Control Valve Reaches Only 88% Travel at Full Output
A 4-inch air-to-open globe valve (fail closed) controls flow to a column reboiler. Nameplate data: spring-and-diaphragm actuator with a bench set of 3-15 psi, electro-pneumatic positioner, 4-20 mA input, 0-100% travel. During a plant uprate, operations reports the loop cannot reach the required high flow; the controller output sits at 100% and the flow is still short. Field data: instrument air header 95 psig; the filter-regulator at the valve is set to 20 psig and its gauge reads 20 psig. Stroking the valve with a calibrated 4-20 mA source gives: 4 mA -> 0% travel, 12 mA -> 51%, 18 mA -> 79%, 20 mA -> 88%. At 20 mA the positioner output gauge is pegged at 20 psig. No air leak is audible. Up-scale readings only were recorded.
1. Using the numbers, determine whether low air supply is the root cause. What does the saturated positioner output tell you about the likely fault?
Read the worked answer
Check the supply against the bench set first. With a 3-15 psi bench set, the actuator should reach 100% travel at 15 psi. Supply is 20 psig, which is 5 psi above the upper bench set - the usual minimum margin. So supply pressure by itself does not explain the failure.
Look at the gain.
- Observed: 12 mA -> 51% and 20 mA -> 88%. Slope = 37% / 8 mA = 4.63% per mA.
- Expected for a correctly ranged valve: 100% / 16 mA = 6.25% per mA.
- The valve is moving only about 74% as far per mA as it should.
Extrapolating the observed slope, 100% travel would require roughly 21.6 mA - beyond the 4-20 mA input range. Practically, the positioner runs out of pressure first: it hits the 20 psig supply limit while the valve is still at 88%, meaning the pressure actually needed for full stroke is above 20 psig, not the 15 psi the nameplate implies.
What the pegged output proves: the positioner is still commanding more pressure and cannot get it, so it knows it is not at setpoint. If the positioner's feedback linkage or cam were mis-calibrated, the positioner would stop driving at the setpoint with its output well below supply. Therefore this is a force/pressure problem, not a positioner feedback problem. The two candidates are: the actuator's real bench set is higher than the nameplate (wrong or replaced spring, wrong actuator), or friction/binding is consuming actuator force so more pressure is needed for the same travel.
2. Describe the tests that separate a wrong bench set from friction or binding, including the acceptance criteria you would apply.
Read the worked answer
Test 1 - Real bench set and friction (positioner out of the picture).
- Block the valve out of process service (or bench it) and take the controller to manual.
- Disconnect the positioner output and drive the actuator directly from a regulated air supply with a calibrated gauge, or place the positioner in its manual/calibration mode.
- Increase pressure slowly and record: (a) the pressure at which the stem first moves (breakaway), and (b) the pressure at full travel.
- Decrease slowly and record the same two points.
Interpretation:
- Breakaway near 3 psi and full travel near 15 psi -> bench set is correct; the fault is in the positioner, tubing, or the signal path.
- Breakaway near 3 psi but full travel needing 22-24 psi -> the spring/actuator is not the nameplate bench set. Compare the actuator nameplate, model, and serial against the datasheet and valve tag.
- Large separation between the up-scale and down-scale pressures at the same travel (>~1-2 psi) -> excessive friction/hysteresis: over-tightened packing, dry or galled stem, bent stem, or a binding guide.
Test 2 - Air capacity, not just gauge pressure. A regulator gauge can read 20 psig while flow is restricted by a partly closed isolation valve, a clogged filter element, or a kinked tube. Measure at the actuator or confirm flow capacity, and inspect the filter bowl and positioner screen.
Test 3 - True travel. Confirm the actual stem travel with a scale or index mark against the positioner's indicated travel to rule out a slipped linkage.
3. The valve is needed immediately and the process is running. What corrective actions and safeguards apply?
Read the worked answer
Corrective action depends on the test result:
- Wrong bench set: install the correct spring/actuator, or, if the actuator is rated for a higher spring range, raise the supply to at least the upper bench set + 5 psi (e.g., 30 psig for a 5-25 psi set) with tubing and a regulator sized for it. Never exceed the actuator case and diaphragm rating just to get travel.
- Friction/binding: re-pack to the manufacturer's torque specification, lubricate per the maintenance manual, and check stem straightness. Re-stroke after repair.
- Restricted air: replace the filter element, open the isolation valve fully, re-terminate or replace kinked tubing.
Verification after repair: stroke 0/25/50/75/100% up and down with the mA source, record travel and hysteresis at each point, confirm linearity within the positioner's specification, and confirm the air-failure action by tripping the supply - the valve must fail closed, with the closure time recorded.
Process safeguards before touching the valve:
- The loop is at 100% output and short of demand, so the column is already below target. Notify the control room and the process engineer.
- Obtain a maintenance work permit and an MOC for blocking out or bypassing the valve; define a manual backup (bypass valve or a fixed manual position) before the valve leaves service.
- Do not raise supply pressure above the actuator rating without engineering approval, and do not leave the positioner in manual with the loop in automatic.
- Record the as-found condition before adjustment for the maintenance history.