The Written Practice a Mine Site Audit Actually Reads

What the document has to contain

A written practice is a controlled procedure that governs people rather than parts. It has to identify the employer and the scope of work it covers, list the methods and levels certified, and set out for each level and each method the education, training and experience required for qualification. It must contain or reference the training topical outlines, describe the general, specific and practical examinations, state how they are composed, administered and graded, and name the certifying agent who issues certificates.

Beyond that core it must address vision acuity — near-vision testing and colour contrast differentiation, with stated criteria and intervals — recertification intervals and the basis on which recertification is granted, technical performance evaluation, the conditions under which certification is suspended, revoked or reinstated, records retention and location, and the use of outside Level III services if any are engaged. Each of these is a place where an auditor can ask a simple question and get an answer that either exists in a file or does not.

The overlooked requirement is the deviation record. SNT-TC-1A is a recommended practice; every recommendation the employer chooses to modify is legitimate provided the modification is documented. A practice that reduces a training hour requirement with a stated technical justification is defensible. A practice that reduces it silently, or that copies the recommendation and then does not deliver it, is not. The deviation list is often the shortest and most valuable section of a well-built practice.

Who owns the document, and who signs the certificates

Two roles get conflated constantly. The Level III is the technical authority for the method: they approve the procedures, write or approve the examinations, grade the practicals, and own the technical content of the practice's method sections. The certifying agent is the employer, acting through a named individual, who actually issues certification. Those can be the same person and often are, but the practice must be unambiguous about which name goes on the certificate.

Where the Level III is an outside consultant — which is the norm for mining groups that do not employ NDT staff full time — the practice must define the split precisely. What the consultant approves, what the employer retains, how the consultant's decisions are communicated and recorded, and what happens if the consultant is unavailable. Critically, the employer must hold a copy of the outside Level III's own current certification in its records. Auditors ask for it, and the answer "the consultant has it" is not an answer.

The failure mode here is quiet and near-universal. A practice names a certifying agent who left the company two years ago. Certificates continue to be issued, signed by whoever is available in the QA office. Every certificate issued after that departure is defective on its face, and remediation means reissuing them, which means re-establishing that the underlying qualification evidence still exists. Reviewing who signs what, against what the document says, is usually the single highest-value hour of a written practice engagement.

Choosing the parent: SNT-TC-1A, CP-189, NAS 410 or ISO 9712

SNT-TC-1A is a recommended practice. It becomes binding only because a referencing document invokes it — a construction code, a client specification, a purchase order. That gives the employer latitude to adapt it, with documented deviations. CP-189 is a standard: its requirements are mandatory as written, it constrains who may act as Level III, and it removes most of the adaptation latitude. Choosing between them is a real decision, and the right one depends on what your clients contractually demand rather than on which is more familiar.

NAS 410 applies where any part of the group supplies aerospace, and it imposes its own training hours, examination structure and Level 3 arrangements. ISO 9712 and its Canadian implementation CAN/CGSB-48.9712 are different in kind: they are third-party central certification schemes, not employer certification, so an employer practice cannot issue them. Mining groups with a fabrication arm serving multiple sectors sometimes need more than one scheme running side by side, and the practice must say which applies to which work.

The trap is invocation drift. A practice written against the 2011 edition of SNT-TC-1A sits in a company whose current client contracts call up a later edition. The document is internally consistent and externally wrong. Checking the edition invoked by every live client specification against the edition your practice claims takes an afternoon, and it is a finding that clients raise early because it is so easy to check.

The trace an auditor actually performs

Auditors do not read a written practice cover to cover looking for errors of doctrine. They pull three to five certification files at random and trace each one back to the document. Does the file contain evidence of the education claimed? Do the experience hours meet what the practice states, in the method claimed, and is there a supervisor attestation? Is there a general examination, a specific examination and a practical, each graded, with a composite result meeting the practice's threshold? Is there a current vision record?

