Views: 0 Author: J-VALVES Publish Time: 2026-11-10 Origin: Site
Most "the valve is stiff" complaints are not valve faults. They are arithmetic faults: an operator chosen against a catalog torque that was measured at the wrong differential pressure, a gear ratio mistaken for mechanical advantage, a stem asked to carry more torque than its material allows, or a multi-turn valve where the effort needed and the turns available simply cannot both be satisfied by a human arm. This guide works through manual operator selection from the torque side - what the valve demands, what a person can deliver, what the stem can survive - with the numbers that belong in the datasheet and in the RFQ.
Walk through any commissioning punch list and the same line appears in different words: gate valve on the pump discharge "very stiff, needs two men"; butterfly valve at the chiller header "won't open against pressure"; the gearbox on a 20-inch line "turns but nothing happens". Each of these has a specific arithmetic cause, and none of them is solved by a longer lever.
The reason is that manual operation sits at the intersection of three independent limits, and a valve only works when all three are satisfied at once:
The load limit. The torque the valve needs at the worst credible differential pressure, after the seat has sat loaded for months, with the packing at end-of-life compression.
The human limit. The torque a person can actually deliver at the rim of a handwheel or the end of a lever, for the duration the stroke takes - not the peak a fitter can produce once.
The structural limit. The torque the stem, key, stem nut and gearbox casing can carry without permanent deformation. This one is a ceiling, not a target.
Selection errors come from mixing them up: applying a safety factor to the human limit (which cannot be increased by specification), treating the structural limit as a comfortable margin (so an oversized operator becomes a failure mode), or quoting a torque figure without the differential pressure at which it was measured, which makes the number meaningless. A 12-inch ball valve might need 300 N m to turn with equalized pressure and 2,800 N m at full rated differential - same valve, same day, an order of magnitude apart.
Published torque data has a shape, and reading the wrong point on it is the single most common selection error. Every quarter-turn valve, and every multi-turn valve with a seated closure, produces three distinct values across one stroke:
Torque value | Where it occurs | What creates it | How it is used |
|---|---|---|---|
Breakaway (break-to-open, unseating) | First movement out of the fully closed, pressure-loaded position | Static friction at the seat, seat adhesion after long closure, stem packing stiction, plus the pressure-driven seat load | The governing number for operator selection. Operators are sized on breakaway at maximum differential, never on running torque |
Running | After the first few degrees, through mid-travel | Kinetic friction at seat, bearings and packing, plus hydrodynamic forces on the closure member | Checks whether the operator can carry the stroke to completion; on soft-seated quarter-turn valves it commonly runs at roughly 25 to 70 percent of breakaway depending on design |
Seating (end-to-close) | Final travel into the closed position | Re-establishing seat contact and elastic seat deflection; on wedge gates, the wedging action itself | Confirms closing and sealing capability; on metal-seated and wedge designs it can approach or exceed breakaway |
Two practical consequences follow. First, torque without a stated differential pressure is not data. A manufacturer's table measured on clean water at ambient temperature with equalized pressure describes a different machine from the one in a Class 600 hydrocarbon line. Ask for the torque at the specified differential - and if the valve will be opened against full shut-in pressure, ask for that case explicitly. Second, if a quotation lists a single "operating torque", ask which of the three it is. Vendors that publish only one number are usually publishing breakaway for a clean-service valve, with no service factor included - which is defensible only if the buyer applies the factor and the differential himself.
