Phased Array Ultrasonic Inspection, Bought as a Service

What Companies Buy PAUT For

Four scopes account for most of the phased array work purchased in the United States. Weld inspection on new fabrication, repairs and turnaround tie-ins, where PAUT proves the joint volume is sound. Corrosion mapping on vessel shells, piping and elbows, where a C-scan grid replaces a scatter of single-point thickness readings. Crack height sizing, where an existing flaw has to be measured accurately enough to be assessed rather than cut out. And high-temperature hydrogen attack screening on hydroprocessing equipment, where the damage begins inside the wall and never breaks a surface.

These are four different jobs. They use different probe frequencies, wedge angles, scan plans and calibration blocks, and they produce different data types. A vendor who quotes PAUT as a single line item without asking which of the four you need is selling a technician and an instrument, not an inspection. The scan plan, which angles, which skips, what coverage of the weld volume and heat-affected zone, and what the calibration proves, is the product. Ask to see it before mobilisation rather than after the report lands.

The reason PAUT keeps taking share from radiography in US plants is not resolution. It is that a phased array crew works alongside other trades. No exclusion zone, no night shift to clear the unit, no source in transit. On a turnaround where every hour of a stopped work front has a number attached, that is often worth more than the inspection itself.

PAUT vs Radiography: Which Method the Code Accepts

The honest comparison is not that one method is better. Radiography images volumetric flaws such as porosity, slag and inclusions, with a picture anyone can read and archive. Ultrasound, phased array included, is far better at planar flaws: lack of fusion, lack of penetration, and cracks lying parallel to the beam path that a film may never show. It also measures through-wall height, which film cannot. If your dominant failure mode is planar or you need flaw height for an engineering assessment, PAUT wins on technical merit. The RT vs UT weld inspection decision works through the trade-off joint by joint.

Where it gets contractual is replacing radiography on coded pressure equipment. ASME Boiler and Pressure Vessel Code Section VIII, Division 1, UW-51(a)(4) permits ultrasonic examination in lieu of radiography when the thinner of the materials joined is at least 0.25 in. The examination routes through Division 2, paragraph 7.5.5, to Section V, Article 4, Mandatory Appendix VIII, and that appendix is where the restriction lives: the system must use automated or semi-automated scanning with computer-based data acquisition and analysis. Our ASME Section V compliance reference sets out the article structure.

Structural steel has its own route. AWS D1.1/D1.1M:2020 added Annex H, covering phased array examination of welds from 3/16 in to 8 in thickness, requiring encoded scanning, and setting acceptance in Table H.2 with no planar indication acceptable at any length. That is a stricter standard than the amplitude-and-length rating in Table 8.2 that conventional ultrasonic testing is judged against, and buyers moving a job from conventional UT to PAUT should expect the rejection profile to change.

Encoded vs Manual PAUT, the Line That Decides Whether the Data Survives

Manual PAUT means a technician holds the probe and watches a live sectorial scan. It is quick, cheap to mobilise, and genuinely useful for screening a weld or hunting a known indication. What it does not produce is a record. When the technician lifts the probe, the evidence is gone and what remains is an opinion typed into a report. If the job is a screening pass to decide where to look harder, manual is the right tool and there is nothing wrong with buying it.

Encoded PAUT puts the probe on a scanner with a position encoder. Every A-scan is stamped with a coordinate, so the file is a spatial record of the component. Six months later an analyst who never visited your site can reopen it, re-measure an indication, and check whether the coverage the procedure promised was actually achieved. That is the difference between an inspection you can defend and one you have to re-run. Our note on what makes an NDT report defensible covers what a reviewer looks for in the raw data.

Buy encoded when the result feeds a repair decision, a fitness-for-service case, an insurer file or a regulator. Buy manual when the result feeds the next inspection decision and nothing else. Stating which one you want in the purchase order removes the single most common dispute in phased array contracting, which is a client expecting an auditable file and a contractor having priced a hand scan.

Corrosion Mapping: Turning Thickness Readings Into a Surface

Conventional thickness surveys sample. A technician takes readings on a grid, and the minimum recorded thickness becomes the number in the file, with no way of knowing whether the true minimum fell between two points. Phased array corrosion mapping raster-scans an area with a linear array and produces a continuous C-scan, a colour map of remaining wall across the whole scanned surface. On a vessel shell or a pipe elbow with localised attack, the difference between a sampled minimum and a mapped minimum is routinely several tenths of an inch.

That matters because remaining-life calculations under API 510 and API 570 are driven by the minimum measured thickness and the corrosion rate derived from it. A sampling survey that misses the deepest point overstates remaining life and pushes the next inspection date further out than the metal justifies. A mapped survey compresses that uncertainty. It also gives a comparable dataset: the same area, rescanned at the next outage, aligned to the same reference, showing whether the corrosion rate is what the model assumed.

Mapped corrosion data has a second life. C-scan grids registered to component coordinates sit on the geometry rather than in a spreadsheet, so wall-loss history accumulates against the asset instead of against a filename. That is the practical bridge from inspection data into digital twins, and it is the difference between having ten years of scans and being able to use them.

