Radiography in Refining, Petrochemical and Upstream Service

What this industry actually radiographs

Oil and gas radiography is overwhelmingly pipe work, and that single fact reshapes the method. A turnaround scope will hand a radiographer two-inch socket welds on instrument tubing, six- and eight-inch process piping tie-ins, nozzle-to-shell welds on vessels, and occasionally heavy-wall reactor or drum seams, all inside the same week. Each of those is a different source, a different technique, and a different geometric problem. A radiographer whose experience is plate butt welds on structural steel has met perhaps one of the four.

New construction and repair drive most of the film. Replacement spools fabricated during an outage, weld repairs on excavated cavities, and tie-ins joining new equipment into existing circuits all require volumetric examination, and where the geometry defeats an angle beam probe radiography is the only practical answer. That is precisely why gamma crews work nightshift during Gulf Coast turnarounds: the shot happens when the fewest people are in the area and the fewest crews are displaced.

A second and often unexpected category is in-service imaging. Tangential and profile radiography measures remaining wall on small-bore lines and images corrosion under insulation without stripping the jacket, and it is used to find erosion at elbows and tees where flow does the damage. This work is not weld examination at all. It produces a dimensional measurement, and the technician must understand that the acceptance question it feeds is a thickness calculation under API 570, not a weld acceptance clause.

Source and energy selection against real wall thicknesses

Source selection in this industry is a practical judgement made against the wall in front of you, the access you actually have, and the exposure time the site will tolerate. Iridium-192 does the bulk of the work because its useful range covers most vessel and heavy piping wall. Selenium-75 has grown in use on thinner sections because it gives better contrast there and its smaller exclusion boundary disrupts fewer adjacent crews, which on a congested unit is worth real money. Cobalt-60 and high-energy X-ray are reserved for heavy wall where nothing else penetrates in a reasonable time.

X-ray gives the best sensitivity on thin material, and where a tube head can be rigged it is the right answer. The constraint is rarely technical. On a live unit, rigging power, positioning a tube on a scaffold, and the physical size of the head frequently make a small gamma source the only workable option even when a tube would give a better radiograph. Understanding that trade-off, and being able to explain it to a client who asks why the sensitivity is not better, is part of the competence this sector expects.

The error technicians import from other sectors is choosing energy for the thin part of the shot rather than the thick part. On a casting, a valve body or a double-wall exposure the density variation limit is what fails the film, and it is set by the range of thickness in the area of interest, not by its average. Section V limits density in the area of interest to no less than minus 15 percent and no more than plus 30 percent of the density through the IQI body, and a shot chosen for the thin section blows straight through that ceiling.

The pipe problem: technique geometry in oil and gas

Single-wall single-image is the technique everyone learns and the one this industry can least often use. It requires the source on one side of the wall and the film on the other, which on closed pipe means getting the source inside. Where a crawler or a source placed through an open end can achieve that, a panoramic exposure examines an entire circumferential weld in one shot and gives excellent quality. That is the ideal case, and on operating plant it is frequently unavailable.

So double-wall techniques carry the work. Double-wall single-image places the source outside the pipe and interprets only the film-side weld, with the source-side weld deliberately projected away from the area of interest. Double-wall double-image, the elliptical technique, superimposes both walls in one image and is used on small diameters, typically with the source offset so the two weld images separate into an ellipse. Each has its own coverage requirement, its own minimum number of exposures around the circumference, and its own IQI rules, and the qualified procedure states which applies.

The recurring failure is a technician executing an elliptical shot on a job whose procedure specifies double-wall single-image, or vice versa, because the geometry looked similar. On a small-bore weld the two techniques produce different numbers of required exposures and different interpretable areas, and a shot count that does not match the procedure is a rejected radiograph regardless of image quality. Reading the technique sheet before rigging is not optional here; on a congested unit rigging twice costs a shift.

IQI selection, placement and the source-side rule

The image quality indicator is the evidence that the radiograph achieved the sensitivity the code demands. Hole-type IQIs are selected by the material thickness being radiographed and are designated by their own thickness, with the essential hole, typically the 2T hole, being the feature that must be visible. Wire IQIs are designated by set and by essential wire, and are often preferred on small-diameter and double-wall work where a plaque will not lie flat. Section V's tables determine which IQI and which essential feature apply, and using an IQI selected for the wrong thickness invalidates the shot.

