PAUT Corrosion Mapping for API 510 Vessels and On-Stream Inspection

Short answer: PAUT corrosion mapping scans an area of the vessel wall with an encoded phased array probe and builds a C-scan thickness map, instead of taking a reading at a single point. API 510 prefers scanning methods where corrosion is localised or wall is near the minimum. That makes mapping a natural tool for on-stream inspection. Whether an on-stream inspection can replace an internal one is decided by the inspector under the code's conditions, not by the method.

This guide is for inspection engineers, API 510 inspectors and reliability leads in the US and Canada who are deciding whether encoded PAUT mapping belongs in a vessel inspection plan. It explains what the technique produces, how API 510 (eleventh edition, October 2022, with errata) treats scanning and on-stream inspection, what limits hot surfaces bring, and what the inspector should expect in the data package. Code points are given in our own words. Confirm them against your licensed copy and your jurisdiction. Interval and substitution decisions belong to the owner and the owner's API 510 inspector.

What PAUT corrosion mapping is

Direct answer: A phased array probe holds many small elements that fire in groups. For corrosion mapping, the instrument steps a 0° beam electronically along the probe while a scanner records position. Each scan line gives a strip of thickness data, and the strips are stitched together into a C-scan: a plan view of the area, coloured by remaining wall.

Conventional UT thickness gauging takes one reading under one transducer. To cover an area by hand, a technician moves the probe across a grid and writes down the lowest value. Automated conventional UT (often called AUT mapping) mounts a single-element probe on a raster scanner and records the full grid. PAUT replaces the single element with an array. One pass of the probe covers a strip as wide as the active aperture, so the scanner can cover an area in fewer passes with a fine data spacing in both directions.

The result is three related views:

• C-scan. A top-down map of thickness or of time-of-flight to the back wall. It shows the shape and extent of thinning and where the minimum is.
• B-scan. A cross-section along a scan line, showing the profile of the back wall. It helps tell general loss from pitting, and real loss from laminations or inclusions.
• A-scan. The raw signal at any point, kept in the file so an analyst can check a reading later.

Because the data is encoded and stored, the same area can be scanned at the next inspection and compared point by point. That repeatability matters more under API 510 than raw speed. A corrosion rate is only as good as your ability to measure the same steel twice. Our corrosion mapping service page covers the hardware options, including manual, semi-automated and fully automated scanners.

Why API 510 favours scanning where corrosion is localised

Direct answer: API 510 says that where localised corrosion is expected, examinations should use scanning methods, such as scanning UT or profile radiography, that show the extent of the damage. It prefers scanning or profile techniques where corrosion is localised or where remaining thickness is close to the required thickness.

The eleventh edition defines scanning NDE as methods meant to find the thinnest spot or all defects in a specified area. Examples given include scanning ultrasonic techniques and profile radiography of nozzles. In its thickness-monitoring guidance, the code says the minimum thickness at a CML can be found by UT or radiography. When localised corrosion is a concern, scanning methods should be used to reveal its scope and extent. Its section on measurement methods goes further. Any suitable NDE may be used if it gives a minimum thickness. Where a method leaves too much uncertainty, A-scan, B-scan or C-scan techniques may be used instead. Scanning or profile techniques are preferred where corrosion is localised or close to the required thickness.

The code also says that either the thinnest reading or an average of several readings within an examination point is recorded and used for corrosion rates. Where detailed thickness grids are needed to assess metal loss, it refers to API 579-1/ASME FFS-1 for how to prepare them. Corrosion mapping is the usual way to produce those grids. The fitness-for-service assessment itself is an engineering exercise the owner commissions, separate from the NDE.

A 2023 laboratory study in the journal Materials compared phased array corrosion mapping with spot thickness gauging on steel plate. It found that mapping gave better coverage and that spot gauging can miss local corrosion. In that study, flat-bottom holes with 40% wall loss were consistently detected and 20% wall-loss holes of 10 to 20 mm diameter were clearly identified. Holes of 5 mm diameter were harder to resolve. Results like these depend on the probe, the setup and the surface, so treat them as an illustration of why scanning matters, not as a performance figure for any given job.