Torque is not one force. It is a sum of friction sources that scale differently with size, pressure and time in service, which is why rules of thumb borrowed from one valve type mislead on another:
Valve type | Dominant torque source | Secondary source | How torque scales | Selection consequence |
|---|---|---|---|---|
Floating ball, soft seat | Seat friction driven by pressure load on the ball (roughly 40 to 60 percent of breakaway) | Stem packing (commonly 20 to 40 percent), stem bearings | Seat load rises with bore area and differential; friction radius rises with bore - breakaway climbs steeply with size and pressure | Light at low pressure, heavy at rating; the design loses viability well before large sizes |
Trunnion ball, soft or metal seat | Seat friction with spring preload plus pressure-assisted seat load | Packing, trunnion bearings | Grows roughly linearly with seat contact area and with differential, not with the square of bore | Manages large sizes at high class; torque still rises with differential |
Butterfly (concentric, rubber lined) | Breakaway from peeling the disc out of an interference-fit liner | Hydrodynamic torque, peaking mid-stroke; bearing friction | Breakaway very sensitive to liner compression and to differential; hydrodynamic component peaks around 60 to 80 degrees of opening | The stroke is short but not uniform: an operator sized on breakaway still has to survive the mid-stroke peak - see the torque discussion in our butterfly versus gate valve selection guide |
Gate (wedge or slab) | Seat and wedge friction converted through the stem nut thread into stem torque | Packing friction, backseat contact, stem thrust | Torque = thrust x thread lead / (2 x pi x thread efficiency) plus packing torque - so it scales with thrust, not with bore diameter alone | Very sensitive to wedging force and to thermal binding after a hot closure |
Globe | Seat load plus unbalanced stem thrust | Packing, thrust bearing, backseat | Same thrust-to-torque conversion as the gate, over a longer stroke | Many turns, high thrust, and a stem that sees both torsion and compression |
The gate and globe rows deserve emphasis because they are where the mental model usually breaks. On a quarter-turn valve the operator applies torque directly to the closure member through the stem. On a multi-turn valve the stem turns a nut, the nut drives the stem axially, and the seating thrust is generated by that screw action. The stem therefore carries torsion and thrust simultaneously, and the torque at the handwheel is set by thread geometry at roughly 30 to 40 percent efficiency for a typical acme stem thread. That is why a gate valve with a modest seat load can still demand a large handwheel - and why the same valve, closed hot and allowed to cool, can demand far more than the drawing predicted.
Manual operation is one equation and one body of ergonomic data. The equation is T = F x R: stem torque equals rim force times the radius at which it is applied. A 400 mm handwheel gives R = 0.2 m, so a 600 N pull delivers 120 N m. Nothing in the valve changes this.
The ergonomic data is the part most datasheets ignore, and it is codified in MSS SP-91, Guidelines for Manual Operation of Valves. The standard's reference operator is a person able to exert approximately 670 N (150 lbf) on a lever of 300 mm effective length at waist level - about 200 N m of torque in the ideal case - and then applies input factors for position, duration and whether an impact device is used:
Input factor | Multiplier | What it means at the handwheel |
|---|---|---|
Position: waist to shoulder height | 1.0 | The reference posture; design the operator for this if you can choose the elevation |
Position: below hip or above shoulder | 0.5 | Half the capacity - the arithmetic reason chainwheels and extended stems exist |
Momentary force (breaking the valve loose) | 1.0 | Available for an instant, not for a stroke |
Short-term force (a fraction of the stroke) | 0.85 | Realistic for the seating and unseating phases |
Uniform force (up to about 5 minutes) | 0.7 | Realistic for continuous turning on a multi-turn valve |
Long-term force | 0.25 | Hours of work: a quarter of peak capacity |
Manual impact device, 300 mm (12 in) | 2.0 | Inertia multiplies the available torque - and the load on the drive train |
Manual impact device, 600 mm (24 in) | 3.0 | Higher stored energy, higher peak load on stem and key |
Manual impact device, 900 mm (36 in) | 4.0 | Effective only on valves with the stem mass and strength to absorb it |
Apply the factors and the comfortable numbers collapse. A 610 mm handwheel (24 in) at waist height, used with short-term force, gives a base capability of roughly 178 N at the rim, multiplied by 0.85 - about 151 N - which is 46 N m at the stem. That is a small fraction of a large valve's breakaway torque, and it explains why the same size valve feels easy on a test bench and immovable on a live line: the bench torque was running torque at low differential, and the line torque is breakaway at full differential.