Placement is where the industry loses the most film. The IQI belongs on the source side, because that is where it experiences the full material path and therefore genuinely represents the quality achieved through the section a flaw could lie in. Section V permits film-side placement when the source side is physically inaccessible, and this is common on pipe, but the code then requires a lead letter F, at least as high as the identification numbers, to appear adjacent to or on the IQI. A film-side IQI without an F is one of the most frequent rejections in the industry, and it is entirely self-inflicted.

Shims are the other detail that arrives from fabrication practice and gets misapplied. When the IQI is placed on the base metal adjacent to a weld rather than on the weld itself, a shim of radiographically similar material may be required so the total thickness under the IQI approximates the weld thickness including reinforcement. Technicians who omit the shim produce an IQI reading through less material than the area of interest, which overstates the sensitivity achieved. An auditor reviewing film for a client will find this, and it calls the whole exposure into question.

Density and geometric unsharpness: the numbers you are held to

Two families of numbers decide whether a radiograph is acceptable before anyone looks for a flaw. The first is density. For single film viewing, Section V requires transmitted density through the radiographic image of the IQI body and the area of interest to be at least 1.8 when the source is X-ray and at least 2.0 when it is a gamma source, with 4.0 as the maximum for either single or composite viewing. Within the area of interest, density may not fall below minus 15 percent or rise above plus 30 percent of the density measured through the IQI body. Those percentages are what fail shots on varying section.

The second is geometric unsharpness. Penumbra arises because the source has physical size, and the resulting blur is calculated as the source dimension multiplied by the distance from the source side of the object to the film, divided by the distance from the source to the source-side of the object. Section V caps it by material thickness: 0.020 inches for material under 2 inches, 0.030 inches for 2 through 3 inches, 0.040 inches for over 3 through 4 inches, and 0.070 inches for material greater than 4 inches. The only lever a field radiographer has is source-to-object distance, and increasing it lengthens the exposure by the square.

That trade-off is the heart of field radiography in a congested plant. A shorter standoff shortens exposure time and shrinks the exclusion boundary, which every night superintendent wants, and it degrades unsharpness, which the code will not allow past a limit. A radiographer who cannot calculate this in the field ends up either rigging shots that fail on unsharpness or standing at a needlessly long distance and blowing the schedule. Being able to do the arithmetic on the platform is a skill this industry actually pays for.

Acceptance: the clause that decides, and how it differs from your last sector

ASME Section V will not tell you whether an indication is rejectable, and technicians who look for the answer there waste time. Section V Article 2 governs technique and image quality and then defers to the referencing code. For pressure vessel construction that referencing code is ASME Section VIII Division 1: UW-51 for full radiography and UW-52 for spot, with Mandatory Appendix 4 covering rounded indications. In outline, cracks and zones of incomplete fusion or incomplete penetration are unacceptable outright, and elongated indications are limited by length against a threshold that scales with the material thickness.

For process piping the governing document is ASME B31.3, and Table 341.3.2 is the table a radiographer in a refinery should be able to find without help. It sets acceptance by weld type and by fluid service category, which means the same indication can be acceptable in normal fluid service and rejectable in severe cyclic service on the same site. For cross-country pipeline girth welds the document is API 1104, which handles the common weld discontinuity types with its own criteria and permits an alternative fracture-mechanics-based acceptance route in its appendix.

This is where sector transfer goes wrong. A radiographer arriving from structural steel is used to AWS D1.1, which treats acceptance differently again and distinguishes statically from cyclically loaded connections. One arriving from aerospace is used to engineering-approved acceptance embedded in a process specification with almost no interpretive latitude. Neither habit survives contact with B31.3. The discipline to identify the referencing code before interpreting the film, every time, is what separates a competent oil and gas radiographer from a competent radiographer.

Imaging service damage, not fabrication defects

Most radiography training is built around fabrication discontinuities: porosity, slag, lack of fusion, incomplete penetration, undercut, cracks. Those matter enormously on new construction and repair welds, and they are what the acceptance tables are written about. But a large share of what this industry actually looks for is damage that developed in service, and it does not resemble the training images. API RP 571 is the reference that catalogues these mechanisms for the refining industry, and reading it changes how a radiographer interprets what they are seeing.

The mechanisms that show up radiographically include preferential weld corrosion, where the weld metal or heat affected zone wastes faster than the base metal and produces a groove along the seam that a technician may read as a fabrication defect. Erosion and erosion-corrosion at elbows, tees and downstream of control valves produces localised thinning best imaged tangentially rather than through the weld. Hydrogen-induced and stress-oriented hydrogen-induced cracking in wet sour service produces stepwise features that are easily missed on a single orientation. Corrosion under insulation produces external wall loss visible on a profile shot without stripping the jacket.