On-stream inspection in lieu of internal inspection: what API 510 says

Direct answer: At the inspector's discretion, API 510 allows an on-stream inspection to replace an internal inspection when entry is physically impossible, or when a list of conditions is all met. These include a known general corrosion rate below 0.005 in. (0.125 mm) per year, more than 10 years of remaining life and at least five years of similar service. The vessel must also have no environmental cracking or hydrogen damage and no non-integral liner.

The eleventh edition sets out the full list for vessels that can be entered. All of these must hold:

1. The general corrosion rate is known to be less than 0.125 mm (0.005 in.) per year.
2. Remaining life is more than 10 years.
3. The corrosive character of the contents, including trace components, has been established by at least five years of the same or similar service.
4. The external inspection found no questionable condition.
5. The shell operating temperature does not exceed the lower limit of the creep-rupture range for the material, as referenced in API 579-1/ASME FFS-1.
6. The vessel is not subject to environmental cracking or hydrogen damage from the process fluid.
7. The vessel has no non-integrally bonded liner, such as strip or plate lining.

If these are not met, the next inspection must be internal. The alternative route is an RBI assessment under API RP 580 that finds the risk acceptable to the owner and shows the external NDE is effective for the expected damage. When an on-stream inspection is used, the inspection plan must specify the type and extent of NDE. The inspector must have enough access to the heads, shell and nozzles to make an accurate assessment.

Two related clauses matter for vessel trains and large columns. The results of an internal inspection on one vessel can support on-stream substitution on a similar vessel in the same or similar service, but not where the vessel is subject to environmental cracking or hydrogen damage. Large vessels with zones that corrode at different rates may treat each zone separately. Our guide to API 510 intervals in wet H2S service shows how the cracking exclusion plays out in sour service.

The code adds that an on-stream inspection may be done with the vessel either pressurised or depressurised. It says techniques are chosen for their ability to find the expected damage from outside and to work in on-stream conditions such as metal temperature. All on-stream work by an examiner must be authorised and approved by the inspector.

Where PAUT mapping fits in an on-stream inspection

Direct answer: Mapping answers the thinning question across defined areas: how thick, where the minimum is, and whether loss is general or pitted. It is the evidence an inspector needs to trust a general corrosion rate and to cover zones where spot CMLs could miss local attack. It does not cover cracking, lining failure or other mechanisms on its own.

The code lists the factors that limit any external technique trying to find internal damage. They include the material, product form (plate, pipe, casting), welds, nozzles and saddles, internal attachments, linings or cladding, access and temperature, the limits of the technique itself, and surface condition. A sound on-stream plan works through each of these for the vessel at hand. In practice that leads to scope decisions such as:

• Where to map. Bottom heads and boots where water collects, liquid-level bands, areas below inlet nozzles, zones near internal attachments, and any area where spot CMLs have shown scatter or a step change.
• How much to map. Enough of each zone to make the general corrosion rate credible. The code's 0.005 in. per year condition is about a known general rate, and a map gives a far better basis for that claim than a handful of spot readings.
• What mapping cannot see. Cracking at welds needs angle-beam methods, such as shear-wave UT, PAUT sectorial scans or TOFD. Damage behind a lining, or loss in cladding, needs a separate approach. Nozzle bores and small connections may need profile radiography.
• Repeatability. Datum marks or reference points so the next survey maps the same area, and the same scan settings.

Weld-crack examination from outside has its own qualification angle. API 510 says the owner shall specify industry-qualified UT angle beam examiners where detection, characterisation or through-wall sizing of defects is required, for example when looking for ID-surface flaws from the OD. That applies to crack work, not to 0° thickness mapping, but on-stream campaigns often include both.

A worked example, described qualitatively

Direct answer: Mapping helps most where spot readings cannot support the rate or remaining-life case for substitution, or where a zone needs better coverage. The inspector, not the NDE crew, decides whether the conditions are met.

Take a carbon steel knock-out drum in hydrocarbon service with no wet H2S, no lining, and well over five years of service history. Spot CML readings over three inspections suggest a low general rate. One bottom-head CML, though, has read lower each time, and the scatter makes the rate hard to defend. The owner wants to know whether the next inspection can be on-stream.

