Valve Partial Stroke Testing: Proving ESD Valve Integrity Without a Full Trip

The Problem with Testing an ESD Valve

Every emergency shutdown valve carries a hidden obligation: it must work on demand, and you must be able to prove it. The traditional proof is a full stroke test — close the valve completely, verify tight shut-off, reopen it. The trouble is that a full stroke test means a process interruption. In continuous plants — refining, petrochemicals, gas processing — a planned trip of an ESD valve is itself a process upset, sometimes a flare event, and occasionally an unplanned trip of the whole unit. Plants respond rationally: they stretch the test interval, they defer the test to the next turnaround, and the ESD valve quietly becomes the least-tested, highest-consequence device in the safety instrumented system.

Partial stroke testing, PST, exists to break that trade-off. It gives maintenance teams a way to exercise most of the valve’s travel — proving that the mechanism actually moves — without ever interrupting flow.

How Partial Stroke Testing Works

In a partial stroke test, the valve is commanded to move a limited distance from its normal open position and then returned. Typical stroke ranges sit between 10% and 30% of travel — the valve might be driven from fully open to 85% open and back. The principle is simple: most dangerous failures of a shutdown valve reveal themselves as motion problems long before they become shut-off problems.

The test can be performed manually, with a technician using the handstation and observing the position indicator, or automatically, with a smart positioner executing a scripted test routine while the process runs undisturbed. The diagnostic version is what modern plants mean by PST: the positioner compares commanded position against measured position throughout the test, calculates stiction and hysteresis, flags deviations, and logs the result.

Choosing the stroke range is an engineering decision, not a checkbox. Stroke too little — under roughly 10% — and the test proves almost nothing about the stem, disc and seat. Stroke too much and you risk disturbing the process, particularly on control loops sensitive to flow changes, or on valves where reopening under high differential can score soft seats. Many operators settle on a bump of 15–20%, adjusted per valve based on its service and loop response.

What PST Can and Cannot Detect

A partial stroke test is genuinely diagnostic, not ceremonial. Performed properly, it exercises the actuator, stem, bearings, disc or ball, and — critically — breaks the static friction between the closure element and its seats. That last point matters: a ball valve parked closed for years in ESD service can seize its seats, and PST forces the ball to break free while there is still time to fix it.

Failure modes detectable by PST:

• Mechanical binding or seizure of stem, trunnions, bearings or gearbox.
• Seat adhesion — the valve cannot break away from its closed or parked position.
• Insufficient actuator force: a degrading cylinder, an air supply problem, or an undersized actuator shows up as slow or incomplete travel.
• Spring degradation in spring-return actuators, seen as sluggish return motion or reduced breakaway torque.
• Air leaks, solenoid faults and positioner problems in the actuation chain.
• Position feedback drift and stuck open/closed indication.

What PST does not prove is tight shut-off. No partial test verifies the seat seal, and no partial test confirms full closure within the required safety time. That limitation is not a weakness of the method — it is simply a boundary that the SIL verification must account for.

Where PST Fits in SIL and SIS Management

Emergency shutdown valves are final elements of a safety instrumented function, and their proof test interval is set during SIL verification. The math is unavoidable: the proof test coverage factor determines how much dangerous undetected failure the valve accumulates between tests. A full stroke test catches nearly everything; a partial stroke test catches most mechanical failures but none of the seat leakage failures.

The recognized approach, reflected in IEC 61511 practice, is to treat PST as a diagnostic layer that reduces the dangerous failure rate of the final element, allowing longer intervals between full proof tests — or, in some cases, supporting a higher SIL claim for the same hardware. For the calculation to hold, the PST itself must be credited properly: it needs defined coverage, a defined frequency, and a defined response when it fails. A diagnostic test that nobody reviews is worth nothing.

Typical arrangements in operating plants:

• PST executed monthly or quarterly on high-consequence ESD valves, with full stroke testing at the turnaround.
• PST integrated into the positioner’s asset management software, with alarms routed to operations when a test is missed or fails its tolerance band.
• Manual PST procedures for valves without smart positioners, using a documented bump test with witnessed position readings.

The Process Impact and Residual Risk

Even a 15% stroke bump moves a valve that is sitting in a live process. On most services the disturbance is negligible — the flow through a 15%-open gap in a large ball valve barely registers on the control system. But not every service is forgiving. Tight control loops may see a transient; compressors with anti-surge valves may react; and valves in very small lines can pass a meaningful flow change at small strokes.

Good practice is to screen each ESD valve before adding it to the PST program:

• Confirm with process engineering that the bump is tolerable, and record the justification.
• Rehearse the response path: if the PST stalls mid-travel, how does the valve return to open? Spring-return valves self-recover on loss of air; double-acting valves need a defined recovery sequence.
• Set interlocks so PST cannot be initiated during startup, shutdown, or other transient plant states.
• Remember that during the test, for those few seconds, the valve’s full protective function is not available. The SIL calculation already accounts for this in diagnostic coverage, but operating procedures should acknowledge it too.

Records and Trend Analysis

The value of PST compounds over time. A single test tells you the valve moved; a trend tells you where it is heading. Leading plants treat PST data like vibration data on rotating machinery:

• Plot breakaway torque or positioner drive current test over test — a rising trend on successive tests is an early warning of seat or bearing degradation long before the valve actually fails a test.
• Log stiction and hysteresis figures against the positioner’s own tolerances, not just pass/fail flags.
• Record every aborted or failed PST, the corrective action taken, and the retest result, so the safety file is complete for audits.
• Feed recurring deviations back into the maintenance strategy — a valve that trends badly may deserve earlier replacement at the next turnaround, or a change to a more suitable closure element.

For new projects, this is one more reason to specify smart positioners with diagnostic capability on ESD service, and to verify during control valve and actuation package review that the PST routines, alarm handling and data export are part of the delivered scope rather than an afterthought.

Limitations, and Why Full Stroke Testing Still Matters

PST is a complement, not a replacement. The seat still needs to be proven tight, and the full emergency closure still needs to be exercised at the interval defined by the SIL study. The right program layers both: frequent partial strokes to keep the mechanism alive and catch mechanical degradation, plus scheduled full closure tests at turnarounds or planned windows to verify shut-off and response time. Together they give the plant what it actually needs — evidence, on a defensible interval, that the last line of defense will close when it is called.