UT Level III Authority for New Orleans River, Marine and Fabrication Work
An ultrasonic testing Level III approves your UT procedures, writes and grades the UT examinations your technicians sit, and signs the technical judgements behind flaw sizing and thickness acceptance. In New Orleans that authority is exercised across barge and towboat fleets, river-corridor petrochemical units and offshore fabrication yards — three regimes with different acceptance criteria and different auditors.
Level III authority in UT is method-specific and narrower than most buyers assume. The Level III approves the written procedure and the calibration approach behind it, prepares and grades the general, specific and practical examinations, and certifies — on the employer's behalf — that a technician can do the work described. That certification is the employer's, not ASNT's, and it does not travel to the next employer. It also does not extend to techniques the practice never named: a technician certified for manual contact UT is not thereby qualified for encoded corrosion mapping, phased array or TOFD, and no Level III should sign as though he were. New Orleans sharpens the point. A thickness reading taken for a Coast Guard internal structural exam, a class survey, an API 653 tank evaluation and a fixed-equipment corrosion loop are four different obligations sharing one instrument, and the Level III is where the difference is supposed to be written down.
Source: Written against ASNT SNT-TC-1A and ANSI/ASNT CP-189 for personnel certification; ASME Section V Articles 4, 5 and 23 (including SE-797) for ultrasonic technique and thickness measurement; ASME Section VIII Division 1 and ASME B31.3 for construction and repair acceptance; API 510, API 570, API 653 and API 577 for in-service evaluation and welding metallurgy; IACS UR Z17 and 46 CFR Subchapter M where marine hull gauging is in scope.
| UT scope | What authorises the technique | What it does not authorise | Who accepts the result |
|---|---|---|---|
| Weld flaw detection, angle beam | A procedure approved by the UT Level III to ASME Section V Article 4, with technique sheets per joint | Corrosion mapping, encoded sizing, or acceptance outside the stated criteria | The authorised inspector or the customer's QA, against the construction code |
| Fixed-equipment thickness, straight beam | A thickness procedure to Section V Article 23 / SE-797 with recorded velocity, probe zero and surface condition | Determining remaining life or setting the next inspection interval | The API 510 or API 570 inspector of record |
| Tank shell and bottom thickness | Written procedure plus the API 653 evaluation basis the readings feed | Signing the tank inspection report or the repair recommendation | The API 653 inspector of record |
| Hull, barge and towboat gauging | Class-approved thickness measurement arrangements and the surveyor's agreed gauging scope | Crediting an internal structural exam on gauging alone | The class surveyor or the Coast Guard marine inspector |
| Encoded corrosion mapping, PAUT or TOFD | A separate technique qualification and a limited certification named in the written practice | Substitution for radiography without the referencing code's own route | Owner engineering plus the referencing code |
What a UT Level III is actually signing
When an inspection manager buys Level III support for ultrasonics, what changes hands is a short list of specific signatures. The Level III approves the ultrasonic procedure — the controlled document that fixes search units, calibration reflectors, scanning patterns, sensitivity, recording threshold and acceptance criteria across a stated range of material, thickness and geometry. He approves the technique sheets that adapt that procedure to a particular joint or component. He prepares, administers and grades the examinations that qualify your technicians. And where your written practice authorises it, he signs the certification decision itself.
None of that makes him the person who accepts the equipment, and the three roles get conflated constantly. The API 510, 570 or 653 inspector authorises and signs the in-service inspection. The NDT technician is certified under the employer's written practice to SNT-TC-1A or CP-189 and performs the examination. The ASNT Level III approves the procedures the technician works to, and the practice under which the technician is certified. A programme that blurs those three is the one that argues with an auditor about who was allowed to decide what.
The practical value of the arrangement is that ambiguity gets removed before a shutdown rather than during one. Most UT disputes are not about whether an indication is real; they are about whether the technique used was ever approved for that configuration, whether the sensitivity basis was transferable to the part, and whether the person holding the probe was certified for the scope. All three questions have documentary answers, and all three are the Level III's to have prepared in advance.
Straight beam, angle beam and two different calibration philosophies
Straight beam and angle beam are not variations of one examination. Straight beam looks along the thickness for laminar and planar reflectors normal to the sound path, and is the basis for thickness measurement and for lamination checks before an angle beam scan is credible. Angle beam interrogates a weld volume obliquely, and its whole geometry — skip distance, beam path, index point, the number of legs required to cover the root — is arithmetic that has to be worked out for the actual wall and the actual joint, not assumed from a wedge stamped with a nominal angle.
DAC and DGS answer the same question by different routes and fail differently. DAC is empirical: a curve drawn from real reflectors in a block, then corrected for the difference between block and component through a transfer measurement. Its weakness is transfer. Surface roughness, coating, curvature, material attenuation and temperature all separate the block from the part, and a DAC used without a recorded transfer correction is a curve that describes a block nobody is inspecting. Its second weakness is extension: once the last drawn point falls into noise, extrapolating DAC beyond it is guesswork dressed as calibration.
