Table of content
- Why method selection matters for CUI
- Pulsed Eddy Current (PEC)
- Guided Wave Testing (GWT / LRUT)
- Profile (Tangential) Radiography
- Passive Infrared Thermography
- Screening vs. quantification: a two-stage framework
- Who publishes in this space
- Common mistakes
- Key takeaways
- FAQs
- References
- Related reading on OneStopNDT
Why method selection matters for CUI
Corrosion Under Insulation (CUI) is hidden by definition: it develops beneath thermal insulation, weather jackets, and fireproofing, out of sight of a routine walk-down, and is a leading cause of unplanned loss-of-containment events on refinery and offshore piping and vessels. No single NDT method both finds CUI over a wide area fast and tells an integrity engineer the precise remaining wall thickness at a flagged point — doing both at scale on an entire unit is not how any of the four methods compared here actually work.
Industry practice, as documented in independent field guidance such as the HOIS/OGTC “Guidance for CUI Inspection” (HOIS G-023, Issue 2), splits the job into two stages: screen broadly and quickly to flag likely CUI locations, then quantify at those flagged locations — typically by removing insulation and running conventional ultrasonic thickness (UT) gauging or radiography. Pulsed Eddy Current (PEC), Guided Wave Testing (GWT), and passive infrared thermography are screening tools; Profile Radiography sits in between, with a real but limited quantification role. Treating any of the three screening tools as a substitute for point quantification risks under-reporting wall loss at exactly the locations that matter most for a fitness-for-service decision.
This article compares the four methods on what they actually measure, their real coverage and speed, their documented limitations, and where each fits in a screening-then-quantification program — sourced to HOIS G-023, API RP 583’s specific thermography guidance, named vendor technical specifications, and OSNDT’s own existing CUI coverage (see Further Reading).
Pulsed Eddy Current (PEC)
Principle. PEC induces eddy currents in the pipe or vessel wall with a pulsed magnetic field and measures their decay rate, which correlates with average wall thickness. It works through insulation, coatings, weather jackets, and (per vendor material) marine growth — but only on ferromagnetic materials (carbon/low-alloy steel, cast iron). It does not work on stainless steel, aluminum, or copper alloys.
Coverage. Eddyfi’s PECA-6CH array probe covers an 18 in (457 mm) swath per dynamic scan pass and can inspect through up to 4 in (100 mm) of insulation, including thick concrete fireproofing, with no surface prep or insulation removal. Eddyfi’s general PEC technology is rated for in-service contact temperatures up to 120°C (250°F) with a probe shoe (direct single-element contact is rated lower, around 70°C) — this figure is a general Eddyfi PEC capability, not a published PECA-6CH array datasheet spec. Manual/rope-access grid scanning is far slower: in one HOIS field case on a live 16 in riser, a rope-access crew managed roughly 4 scans per 6-hour shift at a 90 mm × 40 mm grid spacing, covering about 40 m of pipe over a two-week campaign.
Accuracy and limitations. Published Eddyfi material does not state a specific remaining-wall accuracy percentage for PEC. Independent HOIS trial data found the two best-performing methods tested achieved only around 50% probability of detection for roughly 12% wall loss — a coin-flip level of reliability, not a figure confirmed specifically for PEC — and most methods needed 20% or more wall loss before detection became dependable. The defining limitation is averaging: PEC reads are averaged over the probe’s footprint, so a corrosion patch or pit smaller than that footprint is undersized or missed outright. PEC also cannot distinguish near-side from far-side wall loss, and produces edge effects near nozzles, flanges, welds, and supports. In one documented HOIS riser case, PEC reported 12.7 mm minimum remaining wall where a follow-up phased-array UT scan (after coating correction) found 10.7 mm — a real, sourced example of PEC’s undersizing risk at a specific flagged location. For this reason PEC is best treated as a screening and relative-comparison tool, not a substitute for point-quantification UT, despite some vendor marketing framing it as semi-quantitative.
Guided Wave Testing (GWT / LRUT)
Principle. A ring of transducers launches a guided ultrasonic wave along the pipe axis; reflections from changes in cross-sectional area reveal wall loss along the run. GWT reports a change in cross-section, not a point wall-thickness reading, and multiple independent sources (HOIS, Eddyfi, Nexxis) and Guided Ultrasonics Ltd’s own material agree it cannot give a minimum remaining wall thickness.