Then they go forward instead of back. Take a report the technician signed and check whether the method and technique on that report fall inside the scope of their certificate. This is where mining programmes fail most damagingly, because the technique on the report has moved with the technology while the certificate stayed where it was. A technician certified in ultrasonics in 2018 signing an encoded phased array report in 2026, with no technique-specific training, examination or demonstration in the file, is a finding that calls the data into question, not just the paperwork.

The last step is the certifying agent and the Level III. Is the signature on the certificate the person the practice names? Is the Level III's own certification current, and does it name the method in question? Does the employer hold a copy? Four questions, four documents, and in most first audits at least one of them is missing. Running this trace yourself, on your own files, before anyone else does, is the entire value of a pre-audit review.

The arithmetic buried in the qualification tables

The qualification tables are where good-faith programmes go wrong without anyone acting dishonestly. The first trap is the direct-access route. A candidate progressing Level I to Level II accumulates hours differently from one entering at Level II directly, and the two paths have different totals. Files routinely mix them, taking the lower total from one route and the entry conditions from the other.

The second trap is multi-method hours. The tables carry a provision allowing a reduced total when a candidate qualifies in more than one method, subject to a minimum in each method. This is applied backwards with striking regularity: the same eight hundred hours of general inspection work are counted in full toward UT, MT, PT and VT simultaneously, producing a technician with four certificates and one method's worth of actual experience. The arithmetic is checkable in five minutes and the correction is unpleasant, which is why nobody checks it until an auditor does.

The third is examination grading. A composite of at least eighty percent with no individual examination below seventy is the familiar threshold, and files often show a composite that passes while one paper sits in the sixties — recorded accurately, and nobody applied the second half of the rule. The fourth is experience that is not in the method: time spent on a site as a general inspector, or supervising others, credited as hands-on method experience. The practice should define what counts, and the supervisor attestation should say what the person actually did.

How a good practice drifts into a bad one

Practices are almost never wrong on the day they are issued. They are written by a competent Level III against the work the company did that year, and then the company changes and the document does not. The commonest drift in mining is method and technique scope: the site adds phased array for mill shell welds, magnetic flux leakage for conveyor cords, electromagnetic examination for hoisting rope, or corrosion mapping for slurry lines, and the practice still lists four conventional methods with outlines to match.

The second drift is personnel. The Level III retires, the certifying agent moves, the QA manager who understood the document leaves, and the practice becomes a file nobody opens. Certificates continue to issue on momentum. The third is edition drift, described above. The fourth is site divergence: a group with operations in Nevada, Ontario and Quebec applies the same practice nominally while each site develops its own interpretation of what a practical examination looks like, and no one compares them until a corporate audit does.

The fifth, and hardest to see, is that the practice describes a programme that once ran and no longer does. Technical performance evaluations for recertification, specified in the document, degenerate into a signature. Training that was delivered in-house is now assumed from experience. Records that were kept centrally are now on someone's laptop. None of this is visible from reading the document, which is exactly why a written practice review has to sample files rather than review text. A practice is only as true as its most recent certification file.

Mining's own regime: MSHA, hoisting rope and site training

MSHA regulates United States mines under 30 CFR — Part 56 for surface metal and nonmetal operations, Part 57 for underground — and it says essentially nothing about NDT personnel levels. What it does do is mandate examinations that turn out to be NDT work. Hoisting requirements bring wire rope under a testing and retirement regime, and electromagnetic rope examination under ASTM E1571, producing loss-of-metallic-area and localised fault data, is a specialised discipline that sits awkwardly outside a conventional four-method practice.

This is the industry-specific gap worth naming plainly: mines buy rope testing, act on rope testing, and retire ropes on the strength of it, while the written practice that governs their NDT personnel makes no provision for the people doing it. Either the practice adds rope examination as a defined method or technique with its own training outline, examination and practical, or it states explicitly that rope examination is procured from a specialist under a separate qualification regime and describes how that qualification is verified. Both are defensible. Silence is not.

