Who Approves the UT Procedure — Level III Authority in Corpus Christi

What ultrasonic Level III authority actually covers

Ultrasonics is the method where the gap between a certificate and a competence is widest, because the method spans work of radically different difficulty. Taking a thickness reading on a clean carbon steel pipe and characterising a planar flaw in an austenitic weld are both ultrasonic testing, and both can be performed by the same Level II holding the same certificate. What separates a defensible program from an indefensible one is that the procedures distinguish the two and the certification file shows who has demonstrated what.

The Level III owns the procedure and everything upstream of it. That means selecting the probe frequency and element size against the material's attenuation and grain structure, choosing wedge and refracted angles that actually insonify the weld volume including the root and the far toe, setting the calibration basis and the reference reflector, specifying surface condition and couplant, fixing scanning speed and overlap, deciding whether transfer correction is required and how it is measured, and naming the acceptance standard being applied. Every one of those is a decision that can be argued in an audit, so every one needs a reason recorded.

The Level III also owns qualification. Where a code or an owner specification requires demonstration — encoded techniques, automated ultrasonics on girth welds, examination of materials the base procedure was not written for — someone has to design the demonstration, fabricate or source flawed specimens, run it, and sign the record. That is not a task a Level II can complete on their own authority, and it is the piece most frequently skipped when a technique arrives on site in a hurry.

Straight beam: thickness, laminations, and the arithmetic that goes wrong

Thickness measurement looks like the simplest thing ultrasonics does and generates more bad data than anything else in an inspection program. The single-echo gauge measures a transit time and converts it with an assumed velocity. Put coating in that path and the instrument adds the coating to the steel, reading high by an amount that can exceed the entire corrosion allowance on a thin dock line. Echo-to-echo mode solves it by timing between successive back-wall echoes, but only if the procedure requires it and the technician's instrument supports it. A procedure that is silent on coating is producing a thickness history that trends the paint.

The second trap is what the reading represents. Ultrasound reports the remaining ligament directly under the transducer. On generalised wall loss that is representative; on pitted or grooved corrosion it is a lottery, and the honest answer is that spot readings do not characterise localised loss at all — that is what encoded mapping is for. The evaluation rules then compound the problem. API 510 and 570 both require long-term and short-term corrosion rates to be computed and the more conservative used to set the next inspection date. Averaging readings across a monitoring location, or dropping an anomalous low reading as 'operator error' without re-taking it, quietly moves an inspection interval outward on arithmetic nobody re-checks.

Straight beam also does lamination work — incoming plate to ASTM A435 or A578, mill product at a flat-roll facility, and hydrogen blistering and stepwise cracking in refinery wet H2S circuits. These are area-coverage examinations with sensitivity set against a back-wall reference from sound material, and the acceptance criteria come from the material specification or the order, not from the weld code. Getting the invoked standard right on the report is a surprisingly common finding.

Angle beam: DAC, DGS and the argument about reference reflectors

Shear-wave weld examination under ASME Section V Article 4 is built on a distance-amplitude correction curve constructed from side-drilled holes in a basic calibration block whose thickness and material are matched to the part. That block is not a formality. Its material must acoustically represent the component, its thickness governs which hole positions are used, and it must be at a temperature close to the examination surface — a block sitting in an air-conditioned trailer and a line at process temperature do not calibrate each other honestly.

The DGS or AVG approach replaces the empirical curve with a diagram relating distance, gain and equivalent reflector size for a given probe. It is efficient, it removes the block-machining burden, and it depends absolutely on the probe's published diagram being valid for the wedge and the material in use. European-derived specifications reach for it more readily than ASME work does. Time-corrected gain flattens the response electronically and is the natural companion to encoded scanning. All three are legitimate; the Level III's job is to pick one, justify it against the material and thickness, and make the procedure say which reflector amplitude constitutes the recording level.

Transfer correction is the clause that separates careful procedures from copied ones. Surface condition, coating, curvature and grain structure attenuate sound in the part more than in the polished calibration block, and if that difference is not measured and compensated, examination sensitivity is lower in the field than on the bench by an unknown margin. Requiring a transfer measurement, and stating how it is made and when it must be repeated, costs a paragraph and buys the whole procedure its credibility.

Detection versus sizing — and why the distinction decides a repair

Workmanship codes reject on amplitude and length. An indication exceeding a percentage of the reference curve, longer than an allowed length for the thickness, is rejectable — and nothing in that judgement tells you how tall the flaw is or where through the wall it sits. That is a perfectly rational basis for accepting new fabrication. It is close to useless for deciding whether an in-service component can keep running.

Fitness-for-service under API 579-1/ASME FFS-1 asks entirely different questions: flaw height, through-wall position, orientation, whether it is surface-connected, whether it is planar or volumetric. Producing those numbers requires a sizing technique — tip diffraction, time-of-flight diffraction, encoded phased array with a documented sizing basis — qualified for the geometry in question, and personnel who have demonstrated sizing accuracy against flaws of known dimension. Amplitude data cannot be converted into height by anyone's judgement, however experienced.

This is where a Level III earns an engagement outright. When a turnaround finds a rejectable indication at three in the morning, the choice is repair now at outage cost or assess it and defer. Making that choice defensibly requires the sizing capability to already exist — the procedure written, the technique demonstrated, the analyst certified. Sites that build it in advance make the call in hours. Sites that do not, cut the weld out, because that is the only decision they can defend.

