A valve that never settles wears out its packing, swings the process and makes operators lose trust in the loop. This guide shows how to read the trend, find the real cause and fix it in the right order.
A hunting control valve keeps moving up and down around the setpoint instead of holding a steady position. Reading the trend shape, testing in manual and checking the valve before retuning finds the root cause quickly.
What Is Control Valve Hunting and Why Does It Happen?
Control valve hunting is a sustained oscillation in which the valve keeps opening and closing around the required position, so the process variable cycles above and below the setpoint instead of settling. It is caused either by the valve assembly itself, such as stiction, a worn positioner or an oversized trim, or by the control loop, such as aggressive tuning or an interacting loop.
- Put the controller in manual: if the cycle stops, the cause is inside the loop.
- A square wave PV with a sawtooth controller output points to stiction.
- Smooth sine waves on both PV and output usually mean aggressive tuning.
- Process gain above 3 or below 0.3 suggests a sizing or range problem.
- Fix the valve first, then retune the controller.
Every closed loop has three players: the transmitter, the controller and the final control element. Hunting appears when one of them adds too much gain, too much delay or a dead zone, and the loop then keeps correcting its own corrections.
Please always remember that continuous hunting deteriorates the gland packing, so the problem tends to get worse with time. A valve that moved a few millimetres every minute for a year has done far more strokes than one that held still.
George Buckbee writes in Control Engineering that more than 30 percent of control valves have problems, and many of them can be found with simple online tests. That is why a trend check is the cheapest first step in any hunting complaint.
Control Valve Hunting Causes
The causes of control valve hunting fall into two groups, and the first job is to decide which group you are dealing with. The table below lists the usual suspects found in Indian refineries, power plants and chemical units.
Friction from hard packing, sticky fluid or damaged trim makes the stem stick, then jump.
The valve works near its seat, so a tiny stem movement changes flow a lot.
Worn spool, loose linkage or very high positioner gain makes the valve overshoot.
Too much proportional gain or too little integral time makes the controller itself oscillate.
Low, unstable or wet instrument air slows the actuator in one direction.
Before you touch any tuning parameter, walk to the valve and watch the stem and the positioner gauges for one full cycle. Many hunting complaints end right there with a loose linkage, a leaking tube fitting or a regulator set too low.
Control Valve Hunting Trend Patterns Table
The shape of the trend is the best clue to control valve hunting, and it costs nothing to read. Control Engineering describes the classic stiction signature as a sawtooth controller output with a square wave process variable when the loop is in automatic.
| Trend Pattern | PV Shape | Output Shape | Likely Cause |
|---|---|---|---|
| Stick and jump | Square wave | Sawtooth or triangle | Stiction, tight packing |
| Smooth cycling | Sine wave | Sine wave | Aggressive tuning |
| Cycling near closed | Spiky, large swings | Small steps | Oversized valve or trim |
| Fast chatter | Small, fast ripple | Almost constant | Positioner gain or linkage |
| Cycle remains in manual | Any shape | Constant | External disturbance or interacting loop |
| Overshoot after reversal | Long uneven cycle | Ramps then jumps | Backlash, hysteresis |
Control Station adds that a stiction cycle is less sensitive to load and setpoint changes than a tuning cycle. Plotting PV against output across one cycle gives a parallelogram shape for a sticky valve, while a tuning problem gives a smoother ellipse.
How to Diagnose Control Valve Hunting Step by Step
Work from the cheapest test to the most invasive one, and inform the panel operator before you switch the loop to manual. The DCS manual, auto and cascade modes decide what the operator sees, so agree on safe limits first.
Buckbee recommends at least two small output changes in each direction in manual, because one step can hide backlash. If the output has to move more than 0.5 percent each time before the PV responds, he says there is a problem to fix.
Smart positioners make this much easier, because the valve position feedback is already available over HART. Compare the travel reading with the output on a digital positioner and you can see stick and jump steps without guessing from the PV.
The Emerson Fisher bulletin on dead band plus hysteresis estimation runs a 29 point test with 0.25 percent steps and holds each step for 45 seconds. Its X Y plot shows the band width directly as the combined dead band and hysteresis.
Valve Stiction vs Tuning: Which One Is It?
Stiction is the extra force needed to start a stopped valve moving. Control Engineering lists packing gland torque, fluid viscosity and the plug and seat design as the main contributors, which is why the packing and the trim are always the first places to look.
The controller makes stiction worse without knowing it. While the stem is stuck, integral action keeps ramping the output, and when enough force builds up the valve jumps past the target, so the cycle repeats.
- Square or flat topped PV with sudden jumps.
- Sawtooth or triangular controller output.
- Parallelogram shaped PV versus output plot.
- Cycle continues even with gentle tuning.
- Smooth, sinusoidal PV and output.
- Cycle period changes when gain is cut.
- Elliptical PV versus output plot.
- Valve tracks the output step for step.
Detuning a sticky valve only stretches the cycle and lowers its amplitude, as Control Station points out, but the oscillation does not go away. Follow the methods in how to tune a PID controller only after the valve passes its bump test.
Oversized Valve and Process Gain
An oversized valve spends its life just off the seat, where a 1 percent stem movement can change the flow by many percent. The loop gain becomes so high that any reasonable controller setting starts to cycle, which is a common source of control valve hunting after generous control valve sizing margins.
