What a UT Level I May Do, and What the Written Practice Reserves for Level II
A UT Level I calibrates the instrument, performs specified ultrasonic tests and records results under the direction of a Level II or III. It may accept or reject only where written instructions state the criteria, never on its own judgement. Training hours, experience by method, examination content and the exact scope of the level are fixed by the employer written practice.
Ultrasonics is the method Port Arthur runs on, because the dominant question along the Sabine-Neches corridor is how much wall is left. A Level I on this coast spends most of a shift on thickness: grid readings at condition monitoring locations, dual-element probes on corroded external surfaces, echo-to-echo readings through coating, and long walkdowns of piping circuits producing data someone else will evaluate. Shear-wave work comes next, usually building a distance amplitude correction curve on a basic calibration block before a Level II scans welds. The level is genuinely narrow, and deliberately so: you produce data whose quality can be independently checked, and you do not decide whether an indication is acceptable unless a written instruction has already decided it for you. Understanding exactly where that line sits is what separates a Level I who gets kept on a turnaround crew from one who does not.
Source: Written against ASME BPVC Section V Article 4 and Article 5, ASME B31.3, AWS D1.1 ultrasonic provisions, ASTM E797 and ASTM E317, ASNT SNT-TC-1A (2020) and ISO 9712.
| Task | Level I | Level II | Governed by |
|---|---|---|---|
| Set instrument velocity and zero on a step wedge | Yes | Yes | Written instruction, ASTM E797 practice |
| Build a DAC curve on a basic calibration block | Yes, under a written instruction | Yes, and verifies it | ASME Section V Article 4 |
| Record thickness readings at fixed CMLs | Yes, this is core Level I work | Yes | Owner inspection programme |
| Decide whether a weld indication is acceptable | No | Yes | Construction code, e.g. B31.3 or AWS D1.1 |
| Select the technique, probe angle and frequency | No | Yes, within an approved procedure | Procedure written or approved by Level III |
| Write or approve the ultrasonic procedure | No | No | Level III certifying authority |
| Sign the final examination report | No | Yes | Employer written practice |
The Level I boundary, stated precisely
Most descriptions of Level I get one clause wrong. SNT-TC-1A describes a Level I as qualified to set up and calibrate equipment, conduct tests, and perform evaluations for acceptance or rejection according to written instructions, recording the results, while receiving the necessary instruction and supervision from a certified Level II or Level III. The phrase that carries the weight is according to written instructions. A Level I is permitted to apply a criterion that someone else has already fixed, on a specific task, in a specific form.
What that excludes is everything requiring judgement about the examination itself. A Level I does not choose the probe angle or frequency, does not decide that a technique is inadequate for the geometry in front of them, does not characterise an indication as crack-like or geometric, does not size it, and does not sign the examination report. Those belong to Level II. Writing or approving the ultrasonic procedure belongs to Level III.
This matters commercially, not just administratively. When an owner's representative in Port Arthur asks who performed and who evaluated an examination, the answer has to be consistent with the certificates on file. A report showing a Level I as the evaluating technician on a weld examination is a finding, and it can invalidate the examination, which on a turnaround means rescanning at your own cost.
Velocity, refraction and the numbers a Level I should hold
Longitudinal waves travel through carbon steel at roughly 5,900 metres per second, shear waves at roughly 3,230, so a shear wave arrives at about fifty-five per cent of longitudinal speed. Water sits near 1,480 and the acrylic used in angle-beam wedges near 2,730. These four numbers explain most of what an instrument does. If the velocity entered into the gauge is wrong, every reading is wrong by the same proportion, and it is the single most common systematic error in thickness work.
Refraction at the wedge interface follows Snell's law, and the consequence is the first and second critical angles. Between them, only a shear wave persists in the steel, which is why standard weld inspection wedges are cut for forty-five, sixty and seventy degrees rather than for arbitrary angles. Below the first critical angle both longitudinal and shear waves exist in the part at once, and the resulting confusion of signals is why that region is avoided rather than exploited.
Wavelength is velocity divided by frequency, and it sets what you can resolve. Raising frequency shortens the wavelength, improves resolution and shortens the near field, but increases attenuation and scattering, which is why coarse-grained or cast material forces you down in frequency and why a technician who reaches for high frequency on austenitic weld metal gets a screen full of grain noise. Near field length grows with the square of the crystal diameter, and within it amplitude varies so erratically that reflector comparison is unreliable.
Calibration is the Level I deliverable
The one thing an employer genuinely buys from a Level I is a calibration that another technician can trust. For thickness that means velocity and zero set on a step wedge of the same material, verified across the thickness range you will actually measure, with the verification recorded and repeated at the interval the procedure states. For angle beam it means index point, refracted angle and range set on a standard block, then the reference sensitivity established as the procedure directs.
Under ASME Section V Article 4, the basic calibration block for weld examination is made from material acoustically similar to the component, with a comparable surface finish and heat treatment, containing side-drilled holes and notches sized to the thickness range being examined. The point of those requirements is not bureaucracy. A calibration performed on a block that behaves differently from the part produces a sensitivity setting that is simply wrong for the part.
