IRIS vs eddy current vs APR for tube inspection
Three methods, three different jobs. IRIS maps wall thickness quantitatively but needs a near-bare, flooded bore and time per tube. Eddy current finds cracking and pitting fast in non-ferromagnetic tubes but reads through a probe that must traverse a clean bore. Acoustic pulse reflectometry (APR) trades quantitative depth for coverage: it screens every tube from the open end in seconds per tube, tolerates near-service surfaces, and locates and sizes restrictions and wall loss — which is why current practice screens 100% by APR first and sends the probes only where the screen flags. None of the three replaces the others, and a program built on the wrong one pays either in blind spots or in window time. This page lays out what each method sees, what each needs, and published results cited by document. If a sampled IRIS program already serves your bundle, keep it; when coverage, cleanliness, or the window is the constraint, this is how the methods actually divide the work.
How do the three methods compare?
| Dimension | APR (acoustic pulse reflectometry) | IRIS (ultrasonic) | Eddy current | Source |
|---|---|---|---|---|
| What it measures | Changes in internal cross-section: blockages and deposits (position and size), wall loss, holes | Quantitative wall-thickness map around and along the tube | Cracking, pitting, wall thinning; characterizes defect type | Talcyon published material; standard NDE method descriptions |
| How it reads the tube | Acoustic pulse in the air column, launched from the open tube end; no probe enters | Rotating ultrasonic transducer traverses the bore in a flooded tube | Electromagnetic probe traverses the bore | Talcyon published material; standard NDE method descriptions |
| Cleanliness needed | Near-service surfaces tolerated; the deposit is part of what it detects | Near-bare bore; deposits corrupt the reading | Clean bore for probe passage and signal quality | Talcyon geothermal case study (higher-end cleaning named as the ultrasonic constraint); standard practice |
| Blocked tube | Reads it from the end — locates and sizes the blockage | Probe cannot enter | Probe cannot enter | Talcyon cleaning-validation case studies |
| Tube materials | Material-independent (acoustic) — published cases on carbon steel and titanium | Metals generally, including carbon steel | Non-ferromagnetic alloys; ferromagnetic tubes need adapted techniques (e.g. remote field) | Talcyon cleaning-validation cases; standard NDE method descriptions |
| Geometry | Published application notes on finned, U-tube, and twisted tubes | Straight sections; bends limit the rotating head | Handles bends with flexible probes, within limits | Talcyon resources library; standard practice |
| Published speed results | ~10 s/tube stated; 146 tubes under 1 h; 512 in 2 h; 7,541 in 24 h — varies materially by job | Minutes per tube; the reason IRIS programs sample | Faster than IRIS, still probe-limited per tube | Talcyon brochure and case studies; standard practice |
| Role in a program | 100% screening; cleaning validation; triage for the probes | Quantitative verdicts on flagged tubes; plug/retube decisions | Crack detection on flagged or sampled tubes in condenser alloys | Method roles as described across the cited documents |
APR figures are Talcyon's published statements and single-job case results, cited under Sources; field throughput varies materially with tube geometry, access, and site movement. IRIS and eddy current characteristics are stated at the level of standard NDE method descriptions; a specific campaign's capability is defined by the NDE vendor's written procedure, not by this table.
When is IRIS the right method?
When the decision needs a number. IRIS produces a quantitative wall-thickness map, and that is the data a plug-or-keep ruling, a retube deferral, or a fitness-for-service assessment stands on. The price of the number is preparation and time: a near-bare bore, a flooded tube, and minutes per tube rather than seconds — which is why IRIS campaigns either sample the bundle or consume window time in proportion to coverage. On tubes an APR screen has already flagged, that cost lands exactly where it earns.
When is eddy current the right method?
When the enemy is cracking or pitting in non-ferromagnetic tubes — the titanium, brass, and stainless alloys of condensers and coolers. Eddy current is the fastest of the probe methods and characterizes defect type, which APR does not. Its limits are the probe's limits: the bore must be clean enough to traverse, and conventional eddy current loses sensitivity on ferromagnetic carbon steel, where adapted techniques such as remote field testing take over. Like IRIS, it cannot enter a blocked tube.
When is APR the right method?