Buckbee gives a simple field check: the overall process gain in percent PV per percent output should stay between 0.3 and 3. A value outside that band points to valve sizing or transmitter scaling problems, and his fix ranges from new trim to a new valve.
ΔPV % = PV change ÷ span × 100
Solved example: range 0 to 200 m³/h
Output step 40 % to 45 %, ΔCO = 5 %
Flow 80 to 116 m³/h, change = 36 m³/h
ΔPV = 36 ÷ 200 × 100 = 18 %
Gain = 18 ÷ 5 = 3.6
Above 3: suspect oversized valve
An inherent equal percentage characteristic also changes the gain along the stroke, as explained in flow characteristics of control valves. Always repeat the gain test at the normal operating opening, not at a convenient one.
Process Gain Calculator
If the valve normally runs below about 15 percent open, ask process engineering for the real minimum and maximum flows. A reduced trim is often a cheaper fix than a smaller valve body.
Positioner Gain, Air Supply and Hardware Faults
A positioner is a fast inner loop, and if its gain is set too high it can cycle on its own even when the controller output is steady. Worn spool valves and broken feedback links cause the same fast control valve hunting, as Automation Forum explains.
Air supply is the forgotten cause of control valve hunting. Control Engineering advises checking that supply pressure meets the manufacturer recommendation, and wet instrument air can freeze or corrode small orifices inside the positioner relay.
- Supply pressure at the regulator matches the actuator nameplate.
- No leaks at tube fittings, diaphragm casing or piston seals.
- Feedback linkage tight, with no play at the stem connector.
- Positioner travel calibration done, zero and span correct.
- Packing nuts at recommended torque, not over tightened.
- Valve position feedback trends match the controller output.
- Actuator size gives enough thrust margin over packing friction.
Buckbee also suggests confirming that the actuator and positioner are sized for the required force. A small diaphragm actuator fighting tight graphite packing will always move in jumps.
Hysteresis and Backlash Limits
Backlash is the part of hysteresis caused by lost motion in loose or worn linkages between the positioner, actuator and valve. Control Engineering treats 1 percent or less as normal for a well performing valve, and says anything above 3 percent should be corrected as soon as practical.
The Emerson Fisher bulletin agrees that dead band plus hysteresis is process dependent but typically 1 percent or less for the complete assembly. It also calls dead band the dominant nonlinearity in a control valve, mostly from friction in packing and seals.
Proven Fixes for Control Valve Hunting
Control Station recommends benchmarking loop variability, output travel and time in auto before and after the repair, so the improvement is visible to management. Keep the trends together with the control valve troubleshooting record for the next shutdown planning.
Control Valve Hunting Myths
Fisher Dead Band Estimation Bulletin PDF
Video on Stiction and PID Control
Control Valve Hunting FAQ
It is a sustained oscillation in which the valve keeps opening and closing instead of holding a steady position. The process variable then cycles around the setpoint and never settles.
It can come from the valve assembly, such as stiction or a faulty positioner, or from the control loop itself. Packing wear, product quality swings and operator distrust all grow the longer it continues.
The common causes are stiction, an oversized valve, a faulty or badly tuned positioner and aggressive controller tuning. Low or unstable air supply and actuator leaks can also cause it.
An oscillation coming from another interacting loop is a further possibility. A simple manual test separates loop causes from external ones in a few minutes without any special tools.
To stop control valve hunting, find the cause before changing anything, starting with a manual test and the trend shape. Then repair mechanical faults such as tight packing, loose linkage or air leaks.
Calibrate the positioner and check the process gain for sizing problems. Retune the controller only after the valve tracks small output steps properly.
Look for a square wave process variable with a sawtooth controller output in automatic mode. A plot of PV against output that forms a parallelogram is another strong sign.
In manual, make small output steps in both directions and watch for a delay followed by a jump. Smart positioner travel trends show the stick and jump steps clearly.
Yes, because an oversized valve works close to its seat where a small stem movement changes the flow a lot. This raises the loop gain so much that normal tuning settings start to cycle.
A process gain above 3 percent PV per percent output is a useful warning sign. Reduced trim or a smaller valve usually solves it.
Stiction is one possible cause, while control valve hunting is the symptom you see on the trend. Stiction is static friction that makes the valve stick and then jump past its target.
Hunting is any sustained oscillation of the valve and the loop around the setpoint. It can come from stiction, tuning, sizing, positioner or air supply problems.
No, detuning only lowers the amplitude and stretches the period of a stiction cycle. The oscillation remains until the mechanical cause inside the valve assembly is repaired.
Repack, clean or repair the valve first and confirm it follows small output steps in both directions. Then retune the loop so it responds smoothly without any overshoot.
Related Articles
- Control Valve Troubleshooting
- Control Valve Positioner Working
- PID Controller Tuning Guide
- Control Valve Sizing Basic Requirements
- Flow Characteristics of Control Valves
External References
- Dead Band Plus Hysteresis Estimation with ValveLink Diagnostics, Emerson Fisher
- How to Identify and Troubleshoot Control Valve Problems On the Fly, Control Engineering
- Stiction, Wikipedia
What We Learn Today
- Control valve hunting is a sustained cycle of the valve and process, and a manual test quickly shows whether its source is inside or outside the loop.
- A square wave PV with a sawtooth controller output points to stiction, while smooth sine waves on both signals usually mean aggressive tuning.
- Repair packing, linkage, air supply and sizing faults first, check process gain between 0.3 and 3, and only then retune the controller.
