How to Verify a Control Valve Actuator's Bench Set
Bench set is the air pressure range over which a spring-and-diaphragm actuator moves through its full travel with no process load and no packing friction, the condition it was set up in at the valve shop bench. Verifying it in the field tells you whether the spring still matches the nameplate and whether the actuator can still develop the seat load the service needs. This page covers how to run the check with a regulator and a good gauge, and how to interpret readings taken on an installed valve, which will never quite match the bench numbers.
Verify Actuator Bench Set in one line: To verify an actuator's bench set, disconnect the positioner output, apply air directly to the actuator through a precise regulator, and raise the pressure slowly while watching the travel scale. Note the pressure where the stem first moves and the pressure where it reaches full travel, then compare against the nameplate spring range. On an installed valve, expect packing friction to delay the start and finish by a roughly equal band in each direction; a shifted or stretched range points at the spring, not the friction.
Set Up a Direct Air Feed to the Actuator
You need a manual air regulator with a gauge you trust, teed directly into the actuator's diaphragm port with the positioner output disconnected or bypassed, and the valve's nameplate or datasheet showing the spring range. Common spring ranges follow the classic pneumatic signal spans, 3-15 psi being the historic one, with wider ranges used where more seat load is needed; the number that matters is the one stamped for this assembly.
The valve should be isolated from the process or in a shutdown state, because the check will stroke it fully in both directions. Know whether the assembly is air-to-open or air-to-close, since that decides whether rising pressure lifts the plug off the seat or drives it toward the seat, and therefore which end of the stroke carries the seat load.
Stroke the Actuator Slowly and Record Both Endpoints
Raise the pressure smoothly and slowly from zero, watching the stem and the travel scale, and record the pressure at which the stem first moves. Keep increasing until the valve just reaches full travel and record that pressure too. Then bleed back down equally slowly and record the same two events in reverse. Slow matters: stroke fast and you are measuring air flow dynamics rather than the spring.
The ascending and descending readings will not match, and the difference is information. The gap between the pressure that starts motion going up and the pressure that resumes motion coming down is the friction band, mostly packing. The midpoints between those pairs approximate the true spring behavior with friction averaged out. This is the same physics a valve signature test plots automatically as a pressure-versus-travel loop; doing it manually with a regulator gives you the two endpoints without the instrument.
Compare Bench Numbers Against the Installed Reality
A true bench set is measured with no packing friction and no process load, so an installed valve should bracket the nameplate numbers, not land on them. As a worked example in symbols: if the nameplate spring range runs from a lower pressure L to an upper pressure U, and the friction band you measured is F, expect motion to begin near L plus half of F on the way up, and full travel to arrive near U plus half of F, with the descending pass shifted the other way by the same amount. Friction widens the apparent range symmetrically; it does not move its center.
What friction cannot explain is a shifted center or a stretched span. Travel that starts well above the nameplate range on both passes suggests a spring that has been compressed by a previous adjustment or a wrong spring installed. A span much wider than nameplate, needing far more pressure change than the spring range to complete travel, can mean a damaged spring or binding in the guiding. And on the seating end, remember the plug needs pressure beyond the travel endpoint to develop seat load, which is part of actuator sizing rather than a fault.
Verifying the Result
Cross-check the friction interpretation by repeating the stroke two or three times; friction effects repeat, mechanical damage often does not. If the numbers still disagree with the nameplate after accounting for the friction band, confirm the nameplate itself matches the datasheet for the tag, because springs get swapped during overhauls and nameplates do not always follow.
Reconnect the positioner when done and rerun its calibration if you disturbed the air connections, following the procedure for calibrating a valve positioner. A bench-set check done with the positioner still in the circuit is the most common way this measurement goes wrong, because the positioner's relay adds and bleeds air on its own and your regulator is no longer in charge of the diaphragm.
Common Mistakes
Stroking too fast is the big one: the reading becomes a race between your regulator and the actuator volume, and both endpoints land high on the way up and low on the way down. Reading only the ascending pass is next, because without the descending pass you cannot separate friction from spring shift, and you will condemn a healthy spring for what is actually tight packing.
The subtler mistake is expecting the installed valve to reproduce bench numbers exactly and adjusting the spring until it does. Adjusting spring compression to cancel packing friction moves the true bench set away from its design value, costs seat load on one end of the stroke, and leaves the next technician a valve whose nameplate no longer describes it. If friction is the problem, fix the packing, not the spring.
Frequently Asked Questions
Why do my installed readings not match the nameplate bench set?
Because bench set is defined with no packing friction and no process load, and an installed valve has both. Packing friction delays the start of motion and the arrival at full travel by roughly half the friction band in each direction, widening the apparent range symmetrically. If the center of the measured range has shifted or the span has stretched well beyond the friction band, then the spring or its adjustment is the suspect.
Can I verify bench set without disconnecting the positioner?
Not reliably with a manual regulator, because the positioner's relay keeps adding and venting air to hold its commanded position, so the diaphragm pressure is not the pressure you set. Either disconnect or bypass the positioner output and drive the actuator directly, or use the positioner's own diagnostic signature test, which plots actuator pressure against travel and reveals the same spring endpoints.
Automation services
Need help turning this into a working system?
Merobix integrates SCADA, programs Allen-Bradley and Siemens PLCs, and designs and fabricates industrial control panels.
Meeting requests are reviewed before confirmation.