Ultrasonic Corrosion Mapping: Mapping Wall Loss Instead of Measuring One Point

An insulated line passes twelve thickness readings, all within tolerance, and six months later it leaks between two of those points. The pit that holed it sat right there, in the space the measurement grid never touched. It's the scenario every asset integrity inspector knows, and it doesn't come from a reading error: it comes from the method. An ultrasonic thickness gauge measures one point at a time, and however tight the mesh, some surface always stays unseen between one point and the next. Corrosion mapping changes exactly that. Instead of sampling the wall at a few points, it reconstructs it as a dense matrix of measurements, and the result isn't a number but a map.

The worst spot rarely sits under a reading

Spot thickness measurement is a solid, standardised technique: ISO 16809, in its 2025 edition, defines how thickness is determined from the time of flight of the ultrasonic pulses, by contact or immersion. On general, uniform corrosion, where the wall thins predictably, a few readings at representative points are enough to estimate the rate and the remaining life, and that's how API 570 sets up the monitoring of in-service piping, with registered condition monitoring locations (CMLs) and the interval fixed at the lesser of a maximum and half the calculated remaining life. The picture changes when corrosion isn't uniform. Pitting, localized erosion downstream of an elbow, attack under a deposit: these live on a few millimetres, and a coarse grid averages them out or misses them entirely. Hunting for them with scattered readings is like judging the worst pothole in a road by measuring its depth every ten metres, and the single-point side is covered in the guide to the ultrasonic thickness gauge for corrosion measurement.

From gauge to map: encoder, C-scan, B-scan

The move from point to matrix is mechanical before it's conceptual. A motorised or manual scanner carries the probe over the surface, and an encoder records its position at every measurement, so each thickness value is born with its own coordinate. The software lays the data out in a false-colour plan view, the C-scan, where thinned zones show up as patches, and in vertical sections, the B-scans, which reveal the wall profile along a line. Phased array is today the preferred technique, because a single probe covers a wide strip on each pass, cuts the time and raises the data density: it's the same principle set out in the phased array (PAUT) guide, applied to mapping rather than to hunting cracks in welds.

Density, though, is a choice, not an automatism. A fine grid, around one millimetre, reveals pitting and fine contours; a coarse mesh covers large areas fast but can miss the small indications. And this is where the C-scan alone can mislead. A study in Materials (Tai and colleagues, 2023) on A36 steel showed that simulated defects of 10, 15 and 20 mm in diameter were clearly identified in shape, while the 5 mm one didn't return a clean outline; the measured depth deviation stayed below 0.8 mm all the same. The smaller pits are seen and sized better in B-scans than in the plan-view map: stop at the C-scan and you risk losing them. The technique belongs to industrial ultrasonic testing (PAUT/TOFD), of which mapping is the application dedicated to corrosion control.

Piping, tanks, and corrosion under insulation

The sectors where mapping pays off are those where a failing wall has large consequences: oil & gas, power, chemical, refining. The governing standard changes with the asset, and it helps to have it clear before writing an inspection plan.

• ISO 16809:2025 — ultrasonic thickness determination (contact or immersion), the physical basis of both spot measurement and mapping.
• API 570 and API RP 574 — inspection and thickness measurement of in-service piping; RP 574 covers techniques and extent and reminds that screening methods (guided wave, Lamb wave) count only as a complement to periodic quantitative examinations.
• API 653 — aboveground storage tanks: bottom-side floor corrosion, which can't be seen visually, is searched for with Magnetic Flux Leakage and sized by UT, while shell thickness is verified ultrasonically.
• API RP 583 — corrosion under insulation (CUI): for carbon steel the risk window runs roughly from −12 to 175 °C, with the most severe attack between 77 and 110 °C.

CUI is precisely where the matrix earns its place over the point. The wall under the insulation can't be seen, corrosion is often localized, and the temperature window says where it's likely but not how deep it runs. Opening the lagging at two points and measuring them is a statistical gamble; mapping the at-risk zone, once located, is data to decide on.

Heat exchangers add a case of their own: bundles of hundreds of tubes where shell-side corrosion, inlet erosion and deposits spread anything but evenly, and where a map tells which tubes to plug and which to keep. Across all these assets the matrix has a second value beyond detection: it's a repeatable record. Rescanning the same area months later and overlaying the two maps gives the real corrosion rate across the whole surface, not just at points, and that's the figure the next inspection interval can be set on, on firmer ground than a handful of readings.

When a thickness gauge is still enough

Not everything should be mapped, and saying so is part of the trade. For general, slow corrosion on an accessible surface, a grid of spot readings managed to API RP 574 is enough and costs a fraction: you register the CMLs, compute the two corrosion rates and use the more conservative one to derive remaining life, without a scan. Mapping comes in where the damage is localized or its morphology matters, or where the detailed record serves to compare the asset over time. There's a physical limit to budget for as well: UT mapping wants contact with the bare wall, so under insulation or coating you have to open a window or remove the lagging over the area to be checked. The choice, at that point, is clear-cut: either you accept the cost of exposing the wall where risk is high and map it properly, or you rely on screening to decide where to open, knowing it points to the zone and not to the remaining depth.

From one pit to an unplanned shutdown

A five-millimetre pit missed by a coarse mesh is, on paper, a detail. On the line it becomes the point a leak starts from; on the unit it becomes an unplanned shutdown, with production stopping while the section is isolated and repaired. It's the chain mapping serves to break at the start, while the defect is still a patch on a C-scan and not an escape. Anyone managing the integrity of piping, tanks or exchangers, and today relying on spot readings on critical assets, has a simple way to measure the gap: have a suspect area mapped, ideally one already passed by the gauge, and see what the matrix returns. This is how PITECH sets up the assessment, with analysis of the asset and the expected damage mechanism, a demonstrative scan, and a reasoned choice between spot measurement and mapping, including the honesty to say when spot readings already suffice. To start it, describe the asset and the problem from the contact page.