- Actuator sizing is one of the most commonly under-engineered aspects of valve automation — undersized actuators fail to operate valves under worst-case conditions, while oversized actuators damage seating surfaces and drive up project cost; getting the calculation right requires understanding every torque and thrust component across the full operating envelope.
- The total actuator output requirement is not simply the valve break-out torque — it is the sum of seating torque, unseating torque, running torque, packing friction, stem friction, and dynamic flow forces, all evaluated at the worst-case condition of minimum supply pressure with maximum differential pressure across the closed valve.
- A safety factor on actuator output is not a fudge — it is an essential engineering allowance for packing wear, lubricant degradation, temperature effects on elastomers, and the inevitable variation between catalogue torque figures and as-installed valve behaviour.
Why Actuator Sizing Gets It Wrong So Often
I have seen more automated valve failures caused by poor actuator sizing than by any other single engineering error. The valve leaks, the actuator stalls mid-stroke, the valve will not open against line pressure, the gearbox strips its teeth on the first emergency closure — and in almost every case, when you dig back through the engineering, the actuator was sized against a single torque figure taken from a catalogue without applying any engineering judgment to what that figure actually represents or what the valve will actually experience in service.
Actuator sizing is the calculation of how much force or torque the actuator must deliver to operate the valve reliably across every condition the valve will encounter during its service life — not just at the design point on a dry lab bench, but at line pressure, at temperature, with aged and worn packing, with potential corrosion or fouling of the stem, and with the supply medium at its minimum pressure tolerance. When you frame it that way, the calculation becomes significantly more involved than looking up a single number in a table.
Understanding Valve Operating Torque and Thrust Components
For quarter-turn valves (ball, butterfly, plug), the relevant output quantity is torque measured in Newton-metres or foot-pounds. For linear motion valves (gate, globe, diaphragm), the relevant quantity is thrust measured in Newtons or pounds-force. Some quarter-turn valves with very high stem loads are also characterised in terms of thrust on the stem bearing. The key torque and thrust components that must be accounted for are as follows.
Break-out torque (or break-out thrust): the force required to initiate movement from the seated (fully closed) position. This is typically the highest torque value in the operating cycle because static friction is higher than dynamic friction, the seat sealing load has been applied to the seat surface by stem thrust or spring force, and any thermal cycling since the last operation may have caused the seat and stem to bind. For ball valves, the break-out torque is strongly influenced by seat material: soft seats (PTFE, nylon) have higher static friction coefficients than metal seats, but soft seats also self-adjust — their break-out torque can increase substantially over time as the PTFE seat deforms around the ball surface under sustained contact load.
Running torque: the torque required to continue moving the valve disc or ball through its travel once movement has started. For butterfly valves this varies significantly across the stroke because the disc is asymmetric relative to the flow axis, and hydrodynamic forces from the process fluid create a torque that adds to or subtracts from the actuator requirement depending on the direction of flow and disc orientation. At around 60 to 70 degrees of opening, the hydrodynamic torque on a butterfly valve disc can be at its maximum — and if the actuator is sized only for the seated and end-of-travel conditions, it may stall at this mid-stroke peak.
End-of-travel seating torque: the torque required to seat the valve fully closed and achieve the required seat tightness (ANSI/FCI 70-2 Class IV, V, or VI for soft-seat valves; API 598 leakage rates for metal-seat valves). For soft-seat ball valves, the seating load is largely determined by the seat spring design rather than being an independent variable; for metal-seat butterfly valves and gate valves, the seating torque required to achieve acceptable bubble-tight leakage is substantially higher than the running torque and must be explicitly evaluated.
Packing friction: the axial or rotational force required to overcome packing drag on the stem. This depends on packing type (PTFE, graphite, live-loaded), gland compression, stem diameter, and packing temperature. A PTFE packing set at a 50 mm stem diameter can contribute 200 to 500 Nm of friction torque alone. For HIPPS (High Integrity Pressure Protection System) valves and ESD valves where reliability is paramount, the packing friction contribution to total actuator requirement is formally calculated using measured or conservatively estimated values from the valve supplier.
Differential Pressure Effects: The Most Underappreciated Variable
Differential pressure across the closed valve is the most significant variable in actuator sizing for many valve types, and it is also the variable most frequently evaluated incorrectly. The error is to size against the normal operating differential pressure rather than the maximum credible differential pressure the valve could experience in service.
