Surviving a Pipeline NDT Audit: the Evidence Chain

What a Pipeline Auditor Actually Samples

Whether the auditor is a regulator, a client's own assurance team, a certifying body or a class society on an offshore line, the method is broadly the same and it is not what most NDT contractors prepare for. He does not begin by looking at radiographs and hunting for missed indications. He begins with the as-built weld map for a construction spread or a tie-in campaign, selects a small number of welds, and then asks for everything that touched them. The examination is of the programme, and the welds are just the entry point.

The sample is deliberately small at first, typically five to ten welds, because the auditor is testing whether the chain holds rather than whether the welding was good. If every requested document appears, is internally consistent and is dated correctly, the sample closes. If one link fails, the sample expands, and it expands in the direction of the failure: another spread, another welder, another year, another contractor. Programmes are rarely damaged by the first finding. They are damaged by the expansion.

This is why audit preparation that focuses on tidying the file room misses the point. The auditor is testing traceability under adversarial conditions, on documents produced by several organisations at different times. Building that chain deliberately, and proving it can be walked end to end, is the substance of ASNT Level III consulting in a pipeline setting.

The Evidence Chain That Must Hold, Link by Link

Take a single girth weld and walk it. The weld map identifies the weld by station or milepost, gives a date and names a welder. That welder must have a qualification record under API 1104 Section 6 that was valid on that date and that covers the diameter, wall thickness, position and process actually used. The joint must be covered by a welding procedure specification traceable to a procedure qualification record, qualified under API 1104 or, where invoked, ASME Section IX, consistent with 192.225.

The examination side has the same structure. The NDT report must reference an examination procedure at the revision in force on the date of the weld, not the revision current today. The technician who performed it and the person who interpreted it must hold certification covering that method and level, issued under a written practice, with the supporting examination records and near-vision testing in date. The image or data set must be retrievable, unaltered and identifiable to that weld. Any rejection must lead to a repair under a qualified repair procedure and a re-examination that is itself documented.

Every one of those links carries a date, and dates are where programmes fail. A certification issued three weeks after the report it signs. A procedure revision that post-dates the examination it supposedly governed. A welder requalification recorded after a gap that exceeds the currency limit. None of these mean the weld is defective, and all of them are findings, because the operator cannot demonstrate that the work was performed under a qualified system at the time it was performed.

49 CFR 192.243 and 195.234: the Records That Get Counted

Part 192 requires that nondestructive testing be performed under written procedures, by personnel trained and qualified in those procedures and the equipment used, and that procedures exist for the proper interpretation of each test so that weld acceptability under 192.241(c) is assured. It also drives coverage: the proportion of girth welds that must be nondestructively tested rises with class location, from a modest percentage in Class 1 to full examination in the higher class locations and at specified crossings and proximities. Part 195 imposes a parallel structure for hazardous liquid lines under 195.234, including the requirement that a weld which is tested be tested over its entire circumference.

The clause that generates the most audit traffic is 192.243(f). It requires a record showing, by milepost, engineering station or geographic feature, the number of girth welds made, the number nondestructively tested, the number rejected and the disposition of those rejects, retained for at least five years or as long as the pipeline remains in service, whichever is longer. It sounds trivial. In practice it requires that the NDT contractor's report set, the welding contractor's weld log and the operator's as-built records agree, and on a large spread with multiple crews they very often do not.

Reconciliation is the work. Counting welds is easy; reconciling three independently maintained datasets after the fact is not, particularly when tie-ins, cut-outs, repairs and golden welds have been recorded differently by each party. Programmes that capture NDT results against the weld identifier as the work happens, in a controlled inspection data management system rather than in spreadsheets emailed between contractors, produce this tally on demand instead of rebuilding it under audit pressure.

API 1104, Appendix A and the Automated UT Trap

API 1104 supplies the technical spine: qualification of welding procedures and welders, inspection and testing of production welds, the acceptance standards for nondestructive testing in Section 9, the procedures for the NDT methods themselves, and the alternative acceptance standards for girth welds in Appendix A. Part 192 adopts Section 9 as the acceptance basis and permits Appendix A where it applies and where the weld is unacceptable for a reason other than a crack. Both routes are legitimate. Using them inconsistently across one project is not.

Appendix A is an engineering critical assessment route: acceptance limits derived from fracture mechanics for a specific combination of material, geometry, loading and flaw sizing capability. That imposes obligations that Section 9 does not. The flaw height sizing capability assumed in the assessment has to be demonstrated by the examination technique actually used, and the assessment inputs have to match the pipe and the loading actually installed. An Appendix A criteria set applied by a technician who has never seen the underlying assessment is a finding waiting to be written.

Automated ultrasonic testing is where this concentrates. Mechanised girth welding with zonal AUT is efficient and defensible when the AUT procedure was qualified on the production bevel, the production wall thickness and the production material, using a qualification block with representative reflectors in the correct zones. The recurring failure is a procedure qualified on one geometry and deployed on another after a design change, with nobody treating the bevel as an essential variable. Independent review of the examination data against the qualified procedure catches that before an auditor does.

