A globe control valve fitted with equal-percentage trim typically delivers 50:1 to 100:1 rangeability, versus roughly 10:1 to 20:1 for linear-trim globe valves; a butterfly valve, by contrast, generally remains controllable across only 25% to 50% of its disc travel [S2][S3].
That single number, the slice of travel where the loop gain stays in a controllable band of roughly 0.5 to 2, decides whether a project should be specced around a butterfly valve or a globe valve, and it is the most common reason field loops oscillate or go sluggish after start-up [S3].
How rangeability is defined for each body style
Rangeability, in the ISA sense, is the ratio of the maximum controllable flow to the minimum controllable flow, and it is bounded by where installed gain falls outside the practical 0.5 to 2 window [S3].
For a linear-trim globe, the published number is 10:1 to 20:1; with equal-percentage trim, the same valve pushes to 50:1 to 100:1, which is why equal-percentage trim is the most common inherent characteristic in tight control loops [S2]. A butterfly valve inherently behaves closer to a linear characteristic, and in most applications its useable controllable range sits between 25% and 50% of valve travel, so the practical turndown collapses to roughly 4:1 to 2:1 even when the mechanical stroke is 90 degrees [S3].
Why the butterfly has a narrow control band
At the lower end of butterfly travel, a small change in input signal drives a disproportionately large change in flow, so the loop hunts; at the upper end, gain is so low that the valve becomes sluggish and the controller can no longer hold setpoint [S3].
This is structural: the disc presents a small projected area to the flow at low opening angles, so the Cv curve is steep at the bottom and flat at the top, and the valve is being asked to do modulation in the worst 25% to 50% of its travel, not in a band centred on the equal-percentage sweet spot [S3]. Disc profile redesigns (concentric, double- or triple-offset high-performance geometries) can flatten that curve, but the underlying turndown penalty of a rotary body remains, and it is the reason control valve sizing sheets still flag butterfly for limited turndown [S1][S3].
Globe valve strengths that buy the extra turndown
Globe control valves earn their reputation by handling high temperature, high pressure drop, cavitation, and flashing services, because the tortuous plug-and-seat path drops pressure gradually and resists recovery to a high vena contracta [S2].
That same geometry also accepts a family of interchangeable trims (linear, equal percentage, quick opening) and reduced-trim options, so a 6-inch line-size body can be cut down to a smaller Cv and later re-trimmed upward as plant capacity grows, an option that rotary bodies do not offer cleanly [S2]. The trade-off is weight, envelope, and cost: globe bodies are the heaviest and priciest per line size, which is why process groups default to butterfly for lines above roughly 16 to 24 inches unless the service demands globe-class severity [S1][S2].
Decision matrix: butterfly vs globe on five selection criteria
On cost per line size, butterfly wins, typically the lowest-cost rotary option and the only practical body above 24 inches [S1][S2]. On rangeability, globe wins by a wide margin, 50:1 to 100:1 with equal-percentage trim versus the butterfly's 25% to 50% controllable travel band [S2][S3]. On high pressure drop, cavitation, and flashing resistance, globe wins because the recovery coefficient is far lower; butterfly valves are explicitly called out as not suited to high pressure drops with flashing or cavitation [S1]. On slurry or fibrous service, butterfly wins because the unobstructed bore passes solids that would jam a globe plug [S2]. On package skids where envelope and mass matter, butterfly wins for the same reason, less metal, less support steel, no reducers required because it is line-size [S2][S3].
When butterfly is the right answer, and when it is not
Use a butterfly control valve when the line is large, the pressure drop across the valve is a small fraction of the system drop, the fluid is dirty or fibrous, the loop is secondary (level, pressure letdown, cooling-water tempering), and the operating band covers only a slice of the 25% to 50% controllable travel [S1][S3].
Do not use a butterfly valve when the loop is primary flow or composition control, when turndown above 10:1 is required, when the service flashes or cavitates at design, when the pressure drop across the valve is more than a small fraction of system drop, or when the disc profile is going to spend its life modulating below 25% or above 50% opening, because that is exactly where the loop will either hunt or stall [S1][S3]. For those cases, the answer is a globe valve with equal-percentage trim, sized so the normal operating point lands in the middle of the trim curve, not at the tail [S2].
Failure modes engineers actually see in the field
The two failure modes on a misapplied butterfly are oscillation at low openings and sluggish response at high openings; both are visible in the same loop at different operating points, because the gain curve is so asymmetric across the 25% to 50% band [S3].
On a misapplied globe, the failure modes are noise and trim erosion at high pressure drop, seat-wire-draw on tight shut-off duty, and packing leaks on high-temperature service, all of which the seat-and-plug geometry is at least designed to manage, unlike the disc-to-seat interface of a butterfly, which has less metal to give up before leakage class slips [S1][S2]. Slurry in a globe plug will jam it; a globe in a 36-inch line is rarely a workable option on a structural or cost basis, which is the practical reason butterfly bodies dominate the largest line sizes even with their turndown penalty [S1][S2].
Sizing practice that protects rangeability on either body
Sizing is the lever that turns a nominal 50:1 trim into a real 10:1 installation, or a 25% travel butterfly into a controllable one: the rule of thumb is to pick the valve so the normal operating point lands near 70% to 80% of travel at design flow, leaving equal-percentage trim room to ramp down and the butterfly disc out of the 0% to 25% hunting zone [S1][S2][S3].
Globe bodies tolerate reduced trim and later re-trim, so a 6-inch body with a 4-inch trim can be re-trimmed upward as plant capacity grows, recovering rangeability that a fixed rotary body cannot [S2]. For butterfly bodies, the equivalent move is to change disc profile or move to a high-performance (double- or triple-offset) geometry rather than re-trim, which keeps the project on the rotary platform when envelope and cost drive the choice [S1].
Two trackable signals: ISA-75.01.01/IEC 60534 sizing equations continue to be applied with a required Cv at least 1.3 to 2 times the design Cv for butterfly service to keep the operating point off the bottom of the gain curve; and high-performance double-offset butterfly bodies are now specified as the default for control service in water and HVAC rather than the older concentric designs, both of which shift the rotary-side rangeability number upward at the spec stage rather than at the tuning stage. For a deeper look at how stem retention and disc geometry affect that ceiling on the butterfly side, see this spec deep-dive on blowout-proof stem retention.