UT Level III Authority for Beaumont Refining and Petrochemical Assets

An ultrasonic testing Level III approves your UT procedures, qualifies the techniques behind them, and certifies UT technicians under your written practice. In Beaumont, that authority is what makes thickness data on hot refinery piping defensible: it fixes the calibration block, the temperature correction and the coverage rule before a corrosion rate is ever calculated from a reading.

Beaumont inspection work is dominated by fixed equipment in hot hydrocarbon service, and most of the ultrasonic data taken there is thickness data feeding a corrosion rate. That makes the Level III decisions arithmetic decisions. A reading taken at 380 F against a block calibrated at 70 F reads roughly three percent thick, so the wall looks healthier than it is and the remaining-life calculation inherits the error. Dual-element probes hide a second failure mode on thin walls, where the instrument locks onto a doubled backwall echo and reports twice the true wall. Angle beam work brings its own choice: whether the procedure builds a DAC curve on a block matched to the component material, thickness, curvature and heat treatment, or leans on a DGS diagram valid only for the probe it was drawn for. None of that is settled by the technician on the scaffold. It is settled, in writing, by the Level III who approved the procedure.

Source: Written against ASNT SNT-TC-1A and ANSI/ASNT CP-189; ASME Section V Articles 4 and 14; ASME Section VIII Division 1 and ASME B31.3; API 570, API 574, API 577, API 939-C and API 941; and the Texas Department of Licensing and Regulation boiler rules that reach fixed equipment in Jefferson County.

Technically reviewed by Anoop Rayavarapu — ASNT NDT Level III (UT, RT, MT, PT, VT, ET) · API 653 · ISO 9001:2015 Lead Auditor
Who signs what on a Beaumont fixed-equipment inspection
RoleQualified or certified bySignsDoes not sign
API 510, 570 or 653 inspectorAPI individual certification programme, held by the personThe in-service inspection report, the remaining life and the next inspection dateThe UT procedure or any technician certificate
ASNT NDT Level III in UTASNT examination, then named and scoped in the employer written practiceThe UT procedure, technique sheets, the written practice, personnel certification recordsThe in-service inspection or the fitness-for-service run
UT Level II technicianEmployer certification under SNT-TC-1A or CP-189, examined and recordedThe UT report: readings, indications, instrument settings, coverage achievedApproval of the procedure worked to, or a deviation from it
Owner integrity engineerOwner competency system, not an NDT certificationCorrosion rate acceptance, t-min, inspection interval, escalation decisionsNDT personnel certification or procedure approval
Third-party auditor or client QAClient or registrar appointmentAudit findings against the practice and the proceduresAnything that closes the finding
Two of the four documentation failures we see most often in the Golden Triangle come from one person signing across two of these rows without holding what that row requires.

What a UT Level III actually signs, and why the signature matters

Three roles get conflated on almost every unit in Jefferson County, and the confusion is expensive. The API 510, 570 or 653 inspector authorises and signs the in-service inspection of the vessel, the piping circuit or the tank. The NDT technician is certified under the employer written practice, to ASNT SNT-TC-1A or ANSI/ASNT CP-189, to operate the instrument and report what the screen shows. The ASNT Level III sits behind both of them: he approves the ultrasonic procedure the technician works to, approves the written practice under which that technician was certified, and answers for both when a client or a regulator asks who decided the technique was adequate.

In ultrasonic testing the approval covers a short and unforgiving list. Which reference block the calibration is built on and whether it represents the component. Whether the DAC curve or DGS diagram applies to the material being examined. The scanning gain and the transfer correction. The scan plan and the volume it genuinely covers rather than the volume it claims. The temperature range over which readings may be taken and the correction applied outside it. And the qualification of the individual for the specific technique, which is not the same thing as certification in the method.

Atlantis supplies that technical authority and nothing wider. It is not the API inspector of record and it does not perform process safety management audits. The engagement is written practice, procedure development and qualification, personnel certification within the practice, independent review of inspection data, and representation on technical questions when an audit reaches them.

Hot-service thickness readings and the correction nobody records

Most ultrasonic work on a Beaumont operating unit is straight-beam thickness measurement on lines that are not coming down. Velocity in carbon steel falls as the steel heats, so a probe on a 380 F line calibrated against a block at ambient sees a longer transit time, and the instrument converts it using the cold velocity. The readout is high by roughly one percent per 100 F of difference. Three percent of a 0.375 inch nominal wall is about ten thousandths of an inch, the same order as a year of general corrosion, and it is reported in the direction that makes the equipment look healthier than it is.

The second failure mode is instrument behaviour rather than physics. Dual-element probes carry a V-path error that grows as the wall thins, and on thin or heavily pitted walls the instrument can lock onto the second backwall echo and report a doubled thickness. High-temperature couplant burns off within seconds, so contact time is limited and the operator is tempted to accept the first stable number rather than the lowest repeatable one. A procedure that does not state the couplant, the maximum contact time, the minimum measurable wall and the doubling check has delegated all four to whoever is on the scaffold that morning.