Coverage. Vendor and republished vendor material states that, under favorable conditions (bare or lightly damped pipe), one test position can screen the full circumference up to roughly 100 m in each direction; this figure is not independently confirmed and should be treated with caution. On pipe with typical insulation, coverage is considerably lower — OSNDT’s own existing coverage of the method cites approximately 50 m² per test location on 12 in pipe (roughly 1 m²/min), doubling on 24 in pipe. A real HOIS flowline case illustrates this lower, typical-insulation coverage directly: 33 test positions were needed to screen 740 m of a 16 in line — an average of roughly 22 m per test position, well below the 100 m favorable-conditions figure — with UT/RT needed as follow-up at flagged points. Guided Ultrasonics’ Wavemaker G5 datasheet confirms a scan range of up to 400 m (1,300 ft) in each direction from a single position; a separate, broader marketing claim that GWT achieves “100% volumetric coverage” appears on Guided Ultrasonics’ general product pages, not on the G5 datasheet itself, and should be attributed accordingly.
Limitations. Signal attenuates heavily through damped or heavily corroded insulation systems, bitumen, thick coatings, and at welded supports, flanges, bends, and tees, each of which can require an additional test setup or stop the scan range outright. HOIS notes GWT has real limitations detecting CUI that is circumferentially localized — it is a global, axial screening tool, weaker on small, localized patches than on broad wall loss. Detection threshold is approximately 5% cross-sectional loss in standard screening mode, improving to about 1% in continuous monitoring mode (e.g., gPIMS-type systems). In one HOIS case, GWT estimated roughly 40% wall loss on a 24 in gas line and a follow-up excavation confirmed 38% — a genuine quantitative validation, though GWT estimates still require point confirmation before being used in a fitness-for-service calculation.
Profile (Tangential) Radiography
Principle. A radiographic source on one side of the pipe and a detector or film on the other produce an image of the outer-diameter/wall profile through the insulation; wall loss is inferred from a geometric change in that profile rather than measured directly. This makes profile RT an imaging/profiling technique, not a direct thickness gauge.
Quantification role — real but limited. Sources disagree somewhat on how reliable this is. Vendor material (Eddyfi) states profile RT “cannot measure thickness” directly and that converting an OD-profile change into a true wall-loss figure is difficult and can generate false indications. The independent HOIS G-023 guidance is more nuanced: tangential radiography can accurately size the remaining ligament under some areas of CUI but may significantly undersize it under others — a real, documented quantification role, but with meaningful reliability caveats relative to PEC or UT at a given point.
Access and coverage constraints. Profile RT needs access to both sides of the pipe or vessel (source on one side, detector on the other), which limits its use on large-diameter vessels and in congested piping. Per HOIS G-023’s own pipe-diameter table, LIXI-type profile radiography systems handle pipe diameters from roughly 0.5 to 24 in; a related but distinct technique, real-time imaging radiography (e.g., handheld OpenVision-type devices), is limited to around 20 in. Very large diameters or congested geometries can still require multiple sequential shots, which is time-consuming. HOIS notes that LIXI-type profiler systems have a total steel-equivalent penetration capability of about 30 mm, inclusive of product plus insulation — a hard physical limit on what combined pipe-and-insulation thickness the technique can penetrate at all.
Radiation safety. Unlike the other three methods, profile RT carries genuine radiation-safety and access-control costs: HOIS specifically flags that handheld real-time radiography devices raise radiation-hazard and exclusion-zone concerns under UK regulations (IRR17), meaning area clearance is an operational cost to plan for, not a theoretical one. Conventional film radiography is slow; digital detector arrays improve throughput on long straight runs but still require dual-side access.
Passive Infrared Thermography
Principle. Passive thermography images surface temperature differentials caused by wet insulation, which conducts and retains heat differently than dry insulation. It is an indirect indicator of CUI risk — it locates likely wet zones where corrosion may be occurring — not a direct measurement of corrosion or wall thickness. Both HOIS and Nexxis are explicit on this point; thermography gives zero quantification.
Coverage and speed. Thermography is the fastest first-pass screening method of the four: it is non-contact, can be surveyed from a distance (HOIS cites up to about 20 m away), and can cover large areas of piping and vessels quickly.
Limitations — a real gap between the written standard and field experience. API RP 583 states a minimum temperature differential of approximately 10°C is needed for a reliable survey (confirmed via HOIS G-023, which quotes API 583 directly), but HOIS notes that vendor field experience suggests 25–30°C is more realistic in practice — a documented, citable gap between the published standard and real-world performance. API RP 583 also recommends surveying 2–3 hours after sunset, since wet insulation retains solar heat longer than dry insulation; HOIS questions whether this timing works reliably across all climates and conditions, and Nexxis independently corroborates the same post-sunset survey window. Wet insulation does not necessarily mean active corrosion is underway, so false positives occur; intermittent wetting-and-drying cycles can mislead; and shiny metallic weather-jacket cladding produces low-emissivity readings that degrade the technique.