The practical threshold used across valve specifications, and quoted in test procedures that reference MSS SP-91, is a maximum rim force of about 360 N (80 lbf) for a new valve operated by hand. Above that, a gear operator is expected rather than optional. Some project specifications go stricter, capping rim pull near 222 N (50 lbf) and turning effort near 100 to 200 handwheel turns; others accept the 360 N figure outright. The number to design to is the one written in the project specification - and where none is written, 360 N is the defensible default.
Convert that threshold into available torque and the size at which gearing becomes mandatory falls out directly:
Handwheel diameter | Radius | Torque at 360 N rim force | What that means in practice |
|---|---|---|---|
200 mm | 0.10 m | 36 N m | Small valves, low differential only |
300 mm | 0.15 m | 54 N m | Light duty; typical small gate and globe valves |
400 mm | 0.20 m | 72 N m | Above this, a gear operator usually earns its cost |
500 mm | 0.25 m | 90 N m | Near the practical single-operator limit |
600 mm | 0.30 m | 108 N m | Large enough to foul adjacent piping and handrails |
800 mm | 0.40 m | 144 N m | Common end-user ceiling; some specifications cap the wheel at 800 mm or at the face-to-face dimension |
Two design rules from practice line up with this arithmetic. Lever handles on ball and butterfly valves stay practical to about DN150, with most ranges going gear-only from DN250 upward, because the break-to-open torque of a rubber-lined disc climbs faster with size than a lever can absorb. And a larger handwheel is not a solution to a torque problem: beyond the sizes above it becomes a lever for a cheater bar, fouls the surrounding steelwork, and moves the failure from "hard to turn" to "sheared stem".
The most expensive misunderstanding in manual operator selection is the belief that a 100:1 gearbox gives 100 times the torque. It gives 100 times the turns. What it gives in torque is the ratio multiplied by the gear train's efficiency:
Mechanical advantage = gear ratio x efficiency
Single-start metal worm gear sets - the standard architecture for quarter-turn valve operators - are by design inefficient, in the range of 30 to 40 percent, with about 32 percent a realistic average across commercial manual operators. That inefficiency is the price of self-locking: a worm set efficient enough to exceed roughly 35 percent typically stops being self-locking, which means the valve can back-drive the operator under its own seat load. The trade is not optional, it is a design identity.
So a 80:1 worm operator delivers a mechanical advantage near 25.6, not 80. Two thirds of the theoretical advantage is spent as heat in the gear mesh. Bevel and spur trains are more efficient but usually not self-locking, which is why high-efficiency multi-turn operators add a brake or anti-backdrive device, and why the "which gearbox" question cannot be answered without knowing whether the valve must hold position with the hand removed.
Operator type | Typical ratio | Efficiency | Self-locking | Typical duty |
|---|---|---|---|---|
Worm gear, single start | 30:1 to 200:1 | About 30 to 40 percent (roughly 32 percent average) | Yes, by design | Quarter-turn ball, plug and butterfly valves; the default manual operator above DN200 |
Bevel or miter gear | 3:1 to 12:1 (higher with two stages) | About 70 to 90 percent | Usually no - needs a brake or anti-backdrive device | Multi-turn rising-stem gate and globe valves, where turn count must stay low |
Spur or planetary add-on stage | Adds 3:1 to 5:1 | About 90 to 95 percent | No | Fitted to an existing gearbox when rim force must come down without changing the base unit |
Combined worm plus spur | Multiplies both ratios | Product of both efficiencies | Yes if the worm stage dominates | Very high torque, at the cost of turn count - see the next section |
Typical multi-turn ratios also track valve size, which is worth knowing when a gearbox is being specified from a size table rather than from torque: 6 to 8 inch valves commonly use 3:1 to 6:1, 10 to 12 inch use 6:1 to 12:1, and 14 to 24 inch use 12:1 to 30:1. Those ratios are chosen to keep turn counts tolerable, not to minimise rim force - and that is exactly where the next limit appears.
