A control valve is the final control element in a process loop: it receives a signal from a controller and continuously modulates flow to hold a process variable such as pressure, level, temperature, or flow at setpoint. Unlike an isolation valve, which only opens or closes, a control valve must throttle accurately and repeatably anywhere between its end positions, which is why it carries a characterized trim, a sized actuator, and a positioner.
Selection turns on four physical questions: how much capacity is needed (flow coefficient Cv or Kv), how flow should change with travel (the inherent characteristic), how the trim will survive cavitation or erosion, and how tightly it must seal when closed (the seat leakage class). This guide works through each, referencing the public engineering standards that govern the field.
Photo: Bitjungle, CC BY-SA 4.0, via Wikimedia Commons
This guide is written for procurement engineers and design engineers specifying throttling valves. It covers six chapters from definition and history, body styles, flow characteristics and trim, materials and pressure ratings, spec-sheet decoding, to a structured selection sequence, with 7 selection FAQs and manufacturer comparisons. All parameters reference public standards: the IEC 60534 series (sizing, characteristics, seat leakage, and noise), ISA-75.01, ANSI/FCI 70-2, and ASME B16.34.
Chapter 1 / 06
What is a Control Valve
A control valve is a power-operated device that continuously modifies fluid flow rate in a process control system. It is the most common final control element in process automation: the controller compares a measured variable against its setpoint, computes a correction, and sends an analog or digital signal to the valve, which strokes its closure member (plug, ball segment, or disc) to a matching position. By varying the restriction it presents to the line, the valve changes the flow, and through it the downstream pressure, level, temperature, or composition. This continuous throttling role is what separates a control valve from an isolation valve that only travels fully open or fully closed.
A complete control valve assembly is built from three subsystems. First, the valve body and trim: the pressure-retaining body, plus the internal parts that contact and shape the flow, namely the plug or closure member, the seat ring, the cage or guide, and the stem. The trim is the part that wears and is designed to be replaced. Second, the actuator: the device that converts an air, electric, or hydraulic input into the thrust or torque needed to stroke the valve against process and friction forces. Third, the accessory chain: the positioner that enforces stem position, plus solenoid valves, air filter-regulators, limit switches, and volume boosters as the duty requires.
The engineering vocabulary of control valves was formalized through the second half of the twentieth century. The flow coefficient Cv was introduced in the 1940s by Masoneilan as a single number to express valve capacity, defined as the US gallons per minute of 60 degrees Fahrenheit water that pass through the valve at a 1 psi pressure drop. The metric Kv followed, defined in cubic metres per hour at a 1 bar drop. The Instrument Society of America codified sizing in the ISA-75 series, and the International Electrotechnical Commission published the IEC 60534 series, which today governs sizing equations, inherent flow characteristics, seat leakage, and aerodynamic and hydrodynamic noise prediction on a single international basis.
The pneumatic spring-and-diaphragm actuator that still dominates the field traces to the same era, when the 0.2 to 1.0 bar (3 to 15 psi) instrument air signal became the standard analog command. The 4-20 mA electronic loop and the current-to-pressure converter extended that reach in the 1970s, and the digital valve controller, a microprocessor positioner communicating over HART, arrived in the 1990s, adding self-diagnostics, travel histograms, and partial-stroke testing for safety valves. Despite a century of refinement, the central design problem is unchanged: convert a low-energy command into precise, repeatable, durable throttling of a fluid that may be hot, corrosive, abrasive, or prone to cavitation.
The application scale is broad. Control valves run from fractional DN 8 (quarter-inch) trim valves dosing reagents at a few litres per hour, up to DN 600 (24 inch) and larger valves passing thousands of cubic metres per hour in power, refining, and pipeline service. Pressure ratings span ASME Class 150 through Class 2500 and above, and process temperatures run from cryogenic liquefied gas near minus 196 degrees Celsius to superheated steam beyond 540 degrees Celsius. No single valve covers this range; selection is the discipline of mapping a specific service to a specific body style, characteristic, trim, and material set.