The inspection plan calls for encoded PAUT mapping of the bottom head and the lower shell course up to the normal liquid level, plus a band around the inlet nozzle. It also calls for conventional spot readings at the remaining CMLs and an external visual. The map shows a general thickness consistent with the earlier readings and a small area of shallow pitting near the low point. The deepest pit sits well above the required thickness. The minimum and its coordinates are recorded, and the map is archived as a baseline.

The inspector now has a defensible general rate, a measured minimum in the worst zone, and a baseline for future comparison. With those, the inspector judges whether every condition for substitution is met. If they are, the inspector records the decision in the plan, along with the NDE scope for the next on-stream inspection. If the map had shown deep local attack, the same data would instead point towards an internal inspection or a deeper engineering review. Either way, the decision and the remaining-life calculation sit with the inspector and engineer. Our corrosion rate calculation guide covers how rates are worked out from successive readings.

Hot surfaces and other field limits

Direct answer: On-stream mapping is often done on hot steel. Heat affects couplant, wedges, sound velocity and probe life. API 510 asks for corrective procedures where metal temperature, typically above about 150 °F (65 °C), affects thickness accuracy. Above that, the setup has to be qualified for the temperature.

The eleventh edition says that where metal temperature affects accuracy, instruments, couplants and procedures suited to the temperature should be used. Calibration is usually on hot test plates, or readings are adjusted with a temperature correction factor. For phased array mapping this means:

• Couplant. Water evaporates on hot steel. The 2023 Materials study saw couplant start to vaporise above about 100 °C and image quality degrade between 150 °C and 250 °C. High-temperature couplants and short contact times help.
• Wedges and delay lines. Standard wedges soften or change velocity when hot. High-temperature wedges, water-cooled wedges or delay-line setups are used, with their velocity checked at temperature.
• Velocity in the steel. Sound velocity drops as steel heats, so uncorrected thickness readings are wrong. Calibrate at temperature or apply a documented correction.
• Insulation and access. Insulated vessels need windows or removable panels, which then need resealing, because API 510 warns that open CML plugs can cause CUI. Scaffolding, rope access or crawlers may be needed for upper shells.
• Surface condition. Paint, scale and heavy external pitting affect coupling and can show up in the data. API 510 lists surface condition among the inherent limits of external NDE.

The inspection plan should record the temperature at which mapping was done and the correction method used, so the next survey can be compared like for like.

PAUT mapping vs other ways to measure vessel thickness

Direct answer: Spot UT is the backbone of CML trending, but it only reports one point. PAUT and conventional AUT mapping cover areas. Profile radiography suits nozzles and small connections. Screening methods cover distance but not detail. Most on-stream plans combine several.

For the trade-offs between phased array and conventional UT in general, see conventional UT vs phased array.

What the inspector should receive

Direct answer: A usable mapping package includes the procedure and setup, the location and orientation of each scanned area, the C-scan with minimum thickness and coordinates, notes on indications that are not wall loss, the temperature and correction used, and the raw data files.

• Written procedure reference, approved under the NDE provider's Level III, with the technique sheet: probe, wedge, focal law, scan and index resolution, and gates.
• Calibration or reference block details, and the velocity used at the measured surface temperature.
• A sketch or drawing reference for each scan area, with datum points so the area can be found again.
• The C-scan for each area, with colour scale, minimum thickness and its coordinates, and the extent of any area below a threshold the owner sets.
• B-scan sections through the minimum, with notes on laminations, inclusions or blisters that could be mistaken for wall loss.
• Coverage limits: areas not scanned and why, such as attachments, nozzles, or hot or inaccessible zones.
• Personnel qualifications under the employer's written practice.
• Raw encoded data files in a format the owner can archive and reopen.

Under OSHA PSM (29 CFR 1910.119(j)), inspection and test records for covered equipment must be kept. A mapping campaign that keeps only a PDF summary loses most of its value at the next comparison.

Common mistakes with on-stream mapping

Direct answer: The usual errors are treating mapping as permission to skip internals, mapping the easy areas instead of the likely ones, ignoring temperature effects, and losing the datum needed for the next comparison.