A quarter-turn valve needs 90 degrees of output, which is a quarter of a revolution. A multi-turn gate valve needs 12 to 40 stem revolutions to travel from closed to open. That single difference decides whether a gearbox can solve the problem at all, and the arithmetic is unforgiving.
The quarter-turn case. Take an NPS 12 Class 300 trunnion ball valve with a breakaway torque of 2,800 N m at full differential - a realistic order of magnitude for that size and class. Direct operation needs 14,000 N at a 400 mm rim: not a valve problem, a crane problem. Now fit a 120:1 worm operator at 32 percent efficiency: mechanical advantage is 38.4, so the required input torque is 73 N m. At a 500 mm handwheel that is 292 N of rim force - inside the 360 N line - and the stroke costs 120 x 0.25 = 30 handwheel turns. Thirty turns for one operation is acceptable, and this is precisely why high-ratio worm operators dominate quarter-turn service: the ratio can be large because the output only has to move a quarter turn.
The multi-turn case. Take an NPS 12 Class 600 wedge gate with a stem torque of 1,500 N m and 26 stem revolutions to open - ordinary numbers for that size. To hold rim force at 360 N with a 500 mm wheel, input torque must not exceed 90 N m, which requires a mechanical advantage of 16.7, which at 32 percent efficiency requires a ratio near 52:1. The turn count then becomes 26 x 52 = 1,352 handwheel turns. At a brisk one turn per second that is 22 minutes of continuous winding for one open-to-close cycle - before the operator gets tired enough for the long-term force factor of 0.25 to apply. Nobody operates that valve.
Drop to the ratio that size actually ships with, 10:1, and the arithmetic fails from the other side: mechanical advantage 3.2, input torque 469 N m, rim force 1,875 N at a 500 mm wheel and still 1,172 N at the largest wheel a specification will usually allow. There is no handwheel that closes the gap, which is why large high-class gate and globe valves are supplied with impact handwheels, input shafts for portable power tools, or motor operators - and why a project that specifies "gear operated, handwheel" on a 12-inch Class 600 gate valve has specified something that will be commissioned with a pipe on the wheel.
The practical screen is therefore two-sided, and both sides belong in the specification: rim force at or below the project limit, and total turns at or below the project limit - commonly 100 turns, beyond which a power drive or a dual-ratio gearbox is expected. A gearbox that satisfies only one of the two has simply moved the problem.
Every number so far has been about making the valve move. MAST is about not breaking it while trying. MSS SP-91 defines the maximum allowable stem torque as the maximum torque that can be applied to the stem without permanent deformation or mechanical damage to any component that could affect sealing or operation - and API 6D Annex L lists drive train MAST among the data a manufacturer provides on request.
The physics is a solid circular shaft in torsion, and it is simple enough to check in the field:
tau = 16T / (pi x d^3), rearranged as d = cube root of (16T / (pi x tau_allow))
Work it forwards on a 60 mm stem in 410 stainless (A182 F6a, yield around 550 MPa). Taking an allowable shear of 0.577 times yield with a 1.5 factor of safety gives about 212 MPa, and the corresponding torque capacity is roughly 9.0 kN m. That is the stem's nominal ceiling. Two corrections cut into it:
The keyway or drive flats. A keyway removes material exactly where the stress peaks, and stress concentration factors of roughly 1.6 to 3.0 are normal, with strength reductions of 25 to 50 percent depending on geometry and fillet quality. A keyed 60 mm stem may therefore be good for only 4.5 to 6.7 kN m in practice - half the nominal figure.
The stem nut and the thread. On multi-turn valves the weak point is often not the stem shank but the stem thread or the nut, which is why thread engagement length and nut material belong in the same review as stem diameter.
Now put the ceiling next to the earlier example. The NPS 12 ball valve needed 2,800 N m at the stem and its operator delivered 73 N m of input at 38.4 mechanical advantage - comfortably below a 9 kN m MAST. Fit a 24-inch impact handwheel with a factor of 3 and an energetic operator, and input torque can reach roughly 219 N m, which at the same advantage is 8,400 N m: still inside, but with almost nothing left. Put a one-metre pipe on that handwheel and push with 400 N, and the input becomes 400 N m, the stem torque 15,360 N m - about 1.7 times MAST. Something gives: the key, the stem, the gear tooth, or the casing. Usually the key first, which is why a "gearbox turns but the stem doesn't" complaint is often a maintenance history of cheater bars rather than a manufacturing defect.