Crack Sizing for Fitness-for-Service

When a crack-like flaw is found, the expensive question is not whether it exists. It is how deep it goes. A weld with a 2 mm through-wall flaw and a weld with a 12 mm through-wall flaw can look similar on a screening scan, but one can often be run to the next outage under an engineering assessment and the other cannot. Height sizing is the entire commercial value of the scan, and it is the part most often quoted as though it came free with detection.

PAUT sizes flaw height by tip diffraction, resolving the echo from the crack tip separately from the corner echo, and by focused imaging that concentrates the beam at the depth of interest. Done properly on a calibrated block with a representative reflector, it delivers height with enough confidence to feed a fitness-for-service assessment under API 579. Done as a quick manual look, it delivers an amplitude that says something is there. If the flaw is going to an assessment, the sizing technique has to be qualified for height and that has to be written into the procedure.

Ask the vendor one question: what is your sizing accuracy, and how was it demonstrated? A crew that answers with a qualification record on representative reflectors has thought about it. A crew that answers with a figure from the instrument manufacturer's datasheet has not, and the assessment engineer downstream will send the data back.

High-Temperature Hydrogen Attack Screening

HTHA is why many US refiners keep phased array capability on retainer. In hydrogen service above certain temperature and partial-pressure combinations, methane forms at the grain boundaries of carbon and low-alloy steel and the metal fissures internally. There is no surface indication and no wall loss to measure. By the time damage is coarse enough for a conventional angle-beam probe to see it, the component is well down the road. API Recommended Practice 941 is the governing document for the operating limits, and API RP 586 addresses the inspection.

The techniques API 941 names for detection are time-of-flight diffraction, phased array and the total focusing method. In practice the sequence is a TOFD screen over parent metal and welds to cover area quickly, followed by PAUT and TFM at flagged locations to characterise what was found and separate genuine hydrogen damage from mid-wall inclusions, laminations and geometry. Attempting HTHA detection with a single technique is where most disappointing HTHA campaigns start.

This is the scope where the analyst matters more than the equipment. HTHA data is ambiguous and calling it requires someone who has seen both the damage and the look-alikes. If your organisation does not have that person, buy the review. Our ASNT Level III consulting services cover technique qualification, procedure approval and independent data adjudication on exactly this kind of call.

Who Is Qualified to Run the Scan

Phased array is not a separate NDT method in the certification schemes; it is a technique within ultrasonic testing. Under ASNT SNT-TC-1A, personnel are certified by their employer against a written practice, method by method and level by level. A UT Level II certificate does not by itself demonstrate phased array competence, and a purchaser who accepts one as proof has verified nothing about the technique that will actually be used. What you want is UT Level II or III certification plus documented phased array training plus procedure-specific qualification on the geometry in question.

The written practice itself is worth reading. It tells you the employer's training hours, experience hours and examination requirements, and whether those meet the recommended practice or merely gesture at it. On critical work, ask for the written practice, the certification records of the named technicians, and the procedure they will run, all before mobilisation. That request costs nothing and it filters out a surprising proportion of bidders.

If you are building this capability in-house rather than buying it, method training and technique qualification are the path, and the sequence matters: ultrasonic fundamentals first, phased array second, application qualification third. Our NDT training programmes run Level I, II and III curricula across ultrasonic testing and the phased array technique.

What a Complete PAUT Deliverable Contains

A phased array report consisting of a summary table and a pass/fail column is not a deliverable, it is a receipt. A complete package contains the written procedure and its revision, the scan plan showing beam coverage of the weld volume and heat-affected zone, the calibration record with the block and reference reflector used, the encoder setup and scan increment, the raw data files in a readable format, the indication table with position, amplitude, length and through-wall height, the acceptance standard applied by clause number, and the technicians' certifications.

The reason to insist on raw data is not distrust. It is that raw data is re-analysable. If a repair is challenged, if the flaw grows, if a different engineer wants to reassess against a fitness-for-service criterion rather than a workmanship one, the file answers the question and a summary sheet does not. Where reports arrive without it, NDT report validation is how the gap gets found before an audit finds it.

Ownership of the data should be stated too. Some contractors treat encoded files as proprietary and release only exported images. Put the raw file format and the transfer obligation in the purchase order. Atlantis delivers encoded data with every encoded scope, in the native format and in a viewer-independent export, and hands over the scan plan at quotation stage rather than after the job.

What to Put in a PAUT Scope of Work

Six items turn a vague enquiry into a comparable bid. One: the component, material, wall thickness range and weld configuration. Two: the code and clause that sets acceptance. Three: whether the result must replace radiography, because that decides encoded versus manual before anything else. Four: coverage, meaning full weld volume, heat-affected zone width, and whether both sides are accessible. Five: the sizing requirement, specifically whether through-wall height is needed and to what tolerance. Six: deliverable format, including whether raw files are required and who owns them.

Add access conditions and they will stop being change orders. Scaffold or rope access, surface preparation, component temperature at the time of scanning, insulation removal, and whether the crew can work alongside production. Most PAUT cost overruns in the United States are access overruns dressed up as inspection overruns, and they are visible in advance to anyone who asks the question during tendering rather than during execution.

Send the drawing and the acceptance clause and we will return a scan plan and a defined scope, with pricing quoted on request. Atlantis covers US shops, plants and job sites with encoded and manual phased array, and reviews phased array data collected by others. Start at contact.

Request a consultation