Gate valve vs ball valve

How each one shuts a line

A gate valve does its work along the axis of the stem. Turning the handwheel drives a wedge — or, in parallel-slide designs, a pair of discs — down across the flow until it lands in the seat, and drives it back up out of the way to open. That takes many turns, which is why the mechanism is described as multi-turn. The reward is a bore that is completely unobstructed when the wedge is up: no seat ring standing in the flow, no cavity, nothing for a pig or a brush to catch on. The penalty is that the wedge is exposed to the flow whenever it is partly closed, and the seat faces wear against each other every time it lands.

A ball valve does its work across the flow instead. A solid ball with a bore through it sits between two seat rings; a ninety-degree turn of the lever swings the bore either in line with the pipe or square to it. There is no intermediate travel to speak of, no rising stem, and no ambiguity about state — the lever is either along the pipe or across it.

Tightness: metal on metal, or ball on soft seat

This is where most arguments are actually settled. A gate valve seals metal against metal. That is a robust arrangement — it tolerates heat, it tolerates a little grit, it can be lapped back into service — but it is not a tight arrangement in the way a fitter means when they say tight. A small permitted leakage rate is normal and specified. On a utility header being isolated for maintenance, nobody minds. On a line that has to be proven shut before someone opens a flange, that permitted leakage is a problem.

A soft-seated ball valve seals elastomer or PTFE against a machined ball, and it does so to bubble-tight classes as standard. That is why the ball pattern dominates instrument isolation, sampling points, gas service and anywhere a positive shutoff has to be demonstrated rather than assumed. The trade is the seat material itself: PTFE sets the temperature ceiling, and a soft seat does not enjoy grit. Choosing between the two families — and remembering that the globe valve is the third option for anything being throttled — is laid out in gate, globe or ball: choosing the right isolation valve.

Speed, torque and how often it moves

Ask how many times a year the valve will be operated and most selections answer themselves. A valve that moves twice a decade — a main isolation on a buried water line, a header block kept open for the plant's life — has no use for speed, and every use for a full unobstructed bore and a body that will still be there in twenty years. That is the gate valve's home ground. A valve that moves every shift, or every batch, is a different animal: the gate's seat faces are being scrubbed each time it lands, and the operator is standing there turning a handwheel while the process waits.

Speed cuts both ways. A quarter turn on a long liquid line is exactly how water hammer starts — the column stops in a fraction of a second and the pressure spike goes looking for the weakest joint. Where that risk is real, either gear the ball valve so it cannot be slammed, or accept the gate's slow travel as a feature. The mechanism and the selection rules are set out in our guide to what causes water hammer and how valve selection prevents it.

If it will ever be actuated, decide now

Quarter-turn geometry is what made the ball valve the default for automation. The stem needs ninety degrees of rotation, the mounting pad is standardised, and a pneumatic scotch-yoke or a compact electric unit bolts straight on. A gate valve can absolutely be actuated, but it needs a multi-turn actuator, it needs it to run for the full travel every time, and the assembly is heavier, slower and more expensive for the same line size. Retrofitting actuation onto a line originally specified with handwheels is one of the more common reasons a valve package gets bought twice — the sizing, the pad and the fail position all have to be settled up front. What that involves is covered on our actuators and automation page.

Where each pattern stops being the cheap answer

Two ceilings decide the rest. The first is temperature: a soft-seated ball valve is limited by its seat, not its body, so steam, hot oil and thermal-fluid duties either push you to a metal-seated ball — which is a specialist item at a specialist price — or straight back to the gate, whose limits are simply the limits of its metallurgy. The second is diameter. Below roughly DN50 a ball valve is usually the cheaper and better buy. As the bore grows, the ball and its seats grow with it in three dimensions, and somewhere in the DN150 to DN250 region the gate becomes the economical answer for the same duty and class. Where the money really goes across a whole line list is the subject of why the cheapest valve usually costs the most.

Body material sits underneath both ceilings, and it is the question we get asked most often once the pattern is settled — bronze, cast iron, carbon steel or stainless, each with its own price and its own service envelope.

What we tell engineers who ask

We stock both patterns across our principals — 283 products across 30 categories, with gate and ball figures in bronze, cast iron, carbon steel and stainless — so we have no commercial reason to push either one. The verdict below is the one we give on the phone.

The mistake we see most often is not choosing the wrong pattern — it is choosing on unit price alone and discovering the consequence in downtime, or specifying three classes above what the duty needs and paying for it forever. Both failures come from the same place: a line list priced before anyone asked what the line actually does. Send us the duty — fluid, pressure, temperature, how often it moves and whether it will be actuated — and we will mark up the list with you. That is the whole argument of choosing the right industrial valve, duty first and budget second.