MAST is also the reason an oversized operator is a failure mode rather than a safety cushion, and the reason a second, separate check exists. API 6D Annex L specifies a drive train strength test - when the purchaser asks for it - at the greater of twice the predicted or twice the measured breakaway torque, applied with the closure member blocked for at least one minute, with no permanent visible deformation permitted afterwards. That 2x figure is a strength qualification of the drive train. It is not a sizing factor, and treating it as one produces operators that are simultaneously oversized for the valve and marginal against the stem.
Three devices exist to bridge the gap between what a person can deliver and what the valve needs, and each has a different failure signature when misused.
Impact handwheels. A sliding hammer wheel converts a short, sharp blow into a torque spike, which is why MSS SP-91 credits them with multipliers of 2, 3 and 4 for 300, 600 and 900 mm sizes. They are legitimate engineering devices when the valve and drive train were designed for them - the stem mass absorbs the blow, and the gear teeth see a load the strength test covers. They are a hazard when fitted to a standard operator on a marginal stem, because the multiplier applies to whatever the operator can deliver, and the operator is not measuring.
Chainwheels and extended stems. When a valve sits above shoulder height, the position factor halves the available force before any gearing is considered. A chainwheel brings the operating point back to waist level where the 1.0 factor applies, and an extended stem with a floor stand does the same for buried or elevated installations. This is cheap engineering that removes a whole class of "hard to operate" complaints without touching the valve.
Cheater bars. The pipe on the handwheel is the most common cause of sheared stems in plant history, and the arithmetic shows why in one line: a 1 m bar multiplies the moment arm by four compared with a 500 mm wheel, and delivers a torque no part of the assembly was sized for. It also removes the feedback that protects the valve - with a bar, the operator cannot feel the seat contact that says "stop". If a valve needs a bar, the correct response is to measure the torque, compare it with the datasheet and with MAST, and then fix the operator, the packing or the process - not to lengthen the lever.
Breakaway torque on the day of commissioning is the best case, not the design case. Three multipliers routinely push field torque above the catalog figure, and they compound:
Temperature. Above roughly 150 deg C, published vendor data commonly shows breakaway torque running 20 to 40 percent above the ambient figure, as polymer seats soften, creep and reload against the ball or wedge, and as clearances change. At the cold end the mechanism is different but the effect is similar: grease stiffens, and an operator selected on ambient torque can exceed its capability on the first winter start.
Packing. Gland compression is the variable most under maintenance control and least under design control. Reported field experience has a single extra turn on the gland nuts doubling stem torque; live-loaded packing and lower-friction packing styles reduce it. Every overhaul that "stops a weep" by tightening the gland adds torque the operator was never sized for - and low-emission packing specifications make this worse unless the torque effect is declared.
Idle time and deposits. A valve that has sat closed for a year needs more torque than the same valve cycled weekly: static adhesion at the seat, hardened grease on the thread and stem, and process deposits bridging the seat. Sizing practice adds roughly 1.3x for aged or infrequently operated valves, and severe or fouling service can move the requirement by a factor of two or more over the valve's life.
These are not additive percentages to be sprinkled on at the end. They change which operator is correct, and they are the reason the specifier should ask for torque at the specified temperature and at end-of-life packing condition rather than at ambient with fresh packing.
Because torque is a sum of identifiable contributions, the way a valve resists tells you which contribution grew. The table below maps the common complaints to a measurement and a fix, and it is worth keeping next to the maintenance procedure.
