Ultrasonic authority for Baton Rouge refining, chemical and river work

What a UT Level III actually puts a signature on

Ultrasonics is the method where the operator's choices determine the answer. A radiograph either shows the flaw or it does not, and the film is there to be re-read a decade later. An ultrasonic examination produces a call, and the evidence that the call was capable of being right lives entirely in the procedure, the calibration record and the technician's qualification. That is why the Level III signature on a UT procedure carries more weight than the same signature on a penetrant procedure, and why UT is where audits concentrate.

The signature covers the procedure and every essential variable range inside it, the calibration blocks and reference reflectors, the instrument calibration and linearity regime, the scan plan and the argument that it achieves the coverage claimed, the sizing method and the tolerance actually demonstrated for it, and the examination material used to certify technicians in the method. Each of those is a place where an auditor can ask a question that a generic template cannot answer.

There is a fourth thing the signature covers that rarely appears in the procedure: the boundary of the technique. A good UT procedure states what it will not find. A shear wave technique from the outside surface of a nozzle has geometry it cannot reach. A single-element thickness technique will not resolve stepwise cracking. Writing the limitation down protects the technician who followed the procedure and gives the inspector the information he needs to decide whether the examination answered his question.

Straight beam is four different jobs, not one

The first is wall loss measurement at condition monitoring locations, which is a metrology problem more than a flaw detection one. It needs a defined velocity, a stated coating strategy, a defined probe and a documented location so the next reading is taken in the same place. The second is lamination and hydrogen damage screening on plate and shell material, which needs enough resolution to see mid-wall reflectors and a scan pattern dense enough to find them, and which produces a completely different report.

The third is clad and weld overlay bond examination, where the interest is the interface itself, and where the acoustic behaviour of austenitic cladding on a ferritic base means a technique tuned for wall thickness will simply not resolve disbond. The fourth is thickness measurement through coatings, where a single-element pulse-echo instrument will happily add the paint to the steel unless an echo-to-echo technique is used, giving readings that are consistently high and consistently reassuring.

These four are routinely covered by one procedure titled "Ultrasonic Thickness Measurement" that names one probe, one velocity and one calibration approach. It passes an internal review because everybody knows what it means. It fails a client audit because the document does not distinguish between measuring a wall and finding a laminar flaw, and the acceptance logic for the two is not the same.

Angle beam: DAC, DGS and the transfer correction nobody records

Amplitude evaluation against a distance amplitude correction curve is the ASME Section V default for weld examination. It requires a basic calibration block of the same material specification, product form and heat treatment condition as the component, with side-drilled holes at depths that bracket the thickness being examined, and curvature considered where the component diameter is small enough to matter. The curve is only as honest as the block. Build it in carbon steel and apply it to a clad or high alloy weld and you have manufactured false confidence.

Distance gain size, or AVG, evaluates against a probe-specific diagram and a single reference reflector. It is efficient and it is legitimate, and it is also the technique most often used incorrectly, because the diagram assumes the attenuation and surface condition of the reference. The correction for the difference between the reference and the actual component is the transfer correction, measured on the part with a pitch-catch arrangement and applied in decibels. Ask any contractor for the transfer correction record on a DGS job. The frequency with which that request produces silence is the reason this paragraph exists.

And then there is AWS D1.1, which is not an amplitude-versus-DAC system at all. Structural welds are judged on a decibel indication rating referenced to a 0.060 in side-drilled hole in an IIW type block, with an attenuation factor applied for sound path, and an acceptance table that varies with weld thickness and probe angle. A technician who learned ultrasonics in a refinery and is moved onto a shop structural job will apply the logic he knows, and the arithmetic he produces will not be the arithmetic the code asks for.

Detection and sizing are separate qualifications

A procedure that demonstrates detection has shown that the technique responds to a flaw type at a given sensitivity. That says nothing about how accurately the same technique measures the flaw's through-wall height, and through-wall height is the number that decides whether a component is repaired, derated, monitored or run. As soon as an ultrasonic result feeds an API 579 fitness-for-service assessment, the sizing tolerance becomes an input to a calculation, and an undeclared tolerance is an undeclared error bar on someone's remaining life number.

Sizing qualification means demonstrating the technique against flaws of known dimension in representative material, stating the tolerance achieved, and showing that a second qualified technician reproduces it. Tip diffraction, TOFD and encoded phased array all improve on amplitude-based sizing, and all of them still require the demonstration, because the improvement is in the physics, not in the operator. A procedure that names a sizing technique without a demonstration record has claimed an accuracy it has never measured.

The corollary matters for buyers. When you receive a report giving a crack height to two decimal places, ask what the demonstrated sizing tolerance was and on what material it was demonstrated. If the answer is a shrug, the number is an opinion with a decimal point attached, and it should not be driving a repair decision on a pressure boundary.

The arithmetic that retires the wrong line

Temperature is the first trap. Velocity in steel decreases as temperature increases, so an instrument calibrated at ambient over-reads on hot metal by roughly one percent per hundred degrees Fahrenheit. On a 700F transfer line a genuine 0.470 in wall can present as about 0.500 in. The reading is comfortably above the retirement thickness and the line stays in service for another cycle it did not have.

Velocity is the second. An instrument calibrated on carbon steel and used without change on 300-series stainless is wrong by several percent, because the longitudinal velocity in austenitic stainless is lower than in carbon steel. This bites hardest on circuits that are mostly carbon steel with an alloy spool nobody flagged, and it is exactly the situation API 578 positive material identification exists to prevent. The thickness error and the metallurgy error arrive together.