Then the data meets arithmetic. API 570 expects both a short-term and a long-term corrosion rate, and the short-term rate is dominated by measurement noise when the interval is short. The trap we see most often is location drift: a re-read taken an inch from the original TML centre on a corroded elbow returns a different wall, the difference is booked as metal loss, and a corrosion rate that never existed drives an unnecessary interval cut, or worse, masks a real one elsewhere. The Level III closes that by specifying TML marking, repeat-reading rules and the minimum-of-N convention in the procedure, not in a toolbox talk.

DAC, DGS and the calibration philosophy the procedure has to choose

Angle beam examination needs a reference sensitivity, and the choice between a distance-amplitude correction curve and a distance-gain-size diagram is a procedure decision with real consequences. A DAC curve is empirical: it is built on a block whose material specification, product form, thickness, surface curvature and heat treatment represent the component, using side-drilled holes at the depths the beam will actually traverse. Where any of those attributes diverge from the part, the curve is decorative rather than diagnostic, and that is precisely the finding an auditor writes up.

Transfer correction is the step that gets dropped. The block and the component rarely share surface finish, coating, curvature or grain structure, and the resulting difference in coupling and attenuation is measurable in minutes with a two-probe pitch-catch comparison on both. European practice generally requires the correction to be applied once the difference exceeds a stated threshold, commonly two decibels, and requires it to be recorded. Skipping it means the examination ran at a sensitivity nobody can reconstruct afterwards, which quietly voids the acceptance decision it produced.

DGS is the other philosophy: a theoretical curve set relating echo amplitude to the equivalent size of a flat-bottomed reflector, valid only for the probe it was published for and only after attenuation and transfer corrections are applied. It is efficient on fine-grained, well-characterised material. It is not trustworthy in coarse-grained austenitic weld metal, in nickel-alloy filler, in castings, or inside the near field. Deciding which philosophy the procedure adopts, and writing down where it stops being valid, is exactly the Level III judgement a client cannot buy from a technician.

What a UT certification does not cover

A Level II certificate in ultrasonic testing is not a licence for everything ultrasonic. Encoded corrosion mapping, sectorial phased array scanning and TOFD each carry their own equipment, their own calibration and their own interpretation traps, and codes expect demonstrated ability on the technique in use. If the written practice does not name them, and the personnel file does not show separate training, examination and practical demonstration, the resulting data has no qualification standing behind it. This is the single most common gap we find in contractor files during a Golden Triangle turnaround.

Certification also does not carry across the two very different jobs ultrasonics performs. Measuring remaining wall and sizing a planar flaw through wall are different disciplines with different error structures. Sizing accuracy has to be demonstrated against known flaws in representative material and then stated as a tolerance in the procedure. A technician who is excellent at corrosion surveys may never have sized a crack tip, and a procedure that assumes otherwise produces a number that a fitness-for-service assessment under API 579 will then treat as fact.

Finally, acceptance criteria are code-specific and the practice has to scope them. The same indication is evaluated differently under ASME Section VIII, ASME B31.3, API 1104 and AWS D1.1, and a technician certified in the method is not automatically competent in the acceptance standard the contract invokes. The Level III decides which codes fall inside the certification, and that scope belongs on the certificate itself rather than in somebody's memory.

Beaumont: what the local asset base asks of ultrasonics

Beaumont sits inside the largest refining concentration in North America. The ExxonMobil Beaumont refinery and its adjacent chemical plant, the Port Arthur cluster operated by Motiva, TotalEnergies and Valero, the Golden Pass LNG export terminal at Sabine Pass, Port Arthur LNG, and the Port of Beaumont with its marine and heavy fabrication trade all draw on the same pool of inspection contractors and the same pool of technicians. The equipment population is old, hot, sour in places, and constantly being re-rated or revamped.

Two local realities drive Level III workload. First, coastal humidity makes corrosion under insulation a permanent programme rather than a campaign, which pushes work toward screening techniques and toward hard decisions about when ultrasonic confirmation is required after a screening hit. Second, turnaround labour arrives in waves: travelling technicians turn up holding certificates issued by other employers, which cannot simply be honoured, so the receiving employer needs a written practice and a Level III capable of certifying people quickly and correctly under schedule pressure.

The regulatory frame is layered rather than single. Fixed equipment is designed to ASME Section VIII and ASME B31.3, boilers touch Texas Department of Licensing and Regulation rules, intrastate pipeline segments answer to the Railroad Commission of Texas, and covered processes sit inside OSHA process safety management under 29 CFR 1910.119, where inspection and testing must be performed by qualified personnel following recognised and generally accepted good engineering practice. The written practice and the UT procedure are where the NDT half of that obligation is actually evidenced.