Screening vs. quantification: a two-stage framework
Across every independent and vendor source reviewed for this article, the same two-stage pattern holds: screen broadly and quickly, then quantify at the locations screening flags. This is confirmed, documented practice — not an assumption — demonstrated in HOIS G-023’s own field case studies (screen with GWT or PEC, then follow up with UT or radiography at flagged points) and stated directly in OSNDT’s existing PEC coverage (“PEC screens broadly and ultrasonic thickness testing confirms selectively”).
Method | Primary role | Gives a defensible point wall-thickness reading? | Typical screening coverage | Key constraint |
|---|---|---|---|---|
Pulsed Eddy Current (PEC) | Screening (vendor-marketed as semi-quantitative) | No — averages over probe footprint; documented undersizing risk | ~18 in (457 mm) swath per pass (array); much slower on manual grid scans | Ferromagnetic materials only; cannot separate near/far-side loss |
Guided Wave Testing (GWT) | Screening only | No — reports cross-sectional change, not minimum remaining wall | Up to ~100 m per test position (bare pipe, vendor figure, not independently confirmed); averaged ~22 m per position in a real HOIS field case under typical insulation | Attenuates heavily at welds, flanges, bends, tees, and through damped insulation |
Profile (Tangential) Radiography | Screening, with a real but limited quantification role | Partial — can size accurately under some CUI, may significantly undersize under other CUI | ~0.5–24 in diameter range (LIXI-type systems); wider/congested geometries need multiple shots | Needs dual-side access; radiation-safety/exclusion-zone cost; ~30 mm total steel-equivalent penetration limit |
Passive Infrared Thermography | Screening only (indirect risk indicator) | No — locates likely wet zones, not corrosion itself | Fast; surveyable from ~20 m away; large areas quickly | Needs an adequate temperature differential (API RP 583: ~10°C minimum; field experience: 25–30°C more realistic); false positives from wet-but-not-corroding insulation |
Conventional UT thickness gauging (reference) | Quantification | Yes — the standard point-reading method once a location is flagged and insulation removed | Point measurement only; no screening role | Requires insulation removal at each point |
Where API RP 583 fits. API RP 583 (“Corrosion Under Insulation and Fireproofing”) is a real, directly relevant API recommended practice, and its specific thermography survey parameters (the 10°C differential, 2–3-hour post-sunset timing) are confirmed above via HOIS G-023, which quotes it directly. No evidence was found that it contains a formal side-by-side comparison table across all four methods or ties method choice directly to risk-based inspection (RBI) intervals.
Who publishes in this space
TechCorr is a working CUI inspection and turnkey-projects provider in the refinery/petrochemical sector.
Eddyfi Technologies publishes detailed, named PEC hardware specifications (the PECA-6CH probe) and a comparative “7 inspection methods for CUI” overview that is explicit about PEC’s screening-only limitations (it does not specifically discuss GWT’s limitations) — useful, clearly vendor-sourced technical material, treated as such throughout this article.
Guided Ultrasonics Ltd (GUL) publishes the Wavemaker G5 guided-wave instrument datasheet, citing scan-range and channel-count specifications; pipe-size range and attenuation figures were not published on the datasheet itself.
HOIS/OGTC (the UK offshore/onshore industry body behind HOIS G-023, “Guidance for CUI Inspection, Issue 2”) is the strongest independent, non-vendor source used in this article, with real field case studies (a 16 in riser, a 740 m flowline, a 24 in gas line) that document actual method performance, including a documented PEC undersizing case and a validated GWT estimate.
Common mistakes
- Treating a screening result as a quantification result. PEC, GWT, and thermography are all screening tools. Using a PEC reading directly in a fitness-for-service wall-loss calculation, without confirming it with point UT or radiography, risks the kind of undersizing documented in HOIS’s own riser case (12.7 mm reported vs. 10.7 mm actual).
- Surveying thermography without meeting the temperature-differential requirement. API RP 583’s stated ~10°C minimum differential is itself optimistic per field experience (HOIS: 25–30°C more realistic); surveys run on marginal days produce unreliable results and false negatives.
- Running guided wave across features it cannot see through reliably. Welds, flanges, bends, tees, and heavily damped or bitumen-coated insulation sections attenuate the signal; treating a GWT scan range as uniform across these features overstates actual coverage.
- Assuming profile radiography can quantify everywhere it images. HOIS is explicit that tangential radiography can significantly undersize wall loss under some areas of CUI even while sizing accurately under others — it is not a uniformly reliable point-quantification method.