Chapter 2 / 06
Body Styles and Classification
Control valves are first classified by the motion of the closure member: linear (sliding-stem) or rotary. Linear valves move a plug straight along the seat axis and give fine, predictable throttling. Rotary valves turn a ball segment, plug, or disc through a quarter turn or less and give high capacity for their size and weight. Within these two families, the globe valve is the linear workhorse and the segment ball and high-performance butterfly are the rotary workhorses. The table below compares the main body styles on the metrics that drive selection.
Body style
Motion
Relative capacity (Cv/size)
Recovery factor FL
Typical duty
Globe (single-seat)
Linear
Low to medium
0.85 to 0.92
General throttling, high pressure drop
Cage-guided globe
Linear
Medium
0.85 to 0.90
Severe service, anti-cavitation trim
Angle valve
Linear
Medium
0.80 to 0.90
Flashing, erosive, coking service
Segment (V-notch) ball
Rotary
High
0.60 to 0.70
Slurry, fibrous, high rangeability
Eccentric rotary plug
Rotary
Medium to high
0.70 to 0.85
Compact general throttling
High-performance butterfly
Rotary
Very high
0.50 to 0.60
Large lines, low to moderate drop
Globe valves are the reference linear style. A globe body forces the flow through an S-shaped path past a plug that seats on a ring, and the long, guided stroke makes the plug shape a precise tool for setting the inherent characteristic. Globe valves have a high pressure recovery factor FL, typically 0.85 to 0.92, meaning pressure does not fully recover downstream of the restriction. A high FL is favorable because it resists cavitation and gives the valve good authority over high pressure drops, which is why the globe is the default choice for demanding throttling. The cost is capacity: a globe passes less flow than a rotary valve of the same nominal size, and its tortuous path raises body pressure loss.
Cage-guided globe valves retain the globe body but guide the plug inside a cage that also carries the flow openings. The cage is the most practical platform for severe-service trim: it can be drilled, slotted, or built as a stack of stages to break a large pressure drop into several smaller steps, holding each stage below the cavitation threshold. Cages also balance the plug, reducing the actuator thrust needed and allowing larger valves on smaller actuators. Angle bodies, a globe variant with a 90 degree flow path, are preferred for flashing and erosive service because the eroding stream discharges straight into an expanded outlet rather than impinging on the body wall.
Rotary valves trade throttling finesse for capacity and economy. The segment, or V-notch, ball valve uses a contoured notch in a rotating ball segment to give an inherently equal-percentage characteristic with very high rangeability, and it shears through slurries and fibrous media that would clog a globe. The eccentric rotary plug lifts its plug off the seat as it opens, reducing seat wear and giving good shutoff in a compact body. The high-performance butterfly offers the highest capacity per size and the lowest cost per Cv, but its low recovery factor (FL around 0.5 to 0.6) means pressure recovers strongly downstream, so it is prone to cavitation and noise and is best reserved for large lines at low to moderate differential pressure. IEC 60534-3 governs face-to-face and end-to-end dimensions so that valves of a given class are interchangeable between makers.
Chapter 3 / 06
Flow Characteristics and Trim
The inherent flow characteristic is the relationship between flow coefficient and valve travel at constant pressure drop, set by the contour of the plug or the shape of the cage openings. IEC 60534-2-4 defines the three standard characteristics, summarized below. Choosing the wrong characteristic is one of the most common and most consequential control valve errors, because it determines whether the loop gain stays roughly constant across the operating range or swings wildly, oscillating at one end and sluggish at the other.
Characteristic
Flow vs travel
Typical rangeability
Best when
Quick opening
Most capacity in first 25 to 30% travel
Low (under 10:1)
On-off, self-acting relief
Linear
Flow proportional to travel
~20:1 to 50:1
Valve drop is large, constant share
Equal percentage
Equal % change in flow per unit travel
~30:1 to 100:1
Valve drop falls as flow rises
Equal percentage is the most widely used characteristic in process control. Each fixed increment of travel changes the flow by the same percentage of the current flow, so the curve is steep at high openings and shallow near the seat. This shape compensates for a piped system in which the valve pressure drop falls as flow rises: as the drop available to the valve shrinks, the steepening inherent curve offsets it, and the installed characteristic (flow versus travel as the loop actually sees it) comes out close to linear. The result is roughly constant loop gain, which is what a PID controller needs to stay stable across the full range.
