Turning valve service records into replacement forecasts

Forecast a decision, not a date

A replacement forecast should help plan inspection, budget, isolation and spares. It should not create a precise failure date from weak records.

The useful outputs are decision ranges and priorities. A valve may remain under monitoring, need assessment, enter a planned replacement window or require earlier action.

Keep the uncertainty beside each output. Managers need to know which forecasts rest on measurements and which rest on missing evidence.

The valve integrity management guide provides the programme structure around these decisions.

Build one reliable record per valve

Forecasting fails when names change or histories combine several assets. Use a stable valve identifier across surveys, permits, tests and work orders.

1. Identity. Record valve type, size, service, location, maker data and drive arrangement.
2. Duty. Record normal position, cycling duty, isolation role and process conditions.
3. Movement. Record travel, direction, torque method, readings and actuator events.
4. Leakage. Separate seat leakage from stem, fitting, body or joint leakage.
5. Maintenance. Record product, amount, injection point, pressure response and operating response.
6. Tests. Record method, instruments, conditions, acceptance basis and result.
7. Decisions. Record restrictions, unresolved gaps, owner and next review trigger.

Follow the per-valve records guide so later teams can compare like evidence.

Turn repeated evidence into condition trends

One service visit describes one observed condition. Repeated comparable visits can show direction and rate of change.

Rising torque can indicate increasing resistance. Worsening seat-test results can indicate reduced isolation performance. Neither observation identifies the cause alone.

Shorter benefit after comparable treatment may suggest continuing deterioration. Changed product, amount, method or operating state can also explain the difference.

Record both successful and unsuccessful work. A database containing only completed repairs makes treatment performance look better than reality.

Use field torque measurement and field seat-leak testing to keep trend inputs comparable.

Rank valves using evidence and consequence

Condition alone does not set replacement priority. A rarely used valve and a critical isolation valve can need different actions with similar symptoms.

1. Review confirmed condition and unresolved failure modes.
2. Review the valve’s isolation, safety and production role.
3. Review deterioration across comparable tests and service visits.
4. Review access, outage opportunities, spares and replacement lead time.
5. Assign an action range with stated evidence confidence.
6. Set the event or date that forces the next review.

The valve criticality guide helps separate consequence from condition. Keep evidence from isolation verification visible for critical barriers.

Control the uncertainty in every forecast

Field data often has gaps, method changes and selection bias. Valves receiving frequent attention may look worse simply because more evidence exists.

Different operators, tools and process states can create false trends. Normalise only where the engineering basis is clear and recorded.

Do not invent a failure curve from a small local sample. A statistical model needs enough relevant events, consistent definitions and review by a competent specialist.

Where that basis is absent, use transparent engineering judgement. State assumptions, confidence, alternatives and the evidence that could change the decision.

A forecast should become less confident as it extends beyond observed history. Review it whenever service, duty, product or condition changes.

Use service response without overstating it

Improved movement after cleaner or lubricant can show changed resistance. Better seat-test results after sealant can show changed leakage under test conditions.

Neither response proves hidden damage is absent or predicts remaining life. Record the outcome as one input, not a warranty.

Injection itself needs gates. Fittings must be sound, design and service must suit injection, OEM guidance applies, and operations must permit the work.

Use controlled amounts within equipment and fitting ratings. Stop for abnormal pressure response, leakage at the fitting or worsening operation.

Connect the forecast to an owned action

Every forecast needs an owner, review trigger and practical next step. Otherwise it remains a chart without operational value.

Compare monitoring, assessment, planned repair and replacement through the repair versus replacement guide. Keep safety and isolation needs ahead of cost timing.

Our field service team can collect and structure valve evidence. The operator owns consequence ranking, outage planning and final replacement decisions.