• Treating the method as the decision. Good data supports the inspector's judgement against the API 510 conditions. It does not replace them. A vessel subject to environmental cracking still needs an internal or an RBI case, whatever the map shows.
• Mapping where access is easy. The plan should send the scanner to the zones where damage is expected, not where scaffolding already stands.
• No temperature correction. Readings on hot steel without velocity correction create false thinning or hide real loss.
• Calling laminations wall loss. Mid-wall reflectors look like thinning in a thickness-only C-scan. B-scans and an experienced analyst separate them.
• No baseline discipline. Without datum marks and the same setup, the next map cannot be compared point by point.
• Confusing mapping with crack detection. 0° mapping is not a weld examination.

Jurisdiction and Canada notes

Direct answer: Whether an owner can use API 510 on-stream substitution depends on the jurisdiction as well as the code. Some US states and Canadian provinces set their own internal inspection requirements for pressure vessels. Confirm the rules with the authority having jurisdiction before you plan on-stream credit.

In the US, many states regulate pressure vessels through a boiler and pressure vessel programme. Some accept owner-user programmes based on API 510 or the National Board Inspection Code, with their own intervals and reporting. OSHA PSM treats API 510 as recognised good practice for covered processes. In Canada, provincial regulators such as ABSA in Alberta and TSSA in Ontario oversee pressure equipment under CSA B51 and provincial rules. An owner-user programme there must meet the regulator's requirements, which may limit or set conditions on replacing internal inspections. Our API 510 inspection support page explains how the NDE contractor and the inspector of record work together.

How Atlantis supports this

Atlantis NDT performs encoded PAUT and conventional UT corrosion mapping on pressure vessels, including on-stream surveys with temperature-qualified setups where conditions allow. The work runs under procedures approved by an ASNT Level III, and we deliver C-scans, minimum-thickness locations and raw data to the owner's API 510 inspector. That inspector stays inspector of record and decides whether on-stream inspection can replace an internal, what the corrosion rate is and when the next inspection is due. Atlantis does not set intervals and does not offer RBI or fitness-for-service assessments. See phased array inspection services and pressure vessel inspection services. Request a mapping scope and quote. We respond within 24 hours.

Frequently asked questions

What is PAUT corrosion mapping?

It is an encoded ultrasonic scan of an area using a phased array probe. The result is a C-scan thickness map that shows the extent of thinning and the location of the minimum, rather than a single reading.

Is phased array corrosion mapping better than UT thickness gauging?

For coverage, yes. Mapping records an area, while spot gauging records points and can miss local corrosion between them. Spot UT is still the standard way to trend CMLs over time. Most programmes use both.

Can on-stream inspection replace internal inspection under API 510?

Yes, at the inspector's discretion, when entry is impossible or when all of the code's conditions are met. These include a known general rate below 0.005 in. per year, more than 10 years of remaining life, five years of similar service and no environmental cracking. An RBI assessment is the alternative route.

Does API 510 require PAUT for on-stream inspection?

No. API 510 requires the plan to specify NDE able to detect the expected damage from outside. PAUT mapping is one option for thinning. Spot UT, scanning UT, profile radiography and crack-detection methods are others.

Can corrosion mapping be done on hot vessels?

Often, yes, with high-temperature couplant, wedges and calibration at temperature. API 510 asks for corrective procedures where temperature affects accuracy, typically above about 150 °F (65 °C). Very hot surfaces limit contact time and image quality.

Does corrosion mapping detect cracking?

Not on its own. 0° mapping measures thickness. Cracks need angle-beam UT, PAUT sectorial scans or TOFD, done by examiners qualified for that work.

How is corrosion mapping data used for fitness-for-service?

API 510 refers to API 579-1/ASME FFS-1 for preparing thickness grids when metal loss must be assessed. Mapping is the usual way to produce those grids. The assessment itself is an engineering exercise the owner commissions.

How often should corrosion mapping be repeated?

On the inspection interval the owner's inspector sets under API 510: by default the lesser of half the remaining life or 10 years for internal or on-stream inspection, unless RBI justifies otherwise. Repeat maps should use the same datum and setup.

Who decides whether an on-stream inspection is acceptable?

The owner's API 510 inspector, within the owner-operator's quality system and any jurisdictional rules. The NDE provider supplies the data.

Talk to an NDE Level III about an on-stream mapping plan or send us your vessel list for a mapping quote.