Consider a block valve on a process line with a design pressure of 100 bar. During normal operation, differential pressure across the closed valve might be 10 to 20 bar — the line pressure on one side, near-atmospheric on the other, with the two sides usually equalised before the valve is operated. But under emergency conditions — sudden downstream blowdown, loss of downstream system, emergency isolation — the full design differential of 100 bar may be present across the closed valve at the moment the actuator must operate it. If the actuator was sized for 20 bar differential, it will not be able to move the valve. The HIPPS system or ESD system fails to perform its function at exactly the moment when performance matters most.
The correct approach is to determine the worst-case differential pressure at which the valve must be operable and size the actuator for that condition. For emergency shutdown valves per IEC 61508 and IEC 61511 (functional safety in process industry), this requirement is embedded in the SIL assessment — the actuator must achieve the required PFD (probability of failure on demand) target, which requires demonstrated operability at worst-case conditions with appropriate safety margins.
For butterfly valves in particular, differential pressure creates a hydraulic unbalance force on the disc that translates directly to additional torque. The magnitude depends on the disc offset geometry, the disc area, and the pressure differential. For a 300 mm butterfly valve at 50 bar differential, the hydrodynamic torque component can be several thousand Newton-metres — easily double or triple the seated torque at zero differential pressure. If you are sizing a butterfly valve actuator from a low-differential-pressure test result and then applying a simple safety factor, you may be badly underestimating the actual operating requirement.
Pneumatic Actuator Sizing: The Spring-Return Calculation
Pneumatic actuators for quarter-turn valves (scotch-yoke, rack-and-pinion) generate torque as a function of supply pressure and piston area (or cylinder geometry for scotch-yoke). The available output torque varies across the stroke — for a rack-and-pinion actuator the output is approximately constant; for a scotch-yoke actuator the output torque varies sinusoidally with stroke position, being highest at 0 and 90 degrees and lowest at 45 degrees.
For spring-return actuators, the spring provides the fail-safe return torque but also opposes the pneumatic supply on the active stroke. The output torque delivered by the pneumatic circuit on the active stroke is: T_pneumatic – T_spring_preload. The return stroke torque delivered by the spring is: T_spring_end – T_pneumatic_back_pressure (if any). Both conditions must meet the valve torque requirement with the specified safety factor.
The minimum supply pressure must be used for the pneumatic output calculation, not the nominal operating pressure. If your instrument air system is nominally 7 bar but has a low-pressure alarm at 5.5 bar and a shutdown at 4.5 bar, the actuator must be sized to operate the valve at 4.5 bar supply — the point at which the instrument air is still available. Most good engineering specifications call for actuator sizing at minimum supply pressure minus 10 percent, with a further 25 to 30 percent safety factor applied to the total valve torque requirement.
Safety Factors: Engineering vs Wishful Thinking
The safety factor applied to valve actuator sizing is not arbitrary — it represents specific, quantified allowances for real-world variability. The main contributors are: valve manufacturing tolerance on torque (valve torque figures published by manufacturers are typically the mean value plus a tolerance; the datasheet figure often represents the 50th percentile, and the worst-case individual valve may be 20 to 30 percent higher); packing wear and re-packing (tighter gland nut settings after field re-packing can increase stem friction substantially); corrosion and fouling of the stem thread (increases stem friction for linear valves); lubricant degradation (loss of grease on ball valve trunnion bearings or stem threads over time); and temperature effects on elastomers (PTFE packing friction increases at low temperatures; soft seat materials stiffen at low temperatures).
A safety factor of 1.25 (25 percent) on the valve torque before actuator selection is a common minimum for routine process valves. For ESD and HIPPS valves, 1.5 or higher is more typical, with some operators requiring demonstrated operation of the physical valve at the maximum credible differential pressure with the actual actuator selected, confirmed during FAT (Factory Acceptance Test). For subsea valves, where the actuator cannot be accessed for maintenance and valve operability must be guaranteed over a design life of 20 to 30 years, safety factors of 1.5 to 2.0 are applied, and torque audits (actual torque measurement on the assembled valve-actuator package) are standard practice.
Electric Actuator Sizing: Torque, Speed and Thermal Rating
Electric actuators are sized differently from pneumatic actuators. The key parameters are output torque (Nm), operating speed (degrees per second or mm per second), number of starts per hour (thermal rating), and duty cycle. Electric actuator motors are thermally limited — a motor that can deliver 1,000 Nm continuously may be able to deliver 1,500 Nm for a short-time duty, but the number of operations before thermal cutout engages is limited.
