Control Valve Sizing —
Cv, Flow Coefficient
and Why It Matters

A control valve that is the wrong size does not fail to open or close — it fails to control. An oversized control valve spends most of its operating life throttled nearly shut, working in a narrow band of travel near the seat where small movements produce large flow changes and the valve hunts continuously without settling. An undersized control valve cannot deliver the required maximum flow regardless of how far it opens, making the control loop fight a constraint it cannot overcome. Both failures look, from the outside, like a poorly tuned control loop — and both are frequently diagnosed as such, triggering weeks of PID tuning work that addresses the symptom rather than the cause.

This article covers how control valves are sized, what the flow coefficient represents and how it is used, what the characteristic curve means for control behaviour, and how to avoid the most common sizing errors that produce valves that open and close but do not control.

The Flow Coefficient — Cv and Kv

The flow coefficient is the fundamental parameter that characterises the hydraulic capacity of a valve at a given opening. Two conventions are used:

• Cv (US customary) — the flow rate in US gallons per minute of water at 60°F that produces a pressure drop of 1 psi across the valve at full open. Cv is the dominant convention in North American practice and in most international control valve manufacturer catalogues.
• Kv (SI) — the flow rate in m³/h of water at 5–40°C that produces a pressure drop of 1 bar across the valve at full open. Used in European practice. Kv = Cv × 0.865 (conversion factor).

The flow coefficient is not a fixed property of a valve — it is a function of valve opening. At full open, a valve has its maximum Cv (Cv100 or Cvmax). As the valve closes, Cv reduces. The relationship between valve travel (stem position as a percentage of full open) and Cv is the valve's inherent flow characteristic — one of the most important parameters in control valve selection, and the one most frequently overlooked.

The Sizing Equation

For incompressible (liquid) flow, the fundamental sizing equation is:

Cv = Q / (N₁ × √(ΔP / SG))

Where Q is flow rate (US gpm or m³/h depending on convention), ΔP is the pressure drop across the valve (psi or bar), SG is the specific gravity of the fluid relative to water, and N₁ is a unit conversion constant (1.0 for US units with Q in gpm and ΔP in psi; 0.865 for SI units with Q in m³/h and ΔP in bar).

This equation gives the required Cv to pass the specified flow at the specified pressure drop. The selected valve must have a Cvmax greater than this required value — but by how much is the critical judgement at the heart of control valve sizing.

Compressible flow — gas and steam

For compressible fluids (gases and steam), the sizing equation is more complex because the fluid density changes as it passes through the valve, and because at sufficient pressure drop the flow becomes choked — it reaches the speed of sound at the vena contracta and cannot be increased further regardless of the downstream pressure. The IEC 60534 standard provides the full compressible flow equations accounting for expansion factor Y, specific heat ratio, and choked flow conditions. For steam, the equations account for the degree of superheat and the thermodynamic properties at the valve inlet. For engineering purposes, simplified steam sizing equations are provided in most control valve manufacturer sizing software and in the Fisher Control Valve Handbook — these should be used in preference to hand calculations for gas and steam duties.

Valve Sizing Ratio — The Most Important Number Nobody Specifies

The required Cv calculated from the sizing equation is the Cv needed to pass maximum flow at minimum pressure drop (the worst-case sizing condition). The selected valve's Cvmax must be larger than this to ensure the valve is not fully open at maximum flow — a control valve running fully open has lost all control authority. But by how much?

The valve sizing ratio (also called the installed flow ratio or Cv ratio) is Cv_required / Cv_selected. Industry guidance generally recommends:

• Maximum operating Cv should be 60–80% of Cvmax — the valve should be 60–80% open at maximum operating flow
• Minimum operating Cv should correspond to at least 10–15% valve opening — below this, control is poor in most valve types

A valve sized so that maximum flow requires only 30% of Cvmax is heavily oversized — the entire operating range is crammed into 30% of the valve travel, and the control is consequently poor. A valve sized so that maximum flow requires 95% of Cvmax has almost no headroom — any disturbance that increases the required flow will push the valve fully open and the control loop saturates.

The practical target: size the valve so that the normal operating flow rate corresponds to approximately 60–70% of Cvmax, with the maximum operating flow at approximately 80%, leaving 20% of Cvmax as headroom for transient conditions.