- Skipping the radiation-safety planning cost of profile RT. Exclusion-zone and area-clearance requirements are a real operational cost (documented under UK IRR17 regulations), not a formality to plan around at the last minute.
- Applying a ferromagnetic-only method (PEC) to non-ferromagnetic piping. PEC does not work on stainless steel, aluminum, or copper alloys; method selection must start with the substrate.
Key takeaways
- No single method both screens broadly and quantifies precisely; CUI programs need a two-stage workflow: screen with PEC, GWT, or thermography, then quantify at flagged locations with UT or (with caveats) profile radiography.
- PEC works only on ferromagnetic materials, averages over its probe footprint, and has a documented real-world undersizing risk — treat it as screening/semi-quantitative, not a UT replacement.
- GWT cannot give a minimum remaining wall thickness under any sourced consensus; it is a global axial screening tool, weaker on small, circumferentially localized CUI patches, and heavily attenuated by welds, flanges, bends, and damped insulation.
- Profile radiography has a real but inconsistent quantification role — accurate under some CUI, significantly undersized under other CUI — plus genuine access and radiation-safety constraints.
- Passive thermography is the fastest screening method but is purely an indirect wet-insulation indicator; API RP 583’s stated 10°C minimum differential is more conservative than the 25–30°C field experience HOIS reports as more realistic.
FAQs
Which method should we use first on a large insulated unit? Passive thermography or guided wave testing, for speed of initial broad screening — thermography when a usable temperature differential exists (ideally 25–30°C per field experience, not just the 10°C API RP 583 minimum), guided wave on long straight pipe runs with limited welds/flanges/bends in the scan path.
Can PEC replace UT thickness gauging? Not reliably. PEC is a screening/relative-comparison tool that averages over its probe footprint; independent HOIS field data shows it can meaningfully undersize true remaining wall at a specific location. Confirm any PEC-flagged location with point UT.
Does guided wave testing give a defensible minimum remaining wall thickness? No. Every source reviewed for this article — vendor (Eddyfi, Guided Ultrasonics) and independent (HOIS) — agrees GWT reports a change in cross-sectional area, not a point thickness reading.
Is profile radiography a good quantification method for CUI? Partially. HOIS documents real cases where it sizes the remaining ligament accurately, and other cases under the same CUI where it significantly undersizes. It also requires dual-side access and carries genuine radiation-safety/exclusion-zone costs.
Does API RP 583 compare all four methods in one table? Not as far as could be verified for this article. API RP 583 is confirmed (via HOIS G-023, which quotes it directly) to specify thermography survey parameters — a minimum ~10°C temperature differential and a 2–3-hour post-sunset survey window — but no evidence was found that it contains a formal cross-method comparison table.
References
- HOIS/OGTC. “HOIS/OGTC Guidance for CUI Inspection, Issue 2” (HOIS G-023). esrtechnology.com.
- API. “Recommended Practice 583: Corrosion Under Insulation and Fireproofing.” api.org.
- Eddyfi Technologies. “Pulsed Eddy Current (PEC)” technology page and “Corrosion Under Insulation: The 7 Inspection Methods.” eddyfi.com / blog.eddyfi.com.
- Guided Ultrasonics Ltd. Wavemaker G5 product page. guided-ultrasonics.com.
- Nexxis. “Guidelines for Nondestructive CUI Detection Techniques for Process Pipelines.” nexxis.com.
- OSNDT. “Guided wave testing fits in pipeline inspection program.” onestopndt.com/ndt-articles/guided-wave-testing-fits-in-pipeline-inspection-program.
- OSNDT. “Pulsed eddy current testing.” onestopndt.com/ndt-articles/pulsed-eddy-current-testing.
- OSNDT. “Advancements in NDT inspection: corrosion under insulation,” by Ahmed Montaser. onestopndt.com/ndt-articles/advancements-ndt-inspection-corrosion-under-insulation.
- Asset Integrity Inspection: API 510, 570, 653 and RBI — how CUI screening fits into the broader integrity-inspection program
- Damage Mechanism to NDT Method Matrix (API 571) — CUI in context alongside other damage mechanisms and their detection methods
- API 510, 570 and 653 Inspection Explained — the certification scopes under which CUI inspection is typically performed
- Active Thermography for Composites: Pulsed and Lock-in — background on the thermography technique used for passive CUI screening
- Thickness Gauging in Hazardous Areas: ATEX and IECEx Explained — ATEX/IECEx considerations relevant to PEC and other CUI survey equipment
- Tube Inspection: ECT vs RFT vs NFT vs IRIS — related electromagnetic screening methods for insulated/tubular assets




There are no comments yet!