- Reactive maintenance (run-to-failure) is appropriate for non-critical, easily replaceable valves — but applying it to process-critical or safety-critical valves is a false economy that leads to unplanned shutdowns, emergency repairs at premium cost, and potentially dangerous situations.
- Smart digital valve controllers have transformed predictive maintenance from aspiration to reality — valve signature testing, friction monitoring, and travel deviation alarms generate the leading indicators that allow maintenance teams to intervene before failure, not after it.
- A risk-based maintenance strategy — assigning maintenance approach based on valve criticality — is the most cost-effective model — it avoids both over-maintaining trivial valves and under-maintaining critical ones, directing resources where they actually protect plant reliability and safety.
I’ve seen plants where the maintenance strategy was essentially “wait until it breaks.” I’ve also seen plants where every valve got the same intensive maintenance schedule regardless of whether it was a critical process control valve or a cooling water drain valve. Both approaches waste money. The answer is risk-based maintenance — but implementing it properly requires understanding the failure modes and the tools available to detect them early. That’s what this guide covers.
The Three Maintenance Philosophies
Reactive (Corrective) Maintenance
Operate the valve until it fails, then repair or replace it. This is only appropriate when: the consequence of failure is low (non-critical utility service); spare valves are readily available and replacement is quick; the failure mode is not safety-critical; and the valve is inherently reliable with low failure probability.
Reactive maintenance is appropriate for: drain valves, sample valves, utility isolation valves on non-critical services, and small manual valves where replacement cost is under £100 and downtime impact is zero.
It is completely inappropriate for: emergency shutdown valves, pressure safety valves, main process control valves, and any valve whose failure causes a process shutdown, safety incident, or loss of containment.
Preventive (Time-Based) Maintenance
Maintain valves on a fixed schedule — every 12 months, every 2 years, every turnaround — regardless of actual condition. The schedule is set based on historical failure rates, manufacturer recommendations, or regulatory requirements (e.g., PSV testing intervals).
Preventive maintenance has two failure modes as a strategy: if the interval is too short, you waste maintenance resources on valves that didn’t need attention and you introduce the risk of “maintenance-induced failures” (packing damaged during unnecessary repacking, seats scratched during unnecessary inspection). If the interval is too long, valves fail between maintenance cycles.
Time-based maintenance is appropriate for: pressure safety valves (where proof test intervals are defined by legislation or insurance requirements), emergency isolation valves where a minimum annual proof test is mandated, and valves in services where wear is predictable and consistent.
Predictive (Condition-Based) Maintenance
Intervene based on the actual condition of the valve, using diagnostic data to identify developing problems before they cause failure. The goal is to do maintenance only when necessary — exactly when a valve is showing signs of degradation, not on a fixed calendar that may be too early or too late.
Predictive maintenance requires data — either from online monitoring (smart positioners, vibration sensors) or from periodic diagnostic tests (valve signatures, partial stroke tests). It represents the highest investment in tooling and analysis capability but delivers the best return for high-value, high-criticality valves.
Control Valve Diagnostics: What Smart Positioners Can Tell You
Modern digital valve controllers (DVCs) — Fisher FIELDVUE, Metso Neles ND9000, ABB TZIDC, Siemens SIPART — are the core tools for predictive maintenance on control valves. They continuously monitor:
- Valve travel — actual stem position vs. commanded position; sustained deviation indicates actuator or mechanical problems
- Actuator air supply pressure — low pressure indicates air supply problems, filter blockage, or SOV degradation
- Friction monitoring — the positioner can calculate the friction force experienced during movement by tracking supply pressure vs. travel; increasing friction indicates packing wear, stem corrosion, or mechanical binding
- Travel deviation alarms — if the valve can’t reach commanded position within a tolerance band, an alarm is generated before operators notice poor loop performance
- Cycle count and travel accumulator — track valve usage; high-cycle or high-travel valves may need packing or trim maintenance sooner than the calendar suggests
Valve Signature Testing
A valve signature test strokes the valve from fully closed to fully open (or vice versa) while recording actuator pressure and stem position. The resulting plot of actuator pressure vs. position — the “valve signature” — is a fingerprint of the valve’s mechanical condition.
A healthy valve signature shows:
- Smooth, consistent actuator pressure throughout the stroke
- Clear breakout force at the start of movement (higher pressure at dead band edges) consistent with the specified packing friction
- Consistent return signature on the reverse stroke
- Seating force at closed position within the expected range
Deviations from the baseline signature indicate specific problems:
- Increased breakout pressure — increased packing friction (packing hardening, stem corrosion, gland over-tight)
- Stick-slip pattern — alternating high-friction zones indicating packing damage or uneven stem wear
- High running friction — body seat damage, guide wear, or foreign material in the valve body
- Excessive seating force — seat wear or misalignment causing high closing force
- Actuator pressure dropout at end of stroke — possible air supply problem or actuator seal leak
By comparing current signatures against baseline (as-commissioned) signatures, you can see exactly how the valve condition has changed over time — without any intrusive inspection.
Emergency Shutdown Valve Proof Testing
ESDs are the valves that keep people safe. They may sit fully open for years without stroking, and their functional reliability on demand is critical. But full stroke testing (taking the valve fully closed) requires a process shutdown — which is expensive and often not practical.
Partial Stroke Testing (PST) solves this. By stroking the valve only 10–20% of its travel (enough to verify it can move on demand without actually shutting the process), PST provides a measurable proof test that can be performed with the plant running. The test is documented, generates a signature, and its diagnostic value is credited in the SIL calculation as a reduction in PFD (Probability of Failure on Demand).
PST should be performed at defined intervals (typically quarterly or semi-annually) between full stroke tests at turnarounds. The proof test interval, partial test coverage, and full test coverage must all be consistent with the SIL calculation for the safety instrumented function.
Building a Risk-Based Valve Maintenance Strategy
The framework:
- Step 1: Classify every valve by criticality — safety-critical (ESD, PSV, BMS shutdown), process-critical (major control valves, key isolation), and non-critical (utility services, drain valves)
- Step 2: Assign maintenance strategy by criticality — safety-critical = predictive + mandatory proof tests; process-critical = predictive with smart diagnostics; non-critical = run-to-failure or time-based at turnaround
- Step 3: Implement diagnostic capability for process-critical and safety-critical valves — smart positioners on all control valves; PST systems on ESD valves; HART or fieldbus integration into control system for continuous monitoring
- Step 4: Establish baseline signatures for all monitored valves — document as-commissioned condition for comparison
- Step 5: Review diagnostic data regularly — set alarms for deviation from baseline; investigate deviations before they become failures
- Step 6: Track maintenance history — use the data to refine your maintenance intervals; extend intervals for consistently healthy valves, shorten for problematic ones
The plants that implement this approach properly see dramatic reductions in unplanned maintenance and, more importantly, in unplanned process shutdowns. The data doesn’t lie — and it will tell you which valves need attention long before they fail if you’re listening to it.