Valve Leakage Class Explained: Limits, Testing and How to Specify

The Short Answer

Valve leakage class is a standardized way of expressing how much leakage a valve may pass when closed, expressed as a percentage of rated capacity or as a measured bubble rate. Industrial valves are tested to leakage classes defined by standards such as API 598 for conventional isolation valves and the FCI/ISA control valve leakage classifications (Class I through VI) for control valves. The class number is not a “quality score” — it is a contractual limit that must be matched to the service: a bubble-tight Class VI seat is unnecessary, and sometimes harmful, for a steam isolation valve, while a Class II seat may be completely inadequate for a gas service that requires minimal fugitive leakage.

What Leakage Class Actually Means

When a valve is closed, no practical seat is perfectly leak-free under all conditions. Standards define acceptance limits so that a buyer and a vendor agree on how much leakage is permissible. For process valves, leakage is typically measured during a seat (closure) test with the valve closed and pressurized, using air, gas, or water as the test medium. The measured leakage is then compared with the allowed limit for the specified class, and the valve either passes or fails its acceptance test.

Common Leakage Classes and Their Meaning

Class / Basis	Typical Leakage Limit	Typical Application
API 598 (conventional valves)	No visible leakage during factory shell and seat test (tight shutoff expectation)	Gate, globe, ball, butterfly valves in general process service
FCI 70-2 / ISA Class I	No measurable leakage limit specified; seat is intended to be tight	Control valves where seat tightness is not contractually specified
FCI 70-2 / ISA Class II	0.5% of rated valve capacity	General throttling service with modest shutoff needs
FCI 70-2 / ISA Class III	0.1% of rated valve capacity	Services needing tighter shutoff with throttling duty
FCI 70-2 / ISA Class IV	0.01% of rated valve capacity	Frequent tight shutoff combined with throttling
FCI 70-2 / ISA Class V	0.0005 ml/min per inch of orifice (liquid test)	Critical liquid shutoff, often with metal seats
FCI 70-2 / ISA Class VI	Bubble-tight — a small allowed bubble count per minute (air test)	Soft-seated control valves where minimal leakage is required

The specific numbers, test pressures and procedures are defined in the governing standard editions, so the values above should be verified against the current standard (data needs further verification for your exact spec). The key point is that each class is a measurable, contractual limit — not a marketing term.

How Leakage Is Measured

For conventional isolation valves, the factory seat test per API 598 pressurizes the closed valve (typically with the rated pressure or a specified test pressure) and checks for visible leakage past the seat. For control valves, the FCI 70-2 / ISA test method measures leakage across the closed valve at the specified pressure, either as a percentage of capacity (Classes II–IV) or as a measured bubble count or liquid volume (Classes V–VI). The test medium and pressure must match the class definition; testing a Class VI seat with water instead of air, or at the wrong pressure, invalidates the result.

Application Example: Choosing the Right Class for a Steam Valve

A plant specifies a control valve for a steam pressure-reducing station. The process engineer requests “the tightest seat available” without thinking through the service: the valve throttles most of the time and only occasionally needs tight shutoff. A Class VI soft seat would give bubble-tight closure but limits the valve’s temperature rating and adds cost. The correct specification is a metal-seated Class IV valve, which gives 0.01% leakage — more than adequate for the infrequent shutoff — while surviving the steam temperature and throttling duty. The leakage class, in other words, should be derived from the actual shutoff requirement, not chosen as a maximum. The same logic applies when sizing relief and safety functions — see our pressure relief valve sizing calculation for how the closed-valve tightness interacts with system protection.

Common Pitfalls

One common pitfall is demanding Class VI everywhere “to be safe,” which raises cost, limits materials and can actually reduce reliability in hot or dirty service. Another is testing a valve to the wrong standard — quoting an API 598 result when the specification called for FCI Class IV produces meaningless acceptance. A third is ignoring the test medium: air tests are more sensitive than water tests for the same seat, so a valve accepted on a water test may not meet a gas-service leakage requirement. And seat leakage classes only describe the closed valve; they say nothing about the valve’s throttling performance, which is validated separately.

Practical Recommendation

Define the leakage class in the datasheet using the correct standard and state the test medium and pressure explicitly. Derive the class from the service: soft-seated bubble-tight (Class VI) for clean, low-temperature media needing minimal leakage; metal-seated classes (II–IV) for throttling service with occasional shutoff; and a documented visible-leakage-free seat test per API 598 for conventional isolation valves. On receipt, witness or review the test certificate and confirm the class, medium and pressure match the specification. For valves that must also prove sealing in the field, the procedure in how to test a check valve shows how seat tightness is verified on installed equipment.

Conclusion

Leakage class is a measurable acceptance limit, not a prestige label. Match the class to the real shutoff requirement, test to the right standard and medium, and document the result. Specify too tight and you pay for capability you do not need; specify too loose and you accept leakage you cannot tolerate.

This article provides general educational knowledge only and is not professional engineering advice.

Sources & Further Reading

API, ISA

Tags: Valve Calculations