Symptom | Most likely cause | How to confirm | Correct action |
|---|---|---|---|
Very hard at the first movement, then normal | Seat adhesion or stiction after long closure; breakaway margin too thin | Measure breakaway with a torque wrench on the input shaft, multiply by ratio and efficiency, compare with datasheet | Resize the operator on measured breakaway; review seat material and closure duration |
Hard throughout the whole stroke | Over-compressed or swollen packing; corroded or bent stem; hardened grease | Compare running torque with baseline; inspect stem finish; check gland bolt length against record | Reset gland to specified compression; re-grease with the correct grade; straighten or replace stem |
Gets harder at one point of travel only | Wedge binding from thermal contraction after a hot closure; body distortion; seat misalignment | Map torque against position; correlate with the temperature at which the valve was closed | Re-open hot, or specify a flexible wedge - the mechanics are covered in our gate valve wedge design guide |
Torque rises steadily month after month | Deposit build-up, seat wear debris, packing creep, or grease loss | Trend the measured input torque at each maintenance interval | Clean or flush; review seat and packing specification; add a lubrication interval |
Handwheel turns but the stem does not | Sheared key, stripped stem nut, broken worm wheel, or a cheater bar in the valve's history | Disconnect and inspect the drive train; check stem for twist marks | Replace the failed element; verify MAST against the operator output before re-assembly |
Gearbox input locked solid | Seized gear set, water ingress, or valve jammed mid-travel | Separate operator from valve; test each independently | Overhaul or replace the operator; verify IP rating and breather arrangement |
Valve drifts back after the hand is removed | Operator is not self-locking at its efficiency | Check gear type and efficiency; verify whether an anti-backdrive device is fitted | Add a brake or anti-backdrive unit; confirm the valve does not need to hold position unpowered |
Only the large valves cannot be turned by hand | The turns ceiling, not a fault - ratio and rim force cannot both be satisfied | Compute rim force and total turns for the installed ratio | Fit an impact handwheel, power-tool input shaft or motor operator as designed, not a bar |
One measurement discipline makes the whole table usable: record the input torque at a defined stroke position, at a defined pressure, on a defined date. Without a baseline, "stiff" is an opinion; with a baseline, it is a trend line, and a trend line tells you whether to plan a maintenance intervention or accept the valve as designed.
Manual operator selection fails at the interface between the valve datasheet and the operator datasheet, so the specification has to close that gap in writing. The clauses below are the ones that prevent the field argument later.
# | Specification line | Why it is there |
|---|---|---|
1 | Breakaway, running and seating torque, each with the differential pressure and temperature at which it applies | A torque figure without a pressure is not data; the three values size different parts of the operator |
2 | MAST for the stem as supplied, including the keyway or drive geometry assumed | Defines the ceiling the operator must stay under; makes the keyway penalty visible |
3 | Gearbox output rating at or above breakaway torque times the project sizing factor, stated as a number | Removes the argument about whether the factor is already included in the catalog value |
4 | Maximum rim force at the handwheel, per project limit (360 N default), verified by calculation in the submittal | Converts "suitable for manual operation" into an auditable number |
5 | Total handwheel turns from closed to open, with an upper limit (commonly 100 turns) | The second half of the operability screen; a high ratio can satisfy rim force and fail this |
6 | Operator type and efficiency: worm with self-locking, or bevel with brake / anti-backdrive device | Determines whether the valve holds position with the hand removed |
7 | Handwheel or chainwheel arrangement, elevation, and whether the wheel must be removable | The position factor of 0.5 is a design input, not an installation detail |
8 | Impact device allowance: fitted or not, size, and confirmation that the stem and drive train are rated for it | The multiplier is real; the strength case has to exist before the device is fitted |
9 | Grease grade and temperature range, IP rating, breather and drain arrangement, coating system | Most "seized gearbox" failures are water ingress and stiff grease, not gear teeth |
10 | Mounting interface: ISO 5211 or MSS SP-101/SP-102 flange, stem drive geometry, and access clearance for the wheel | Adaptation is the most common assembly error, and clearance is the most common site surprise |
Two closing observations for the buyer. First, manual operation is not the low-technology option - it is the option with the tightest arithmetic, because the human limit cannot be re-engineered. Second, the operator decision interacts with the valve design: a trunnion-mounted ball valve, a flexible wedge, a double-offset disc and live-loaded packing all change the torque requirement, often by more than a change of operator would. Seat friction and preload behavior are the mechanics behind the first of those and are worked through in our ball valve seat design guide; where the duty is actuated rather than manual, the margin logic carries over directly to our pneumatic actuator sizing and fail-safe guide. Product ranges for the valve types discussed here are on our ball valve, gate valve and globe valve pages, and the wider engineering library is in our technical article collection.