For on/off service, the electric actuator is sized to the peak torque requirement (typically break-out or end-of-travel seating) with the standard safety factor, and the thermal rating is checked against the expected cycle frequency. For modulating control service, the duty cycle is more demanding — the motor runs more frequently and for longer periods, and the thermal rating becomes the governing constraint rather than the peak torque. Modulating electric actuators use motors with higher duty ratings and often incorporate forced cooling.
Speed selection for electric actuators is a separate consideration from torque. For most process valves, the operating speed is not critical — the valve simply needs to complete its stroke within a reasonable time. For emergency shutdown applications, however, valve closing speed can be critical: a valve that takes 30 seconds to close under emergency conditions may allow significant hydrocarbon release compared to one that closes in 5 seconds. ESD closing time requirements are typically defined in the process safety studies (HAZOP, SIL assessment) and become a specification requirement for the actuator. The trade-off is that faster closing speeds create greater water hammer risk (hydraulic transients) in liquid service — a factor that must be evaluated in the hydraulic system design.
Actuator Sizing for Specific Valve Types
Ball valves with soft seats have relatively low and predictable torque requirements at low to moderate pressures. At high pressure and large bore, the seat contact load becomes significant and torque can be high. Trunnion-mounted ball valves have lower operating torque than floating ball valves at the same pressure and bore size because the trunnion supports the ball and reduces the stem load; this is one of the key reasons trunnion mounting is standard for large bore and high-pressure service.
Butterfly valves require the most careful torque analysis of any quarter-turn valve because the hydrodynamic torque from the process fluid is significant and directional. Double-offset (double-eccentric) and triple-offset butterfly valves have different torque characteristics from concentric (resilient-seated) designs; the offset geometry creates a cam action that can assist closing (reducing actuator torque requirement) but adds to opening torque. This must be accounted for in the sizing calculation, and the actuator must provide adequate torque in both directions across the full stroke.
Gate valves use linear thrust actuators (pneumatic cylinder or electric actuator with stem nut). The key torque components are: stem thrust to overcome packing friction and move the gate against differential pressure, plus the additional thrust required to seat the wedge gate firmly against the seats at end of travel. For rising stem gate valves, the actuator must also support the weight of the stem and gate — not significant for small valves but substantial for large, heavy trim valves. For parallel slide gate valves, the seating load is generated by the differential pressure itself (the downstream seat sees line pressure pushing the slab gate against it), which means the closing thrust requirement is dominated by packing friction rather than seat load, but the valve is inherently differential-pressure-assisted to seat — a useful characteristic for emergency isolation.
Documentation and Torque Audit Requirements
A complete actuator sizing calculation should be documented with: valve make, model, size, and pressure class; valve torque or thrust requirement at each point in the stroke (break-out, running, end-of-travel) from the valve manufacturer’s certified torque data; differential pressure at worst-case operating condition; packing type and estimated friction contribution; total valve torque requirement including safety factor; supply medium minimum pressure; actuator model and rated output torque at minimum supply pressure; actuator-to-valve output ratio (safety margin confirmation); and fail position with spring torque at each point of the stroke for spring-return actuators.
For HIPPS and SIL-rated valve packages, actuator sizing documentation forms part of the safety case and is subject to independent review. The actual torque of the assembled valve is measured during FAT using a calibrated torque meter, and the measurement is compared to the sizing calculation to verify that the safety factor assumption was conservative. If the measured torque exceeds the calculated value, the actuator selection must be reconsidered.
Common Mistakes and How to Avoid Them
The most common mistakes in actuator sizing are: using the nominal supply pressure rather than minimum supply pressure for the pneumatic output calculation; applying the safety factor to the wrong parameter (applying it to the actuator output rather than to the valve torque requirement, which gives the same mathematical result but different engineering thinking and often leads to overlooking specific worst-case conditions); not accounting for hydrodynamic torque on butterfly valves at high differential pressure; sizing for mean valve torque from the catalogue rather than maximum-tolerance valve torque; not considering the effect of temperature on seat friction and packing friction at minimum ambient temperature; and forgetting that the safety factor must account for the entire service life of the valve, not just the as-new condition.
The consequences of getting it wrong range from nuisance to catastrophic. An undersized actuator that stalls an isolation valve under emergency conditions is not a maintenance problem — it is a process safety failure. The engineering investment to get actuator sizing right is modest; the cost of getting it wrong can be enormous. Treat the calculation with the rigour it deserves.