Choosing an actuator is not just picking one that bolts onto the valve; it is making sure the actuator can produce enough force to move that valve under the worst conditions it will ever see. Actuator sizing is the calculation that adds up every force the actuator has to overcome, the fluid pushing on the plug or disc, the load needed to seal the seat, the friction in the packing and bearings, and for rotary valves the dynamic torque as flow tears through a partly open valve, and then confirms the actuator delivers more thrust or torque than that total, with margin. Get it wrong and the actuator stalls at high differential pressure, leaving the valve stuck part-open exactly when it needs to move. This page walks through what goes into the force budget.
Control Valve Actuator Sizing in one line: Control valve actuator sizing calculates the total thrust (for linear valves) or torque (for rotary valves) the actuator must deliver to stroke the valve under worst-case conditions, then selects an actuator that exceeds it with a safety margin. The force budget includes the unbalanced fluid force across the plug or disc, the seat load for shutoff, packing and stem friction, and, for rotary valves, dynamic torque at partial opening. A fail-safe spring changes the budget on both strokes.
The first and often largest term is the unbalanced force: the pressure difference across the valve acting on the exposed area of the plug or disc. When a valve is closed against a high differential pressure, the fluid presses on the closure member and the actuator has to overcome that push to move it. For a linear globe valve this shows up as thrust along the stem; for a rotary valve it becomes a torque trying to hold the disc or ball shut or to slam it open. Because this force scales with the differential pressure, the worst case for sizing is usually the highest dP the valve will see, often at or near the closed position, which is exactly where an undersized actuator gets into trouble.
Seat load is the next term, and it is the force the actuator must apply beyond simply reaching the closed position to actually seal the valve to its required leakage class. Pushing the plug into the seat hard enough to achieve tight shutoff, a Class V or VI seal, takes real force, and the tighter the required shutoff, the more seat load the actuator must supply. This is why shutoff class and actuator sizing are linked: specifying a tighter seat raises the thrust the actuator has to deliver at the end of the stroke, and an actuator sized only to move the valve, without accounting for seat load, may reach the seat but fail to seal it.
Friction rounds out the static budget. The packing that seals the stem grips it and resists motion, bearings and guides add their own drag, and all of that friction opposes the actuator in whichever direction it moves. Packing friction in particular can be substantial, and it grows over the valve's life as the packing ages or is tightened, so a prudent sizing allows for more friction than a brand-new valve shows. Add the unbalanced force, the seat load, and the friction together and you have the thrust or torque the actuator must beat at the hardest point in the stroke.
Rotary valves add a term that linear valves largely do not have: dynamic torque. As a butterfly disc or a ball rotates partly open, the flow rushing past the closure member exerts a fluid-dynamic torque that varies with the opening angle and the flow, and for some disc geometries it peaks not fully open or fully closed but somewhere in between. That mid-travel peak can be the true worst case for a rotary actuator, and sizing only for the seated, unbalanced condition can leave an actuator that seats and unseats fine but stalls partway through the stroke when dynamic torque spikes. Getting the dynamic-torque curve from the valve maker and sizing for its peak is what prevents that.
On top of the calculated worst-case force, actuator sizing applies a safety factor. The margin covers the things the neat calculation cannot fully capture: friction that rises with age, supply pressure that sags below nominal, wear, and simple uncertainty in the numbers. Sizing an actuator to exactly the calculated force leaves nothing for reality, so a sensible margin ensures the valve still strokes when the packing is old, the air header is a little low, and the process is at its worst all at once. The margin is not padding for its own sake; it is what keeps a valve that is right at the edge from becoming a valve that will not move.
A fail-safe spring return reshapes the whole budget because the actuator now has two strokes with opposite demands. On the powered stroke the air or motor must overcome the spring in addition to the process forces, so the spring adds to the load. On the fail stroke the spring alone, with no power, must overcome the process forces and friction to drive the valve to its safe position, so the spring has to be strong enough to close, or open, the valve against the unbalanced force and friction with nothing helping it. Sizing a spring-return actuator therefore means checking both directions: enough powered force to move against the spring and the process, and enough spring force to reach fail-safe against the process on its own. The fail direction is frequently the harder constraint and is exactly the one you cannot afford to get wrong on a safety valve.
The classic consequence of undersizing is a valve that stalls at high differential pressure. Everything checks out during commissioning, when the process may be at low dP, but when the plant reaches a condition where the pressure across the valve is high, the unbalanced force exceeds what the actuator can produce and the valve stops moving, or refuses to close the last bit into the seat, or cannot break away from the seat to open. It is a failure that hides until the worst-case process condition arrives, which is the worst possible time for a final control element to be immovable. Sizing for that worst-case dP, not the everyday condition, is what prevents it.
In service that stall is visible in the mismatch between what the controller is commanding and what the valve is actually doing. When a control loop drives its output hard toward open or closed and the valve position does not follow, or hangs short of the demand, the actuator is being asked for more force than it has. That divergence between commanded position and actual position, or between output and the resulting flow, is precisely what a monitoring system can capture even though it never measures actuator thrust directly.
A cloud SCADA platform such as Merobix historizes controller output, valve position feedback, and the pressures across the valve, so an actuator that is running out of muscle shows up as a valve that lags its command when the differential pressure is high, or one that repeatedly fails to reach full seat at the worst process conditions. For remote and unmanned oil and gas sites this is how a sizing shortfall gets caught without someone standing at the valve, and the historized pressures let an engineer confirm that the stalls line up with high-dP episodes, pointing straight at an undersized or weakening actuator rather than a controls or sizing problem elsewhere. The actuator is sized once on the calculation; the monitoring layer is how you learn whether that calculation held up against the real process.
You add up every force the actuator must overcome at the worst point in the stroke: the unbalanced fluid force from the differential pressure across the plug or disc, the seat load needed to seal to the required shutoff class, and the packing and stem friction, plus dynamic torque for rotary valves. Then you choose an actuator whose thrust or torque exceeds that total with a safety margin. For a spring-return actuator you check both the powered and the fail strokes separately.
Because the unbalanced force across the closure member scales with the differential pressure across the valve. At low dP the actuator has enough force to move the valve, but when the process reaches a high-dP condition the fluid pushing on the plug or disc can exceed what the actuator can produce, and the valve stops moving or fails to seat. The failure hides until the worst-case pressure arrives, which is why sizing must use the highest differential pressure the valve will see, not the everyday one.
It splits the sizing into two strokes with opposite demands. On the powered stroke the actuator must overcome the process forces plus the spring, so the spring adds load. On the fail stroke the spring alone, with no power or air, must drive the valve to its safe position against the process forces and friction, so it has to be strong enough to do that unaided. The fail direction is often the harder constraint and is the one you cannot get wrong on a safety valve.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.