When the question is coverage: is every tube open, how clean is the bundle really, which tubes carry restrictions or wall loss, and where should the probes go. APR answers from the open tube end in seconds per tube, on near-service surfaces, on any tube metal, including geometries probe methods struggle with. Talcyon's published cases show the span — a 146-tube carbon steel exchanger validated in under an hour, a 512-tube finned geothermal condenser screened in 2 hours, a 7,541-tube titanium condenser in 24 hours. What APR does not do is produce IRIS's quantitative wall map or eddy current's crack characterization, and a program that needs those verdicts still books the probes — on the tubes that deserve them.
Why screen first, then probe?
Because sampling is a bet and coverage is a record. A probe program that inspects 10 or 20 percent of a bundle bets that the sample catches the bundle's condition; a screen that answers for every tube removes the bet, then concentrates the probe budget on the flagged few. Talcyon's geothermal case shows the full pattern: 512 tubes screened in 2 hours, 23% flagged with corrosion, 4% plugged on 50–60% wall loss, and the rest put on a 24-month monitoring cycle — every tube dispositioned, inside a window ultrasonic coverage alone could not meet. The same logic runs the cleaning-validation loop after jetting: screen all tubes, re-clean only the flagged, re-test, hand over the record. The full cleanliness side of that story — production clean vs inspection clean and what each method needs — is at tube inspection after cleaning.
Where does Dezu fit in this work?
Dezu is the cleaning contractor in this sequence, and says so plainly. The delivered record is heat exchanger tube cleaning — six recurring shutdowns at an HPAL nickel refinery since 2023, every tube verified one by one on tube maps — and that cleaning is what decides whether probe methods can read a bundle at all. On request, Dezu supports APR screening as part of a cleaning scope using the Talcyon APRIS system, with completion estimates issued after scope review; Dezu publishes no tube-per-day capacity until it holds verified field actuals of its own. Tube integrity verdicts by IRIS or eddy current remain NDE-vendor work — Dezu's role there is the inspection-clean preparation those probes require. The cleaning service is at heat exchanger tube cleaning.
Does APR replace IRIS or eddy current testing?
No. APR is a screening method: it answers fast, for every tube, where a restriction or wall loss sits and how big it is, but it does not produce the quantitative wall-thickness map IRIS gives or the crack characterization eddy current gives. Current practice uses them in sequence — screen 100% of tubes by APR, then send IRIS or eddy current only to the tubes the screen flags. The screen buys coverage; the probes buy precision where precision is needed.
Which tube inspection method works on carbon steel tubes?
Conventional eddy current works best on non-ferromagnetic tubes such as titanium, brass, and austenitic stainless — ferromagnetic carbon steel needs adapted techniques such as remote field testing. IRIS works on carbon steel but demands a near-bare, flooded bore. Acoustic pulse reflectometry is material-independent because it reads the air column, not the tube metal: Talcyon's published cleaning-validation case ran APRIS on a 146-tube carbon steel exchanger, and its condenser case on thin-wall titanium — the same method on both metals.
How fast is each tube inspection method?
Published APR results range from 146 tubes in under an hour to 7,541 tubes in 24 hours, with Talcyon's brochure stating about 10 seconds per tube depending on size, length, and configuration — field throughput varies materially with geometry and access, so treat single-job figures as results for those jobs. Probe methods are slower per tube because a probe must physically traverse every bore — which is exactly why probe programs sample instead of covering every tube, and why screening first changes the economics.
Sources
- Talcyon APRIS brochure (method description; ~10 seconds per tube) — talcyon.com material, mirrored
- Talcyon case study, Geothermal Plant (Turkey) — Isopentane Condenser Tube Inspection — ndt.talcyon.com
- Talcyon resources library (application notes on finned, U-tube, twisted-tube, boiler, furnace, and cooler inspections) — ndt.talcyon.com/resources
- Talcyon Pte Ltd case studies, Cleaning Validation — Heat Exchanger (Petrochemical) and Cleaning Validation — Condenser (Power Plant) — Talcyon-produced documents on file with Dezu; not in the public resources library as of August 2026
- IRIS and eddy current characteristics — standard NDE method descriptions; campaign-specific capability is defined by the NDE vendor's written procedure
Deciding how to inspect a bundle this shutdown?
Send the exchanger list, tube material and sizes, what the inspection must decide, and the window. We reply with the cleanliness grade the method needs, whether 100% screening fits, and a scope-based quotation for the cleaning side.