Linear characteristics suit systems where the valve absorbs a large and roughly constant fraction of the total system pressure drop, so the available drop barely changes with flow. Level control, where the valve faces a nearly fixed head, and many bypass and blending loops fall into this category. Because flow tracks travel directly, the installed and inherent curves stay close, and the loop gain is predictable. Quick opening reaches most of its capacity within the first quarter of travel and then flattens; it is used where the valve is effectively on or off, such as self-acting pressure relief or batch fill, not for fine throttling.
Rangeability, the ratio of maximum to minimum controllable flow coefficient, expresses how far down the valve can throttle while still controlling. It is conventionally measured between the openings where the installed gain remains usable. Equal-percentage globe trims typically reach 30:1 to 50:1, and segment ball valves can exceed 100:1 because their contoured notch maintains controllable gain to a very small opening. A valve operated below its minimum controllable flow chatters between seated and barely open, so oversizing, which pushes the normal operating point toward the bottom of the travel, is a frequent and avoidable cause of poor control.
Trim is the set of internal parts that throttle the flow and take the wear: the plug or closure member, the seat ring, the cage or guide, and the stem. On any high pressure drop liquid service the trim is also where cavitation and flashing attack. Cavitation occurs when the local pressure at the vena contracta, the narrowest, fastest point of the stream, drops below the liquid vapour pressure and forms vapour bubbles, which then collapse as the pressure recovers downstream. The collapse generates intense localized pressure spikes that pit hardened metal, plus noise and vibration. Flashing occurs when the downstream pressure stays below vapour pressure, so the vapour persists and the outlet carries a two-phase mixture; flashing cannot be cured by trim and is managed with hardened materials and expanded outlets. Anti-cavitation trim stages the pressure drop across multiple steps (drilled cages or stacked discs) so no single step crosses the vapour-pressure line, and the methods are codified in the IEC 60534-8 noise-prediction parts. The screening tool during sizing is the pressure recovery factor FL together with the liquid sizing equations of IEC 60534-2-1.
Chapter 4 / 06
Materials, Pressure Ratings, and Seat Leakage
Body material and pressure rating are governed by ASME B16.34, which sets pressure-temperature ratings for flanged, threaded, and welding-end valves. The standard defines the pressure classes 150, 300, 600, 900, 1500, 2500, and 4500, and groups materials so that each combination of class and material group carries a defined allowable working pressure that falls as metal temperature rises. The critical discipline is to confirm the rating at the highest metal temperature the valve will see, not at ambient: a Class 300 carbon steel body rated for roughly 49 bar at 38 degrees Celsius is derated substantially at 400 degrees Celsius. The body must also satisfy the corrosion demand of the process, which is a separate question from the pressure rating.
WCC cast carbon steel is the default body material for non-corrosive water, steam, air, and hydrocarbon service and covers the majority of installations. CF8M, the cast equivalent of 316 stainless steel with 2 to 3 percent molybdenum, is selected for general corrosion resistance and moderate temperatures, and is standard in chemical and many utility services. For aggressive chloride, strong acid, or sour service, nickel alloys such as Hastelloy C and reactive metals such as titanium are used, at a multiple of the stainless cost. Low-temperature and cryogenic service uses LCC or LCB low-temperature carbon steel or austenitic stainless with verified impact toughness at the minimum design temperature.
Trim is specified separately from the body because it sees the highest velocity and absorbs any cavitation or erosion. 316 stainless and 17-4PH precipitation-hardening stainless are common base trim materials, 17-4PH offering higher strength and wear resistance for plugs and stems. For erosive, cavitating, or high pressure drop service the sealing surfaces are hardfaced with Stellite 6, a cobalt-chromium alloy that resists erosion and galling and retains hardness at high temperature; the hardfacing is applied to the plug seating edge and the seat ring. The most severe duties use solid hardened trim or tungsten carbide. A practical caution: aggressive hardfacing or solid-hard trim can roughen the sealing line, so confirm that the required seat leakage class is still achievable after the hardening process is applied.
Shutoff tightness is graded by ANSI/FCI 70-2, harmonized with IEC 60534-4, which defines six seat leakage classes. The class is an explicit specification, not a quality grade: higher classes cost more, load the seat harder, and wear faster, so the correct practice is to specify the loosest class the process tolerates. The table below summarizes the classes.
