KPIs for a valve maintenance programme

1. Two kinds of indicator

A lagging indicator records an event that has already happened. Seat leaks found at surveys, seized valves and emergency call-outs are all lagging measures. They are credible because they count real events, but they arrive late. By the time they move, the cost is already incurred.

A leading indicator moves before the failure has time to develop. Time since last service, cycling compliance and injection pressure trend are common examples. They give you time to act, but they are noisier and easier to argue about.

A workable valve programme therefore needs both kinds of measure. Lagging measures tell you whether the programme works; leading measures tell you whether it will keep working. Weight the mix towards the failure modes that dominate your own population.

2. Lagging indicators worth counting

Start with counts you can defend from records you already hold. A condition survey supplies most of the raw material, because it records each valve on a dated visit.

1. Seat leaks found per survey. A risk-based valve integrity study of upstream and midstream sites, published by Valve World, reported a clear split. Passing valves were 61 per cent of failures and seized valves 28 per cent. Populations vary, so check your own split before weighting.
2. Valves freed versus valves replaced. Seizures resolved in service indicate the programme is working. Replacements show where intervention came too late.
3. Emergency call-outs. A falling count suggests routine work is reaching the right valves before they fail.
4. Valves passed on for assessment. Count these honestly. A programme that never escalates anything is probably not looking hard enough.

Record the service of each valve, not only the fault. Line size, product and last intervention turn raw counts into something you can analyse.

3. Leading indicators that warn earlier

Leading measures come from the maintenance record itself, rather than from failures. Each one needs a clear definition, a named owner and a review date. Without these, the trend quietly goes stale.

1. Time since last clean or lubrication. Rising age on critical valves signals risk before symptoms appear.
2. Cycling compliance. Record whether each valve was operated on schedule, against the interval your site has set.
3. Injection pressure response. A steadily rising response at the fitting suggests blocked pathways or hardened deposits. Treat it as screening evidence, not as proof.
4. Stem and body observations. Note any visible stem leakage when it is seen, with the date.
5. Training and equipment readiness. Trained operators and serviced injection guns keep the programme able to act.

An operator's report of a pressure response carries less weight than a witnessed measurement. Our article on evidence grades explains the difference.

4. Turning indicators into targets

Targets should come from your own baseline, not from a supplier's typical figure. Survey first, count what you find, and set improvement against that starting point. Dwell times, top-up amounts and service intervals vary by valve, product and site, so treat any figure as reported practice.

Give every indicator an owner, a review cadence and a decision rule. A number that nobody acts on is decoration.

One fair lagging measure for outside help is first-pass success of injection work. It assumes serviceable fittings, a design that suits injection, and a permit that allows the work. Where OEM documents set limits or warranty terms, consult them. Count the failures as well as the wins.

For a controlled comparison, torque measured before and after treatment gives a defensible record. A rising replacement count is not failure; it may show the survey is finding hidden problems.

5. What these numbers cannot prove

Trending indicators are screening evidence. A pressure decay observation or a passing symptom suggests a problem; it does not confirm one. The confirming step is a seat-leak test under agreed conditions. Examples include pressure decay across the closed valve, double block and bleed monitoring, or a witnessed leak test.

Factory acceptance standards belong to manufacturing, not to your field records. API 598 and ISO 5208 describe seat leakage testing and leakage classes at manufacture. Do not score field survey results against them. Witnessed test records, showing who tested what and when, carry more weight than a copied pass mark.

Indicators also cannot compare providers fairly unless the valve sets and conditions match. The honest next step is a small trial on a defined group, with results recorded and signed. That record is worth more than any promise on paper when choosing an online valve maintenance provider.