Starting a valve maintenance programme from zero

1. Start with a register, not a purchase

A programme begins with a list, and everything else builds on it. You cannot plan work for valves you cannot name or find.

Build it from drawings, line lists and a physical walkdown. Record the tag number, location, size, pressure class, body type, service and the fittings present.

Fittings matter most, because they decide whether in-service treatment is possible at all. Expect gaps: old plants often hold valves that appear on no drawing.

Keep the format simple enough that your people actually maintain it. Building a valve register covers fields, sources and upkeep in detail.

2. Survey condition before spending money

Next, learn what state the valve population is actually in. A condition survey records operation, seat tightness, external leakage and fitting condition.

Field observations at this stage are screening evidence, never proof. A valve that passes suggests a seat problem, and a weeping stem suggests packing wear.

The confirming test for a passing valve is a seat-leak test under agreed conditions. Pressure decay across the closed valve, or double block and bleed monitoring, are common methods.

You can survey everything, or you can start with a sample. Samples cost less but can miss local problems, such as valves in one buried corridor. Full surveys take longer but give a baseline you can defend.

What a valve condition survey covers explains the scope in plain terms. You can also engage a field survey team for the first pass.

3. Rank priorities you can defend

Now rank the survey findings by consequence and likelihood together. Consequence asks what happens if the valve fails to close or contain. Likelihood follows condition, service severity and the valve's own history.

Where records are blank, treat the valve as unknown and baseline it. Baselining valves with no service history covers that first pass.

The figures come from a published source, not from our own work. A risk-based valve integrity study of upstream and midstream sites, published by Valve World, reported them:

1. Passing seats: 61%
2. Seized mechanisms: 28%
3. External leakage: 6%
4. Other modes: 5%

Distributions vary by population, so read the list as a prompt, not a promise. Your own survey remains the evidence that should set priority.

4. Know the gates before any treatment

Cleaner, lubricant and sealant are the usual in-service treatment tools. Each has conditions, and the gates come before any recommendation:

1. The valve has injection fittings, and they are in good condition.
2. The valve design and the service suit injection.
3. Site rules and the permit to work allow the job.
4. OEM documents, where they exist, have been consulted for limits and warranty terms.

Valves built with injection points are designed to be maintained in service. Even so, in-service treatment is not universal, and some designs and services rule it out.

When injecting, dose in controlled amounts within equipment and fitting ratings. Watch the pressure response, and stop work if it behaves abnormally.

Valves that fail these gates need a proper assessment, or replacement. Recording that conclusion is a result, not a programme failure.

5. Keep records that survive staff changes

Every job goes back into the register on the same day. Record the date, the treatment, the amounts, the pressure response and the observed result.

Note the confirming test as well as the treatment performed. A seat-leak test result, with its method, is the evidence that treatment worked.

Physical tags make the linkage practical when work happens in the field. A tag number links the physical valve to its record, and photos show condition over time. Tagging valves for records describes a simple scheme that travels well.

Paper logs tend to disappear when people change roles or leave. A shared spreadsheet, kept current, lets the next engineer see what was done and when.

6. Review the programme every year

A programme that never changes is only a schedule, not a plan. Review the whole plan yearly against what the records show.

Look for repeat offenders: valves that pass again soon after treatment, or fittings that block repeatedly. Common causes include service severity, product mismatch, or a valve nearing replacement.

Full travel also deserves a place in the yearly plan. Industry practice is to cycle valves at least twice a year to prevent seizing. Let your own records confirm the cadence your site needs.

The review also resets priorities and budgets for the year ahead. This yearly loop, from register to review, is the core of valve integrity management.

Some sites run this routine with their own trained crews. Others move it to an annual maintenance contract after the first year.