Maintenance evidence that stands up to a regulator or auditor

1. What an auditor asks first

An auditor or regulator reads your valve file and tries to reconstruct the work without interviewing anyone. The questions stay the same: what was found, what was done, on what authority, by whom, and what changed afterwards. If those answers live only in memory, the file shows nothing.

Evidence has three parts. The record states what was observed and done, valve by valve. Traceability links materials, equipment and people to each entry. The test basis explains how a finding or result was established. Weak programmes show activity without evidence: visits, invoices and consumables, but no result per tag.

Depth should follow consequence. High-consequence valves deserve fuller entries than minor ones, and valve criticality ranking sets that order before an audit asks for it.

2. The minimum each record entry carries

Traceability starts with a fixed set of fields on every entry. One valve tag opens the record, and the same tag closes it with the result.

1. Valve tag, line and location, the date, and the service conditions at the time.
2. Condition found in plain terms: passing, seized, external leakage, or operational but stiff.
3. Action taken, with product type, quantity and batch reference, plus the equipment used and its rating.
4. Who performed the work, who witnessed it, and under which permit.
5. The result after the action, and how that result was established.

Grouped entries fail review. A sheet that says forty valves were serviced is a schedule, not evidence, because no tag traces to a finding or an outcome. The same test applies to contractor reports; see what good survey evidence looks like before accepting one.

3. Screening results and confirmed results

Each entry should say how a conclusion was reached, because screening and confirmation carry different weight. Pressure rise across a closed valve, flow indication, sound and bubbles are screening evidence. They indicate a condition; they do not prove it. A file built only on screening signals reads as unverified opinion.

The confirming test for a passing valve is a seat-leak test under agreed conditions. That may be pressure decay across the closed valve, double block and bleed monitoring, or a witnessed leak test. Record the method, the acceptance limit agreed before the test, and the outcome against it.

Standard references need care. API 598 and ISO 5208 are factory acceptance references for leakage testing, not field results, and must not be cited as if they were. Name the basis you actually used and nothing more.

4. The basis for intervals and methods

Auditors also test the basis for your intervals and methods. Company practice alone is a weak answer. The stronger file shows OEM documentation where it exists, because ASME B31.8 defers operating and maintenance intervals to the valve manufacturer, alongside any rule that binds your sector.

Coverage differs by rule. PNGRB T4S, the Indian city gas standard, sets annual tests for regulators, slam-shut valves and relief valves, and is silent on intervals for mainline ball valves. Where the rule is silent, you own the justification, and your recorded findings are the evidence behind it.

This is where a decision log earns its place. Each choice about a valve links to the finding that caused it, with dates and names. The companion article on turning survey findings into an approved plan shows one workable structure.

5. Common gaps that fail reviews

Common gaps include entries with no result, products with no quantity or batch reference, unwitnessed work on critical valves, and screening signals written up as findings. Each one turns real work into an unverifiable claim.

Plausibility matters as well. A risk-based valve integrity study of upstream and midstream sites, published by Valve World, reported passing as the leading failure mode at 61 per cent, seized at 28 per cent and external leakage at 6 per cent; distributions vary by population. A large, ageing population with no recorded seat leakage is therefore a warning sign, not proof of good condition.

If the file needs rebuilding, start with a documented condition survey on the highest-consequence valves rather than retrospective paperwork. Our valve survey service produces per-valve entries built for that file, and the guide to choosing an online valve maintenance provider lists the evidence a competent provider should supply.