Sulfidation, HTHA and why the coverage rule beats the reading

High-temperature sulfidation is the damage mechanism that made this industry stop trusting single-point thickness readings. Carbon steel components with low residual silicon corrode substantially faster than adjacent components of the same nominal specification, so a piping circuit can be uniform on paper and wildly variable in the rack. API 939-C exists because of that. A procedure authorising one reading per elbow will eventually place that reading on the thick side of a component thinning on the extrados, and no amount of instrument accuracy rescues it. Component-by-component scanning coverage is the countermeasure, and it is a procedure decision, not a field decision.

High-temperature hydrogen attack is the mirror problem: the damage is volumetric micro-fissuring rather than wall loss, and conventional pulse-echo thickness work is insensitive to it until it is well advanced. API 941 sets the operating envelope; early detection leans on advanced ultrasonic approaches such as full matrix capture with total focusing, velocity ratio measurement, backscatter and spectral analysis. Each needs a qualified technique, a representative demonstration specimen and an explicit statement of what a negative result actually means. A Level III who cannot state the detection limit of the technique should not be signing the report that clears the vessel.

In both cases the Level III value is not in reading the screen. It is in writing the coverage rule and defending it: how many readings, over what area, on what grid, at what scan speed and index offset, with what escalation trigger when a reading falls below the previous minimum. That rule is what an auditor compares the field data against, and it is the first document a client asks for when a leak has already happened.

Working with an outside Level III, and what to ask before you appoint one

Most Beaumont-area service companies, and several owner organisations, do not employ a full-time Level III in every method they sell. Using an external Level III is entirely normal and entirely acceptable, provided the written practice names the individual, states the methods and levels he is authorised to act in, and defines how his review and approval are recorded. The failure mode is scope creep: a Level III certified in three methods quietly signing certificates in a fourth, which an auditor finds in ten minutes by comparing the certificate list against his own credentials.

Five questions separate a usable arrangement from a signature service. Which methods and techniques are you certified in, and can I see the evidence. Who writes the examinations my technicians will sit, and how is the question bank controlled and refreshed. What is your response time during a turnaround, and will you attend site or review remotely. Will you sit in the room when a client audits the practice you approved. And what happens to my personnel certifications if our arrangement ends, which is the question almost nobody asks until it matters.

Engagements are usually shaped as a procedure package, a written practice build or rewrite, ongoing certification support, or independent review of inspection data a client has challenged. Atlantis is affordable, accessible and fully customisable in that scope, and the work is defined after a technical conversation rather than picked off a menu. To start one, contact info@atlantisndt.com and describe the methods, the asset population and the code your contract invokes.

Does an ASNT Level III certification in UT automatically cover phased array and TOFD?

No. Phased array and time-of-flight diffraction are techniques inside the ultrasonic method, and codes expect personnel to demonstrate ability on the technique actually used, not merely on the method. Your written practice must state whether PAUT, TOFD and encoded corrosion mapping are separately trained, separately examined and separately recorded. If the practice is silent, an auditor reads that silence as a gap and treats the encoded data as unqualified.

How much error does an uncorrected hot thickness reading carry?

Sound velocity in carbon steel falls as temperature rises, roughly one percent per 100 F above the calibration temperature. Because the instrument keeps applying the cold velocity to a longer transit time, the readout comes out high. A 380 F line calibrated at 70 F reads about three percent thick. The error is non-conservative: it hides metal loss, and it compounds whenever the previous survey was taken at a different metal temperature.

When is a DGS diagram acceptable instead of a DAC curve?

Only where the referencing code permits it, the diagram belongs to the exact probe in use, and attenuation plus transfer-loss corrections have been applied and recorded. DGS is a theoretical amplitude model built for fine-grained, low-attenuation product. It degrades badly in coarse-grained austenitic weld metal, in nickel-alloy filler, in castings and close to the near-field limit, which is where a lot of refinery angle beam work actually sits.

Can a travelling technician certified by another employer work under our practice during a turnaround?

Not by transfer. Employer-based certification belongs to the employer that issued it and does not move with the person. The receiving employer must certify the individual under its own written practice: verify documented training and experience, administer the required examinations, confirm the vision examination is current, and record all of it before the technician touches an asset. Turnaround surges are exactly when that record gets skipped and later gets found.

Who carries the consequence when a corrosion rate computed from UT data proves wrong?

Practically the owner does, because the inspection interval was set from it. Technically the trail runs backwards: the integrity engineer used the rate, the API inspector accepted the data, the technician took the readings, and the Level III approved the procedure and the coverage rule that decided how many readings were taken and where. A defensible procedure is the point at which that chain stops moving backwards.

Is API 510, 570 or 653 inspector training part of this offer?

No. Atlantis does not sell API inspector certification training and does not act as the API inspector of record. The scope here is NDT technical authority: written practice, procedure development and qualification, personnel certification within your practice, independent review of inspection data, and representation on technical questions when a client or registrar audit reaches them. Consultation and a written scope are available on request.

Request a consultation