The third is statistical rather than physical. Take a grid of readings at a condition monitoring location, record the minimum, then repeat a year later with a different technician, a different couplant and a slightly different grid, and record the new minimum. The difference between two minima of two noisy populations is not a corrosion rate. It generates a short-term rate that panics an inspection group, or a negative one that gets quietly deleted. API 570 and API 653 both want long-term and short-term rates considered, and the reason is that a single pair of numbers taken this way is not evidence of anything.

Where UT authority is actually consumed on this corridor

In the refinery and chemical units around Baton Rouge, Port Allen, Geismar and Plaquemine, ultrasonics is doing damage mechanism work rather than construction work. High temperature hydrogen attack in hydroprocessing service, judged against the Nelson curves of API RP 941, has moved from crude detection methods to advanced ultrasonic techniques whose results are only as good as the qualification behind them. Wet hydrogen sulfide cracking, hydrogen induced cracking and stress-oriented HIC in sour service demand shear wave and mapping techniques with a defined capability against ID-connected cracking. API RP 571 is the reference that connects the mechanism to the technique, and a UT procedure that has not been written with a damage mechanism in mind is a workmanship procedure being asked to do integrity work.

On the river terminals, API 653 tank work has its own pattern. Floor scanning is a magnetic flux leakage job with ultrasonic prove-up, shell course thickness is a straight beam job with a documented CML pattern, and critical zone and shell-to-bottom weld examination is where the shear wave technique earns its keep. Barge and hull gauging at the port and midstream fleeting sits under classification society and Coast Guard requirements rather than ASME, with its own reporting conventions and allowable diminution tables.

North of the city, nuclear in-service inspection under ASME Section XI operates in a different world entirely. Ultrasonic examination of the components covered by Appendix VIII requires performance demonstration through an approved program, inside a 10 CFR 50 Appendix B quality assurance system. Technicians move between that market and the refinery market, and the qualifications do not transfer in either direction. A written practice that treats a performance-demonstrated examiner as automatically qualified for refinery work, or the reverse, is misreading both systems.

What a UT certification does and does not cover

Certification is by method, by level and, in a well-written practice, by technique. A Level II in ultrasonics is qualified to set up the equipment, perform the examination to an approved procedure, interpret and evaluate against the acceptance criteria, and report the result. He is not qualified to write the procedure, to change an essential variable, or to decide that a technique the procedure does not cover is close enough.

Encoded and array techniques need to be handled explicitly. Phased array, TOFD, full matrix capture and automated corrosion mapping each involve setup decisions, focal law construction, wedge delay and velocity calibration, and data interpretation on imaged rather than A-scan displays. Practices that fold all of this into a generic UT certification are the ones that get challenged when a client's supplier quality auditor asks what the practical examination consisted of.

The other boundary is the eye and the ear of experience. A Level II who has spent five years on tank floors has a genuine skill set that does not automatically make him competent on austenitic weld metal at 2.25 MHz, where beam skewing and grain scatter change what the display means. The practical examination should reflect the work the technician will actually be sent to do, and the specimens should look like the plant.

Instrument and calibration housekeeping that fails audits

ASME Section V requires screen height linearity and amplitude control linearity checks on ultrasonic instruments at intervals not exceeding three months, or prior to first use thereafter, with records. This is one of the easiest findings an auditor can write, because either the record exists or it does not. Contractors running a fleet of instruments across multiple crews frequently cannot produce a current linearity record for the specific serial number that appears on a report.

Calibration blocks are the second housekeeping failure. The block must correspond to the component in material specification, product form and heat treatment condition, and blocks get borrowed, damaged and substituted on site. A report that names a block by number, tied to a block certificate held under control, is defensible. A report that says "IIW block" is not, because there is no way to establish afterwards which block was used or whether it was appropriate.

Couplant is the third, and it matters more here than in most regions because of the volume of stainless and nickel alloy equipment in the chemical corridor. Section V limits residual sulfur on nickel base alloys and residual halides on austenitic stainless steel and titanium, and the way you satisfy that is a couplant certificate on file matched to the batch in use. It costs nothing to hold and it is written up every time it is missing.

Scope, limits and how to engage

Atlantis supplies UT Level III authority in three forms. As a named Level III of record on the written practice, covering ultrasonics alongside whatever other methods are agreed. As procedure work only, developing and qualifying specific UT procedures with the demonstration records behind them. Or as independent review, where an owner wants a second technical read on a contractor's ultrasonic data, a disputed indication, or the sizing that is about to feed a fitness-for-service assessment.

What is not offered is the API inspector of record role. The authorized inspector signs the in-service inspection and sets the interval, and independence is worth nothing if the same party does both. Process safety management auditing is likewise outside scope; where mechanical integrity questions touch ultrasonics, we supply the technical answer the programme consumes rather than auditing the programme.

The work is led by an ASNT NDT Level III certified in multiple methods who also holds API 653 authorized inspector certification, so the interface between what the ultrasonic data says and what the in-service code does with it is familiar territory. For a Baton Rouge, Port Allen, Geismar, Plaquemine or Gonzales operation, or for an NDT contractor bidding into those plants, request a consultation at info@atlantisndt.com. Scope is written against your procedures and your equipment, and a quotation follows the scope.

Request a consultation