The interpretive shift matters. A fabrication defect is compared to a table and accepted or rejected. Service damage is measured, trended and fed into a remaining-life calculation by an inspector working under API 510 or API 570. The radiographer's job in that second case is to produce an image someone can take a reliable dimension from, with the geometry and technique documented well enough that a shot taken two years later can be compared to it. Technicians who do not understand that their output feeds a trend produce images nobody can use again.

Radiation safety and the authorisation the law requires

Method certification and radiation safety authorisation are two separate credentials, and a technician needs both. RT Level I or II certification comes from the employer's written practice under SNT-TC-1A or CP-189, or from an ISO 9712 certification body, and it says you can perform and interpret the examination. It says nothing about your right to operate a radioactive source. That comes from the regulator: 10 CFR Part 34 under the NRC, or the equivalent rules of an Agreement State, which both Texas and Oklahoma are. Those rules require prescribed training, documented on-the-job training under a qualified radiographer, a demonstration of competence, and certification through a regulator-recognised radiographer certification programme.

The operational rules are specific and they are enforced. Personnel monitoring requires a direct-reading dosimeter, an alarming ratemeter and a personal dosimeter processed by an accredited processor, and all three are required, not any one of them. The radiography area boundary is established and posted where the radiation level would exceed the regulatory threshold, and it is surveyed, not estimated. Occupational dose limits under 10 CFR Part 20 apply on top of the licensee's own ALARA programme, which on most refinery sites is tighter than the regulation.

The single most important habit is the post-exposure survey. Every exposure ends with a survey using a calibrated meter to confirm the source has fully returned to the shielded position, performed on the way back to the camera, not from a distance. The classic industrial radiography overexposure incident is a source that did not fully retract into a camera whose crank and guide tube looked normal, approached by a radiographer who assumed retraction because the crank reached its stop. Assuming instead of surveying is how radiographers get hurt, and every serious incident review in this industry says the same thing.

Digital, computed and film: what changes and what does not

Computed radiography using phosphor imaging plates and digital detector arrays have taken significant ground in this industry, driven by turnaround schedule pressure. Eliminating darkroom processing shortens the cycle from exposure to interpretation dramatically, and on a critical-path tie-in that time is the whole argument. Digital detectors also offer a wider dynamic range, which helps considerably on the varying-section shots that fail film on density variation.

What does not change is the physics or the code framework. Source and energy selection, geometric unsharpness limits, IQI requirements and acceptance criteria all continue to apply. What changes is how image quality is demonstrated: spatial resolution and contrast sensitivity are evaluated against the requirements the applicable article and the qualified procedure set for the digital technique, rather than against a film density measurement. Technicians who assume digital removes the quality requirements produce images that fail review for different reasons than film does.

There is also a records dimension the industry has been slow to absorb. Digital images are data, and data needs a retention scheme, a controlled viewing environment, and protection against post-acquisition manipulation. Owners are increasingly specific about file format, storage and audit trail, because an image that cannot be shown to be the original is not evidence. A radiographer moving to digital should expect the client's requirements on data handling to be as detailed as their requirements on exposure technique.

What technicians get wrong when they arrive from another sector

The pattern is consistent enough to be predicted. Radiographers arriving from structural fabrication know how to produce a clean image on accessible plate and struggle with double-wall geometry, with the shot counts elliptical technique requires, and with the fact that B31.3 acceptance changes with fluid service. Those from aerospace bring excellent discipline and traceability and find the interpretive latitude of a piping code uncomfortable. Those from casting work bring good instincts about density variation and underestimate radiation safety on a congested operating unit.

The common root is the same in every case: the physics transfers and the framework does not. Every one of those technicians can expose and process a radiograph. What they have to learn is which document governs the job in front of them, how that document's acceptance criteria differ from the one they know, what a live process unit does to access and to source selection, and what service damage looks like when it is not a fabrication defect. That is a matter of weeks of directed training, not years, but it does not happen by osmosis.

Atlantis delivers radiographic testing training and examination preparation to SNT-TC-1A and ISO 9712 at Levels I, II and III, in classroom, on-site corporate and blended formats, built around oil and gas geometry, codes and damage mechanisms rather than generic method theory. ASNT Level III consulting is available for employers who need a Level III of record to author a written practice, build and grade examinations and stand behind the personnel programme under client audit. Atlantis does not deliver API 510, 570 or 653 inspector certification training and is not the API inspector of record on any equipment. Demonstrations, consultations and quotations are available on request at info@atlantisndt.com.

Request a consultation