Inherent Flow Characteristic — The Shape of the Curve

The inherent flow characteristic describes how Cv varies with valve travel in a constant pressure drop test. Three characteristics are standard:

Linear

Cv increases linearly with valve travel — a 10% increase in opening produces a 10% increase in Cv at any position on the curve. Linear characteristics are used where the system pressure drop is dominated by the valve (the valve takes most of the available pressure drop) and where the process gain is otherwise constant. Not commonly specified in practice — equal percentage is the default for most process control applications.

Equal Percentage

Each incremental increase in valve opening produces the same percentage increase in Cv, regardless of where on the curve the valve is operating. A 1% increase in travel from 20% open produces the same percentage increase in Cv as a 1% increase from 80% open — the absolute Cv change is much smaller at low opening than at high opening. The result is a logarithmic curve that provides inherent rangeability and makes the valve's effect on flow approximately uniform across the operating range. Equal percentage is the default characteristic for most process control applications, particularly where the system pressure drop varies with flow (as it does in most real piping systems — higher flow means more friction loss in the pipe, less pressure drop available for the valve).

Quick Opening

Large Cv change per unit travel near the closed position, flattening off as the valve approaches full open. Used for on-off applications (dump valves, relief bypass, level control by flooding) where the requirement is to open quickly to maximum flow — not for throttling control.

Installed Characteristic — What Actually Matters

The inherent characteristic (measured at constant pressure drop across the valve) is not the same as the installed characteristic — the actual relationship between valve travel and flow in a real piping system where pressure drop varies with flow. In most piping systems, as flow increases, friction losses in the pipe increase, reducing the pressure drop available for the valve. This distorts the inherent characteristic: an equal percentage valve installed in a system where the pressure drop falls significantly with increasing flow may exhibit a nearly linear installed characteristic. A linear valve in the same system may exhibit a quick-opening installed characteristic — poor for control.

The valve authority (β) quantifies this effect:

β = ΔPv,min / ΔPsystem

Where ΔPv,min is the pressure drop across the fully open valve at maximum flow, and ΔPsystem is the total system pressure drop (including the valve) at maximum flow. High authority (β close to 1.0) means the valve dominates the system — its inherent characteristic is approximately its installed characteristic. Low authority (β < 0.3) means the system resistance dominates — the installed characteristic diverges significantly from the inherent characteristic and control is poor regardless of how well the inherent characteristic was selected.

A valve authority below 0.2 should trigger a review of the system design — either the valve is undersized for the system, or the system resistance is too high relative to the valve drop, both of which degrade controllability.

Choked Flow and Cavitation

Choked flow in liquids — cavitation and flashing

In liquid service, as the pressure drop across a valve increases, the local pressure at the vena contracta (the minimum cross-section inside the valve body, downstream of the plug) falls. If this pressure falls below the vapour pressure of the liquid, vapour bubbles form — the liquid cavitates. If the pressure recovers downstream above the vapour pressure, the bubbles implode violently — cavitation. If the downstream pressure remains below vapour pressure, the liquid has flashed and the flow becomes two-phase.

Cavitation is destructive — the bubble implosion produces local pressure spikes that erode the valve trim, the body, and downstream pipework. A cavitating control valve produces characteristic noise (described as gravel in a pipe), vibration, and progressive erosion damage. Operating a valve in sustained cavitation service without appropriate anti-cavitation trim will destroy the valve.

The liquid pressure recovery factor FL (provided by the valve manufacturer) characterises the valve's tendency to cavitate. High FL (close to 1.0, typical of globe valves) means less pressure recovery downstream of the vena contracta — less tendency to cavitate. Low FL (butterfly, ball valves) means more pressure recovery — higher cavitation risk at equivalent pressure drop. Anti-cavitation trim (multi-stage pressure reduction designs) reduces the pressure drop per stage, keeping the local pressure above the vapour pressure throughout.

Choked flow in gases

In gas service, flow becomes choked when the pressure ratio across the valve (P2/P1) falls below a critical value (typically ~0.53 for air and diatomic gases). Below this ratio, flow cannot be increased further by reducing downstream pressure — the valve has reached its maximum gas flow capacity. Sizing must ensure the required maximum flow is achievable before choked conditions are reached, or explicitly account for choked flow in the sizing calculation.