Class IV at 0.01 percent of rated capacity is the standard expectation from a metal plug seating on a metal ring and suits most throttling duties. Class V is a tight metal-seat grade defined volumetrically (about 0.0005 millilitres per minute of water per inch of seat diameter per psi of differential) and is achieved by lapping the plug and seat together. Class VI is the soft-seat grade, reached with a PTFE or elastomer insert in the plug or seat, and is tested with air to an allowable bubble count that scales with seat diameter, roughly one bubble per minute for a 25 mm (1 inch) seat. Soft seats give the tightest shutoff but limit temperature and abrasion tolerance, so they are not a free upgrade.
Chapter 5 / 06
Key Specification Parameters
A control valve datasheet can list dozens of fields, but a manageable set drives the selection decision: rated flow coefficient, characteristic, rangeability, pressure recovery and choked-flow factors, actuator thrust and fail action, positioner and signal, seat leakage class, and the body rating already covered in Chapter 4. Each is decoded below so the numbers can be read against a real service.
Flow coefficient (Cv / Kv) is the headline capacity number. Cv is the US gpm of 60 degrees Fahrenheit water at 1 psi drop; Kv is the cubic metres per hour at 1 bar drop; the IEC 60534 conversion is Cv = 1.156 x Kv. The rated value is quoted at full open, and the sizing task is to pick a valve whose rated Cv comfortably passes the maximum flow while the normal flow lands in the controllable middle of the travel, typically with the normal opening near 60 to 80 percent for an equal-percentage valve. Undersizing starves the maximum flow; oversizing pushes normal operation down near the seat where control is poor and trim wear concentrates.
Pressure recovery factor FL and the liquid pressure recovery factor FF determine the onset of choked flow and cavitation. FL is the ratio that describes how much of the pressure drop is recovered downstream of the vena contracta; a high FL (globe valves, 0.85 to 0.92) recovers little and resists cavitation, while a low FL (butterfly, around 0.55) recovers strongly and is cavitation-prone. These factors feed the IEC 60534-2-1 liquid sizing equations to compute the allowable pressure drop before flow chokes, which caps the usable capacity regardless of how large the Cv looks on paper.
Actuator thrust or torque and fail action describe whether the actuator can stroke the valve under worst-case forces and where it goes on air or signal failure. Thrust must cover the unbalanced fluid force at the maximum differential pressure plus packing friction plus seat load, with margin. Fail action is set by the spring: fail-closed (air-to-open), fail-open (air-to-close), or fail-in-last-position with a lock-up valve. Fail action is a safety decision driven by what state is safe for the process, not a convenience setting.
Positioner, input signal, and protocol define the interface and the diagnostics. The legacy pneumatic command is 0.2 to 1.0 bar (3 to 15 psi); the electronic command is 4-20 mA, usually with HART overlaid for configuration and diagnostics; new installations may use FOUNDATION Fieldbus, PROFIBUS PA, or Ethernet-APL. A digital valve controller adds travel and cycle histograms, friction and air-mass trends, and partial-stroke testing that lets a safety shutdown valve prove its ability to move without a full trip.
The remaining datasheet fields complete the picture:
Seat leakage class: ANSI/FCI 70-2 Class I through VI, as decoded in Chapter 4. Specify the loosest class the process tolerates.
Rangeability: maximum to minimum controllable Cv ratio, commonly 30:1 to 50:1 for globe and over 100:1 for segment ball.
End connection and rating: raised-face or ring-type-joint flange, butt-weld, or threaded, at the ASME B16.34 class; face-to-face per IEC 60534-3 for interchangeability.
Packing and emissions: PTFE, graphite, or live-loaded packing; low-emission certification to ISO 15848-1 where fugitive emissions are regulated.
Noise prediction: predicted sound pressure level per IEC 60534-8-3 (aerodynamic) or 60534-8-4 (hydrodynamic), with 85 dBA the common design ceiling.
Read these fields together, not in isolation. A valve with a large rated Cv but a low FL may choke and cavitate long before that capacity is reached; an actuator sized for normal differential pressure may stall at the higher shutoff differential; a Class VI soft seat may not survive the service temperature. The datasheet only protects the buyer when every field is checked against the worst-case point of the actual duty.