Structural work adds another convention. The ultrasonic provisions of AWS D1.1 use an indication rating in decibels, derived from the indication level, the reference level and an attenuation factor computed from the sound path. A Level I working structural steel around the terminals and racks in Port Arthur will be asked to record those numbers correctly long before they are asked to interpret them, and arithmetic errors in that rating are among the most common report defects.
DAC, TCG and DGS: three answers to attenuation
Amplitude from an identical reflector falls as the sound path lengthens, through beam spread and material attenuation. A distance amplitude correction curve handles this empirically: peak the response from equal side-drilled holes at increasing metal paths, mark each peak, and join them. Everything is then judged relative to that curve rather than to a fixed screen height. Time corrected gain does the same job by applying increasing gain with time so the curve becomes a flat line, which is easier to read but hides what is happening underneath.
The distance-gain-size approach is calculated rather than measured. A diagram specific to the probe relates the echo from a flat-bottomed reflector of given size to distance and gain, referenced to a backwall or another known reflector. It saves blockwork and gives an equivalent reflector size directly, but it is only valid if the probe genuinely matches the diagram and the reference is set correctly, and it assumes a reflector geometry that real flaws rarely have.
Whichever is used, transfer correction is the step most often skipped. The calibration block is smooth, flat and clean; the pipe in the field is rough, curved, coated and hot. The difference in coupling and attenuation between the two is measured in decibels and added to the reference sensitivity. Omitting it means scanning at lower effective sensitivity than the procedure requires, and it is invisible in the report unless the transfer value is recorded.
Thickness work: where a Port Arthur Level I earns their keep
The bulk of Level I ultrasonic employment on this coast is corrosion monitoring: repeat readings at fixed condition monitoring locations so that wall loss trends can be calculated. The discipline that makes the data valuable is unglamorous. Locate the point exactly where it was located last cycle, prepare the surface consistently, take several readings rather than one, record the minimum honestly, and note the surface condition and temperature.
Dual-element probes are standard for corroded surfaces because the separate transmit and receive crystals with their roof angle suppress the near-surface dead zone and cope with rough, irregular backwalls. They carry a V-path error that matters most on thin sections, which is why calibration across the actual thickness range is required rather than a single-point calibration. Single-element probes give better accuracy on clean material and are the tool for precision work on machined surfaces.
Two failure modes are worth memorising. Doubling occurs when the gauge locks onto the second backwall echo on thin material and reports twice the true thickness, and it is dangerous precisely because the number looks plausible. Coating error occurs when a standard-mode reading through paint adds apparent wall; echo-to-echo mode eliminates it. Elevated temperature adds a third: sound slows as steel gets hotter, roughly one per cent per hundred degrees Fahrenheit as a working rule, so a hot line read against an ambient calibration reads thick unless the correction is applied.
Couplant and surface condition: the variables that quietly ruin data
Ultrasound will not cross an air gap, so the couplant is not an accessory. Too little and the signal drops out intermittently, which on a corrosion survey looks like pitting. Too much on a delay line changes the delay. Viscosity has to suit the orientation, since water-based couplant runs off a vertical line and gel does not, and high temperature work needs a couplant rated for the surface it will sit on rather than one that boils off.
Chemistry matters on specific materials. Specifications routinely limit residual sulphur and halogen content in couplant used on austenitic stainless and nickel alloys, because residues left in service contribute to stress corrosion cracking later. On the LNG and cryogenic piping in this region that limit is not advisory. A technician who uses the general-purpose drum because it was nearest has created a materials problem that will not surface for years.
Surface condition sets the ceiling on data quality. Loose scale, weld spatter, flaking paint and heavy roughness all scatter energy and lift the noise floor. Curvature on small-diameter pipe means a flat probe contacts on a line rather than a face, so a curved shoe or a smaller element is needed. None of these are exotic problems, and all of them are recorded on a good report so the person evaluating the data knows how much to trust it.
Training, examination and the practical you will actually sit
Qualification runs through three examinations. The general covers the physics and the instrument. The specific covers your employer's own equipment, written procedures, techniques and report forms, which is why it cannot be bought off the shelf without editing and why generic specific examinations are a routine audit finding. The recommended grading rule is a composite of at least eighty per cent with no individual examination below seventy, so a strong general result does not rescue a weak specific one.
The practical examination is the one candidates underestimate. It is not a demonstration that you can hold a probe. Expect to be handed an instrument you did not set up, asked to calibrate it correctly for a stated task, to run the examination on prepared specimens, and to record results on the employer's own form, with the assessment covering a series of checkpoints across the variables a technician must control. Poor handwriting on a data sheet has failed candidates who scanned perfectly well.
