Butterfly valve offsets: from centric to 5-offset
A butterfly valve is a simple principle at heart: a body, a main shaft, and a disc that turns ninety degrees between open and closed. Within that principle, a development towards ever higher-grade valves has been running for decades: better sealing, tighter tolerances, lower operating forces, more controlled opening and closing. That development is about materials, about production techniques — and about design. The offsets are the design line. This note explains what each step does, and what it does not do.
Centric butterfly valves
In a centric butterfly valve the shaft runs through the centre of both the disc and the seat. That is not a failed attempt at tight shut-off; it is a valve for different work. Centric valves are deliberately used metal-to-metal with a certain leakage, and two versions are common.
Without a seat, the disc does not close against anything. The end positions are then limited by the actuator — on a pneumatic actuator via the adjusting bolts, and on hydraulic, electric or manual drives via their own end points. A positioner usually monitors that position with feedback to the control system. An end stop via the main shaft is technically possible, but not common. The closed position of such a valve therefore usually sits in the actuator, not in the valve. That is a deliberate design choice for dust-laden gaseous media: no seat means no place for dust to build up. Such a valve has a fixed leakage in the order of 1% of the Kv at 90° valve position, so with the valve fully open. For a 100% control duty — a valve that never has to shut tight — that is a good solution.
Against stop strips the disc does close against something, if required with an AlSi cord in those strips as an additional seal. Leakage is lower, but leakage remains.
In both versions it comes down to that word fixed: the leakage does not increase. It stays stable and predictable, and that is what makes such a valve usable in control duty. If one does deteriorate, you replace it. Centric butterfly valves with a rubber liner also exist, for instance in drinking water; that is not our field.
The development of butterfly valves
The development of butterfly valves towards ever higher-grade versions turns on a handful of axes that are continually sharpened: sealing quality, tolerances, operating forces, and the opening and closing itself.
That last one needs explaining. On a valve with a high break-away torque, the disc does not come free of the seat calmly — it shoots out. If that valve also has a control function and is pneumatically actuated, that is awkward: a positioner cannot control finely on a valve that jumps loose. With electric or hydraulic actuation this plays a considerably smaller role.
And metal-to-metal is not an ambition in itself — it follows from the application. You end up there where the usual sealing materials, often flexible in nature, do not survive: at extreme temperatures, with aggressive media, or where the medium is too abrasive.
That improvement is a continuous process, and it runs on three fronts at once: materials, production techniques and design. Within the last of these, the approach via multiple offsets is the dominant line — and at present the 5-offset is the most recent standard available on the market.
Double eccentric: rightly a high performance butterfly valve?
Two steps at once. First the shaft moves behind the sealing face, so the sealing ring is no longer interrupted: the shaft no longer passes through it. Then the shaft also moves out of the centreline of the pipe. As a result the disc moves away from the seat on opening, and the valve only wrings briefly in the area of the shaft passage. The intermediate step with only the first shift is geometrically relevant but hardly found as a product; in the market it starts at double eccentric.
This is the valve the industry calls a “high performance butterfly valve”. The valve is position-seated: the sealing force comes from the end position in which the disc wrings itself into the seat. That wringing character calls for flexible sealing materials, and that has three consequences. Wear: intensive operation produces a high degree of wear, and leakage as a result. A fairly high break-away torque: fine control with a pneumatic actuator becomes demanding — this is the valve that shoots out of the seat. And weak bi-directional tightness, due to the generous tolerances required and the flexible sealing materials needed.
And the second offset cannot simply be enlarged to get more out of it. The further out it is placed, the higher the operating torque — and so a thicker main shaft, heavier bearings, costlier production. That offset is therefore not chosen generously; it is a trade-off, not a setting. For less demanding duties that may serve perfectly well. But “high performance” is here rather a name than a measure of performance.
Triple offset (1960): the tilted cone
A conical seat is not new to the third step: a double eccentric valve can already work from a cone. In that case, however, the axis of that cone lies parallel to the axis of the pipe.