Chapter 6 / 06
Selection Decision Factors
To turn the preceding chapters into a specified valve, work the decision sequence below in order. Most selection failures come not from a single wrong number but from deciding a downstream parameter before an upstream one is fixed, for example choosing a body style before the cavitation screen, or sizing the actuator at normal rather than worst-case differential. The eight steps double as a fixed RFQ template.
Define the process data set: fluid, phase, and density; minimum, normal, and maximum flow; inlet pressure and the valve pressure drop at each of those flows; operating and design temperature; and vapour pressure for liquids. Every later step references this set, so capture all three flow cases, not just the normal one.
Size the flow coefficient: compute the required Cv (or Kv) at minimum, normal, and maximum flow using the IEC 60534-2-1 equations, then check that the normal case sits in the controllable middle of the travel and the maximum case is reached below full open with margin. Run the choked-flow check using FL before trusting the rated Cv.
Screen cavitation and flashing: compare the service pressure drop against the cavitation and choking limits set by FL and vapour pressure. If the service cavitates, move to anti-cavitation multistage trim or an angle body; if it flashes, plan hardened trim and an expanded outlet, because no trim cures flashing.
Choose body style and characteristic: select globe, cage-guided globe, segment ball, or butterfly per Chapter 2 capacity, recovery, and media constraints, then choose linear or equal-percentage per the ratio of valve drop to system drop. Confirm the installed gain stays usable across the full flow range.
Specify materials and pressure rating: pick the body material for corrosion and the ASME B16.34 class for pressure-temperature, verified at the highest metal temperature, then specify trim base material and any Stellite hardfacing per Chapter 4. Confirm the seat leakage class is achievable with the chosen trim.
Size the actuator and set fail action: compute required thrust or torque from the unbalanced force at the maximum (shutoff) differential plus friction and seat load, with margin, then choose spring-and-diaphragm or piston accordingly. Set fail-closed, fail-open, or fail-last from the process safety case.
Select positioner, signal, and accessories: 4-20 mA with HART is the default; choose Fieldbus, PROFIBUS PA, or Ethernet-APL for digital plants. Add the solenoid, air filter-regulator, limit switches, and partial-stroke capability the safety and control duty requires.
Add certifications and total cost of ownership: hazardous-area ATEX / IECEx / NEPSI for the positioner, SIL capability for safety service, ISO 15848-1 for emissions, and PED 2014/68/EU where applicable. Weigh purchase price against trim replacement intervals, calibration labour, and downtime, since a control valve is repaired and recalibrated over a one-to-two-decade life.
A frequently overlooked final dimension is serviceability. A control valve is not a sealed unit; its trim wears, its packing is re-tightened, its positioner is recalibrated, and its actuator diaphragm is replaced. Local spare-trim inventory, field calibration service, positioner diagnostics support, and the availability of replacement plugs and seats in the body class all determine repair response time years after purchase. Established makers including Emerson Fisher, Samson, Flowserve, Valmet Neles, Baumann, and Spirax Sarco maintain regional service centres and spare-parts stock, which is a material consideration for any valve expected to stay in production for a decade or more.
FAQ
What is the difference between a control valve and an on-off (isolation) valve?
A control valve modulates flow continuously to a position commanded by a controller, holding any opening between fully closed and fully open with defined accuracy and repeatability. An on-off or isolation valve only travels between the two end positions and is built for tight shutoff, not throttling. The throttling duty drives the differences: a control valve carries a characterized trim (the plug, cage, and seat that shape the flow versus travel curve), a sized actuator, and almost always a positioner to enforce position against process forces. Throttling also subjects the trim to continuous velocity, cavitation, and erosion that an isolation valve rarely sees, so trim hardening and replaceability matter much more on a control valve.
What do Cv and Kv mean, and how do I convert between them?
Cv and Kv are flow coefficients that quantify a valve's capacity at a given opening. Cv is the number of US gallons per minute of 60 degrees Fahrenheit water that flows through the valve at a pressure drop of 1 psi. Kv is the metric equivalent: cubic metres per hour of water at a pressure drop of 1 bar. The conversion standardized in IEC 60534 is Cv = 1.156 x Kv, or Kv = Cv / 1.156 (approximately Cv divided by 1.16). Both are quoted at full open as the rated coefficient, and the inherent characteristic curve describes how the coefficient grows from near zero to rated as the valve strokes. Sizing means choosing a valve whose rated Cv covers the maximum flow while keeping the normal operating opening in a controllable band.
