Contact Us
Home / Glossary Terms /

Partial Valve Stroke Test (PVST)

A partial valve stroke test (PVST) moves a safety shutoff valve through a short portion of its travel, typically 10 to 20 percent, and then returns it to position. It does disturb the process, because the valve genuinely moves, but the excursion is small and brief enough that a stable unit rides through it. The point is to catch the failure mode everyone worries about on a valve that has sat untouched for years, which is a seized or stuck actuator that will not begin to move on demand. Because that disturbance is minor, a PVST can be triggered automatically by the logic solver or by a smart positioner on its own schedule, and that automation is the whole reason the technique is attractive.

This means a PVST likely carries its own test interval, separate from and shorter than the full proof test interval. In the PFD math it is a separate test module with its own coverage, sitting alongside the full test rather than changing it. A site might partial stroke monthly or quarterly while the full proof test stays on a one to three year cycle tied to a planned major shutdown. The trade is coverage. A partial stroke tells you the valve is not seized, but nothing about full closure, seat tightness, or stroke time to the seat, so its proof test coverage (Cpt) is low, and it must always come out lower than the Cpt of the full proof test. Vendors commonly apply around 0.7 and some consultants argue for 0.7 to 0.8, but published coverage data for final elements is thin. NAMUR NA 106 (the 2018 revision of NE 106) now publishes conservative generic bands for partial and full stroke tests, and real-world coverage still falls well below those numbers in dirty service, or where the fluid can coagulate at a seat that sits fully open in normal operation.

The common mistake is treating a PVST as a substitute for the full proof test rather than a supplement to it. A PVST earns a modest improvement in PFDavg between full tests, and that improvement is real. It does not earn permission to push the full proof test interval out indefinitely, because the failure modes a PVST cannot see keep accumulating on exactly the same clock they always did.

Key Points

  • A PVST moves the valve through a small fraction of its travel, usually 10 to 20 percent, disturbing the process only slightly.
  • Because that disturbance is minor the test can be automated, a PVST likely runs on its own interval that is shorter than the full proof test interval.
  • Its proof test coverage (Cpt) is low and must always come out lower than the full proof test’s, since a partial stroke cannot verify full closure, seat tightness, or stroke time to the seat.
  • A PVST supplements the full proof test rather than replacing it, so the full test interval still has to be justified on its own terms.

Example

A 1oo1 emergency shutdown valve on a gas export line is fully proof tested every 24 months, during the planned major shutdown when the line can finally be taken out of service. Between those shutdowns the smart positioner strokes the valve 15 percent every three months, triggered automatically by the safety logic solver, and logs the result to the historian. The SIL verification models that quarterly partial test at a proof test coverage (Cpt) of 0.6 against the stuck valve failure modes only, with the remaining dangerous undetected failures still governed by the 24 month full test interval which has a Cpt of 0.9.

See Also: proof test coverage (Cpt), proof test interval, final element, DTI

Cited Sources

Part Of: testing related category