Rangeability

Rangeability is the ratio of maximum controllable flow to minimum controllable flow for a given valve — in effect, the span of the operating range the valve can cover while maintaining adequate control. Inherent rangeability is typically 50:1 for quality globe valves and around 30:1 for rotary valves (ball, butterfly). Installed rangeability is always lower than inherent rangeability, because the installed characteristic distortion reduces the usable lower end of the range.

Where the required flow rangeability exceeds what a single valve can provide — common in processes with large turndown requirements — the standard solution is a split-range valve arrangement: two valves in parallel (a large valve for high flow, a small valve for fine control at low flow), controlled by a split output from the controller so that the small valve operates through the lower portion of the controller output signal and the large valve through the upper portion.

Body Style Selection

The Actuator and Fail Mode

Control valve sizing determines the valve body and trim. The actuator — pneumatic, electric, or hydraulic — must be sized to provide sufficient force or torque to open and close the valve against the process fluid forces at the maximum operating pressure differential. Under-actuated valves fail to close against high-pressure differentials or are slow to respond, both of which degrade control loop performance.

The fail mode must be specified explicitly and is a safety decision, not a controls decision:

• Fail closed (FC) — spring returns the valve to the closed position on loss of instrument air or power. Specified where the safe state on failure is to stop flow — fuel supply to a burner, reactant to a reactor, coolant bypass.
• Fail open (FO) — spring returns to the open position on failure. Specified where the safe state is maximum flow — cooling water to a heat exchanger, quench flow to a reactor.
• Fail in last position (FL) — valve locks in position on failure. Used where neither fully open nor fully closed is the safe state, typically where position retention is less critical than avoiding a step change.

Common Sizing Errors

Sizing to maximum possible flow rather than normal operating flow

The single most common oversizing cause. The process engineer specifies maximum conceivable flow (pump maximum, system flooded, all other valves open) as the sizing basis. The control valve manufacturer sizes to this and adds a margin. The result is a valve four times larger than needed for normal operation, spending its entire life working in the bottom 25% of its travel. Specify the valve for the normal operating flow range, with a stated maximum flow for headroom confirmation, not as the primary sizing basis.

Not specifying pressure drop across the valve

The pressure drop across the control valve is not fixed — it depends on the system design and the position of other elements in the circuit. A valve sized on the total available system pressure drop (pump head minus static head minus pipe friction) will be massively oversized because in normal operation the valve takes only a fraction of the total system drop. Specify the valve pressure drop at normal flow conditions, not the total system head.

Ignoring the installed characteristic

Specifying an equal percentage valve in a system with high valve authority (β > 0.7) can produce a near-linear installed characteristic where a linear valve would have been appropriate. Understanding the system resistance curve before selecting the inherent characteristic prevents mismatches between the selected characteristic and the system behaviour.

Specifying a valve one line size down "to save space"

Reducing the valve body size relative to the line size increases velocity through the valve body, which is sometimes specified to increase the pressure drop available for control. However, it also increases erosion risk, noise, and cavitation potential. The reducer and expander flanges required to install a reduced-body valve can cause turbulence, vibration, and flow measurement errors upstream if there are instruments close to the valve. Size the valve body for the process conditions, not for physical convenience.

The Valve Datasheet

A correctly completed control valve datasheet is the primary document for valve procurement. It should specify, at minimum:

• Fluid identity, phase, density, viscosity, vapour pressure at operating temperature
• Normal and maximum flow rates with associated pressure drops across the valve
• Inlet and outlet pressure at normal and maximum flow
• Operating temperature (normal, maximum, minimum)
• Required Cv at normal flow and at maximum flow
• Inherent characteristic required (equal percentage, linear)
• Valve body style, end connections, flange rating
• Body and trim material (to suit fluid and corrosion requirements)
• Actuator type and instrument air supply pressure
• Fail mode (FC, FO, FL)
• Shutoff class (ANSI/FCI 70-2 Class I through VI — Class VI for bubble-tight shutoff)
• Any special requirements: cavitation service, low noise trim, ATEX, fire-safe certification

Summary

A control valve is correctly sized when it operates in the middle third of its travel at normal flow, has 20% of Cvmax remaining at maximum flow, maintains adequate valve authority against the system resistance curve, and its inherent characteristic is compatible with the installed characteristic distortion produced by that system. A valve that is too large throttles in the bottom of its travel and hunts. A valve that is too small saturates at high flow. Both look like a tuning problem and are frequently treated as one — the correct diagnosis requires understanding the sizing, not turning the PID gains.