Which flow characteristic should I choose: linear, equal percentage, or quick opening?
Equal percentage is the default for most process control because each fixed increment of travel changes flow by the same percentage of current flow, which compensates for the falling valve pressure drop as flow rises in a piped system. Use linear when the valve pressure drop stays a large and roughly constant share of total system drop, such as level loops or bypass control, because then flow tracks travel directly. Use quick opening for on-off duty and self-actuated relief, where roughly 90 percent of capacity opens within the first 25 to 30 percent of travel. The governing definitions are in IEC 60534-2-4. The installed characteristic, which is what the loop actually sees, depends on the ratio of valve drop to system drop, so always check installed gain across the operating range, not just the inherent curve.
What is the difference between cavitation and flashing, and how do I handle them?
Both occur when the local static pressure at the vena contracta drops below the liquid vapour pressure, forming vapour bubbles. In flashing the downstream pressure stays below vapour pressure, so the bubbles persist and the outlet carries a two-phase mixture; flashing erodes the body and downstream pipe and cannot be cured by trim, only managed with hardened materials and expanded outlets. In cavitation the downstream pressure recovers above vapour pressure, so the bubbles collapse violently inside the valve, generating noise, vibration, and micro-jet pitting of trim surfaces. Cavitation is mitigated by choosing a high pressure recovery factor FL, by anti-cavitation multistage trim that stages the pressure drop, and by hardened trim such as Stellite-faced seats. The sizing screen uses FL and the liquid pressure recovery factors in IEC 60534-2-1.
What do the ANSI/FCI 70-2 seat leakage classes mean?
ANSI/FCI 70-2, harmonized with IEC 60534-4, defines six shutoff classes. Class I is an agreed value with no test. Classes II, III, and IV are metal-seat grades expressed as a percentage of rated capacity, with Class IV the standard metal-to-metal result at 0.01 percent of rated Cv. Class V is a tight metal-seat grade defined volumetrically, roughly 0.0005 ml per minute of water per inch of seat diameter per psi of differential. Class VI is the soft-seat grade, tested with air and expressed as allowable bubbles per minute that scale with seat size, for example about 1 bubble per minute for a 25 mm (1 inch) seat. Higher classes cost more and wear faster, so specify the loosest class the process tolerates rather than defaulting to Class VI everywhere.
How do I size the actuator and why is the positioner needed?
The actuator must deliver enough thrust or torque to stroke the valve against the worst-case combination of fluid forces, packing friction, and seat load, with margin at the highest expected differential pressure, not just the normal one. Spring-and-diaphragm actuators are the workhorse: they are simple, give a defined fail position from the spring, and respond cleanly to a 0.2 to 1.0 bar (3 to 15 psi) signal. Piston actuators give higher thrust for high pressure or large valves. The positioner is a local position controller that compares the commanded signal against measured stem travel and modulates supply air until they match, overcoming friction, unbalanced forces, and long air-line lag. A digital valve controller such as the Fisher FIELDVUE DVC6200 or Samson series 3730 adds HART diagnostics, travel histograms, and partial-stroke testing.
How do I select body and trim materials for an aggressive or high-temperature service?
The body must satisfy the ASME B16.34 pressure-temperature rating for the chosen class and material group across the full operating envelope, so confirm the rating at the highest metal temperature, not at ambient. WCC carbon steel covers most non-corrosive services; CF8M (cast 316) handles general corrosion and moderate temperature; alloys such as Hastelloy C and titanium handle chlorides and strong acids. Trim is selected separately because it sees the highest velocity and any cavitation: 316 and 17-4PH are common base trims, and Stellite 6 (cobalt-chromium) hardfacing on the plug and seat resists erosion and galling and holds hardness at temperature. For flashing, abrasive, or high differential service, specify solid hardened or tungsten-carbide trim and verify the seat leakage class is still achievable after the hardening process. Established makers including Emerson Fisher, Samson, Flowserve, Valmet Neles, Baumann, and Spirax Sarco supply these material and trim options and maintain regional spare-trim and calibration support.