Alongside that runs site training. Part 46 and Part 48 govern miner training depending on the type of operation, and every contractor technician who walks onto a mine site needs the applicable training regardless of their NDT credentials. It has nothing to do with SNT-TC-1A and it will be checked at the gate. A written practice for a mining contractor should acknowledge the interface — that NDT certification and mine site training are separate requirements, both mandatory, tracked in the same personnel file so nobody arrives at a remote site current in one and not the other.

Canada: where the employer practice stops

A Canadian mine changes the certification question fundamentally for one category of work. Pressure equipment — an autoclave in a pressure oxidation or high-pressure acid leach circuit, an acid plant vessel, registered piping, a fired boiler — falls under provincial safety authorities. Alberta's ABSA, Ontario's TSSA, Technical Safety BC and their provincial counterparts operate registration and quality programme regimes, and they look to certification under CAN/CGSB-48.9712 administered through the national certification body rather than to an employer's own certificates.

That means an employer written practice, entirely valid for structural, conveyor and rotating equipment work at the same site, does not create a person qualified to perform code NDT on the autoclave. Companies discover this when a jurisdictional inspector asks for certification and receives a company certificate. The remedy is not a better practice; it is CGSB-certified personnel for that scope. What the practice must do is state the boundary explicitly, so nobody assigns work across it by accident.

Cross-border groups feel this most. A contractor operating in Nevada and Ontario under one document has employees certified under an employer practice in one jurisdiction and requiring third-party certification in the other, for work that looks identical from the office. A well-built practice for such a group carries a jurisdiction table: scope, applicable regime, certification accepted, who verifies. It is one page and it prevents the single most expensive mistake in Canadian mine NDT — mobilising a crew that cannot legally sign the work.

Contractors, remote sites and the acceptance clause

Most mine NDT is bought rather than staffed. The site's own quality system therefore depends on certificates it did not issue, under practices it has never read, signed by Level IIIs it has never verified. The written practice needs a clause that closes this, and in the great majority of documents it simply is not there. The clause should state what evidence is required before a contractor's data is accepted, who reviews it, what record the review produces, and what happens when the evidence is incomplete.

The evidence list is short: the contractor's written practice, the technician's certificate showing method and level, the identity and authority of the certifying agent, current vision records, and demonstration records for any specialised technique the scope involves. Verification takes twenty minutes per technician and can be done before mobilisation. On a remote site, doing it after arrival means either accepting the work or standing down a mobilised crew, and everyone knows which of those actually happens.

Remoteness compounds every weakness in the chain. Records are created on paper in a site office and scanned late or never. A technician who fails a technical performance evaluation is three flights from a replacement. A revision to the practice issued from head office reaches four sites and not the fifth. Building the practice with distribution, acknowledgement and offline record capture in mind is not administrative fussiness; it is the difference between a document that governs and a document that exists.

Revision control, retention and how the work is scoped

A written practice is a controlled document and must behave like one: numbered revisions, a change record, an approval signature, a distribution list, and evidence that the current revision reached every site and every person who certifies or is certified under it. Superseded revisions must be retained, because a certificate issued in 2022 was issued under the 2022 revision and must be judged against that text, not against today's.

Records retention needs a stated period and a stated custodian. Certification files, examination papers, practical records, vision records, technical performance evaluations and the Level III's own credentials all have to be produced on demand years later. The commonest gap is examination papers: the certificate exists, the grade is recorded, and the marked paper that justifies the grade was discarded. Without it there is no evidence the examination occurred at the standard claimed.

Atlantis develops and remediates written practices for mining and minerals operators and the contractors that serve them, runs the file trace an auditor would run before the auditor arrives, and provides Level III technical authority and audit representation thereafter. Work is scoped to your method list, your jurisdictions and your contractor chain rather than sold as a template document. For a consultation on your certification programme, or a quote covering written practice development and certification file review, contact info@atlantisndt.com.

Request a consultation