Where ultrasound struggles: austenitic, duplex, clad and coarse-grained materials

Austenitic stainless weld metal solidifies with coarse columnar grains that are large relative to the wavelength. The result is scatter that raises the noise floor, attenuation that reduces sensitivity, and anisotropy that skews and bends the beam so the geometric assumptions behind the sound path stop holding. An indication may appear displaced from its true position; a real flaw may be lost in grass. Duplex behaves similarly with its own character, and clad or weld-overlaid surfaces add an acoustic interface that generates its own signals.

The technical answers are known: lower frequencies to reduce scatter, dual-element transmit-receive longitudinal-wave probes to suppress interface noise, sometimes creeping-wave or specialised angle configurations, phased array with a modelled focal law rather than a fixed wedge. What is not negotiable is the qualification. A procedure for austenitic weld examination has to be demonstrated on a mockup made by the same welding process, with the same joint configuration and the same grain structure as the production weld, containing representative flaws. A carbon steel calibration block proves nothing here.

For Corpus Christi this is not a theoretical concern. LNG liquefaction and cryogenic storage bring austenitic piping and 9 percent nickel tankage; petrochemical and refining bring clad and overlaid reactors and stainless process lines; the fabrication yards see duplex on offshore scope. A UT program written for carbon steel piping — which is most of what gets inherited — is not qualified for any of it, and the gap tends to surface during a client audit or after a leak rather than during procedure review.

Ultrasonic demand across the Coastal Bend industrial base

The three refineries in and around the city — CITGO, Valero and Flint Hills Resources — generate the steadiest ultrasonic workload: thickness monitoring on process circuits, fired heater tube examination, weld examination during turnarounds, and the damage-mechanism-driven scope that comes out of an API 571-informed review. The export side of the port, the largest crude export gateway in the country, adds dock and jetty piping, loading manifolds and above-ground storage tanks, where coating, marine atmosphere and access dominate the technique choice.

North of the harbor the material mix changes. The ethane cracker corridor at Gregory brings high-temperature furnace and transfer-line scope; LNG liquefaction on the La Quinta Channel brings cryogenic austenitic and 9 percent nickel work; the Steel Dynamics mill at Sinton and the hot briquetted iron plant at Portland bring plate lamination examination, structural welding and mill-support pressure equipment. Ingleside adds heavy offshore fabrication with thick-section and tubular joint ultrasonics under AWS D1.1 and owner structural specifications.

And Corpus Christi Army Depot at the naval air station is a reminder that not all local ultrasonics lives under the same rulebook — depot rotorcraft work qualifies personnel under NAS 410 with a named responsible Level 3, examines aluminium and bonded structure, and should never be folded into an industrial written practice. A regional service company chasing both markets needs two documents, not one flexible one.

Sour service and wet H2S — a scan strategy, not a flaw hunt

Refining on this coast has run heavy sour crude and, more recently, light shale barrels through the same assets, and the damage mechanisms follow the feed. Wet hydrogen sulfide service produces hydrogen blistering, hydrogen-induced cracking and stepwise cracking in the plate itself, and stress-oriented hydrogen-induced cracking near welds. None of these look like the planar weld flaw a shear-wave procedure is written around.

The examination strategy has to change accordingly. Blistering and HIC are area phenomena found with straight-beam scanning of the base material, often with C-scan mapping, with sensitivity and evaluation rules referenced to the mechanism rather than to a weld acceptance standard; AMPP/NACE TM0284 supplies the evaluation vocabulary for material assessment. SOHIC and near-weld cracking need angle beam from both sides with attention to the heat-affected zone. The reporting convention matters too — extent, density and location, not a list of individual indications.

This is a place where a procedure copied from a construction code actively misleads. A technician following an ASME Section V Article 4 weld procedure on a sour water circuit will scan the weld competently and walk past the blistered plate either side of it. The Level III's contribution is insisting the procedure set be organised around the damage mechanisms the corrosion engineer has identified, and writing the scan strategy that finds each one.

Working with an independent UT authority

A typical ultrasonic engagement covers procedure development and approval across the applications you actually run, technique qualification where a code or a client demands demonstration, calibration and sizing basis decisions with the reasoning recorded, examination and certification of UT technicians under your written practice — including the specific and practical examinations that give a UT certificate meaning — and independent review of ultrasonic data when a reading or an indication is disputed between an owner and a contractor.

The independent review element is worth naming separately, because it is quietly the most requested. When a rejectable indication appears at the wrong moment, when a thickness history stops making sense, or when two contractors disagree about the same weld, someone with method authority and no stake in the outcome has to read the data. That is a technical opinion, not an inspection decision: Atlantis does not act as the API inspector of record and does not sign the in-service inspection.

Founder-led, ASNT NDT Level III in multiple methods and API 653 Authorized Inspector. Affordable, accessible and fully customizable, mobilised to Corpus Christi and the wider Coastal Bend for on-site qualification and examination work, and handled remotely for procedure and record work. To discuss a UT procedure review, a technique qualification, or examiner certification ahead of a turnaround, write to info@atlantisndt.com and request a consultation.

Request a consultation