DGS is a model — an equivalent flat-bottom-hole size derived from the probe's own beam profile, with the diagram supplied for that probe as manufactured. Its strength is that it needs no side-drilled-hole block for every thickness. Its weakness is that it silently stops being true when the probe changes. A replaced wedge, a worn face, a substituted crystal or an off-catalogue frequency invalidates the diagram, and nothing on the instrument screen tells the operator. That is precisely the failure a Level III exists to prevent, by writing which probes may be used with which curve set and requiring verification at defined intervals.
The thickness readings that quietly go wrong
Ultrasonic thickness measurement looks like the simplest thing an NDT department does, which is why it produces the most consequential errors. The first is velocity. The instrument computes thickness from transit time and an assumed velocity; set for carbon steel and applied to stainless, a clad wall, a copper-nickel line or a cast component, it returns a confident number that is wrong by a known percentage nobody applied. A procedure that does not require velocity verification on a reference of the same material has already conceded the point.
The second is what the gate is actually seeing. In thin wall the instrument may lock onto a second backwall echo and halve or double the reading; in corroded wall it may read to the top of a pit shoulder rather than the pit floor; with a delay-line probe an unzeroed delay adds a fixed offset to every reading in the survey. In laminated or hydrogen-damaged plate the first strong echo is not the backwall at all, and the operator records the depth of a defect as the remaining wall.
The third is arithmetic done after the readings. API 510 and API 653 evaluations use measured minima and, in defined circumstances, averaging over a specified length — and the way a corroded region is averaged changes the answer materially. A crew that reports a spot minimum where the evaluation calls for an averaged thickness, or averages where the code wants the local minimum, produces a remaining-life figure that is defensible only until somebody checks. Deciding which convention applies, and writing it into the procedure so the field crew cannot choose, is Level III work.
New Orleans: one instrument, four different obligations
New Orleans concentrates an unusual amount of ultrasonic demand into a small geography, and the demand is not homogeneous. Barge and towboat fleets working the Lower Mississippi and the Gulf Intracoastal Waterway generate a constant stream of hull and structural gauging. The industrial corridor running upriver toward Baton Rouge holds refining, petrochemical and chemical process units with fixed-equipment corrosion programmes, tanks and pipe racks. Downriver, Plaquemines Parish LNG export construction and Port Fourchon's offshore service base drive fabrication and repair scope. Port of New Orleans terminals, grain elevators and dock steel add structural work. The Eighth Coast Guard District is headquartered here, which means marine regulatory attention is not remote.
Those are four acceptance regimes sharing one labour pool. Marine gauging for class survey and for a Coast Guard internal structural examination under 46 CFR Subchapter M answers to class rules and the surveyor's agreed scope, and gauging performed for class credit typically has to come through arrangements the society recognises — a supplier approval question, not a technician question. Process-unit thickness answers to API 510, 570 or 653 and feeds an inspector of record. New construction answers to ASME Section VIII or B31.3 and the customer's specification. The reporting threshold, the retention period and the person who accepts the number are different in each.
The failure this produces is predictable: a competent crew, correct instruments, and a report written to the wrong regime. A hull gauging report formatted like a corrosion-loop survey satisfies nobody at the survey table; a tank floor scan reported without the datum and orientation scheme an API 653 evaluation needs has to be repeated. Writing one procedure per regime, and certifying technicians against the scope they actually work, costs a fraction of one repeated mobilisation.
Where UT certification stops
A UT certification is bounded in three directions and buyers routinely test all three. It is bounded by employer: certification issued under one company's written practice does not transfer to another company. A new employer may accept prior training and experience as input, but the certification decision is theirs and has to be made through their own practice. Filing another firm's certificate and putting the technician to work is one of the most common findings in this market, particularly among contract crews moving between yards.
It is bounded by technique. A certificate that says UT covers what the practice and procedure say it covers. Encoded corrosion mapping, phased array, time-of-flight diffraction, guided wave screening and creeping wave techniques are not implied by a general UT certification, and each carries its own interpretation failure modes. Practices that handle this well use limited certifications or explicit technique qualifications with a documented demonstration; practices that handle it badly say nothing and rely on the assumption that UT means all of UT.
It is bounded by authority. A Level II certification authorises examination and interpretation to written criteria. It does not authorise accepting a condition outside those criteria, waiving a scan direction because access is poor, changing sensitivity to control noise, or deciding that an indication is geometry. Every one of those is a procedural decision, and every one of them shows up in field reports written by technicians who were never told they could not make it. The remedy is not more discipline; it is a procedure that anticipates the situation and a Level III reachable when it arises.
Sizing a flaw is a different problem from finding one
Detection and sizing are separate capabilities with separate demonstrations, and conflating them is where fitness-for-service work goes wrong. Amplitude-based criteria are designed to sort indications for acceptance or rejection against a reference level. They were never intended to produce the through-wall dimension an engineering assessment consumes. Feeding an amplitude-derived length and a guessed height into a fracture or remaining-strength calculation produces a number with a false precision that survives right up to the point somebody excavates the weld.