How do I calculate the handwheel force needed to open a valve?
Work backwards from breakaway torque. Required rim force equals input torque divided by the handwheel radius, and input torque equals breakaway torque divided by mechanical advantage, where mechanical advantage is gear ratio times efficiency - not the ratio alone. For a valve needing 2,800 N m at full differential, a 120:1 worm operator at 32 percent efficiency gives an advantage of 38.4, so the input is about 73 N m; at a 500 mm handwheel (radius 0.25 m) that is roughly 292 N of rim force, inside the 360 N line usually specified for manual operation. Without a gearbox the same valve would need 14,000 N at a 400 mm rim, which is the arithmetic reason large valves are geared.
What is MAST and how do I check it on a valve stem?
MAST is the maximum allowable stem torque - the torque that can be applied to the stem without permanent deformation or mechanical damage to any component that could affect sealing or operation, as defined in MSS SP-91 and listed as drive train data in API 6D Annex L. Check it as a solid shaft in torsion: tau = 16T / (pi x d^3), so the allowable torque rises with the cube of stem diameter. A 60 mm stem in 410 stainless, at an allowable shear of about 212 MPa, carries roughly 9 kN m before the keyway is considered; a keyway with a stress concentration factor of 1.6 to 3.0 can take 25 to 50 percent of that away. MAST is a ceiling, so an oversized operator is a failure mode, not a safety cushion.
When should a gear operator replace a plain handwheel?
Three triggers, and any one of them is enough. The first is rim force: about 360 N (80 lbf) is the commonly specified limit for a new valve operated by hand, so with a 400 mm wheel any valve needing more than roughly 72 N m should be geared, and at 500 mm the ceiling is about 90 N m. The second is size practice: lever handles stay practical to about DN150, and most ranges go gear-only from DN250. The third is turn count - if meeting the rim force limit would push the stroke past roughly 100 handwheel turns, the valve needs an impact handwheel, a power-tool input shaft or a motor operator rather than a bigger gearbox.
Why does a geared gate valve need so many handwheel turns?
Because a gear ratio multiplies turns exactly as it multiplies torque, and mechanical advantage only multiplies torque. A quarter-turn valve needs 90 degrees of output, so a 120:1 operator costs 30 handwheel turns and is perfectly practical. A gate valve needs 12 to 40 stem revolutions, so the same ratio costs 1,500 to 5,000 turns. That is why multi-turn gearboxes use lower ratios - commonly 3:1 to 6:1 at 6 to 8 inch, 6:1 to 12:1 at 10 to 12 inch - which is exactly why large high-class gate valves cannot be brought inside the 360 N rim force limit by gearing alone, and why they ship with impact handwheels, input shafts or motor operators.
Can I use a cheater bar or an impact wrench to open a stiff valve?
A cheater bar should never be used, and the reason is arithmetic rather than policy: a 1 m pipe on a 500 mm handwheel multiplies the moment arm by four, so a 400 N push becomes 400 N m of input and, at a mechanical advantage of 38, about 15 kN m at the stem - roughly 1.7 times the MAST of a 60 mm stem. Something yields: the key, the stem, a gear tooth or the casing. A purpose-built impact handwheel is different, because the valve and drive train were qualified for it - API 6D Annex L specifies a drive train strength test at twice breakaway torque with no permanent deformation, when the purchaser asks for it. That 2x figure qualifies the drive train; it is not a sizing factor.
