Control Valve Stiction and Low Air Supply on a Temperature Loop
A temperature control loop on a shell-and-tube exchanger controls product outlet at 120 °C by modulating a 4-inch pneumatic globe valve with a digital positioner. The controller output is 4–20 mA; the valve has spring-return fail-closed action and a 0–100% position feedback signal to the DCS. The loop cycles about ±8 °C with a period of roughly 4 minutes. When the operator puts the controller in manual at 50% output, the valve feedback shows 43%. A step test from 50% to 55% produces no stem movement until the output reaches 58%, after which the valve jumps to 61%. A hysteresis check shows about 9% deadband. The valve was rebuilt 14 months ago, and the air supply header is at 65 psig; the positioner nameplate requires 80 psig minimum. The exchanger has a manual bypass, but product quality is critical and the unit is at 95% rate. The previous shift changed the PID gains to reduce the cycling.
1. Is the cycling primarily a tuning problem? Use the step-test and air-supply data to justify your answer.
Read the worked answer
No. The step test shows a physical deadband: a 5% output change produces no stem movement, then the valve jumps 3% when it breaks free. A 9% hysteresis check and a 43% feedback at 50% command point to friction or insufficient actuating force, not to PID tuning. A controller cannot remove valve deadband; it can only react after the valve moves. Detuning may slow the cycle, but it will not eliminate the limit cycle and may increase offset and product-quality variation. The air supply is also below the positioner nameplate minimum of 80 psig, which can reduce available thrust and make packing friction or stem sticking worse. The correct diagnosis is mechanical and pneumatic. Changing PID gains to mask a 9% deadband is treating the symptom, and it leaves the loop vulnerable when the valve sticks in a different position.
2. What checks and temporary actions would you take this shift, and what safety constraints apply?
Read the worked answer
Place the loop in manual and hold a safe output while monitoring product temperature. Check the air supply at the header, filter regulator, and positioner inlet; inspect for tubing leaks, a clogged filter, and a failed regulator. Verify the positioner zero and span, the I/P converter, and the feedback linkage. Perform a stroke test and, if available, capture a valve signature to separate packing friction from actuator or positioner problems. Do not force the stem or adjust packing blindly on a hot line. The manual bypass may be used only with operations approval and a documented MOC; product quality and exchanger tube integrity must be monitored. If the air supply can be restored to at least 80 psig, retest the step response before returning to automatic. If not, keep the loop in manual, schedule repair, and escalate the product-quality risk rather than detuning the controller.
3. What records and follow-up would you require after the repair?
Read the worked answer
Record the as-found condition: air-supply pressure, positioner input versus stem position, deadband and hysteresis, valve signature if available, and the parts or adjustments made. After repair, repeat the step test at 25, 50, and 75% and record as-left deadband and hysteresis. Calibrate the positioner and verify fail-closed action. Return the loop to automatic only after the mechanical problem is corrected and the temperature response is stable; then evaluate whether the earlier PID changes should be restored. Update the CMMS history, add an air-supply and filter-regulator check to the PM, and note the temporary manual operation and MOC in the shift log. If packing was tightened, monitor stem friction at operating temperature. The goal is to prove the valve, not just the controller, is healthy.