systems such as plugs should be kept in place to keep moisture out. Reseal every plug after each reading.

Can monitoring replace CUI inspection?

No. Monitoring helps target and time inspection, and API 510 recognises continuously monitored sensors as useful in some circumstances, but code-based inspection and documentation still apply.

What NDE follows a CUI moisture alarm?

Typically profile radiography or guided wave screening through insulation, then UT thickness or corrosion mapping after removing insulation at the indicated area; surface examination for stainless steel cracking.

Is neutron backscatter safe to use for CUI screening?

It uses a radioactive neutron source, so it must be operated by licensed, trained personnel under the applicable radiation regulations. Used that way, it is an established wet-insulation screening method.

Have a list of IR anomalies or sensor alarms waiting for follow-up? Send it to us and we will scope the prove-up NDE, or ask about a UT survey at your inspection ports.

Sources: API 510, 11th edition (2022); API RP 583 (current edition); API 570 (earlier edition, for historical wording on insulation plugs); Sirris InnovateCUI project; OSHA 29 CFR 1910.119. Related: CUI glossary, corrosion rate calculation.