The Unwritten Clause: Making Engineering Judgement Traceable When the Standard Goes Quiet

Standards are written for the many, not the every. A British Standard, a customer specification or an approved code of practice can define tolerances, test regimes and responsibilities with great precision, yet a shop-floor question will still arrive that none of them answers directly. The question is not whether such gaps exist. It is what the organisation does when one is reached.

For members, the commercial risk is rarely the absence of a clause in itself. It is the absence of a record showing that the gap was recognised, assessed and decided by competent people. Auditors, customers and insurers do not expect standards to cover every eventuality. They do expect a defensible position. This briefing proposes a member-oriented discipline for capturing engineering judgement in the gaps, using the ordinary tools of risk assessment, peer review and shared experience.

Silence is a finding, not a blank

An unanswered standard question should be logged as a finding, not absorbed into routine practice. That distinction matters. If a welder, designer or maintenance engineer resolves an ambiguity by experience alone, the decision may be sound. But if the reasoning is not recorded, the organisation has no way to show that alternatives were considered, that relevant standards were checked, or that the decision was reviewed by someone with the right competence.

A useful starting point is a short gap note. It need not be a lengthy report. It should capture the question in plain terms, the product, process or installation affected, the standards and specifications reviewed, and the reason no clause applies directly. It should also record the date, the author and the person who accepted the decision. The gap note is not a substitute for a risk assessment. It is the trigger for one.

Risk assessment proportionate to consequence

Once a gap is identified, the next question is consequence. A minor finishing tolerance on a non-safety-critical component may require a light-touch assessment and a single reviewer. A modification to a lifting accessory, pressure system or electrical control circuit demands a more formal approach, with the relevant statutory duties and industry guidance taken into account.

The risk assessment should state its assumptions clearly. It should identify who or what could be harmed, the severity and likelihood before controls, the controls proposed, and the residual position. Where the gap concerns a safety-related matter, the assessment should show how the organisation has sought to reduce risk so far as is reasonably practicable, in line with UK expectations. Where the gap concerns performance, durability or interchangeability, the assessment should address the evidence available and the limits of that evidence.

The record should also say what was not done. An assumption that a material behaves in a particular way, or that a legacy installation matches a drawing, is part of the decision. If it is not written down, it cannot be tested later.

Peer review with named competence

Engineering judgement is strengthened by review, but only when the reviewer is competent and independent enough to challenge the decision. A review that consists of a brief email exchange between colleagues who share the same assumptions is unlikely to satisfy an auditor or an insurer. A better approach is to name the reviewer, state their relevant experience, and record any dissent.

For SMEs, peer review need not mean a large committee. It may be a second engineer, a technical authority from another department, or an external specialist engaged for a specific question. What matters is that the reviewer has the knowledge to assess the judgement and the standing to say no. Where a decision is accepted against a reviewer's advice, the record should show why.

Draw on member experience without copying it

BEMSC's membership includes organisations that have met similar gaps before. Member helpline themes, sector group discussions and anonymised case notes are valuable sources of practical experience. They can shorten the search for a sensible approach and reveal pitfalls that a single site might miss.

But experience is evidence, not authority. A solution that worked in one workshop, on one material batch, or under one customer's concession does not automatically transfer to another. The gap record should show how member experience was used: whether it informed the risk assessment, supported a technical assumption, or suggested a review trigger. It should not simply cite a conversation as justification.

Make the record traceable and live

A judgement record has little value if it sits outside the systems that govern the product or asset. It should be linked to the design file, the change control record, the supplier approval or the maintenance schedule as appropriate. Where the decision affects a drawing, a work instruction or an inspection plan, those documents should point back to the judgement record.

The record should also have a review date or a set of triggers. A new edition of a standard, a change of supplier, a repeat defect, a customer complaint or a shift in operating conditions may all reopen the question. Traceability means that the organisation can find the decision when circumstances change, not only when an auditor asks.

A short test before the file is closed

Before accepting a judgement record, members may find it useful to apply five questions. Was the question identified before the decision was made? Were the relevant standards and specifications checked? Was a proportionate risk assessment carried out? Was the decision reviewed by someone with named competence? Can the record be retrieved and linked to the affected work?

If the answer to any of these is no, the record is not yet defensible. It may still describe a sound engineering decision, but it does not demonstrate one. The difference matters when a customer, insurer or regulator asks for the basis of the position.

Conclusion

Standards are necessary, but they are not complete. The gaps between clauses are part of the engineering landscape, and they will not close simply because an organisation wishes them to. What can be managed is the response: a logged question, a proportionate risk assessment, a competent review and a traceable record. That is how engineering judgement becomes visible, and how members can stand behind it when the standard is silent.

BEMSC will continue to gather member experience on these records through its helpline and sector groups. Members are invited to share anonymised examples of gap records that have worked well, so that the wider community can learn from practical judgement as well as from published text.