The third offset tilts the cone. The apex is displaced so that one side of the cone runs parallel to the inner wall of the body. At that point the seat has a nearly flat angle, and 180° opposite it lies the maximum seat angle. The shorter the cone is chosen, the larger that maximum angle. An oblique section through a cone approximates an ellipse: the seat is therefore not round, and neither is the sealing ring in the disc.
That changes the way the valve closes. The disc comes free of the seat over practically the whole stroke, and the faces only meet in the last part of the closing movement. There is no longer any wringing to reach tightness: tightness follows from the applied torque, not from elastic deformation of the sealing material. That is what makes metal-to-metal sealing practically possible — no flexible material is needed to absorb the wringing. And the break-away torque largely falls away.
Largely — not entirely. A conical seat does not make the valve frictionless, and certainly not at extreme temperatures, where body and disc do not contract or expand at the same rate. This principle was developed and patented in Germany in 1960. Sixty-five years on it is still an important standard worldwide.
The angle progression of the seat
The angle of the seat in the body varies around the circumference: from the maximum angle on one side to the nearly flat point where the side of the cone runs parallel to the inner wall of the body. Published angles in this class lie between fifteen and thirty degrees, with twenty-five as a common value on the generous side.
That angle determines how much of the disc’s movement goes away from the sealing face, and how much along it. At a generous angle the seal comes free sooner; where the angle has fallen off, the faces stay nearly parallel for longer and the seal moves across the surface rather than away from it.
A valve design can be based on a chosen cone length, and with it the maximum angle (the longer the cone, the flatter the angle) — but not, independently of that, the resulting progression of this angle across the whole seat: that follows from the cone chosen. This also fixes the angle that remains at the shaft passage.
The 5-offset design
Two things come together in the 5-offset geometry. The base of the cone is no longer a circle, but a variation on an ellipse. And that base is chosen such that the section at one particular angle — twenty-five degrees, for instance — mathematically yields a perfect circle. The first three offsets remain in place. What changes is the result: a circular seat. Geometrically this is a stack of circles, all aligned on a single point of their outer edge, whose centres together form the axis of the cone.
That is what makes an O-ring possible: an Inconel O-ring for a fully metallic, graphite-free seal, or PTFE where the chemistry calls for it. On a seat that is not round, that cannot be done.
And the angle progression of the seat becomes shapeable, while keeping that circle. The angle therefore no longer has to fall off to a nearly flat point where that is least convenient. This carries through into the actuator: because the angle can be chosen more generously, engagement and sealing on the seat are also assured at a lower second offset — and a lower second offset means directly a lower operating torque. That is where the room lies to select a smaller actuator (Md).
This brings a number of advantages in reliable operation and in a wide range of applications: cryogenic and high-temperature duty, improved fine control from the closed position with pneumatic actuators, simplification of local maintenance work, longer service life, and less wear on the seal — even at high switching frequencies.
Face-to-face length and flanges: replacement without pipework changes
At installation, as a rule, nothing changes. The versions follow the usual standards: face-to-face lengths to EN 558-1, DIN 3202 and ISO 5752, designs to API 609 and ASME B16.10, and an ISO 5211 top flange for the actuator.
A valve due for renewal can therefore in virtually all cases be replaced by a 5-offset butterfly valve without modifying pipework or flanges. The change lies in the higher sealing technology, the lower operating torque required and the reliable behaviour under extremes — not in the way it is installed, which follows all the required standards.
Which valve, when
A centric valve fits where control is needed and tightness is not the requirement.
As soon as the application calls for a valve that has to shut tight metal-to-metal, the 5-offset butterfly valve comes into view: cryogenic media, thermal cycling, high temperatures, thermal shock, high switching frequencies, broad certification requirements. On projects of size it also counts that a lower actuator per valve will do; in large quantities that makes a difference to the investment.
In standard applications too, where unplanned downtime is to be avoided as far as possible and operation has to stay reliable, the 5-offset butterfly valve may be worth considering. And if your process or design conditions are anything but standard, we would like to hear about it and will advise you on what is possible.
