Budgeting valve maintenance across a network

1. Start with a usable asset population

A budget cannot be stronger than its valve register. Each asset needs a stable identity, location, service and required function.

Add valve type, size, operating method, injection fittings and access limits where records support them. Mark unknown fields instead of filling them by assumption.

Unknown-history valves need a baseline survey before detailed work quantities are trusted. The valve integrity management guide sets the wider record structure.

2. Rank consequence, likelihood and evidence quality

Criticality should reflect what happens if the valve cannot perform its required role. Isolation duty, customer effect, process consequence and access all matter.

Likelihood depends on condition and history, not age alone. Failed tests, rising torque, repeat leakage, corrosion and damaged fittings can increase concern.

Evidence quality deserves its own field. A confirmed seat-test failure carries different weight from sound, pressure rise or a single instrument indication.

1. High consequence, weak evidence. Fund confirmation and inspection promptly.
2. High consequence, confirmed fault. Fund engineering response and contingency planning.
3. Lower consequence, worsening trend. Fund planned assessment before the fault limits options.
4. Stable condition, good evidence. Keep proportionate monitoring and records.

The valve criticality guide explains how to keep these factors separate.

3. Budget by work category

One price per valve hides major differences in access, fault and required evidence. Build the budget from named work categories instead.

1. Data and survey. Asset verification, visual inspection, condition baseline and record correction.
2. Functional evidence. Approved movement checks, seat tests and investigation of abnormal indications.
3. Online treatment. Engineering review, products, equipment, field labour, operations support and post-work testing.
4. Monitoring. Follow-up inspection, retesting and temporary-seal management where required.
5. Planned isolation. Engineering, parts, access, shutdown coordination, repair or replacement and restart evidence.
6. Capability. Training, controlled equipment care, procedures and record management.

The repair-versus-replacement guide shows how these categories change between routes.

4. Do not assume every valve needs injection

Inspection and testing are maintenance activities even when no product enters the valve. Treatment follows a diagnosed need.

Before injection, the valve needs suitable fittings in good condition. The design and service must suit the product, OEM guidance applies, and operations must permit work.

Periodic cycling or top-up practices vary by valve and site. They should be framed as operator practice, not universal intervals.

Severe damage, broken parts or unsuitable systems need planned isolation. Repeated field attempts should not consume the replacement reserve.

5. Keep shutdown and replacement readiness visible

Online work can support suitable valves while they remain in service. It does not remove every future shutdown category.

Keep engineering, procurement, access and contingency needs visible for high-consequence assets. Temporary sealing should carry monitoring and a permanent repair trigger.

Use the shutdown-cost case structure to define operating boundaries and assumptions. Do not book estimated avoided cost as guaranteed savings.

6. Review the budget when evidence changes

A network budget is a controlled forecast, not a fixed annual shopping list. Update it after surveys, tests, failures and completed work.

Compare planned work with actual condition findings, not only expenditure. A survey that rules out unnecessary treatment can be a sound engineering result.

Track overdue confirmation, temporary repairs, repeat faults and unavailable spares. These items show where the next budget faces hidden pressure.

BTR VSI provides valve surveys, field services and training. Operators should retain ownership of criticality, acceptance and budget assumptions.