Height sizing needs a technique built for it — tip diffraction, mode conversion, TOFD, or a focused phased array setup — and needs to have been demonstrated on flaws resembling the ones being sized, in the same material, thickness and surface condition. The demonstration is the whole point: sizing accuracy is not a property of the equipment, it is a property of the technique plus the operator plus the geometry. A Level III who signs a sizing procedure without a demonstration record has signed an opinion.
This matters commercially as well as technically. When a sizing result triggers a repair, the cost of that repair is set by the number reported. Over-sizing forces excavation that was never needed; under-sizing leaves damage in service with a documented clearance. Both are expensive, and both are avoidable by deciding in advance which indications will be sized, by what technique, and against what demonstrated tolerance.
The audit findings that recur in UT programmes
Across shops and plants, UT programme audits converge on the same handful of findings. The procedure cites a superseded edition of the referencing code. Calibration blocks are not traceable, or are not of a material and heat treatment representative of the component. Transfer correction is required by the procedure and never recorded on the report. Instrument linearity verifications are performed but not retained, or retained without identifying the instrument. Technique sheets exist in the field at a revision the controlled procedure never issued.
On the personnel side: practical examinations graded with a score and no record of the specimen or what the candidate was asked to find; experience recorded as calendar employment rather than method hours; vision examinations undated or administered by someone the practice never authorised; contract technicians working under certificates the plant filed but never accepted through a written route. None of these are technical failures. All of them make the technical work unsupportable if it is ever challenged.
The finding that costs most is scope creep without documentation: the department bought a phased array instrument two years ago, uses it productively, and the written practice and procedure still describe manual contact UT only. Everything examined with that instrument since is, on paper, outside the programme. Correcting it retroactively is possible but laborious. Anticipating it costs a procedure revision and a demonstration.
How a Level III engagement runs
Most engagements begin with a gap review rather than a document delivery: what methods and techniques are actually in use, what the written practice says, what the procedures cover, and where the three disagree. That review is usually short and uncomfortable, and it sets the order of work — because writing a beautiful UT procedure for a company whose written practice does not authorise the method is effort spent in the wrong sequence.
From there the work is concrete: UT procedures and technique sheets to the referencing code, a calibration and sizing basis that matches the components in front of you, examination sets built to the topical outlines your practice adopts, practical specimens that reflect real scope, and certification recommendations the employer signs. Where an outside Level III is named in the practice, the practice must also state what that person decides and what remains the employer's — an ambiguity auditors find quickly.
Ongoing support is where the value compounds: a reachable technical authority when a field crew hits geometry the procedure does not cover, review of contested reports before they reach the customer, and representation at the table when a surveyor, an authorised inspector or a client's technical authority reads your records. To discuss scope, request a consultation at info@atlantisndt.com.
What exactly does a UT Level III approve that a Level II cannot?
The Level III approves the written UT procedure and every technique sheet derived from it, and selects and validates the calibration basis — reference block, DAC construction, transfer correction, DGS curve set, recording threshold. He also prepares, administers and grades the general, specific and practical examinations behind certification. A Level II performs the examination and interprets results against the procedure's acceptance criteria. He cannot write those criteria, approve a deviation from them, or authorise a technique the procedure does not describe.
Does an ASNT Level III certificate by itself certify our technicians?
No. An ASNT Level III certificate is evidence that one individual holds Level III knowledge in a method. Certification of your technicians is issued by you, the employer, against your own written practice, and the Level III's job is to make that defensible — approving the practice, building and grading the examinations, and signing the technical judgement. If your practice names an outside Level III, it must also state precisely which decisions that person is authorised to make.
Is API 510, 570 or 653 inspector training part of this offer?
No. Those are inspector certifications administered by API, and they authorise a named individual to sign in-service inspection of pressure vessels, piping and storage tanks. What is supplied here is NDT technical authority instead: the ultrasonic procedure, the calibration and sizing basis, the personnel certification programme under your written practice, and independent review of the data an API inspector relies on. The inspector of record stays yours, and so does the signature.
When does DGS sizing stop being valid on a job?
DGS is a model tied to one probe's beam profile, so it fails the moment the probe stops matching the diagram it was built from — a re-shoed wedge, a worn contact face, a replacement crystal, or a frequency substitution made on the barge deck at midnight. It is equally invalid where material attenuation and grain structure differ from the model's assumptions, which is why coarse-grained, clad and austenitic components usually revert to DAC with a transfer correction that is actually recorded.
Our technicians are certified in UT — can they run encoded corrosion mapping?
Only if your written practice and procedure say so and they have demonstrated it. Manual contact UT and encoded corrosion mapping share a method designation and very little else: positional encoding, index resolution, gate and gain strategy, C-scan interpretation and data-file integrity are separate skills with separate failure modes. Most sound practices handle this with a limited certification or a documented technique qualification. Silence in the practice is the finding an auditor writes down first.
Why does New Orleans work need more than one acceptance basis?
Because one crew can cross regimes inside a single week. Hull, barge and towboat gauging answers to class rules and Coast Guard requirements; river-corridor process units answer to ASME and API; new fabrication answers to the construction code and the customer's specification. The instrument and the technician do not change. The reporting thresholds, the minimum-thickness arithmetic, the retention requirements and the person who accepts the result all do — and each of those belongs in a different procedure.