Under ISO 9712 the same three-part structure applies, but the examination is set and marked by a certification body through an approved examination centre, experience is expressed in months of documented industrial work, and the resulting certificate belongs to you rather than to your employer. Which route suits you depends on whether you intend to stay with one Gulf Coast contractor or to work internationally.
Where UT Level I work exists in and around Port Arthur
Port Arthur sits at the mouth of the Sabine-Neches waterway with one of the densest concentrations of refining capacity in North America, including the Motiva, Valero and TotalEnergies refineries, along with associated chemical units, coke handling, tank farms, marine terminals and the LNG export activity at Sabine Pass. Nearly all of that infrastructure is carbon steel in corrosive service, which is exactly the condition that generates continuous ultrasonic thickness work.
The demand splits into a steady base and a seasonal peak. The base is in-service corrosion monitoring across piping circuits, vessels and tankage, running all year on inspection intervals, and it is where most Level I positions actually sit. The peak is turnaround season, when crews mobilise across Jefferson and Orange counties and into Calcasieu Parish, and when a Level I who can produce clean, repeatable data at volume is genuinely valuable to a crew whose Level IIs are fully committed to weld evaluation.
Beyond the plants there is marine and barge repair on the waterway, structural steel in racks and terminal infrastructure, and pipeline and storage facilities feeding the export terminals. Each brings a slightly different governing document and a different reporting convention, and each expects the same underlying calibration discipline. A technician who can move between them without confusing the acceptance frameworks becomes hard to replace.
Getting from Level I to Level II without stalling
The Level I to Level II transition fails for a documentation reason far more often than for a technical one. Level II requires accumulated ultrasonic experience, and if nobody logged your hours by method against job numbers, the hours effectively did not happen. Start a signed log on your first day, have your supervising Level II or Level III countersign it monthly, and keep a personal copy. It takes minutes and it is the difference between a straightforward step up and an argument you cannot win.
Technically, the step up is from producing data to defending it. A Level II is expected to interpret and evaluate against a code or specification, to recognise when geometry rather than a flaw is producing a signal, to size an indication, and to write a report that stands on its own. Deliberately learning the acceptance criteria you are already collecting data against, while still at Level I, shortens that transition considerably.
Atlantis delivers ultrasonic training and certification preparation at Level I, Level II and Level III to ASNT SNT-TC-1A and ISO 9712, extending into phased array and time-of-flight diffraction, alongside RT, MT, PT, ET and VT, in classroom, on-site corporate and blended formats. Employers can also have the specific examination, practical specimens and reporting forms built around their own procedures. Training is affordable, accessible and fully customisable; arrange a consultation or a quotation through info@atlantisndt.com.
What is a UT Level I actually allowed to do?
Set up and calibrate the instrument, perform the specified test, and record the results, all under the direction of certified Level II or Level III personnel. SNT-TC-1A allows a Level I to make acceptance or rejection determinations where a written instruction supplies the criteria, which is narrower than it sounds: someone else has already decided what acceptable means. Selecting technique, sizing and characterising indications, and signing the report remain above the level.
How much training and experience does UT Level I require?
Ultrasonics carries one of the heavier classroom requirements in the recommended tables, and the experience requirement is counted in two ways at once: minimum hours performing ultrasonic testing, and minimum total hours across NDT generally. The exact figures depend on which document governs, on your education tier, and on what your employer wrote into its own practice. Under ISO 9712 the same requirement is expressed in months of documented industrial experience rather than hours.
What do the Level I examinations test?
Three examinations. The general covers ultrasonic principles: velocity, wavelength, refraction and mode conversion, near field, attenuation, transducer construction and instrument controls. The specific covers the employer's own equipment, procedures, techniques and reporting forms. The practical is hands-on, typically framed as a set of checkpoints across one or more specimens, requiring the candidate to calibrate, scan and record correctly. A composite grade of at least eighty per cent is recommended, with no individual examination below seventy.
Why does a coated surface change a thickness reading?
Sound travels slower in paint or coating than in steel, so a single-element gauge in standard mode measures the coating and the steel together and, because it applies the steel velocity to the whole transit time, reports a wall thicker than it is. Echo-to-echo mode measures between successive steel backwall echoes and ignores the coating entirely. Choosing the wrong mode on a heavily coated line is one of the most common data errors on this coast.
What is the difference between DAC and DGS?
Both compensate for amplitude falling with distance. DAC is empirical: you record peak responses from identical reflectors at increasing metal paths in a reference block and draw the curve through them, then apply a transfer correction for the real component. DGS is calculated: a probe-specific diagram relates equivalent reflector size to distance and gain, referenced to a backwall or known reflector, so it needs less blockwork but depends on the probe matching its diagram.
Does a UT Level I certificate transfer to another employer?
Not automatically. Under SNT-TC-1A certification is employer-based, so leaving a Port Arthur contractor effectively ends the certificate. What transfers is the evidence: your training record with hours and content, your examination score sheets, and a signed log of ultrasonic hours by job. The next employer's written practice decides whether it certifies you on review of that evidence or re-examines you. Keep personal copies of everything from day one.