Findings That Recur in Pipeline NDT Programmes

The same findings appear across operators, contractors and jurisdictions. A written practice that references an edition of SNT-TC-1A older than the certifications issued under it, or that describes an examination and certification process the organisation has not actually followed for years. Level III certification lapsed, or held in a different method than the one being approved, so the procedures carry a signature that does not cover them. Near-vision examinations expired, which quietly invalidates every interpretation made after the expiry date.

On the procedure side: examination procedures not qualified for the production geometry; radiographic technique sheets missing image quality indicator selection and placement rationale or the required density or signal-to-noise checks; digital images stored in a location where they can be overwritten, so unalterability cannot be demonstrated; repair welds re-examined under a procedure that was never qualified for repairs. On the records side: the 192.243(f) tally that cannot be reconciled to the weld map, and sub-tier NDT vendors used across a spread who were never audited and whose written practice nobody has read.

The common shape of all of these is that they are systemic rather than technical. One expired vision test does not mean one bad call; it puts every interpretation in that period into question. That asymmetry is why audit exposure is best measured by sampling the chain rather than by counting reports, and why the fix is usually a written practice and procedure correction plus targeted requalification, not a mass re-examination.

What the Level III Signs, and What He Does Not

The scope of the NDT Level III in a pipeline programme is precise and worth stating in writing before an audit rather than during one. He owns the employer written practice, develops and qualifies the examination procedures, establishes and administers the qualification examinations from which personnel certifications are issued, certifies personnel within that practice at the appropriate levels, provides technical justification for deviations, performs independent review of examination data, and answers technical questions on behalf of the programme when the auditor asks how a procedure was qualified or why a call was made.

The boundaries matter just as much. Atlantis is not a process safety management auditor and does not perform PSM compliance audits. It is not the API inspector of record for any asset and does not sign inspection reports in that capacity. It does not sell API 510, 570 or 653 inspector certification training, and no part of this service should be read as offering it. It does not certify the operator's operator qualification programme, and it does not issue statements of overall regulatory compliance, which belong to the operator.

Keeping those lines visible is protective in both directions. An auditor who finds a Level III signature on a document outside the Level III scope will pull that thread, and an operator who has quietly delegated compliance responsibility to a technical contractor has a governance finding independent of any NDT issue. Where NDT personnel qualification within the written practice needs to be built or rebuilt, that is delivered through structured training and certification tied to the practice rather than through generic courses.

Operator Qualification Is Not NDT Certification

Operator qualification under 49 CFR Part 192 Subpart N and Part 195 Subpart G addresses individuals performing covered tasks on an operating pipeline, evaluated for the task and for recognising and reacting to abnormal operating conditions. It is a task-based regime tied to operations and maintenance. NDT certification is a method-and-level regime issued by an employer under a written practice built on SNT-TC-1A, CP-189 or ISO 9712, supported by training hours, experience, examinations and vision testing.

The two get conflated in both directions, and both directions produce findings. Operators sometimes list NDT as a covered task in the OQ programme without the supporting task analysis, then treat the OQ card as evidence of NDT capability. Contractors sometimes present NDT certifications when asked for OQ evidence on operating assets. The cleanest position is to define which activities on the asset are covered tasks, keep NDT certification governed entirely by the written practice, and cross-reference the two documents without merging them.

This distinction becomes sharp during in-service work. A direct examination dig in a high consequence area may involve individuals performing covered tasks alongside NDT technicians performing examinations under the written practice, with both evidence sets requested in the same audit. Having a clear map of who is qualified under which regime, and for what, prevents a straightforward dig record from turning into a personnel qualification finding.

Integrity Management Evidence and Direct Examination

Integrity management brings a second body of records into audit scope. Under Part 192 Subpart O for gas transmission and 195.452 for hazardous liquid lines in high consequence areas, operators must identify threats, complete baseline and periodic assessments, and remediate discovered conditions within defined schedules. Assessment is frequently in-line inspection, and in-line inspection findings become NDT work: a dig sheet, an excavation, surface preparation, and direct examination using ultrasonic wall thickness measurement, phased array, magnetic particle or other methods depending on the anomaly type.

The evidence expectations at the dig are exacting and often underestimated. The examination procedure must be qualified for the geometry and the flaw type being characterised. The technician must be certified for the method used. The measured dimensions must be recorded in a way that supports the remaining-strength calculation applied afterwards, with instrument calibration records for the day and the reference blocks used. The comparison between the in-line inspection call and the field measurement feeds tool validation, which is itself audited. And the remediation must close within the applicable schedule, with the closure documented.

The chain here is longer than in construction because it crosses organisations: the tool vendor, the analysis provider, the excavation contractor, the NDT crew and the operator's integrity engineer. It is also the chain where independent review pays best, because a wall-thickness dataset that does not support the remaining-strength assessment built on it is a problem that surfaces at the worst possible time. If an audit is scheduled or a finding is already open, start with a technical gap review rather than waiting for the response deadline to set the scope.

Request a consultation