Valve Seat Leakage Class: Class I to VI Explained

Choosing the right valve seat leakage class is a quiet decision on a specification—and an expensive one to get wrong. When your process needs a control valve to hold tight, the leakage class on the data sheet decides how much fluid may pass once the plug is seated. This guide explains the six ANSI/FCI 70-2 leakage classes so you can match Class I through VI to your real operating duty. You will also see where soft and metal seats change the math, and when bubble-tight Class VI is worth the cost to your business.

What Are the 6 Leakage Classes?

The six leakage classes are defined in ANSI/FCI 70-2 (formerly ANSI B16.104), with IEC 60534-4 as the equivalent international reference. The standard applies to control valves—the modulating globe, rotary, and specialty designs your plant uses to throttle flow—not to on/off isolation valves. Its job is simple: it turns the vague instruction "the valve must close tight" into a measurable, contractually enforceable acceptance test, with a defined test medium, test pressure, and maximum allowable leakage for each class.

For Classes II through IV, the allowable leakage is expressed as a percentage of the valve's rated capacity (its Cv at full rated travel). That means the permissible leak scales with valve size and trim, so a large Class IV valve is allowed more absolute leakage than a small one. Classes V and VI break that pattern: their limits are fixed volumes based on port diameter and—for Class V—actual differential pressure. Understanding that shift is the key to reading a data sheet correctly.

In plain terms, the ladder runs from "no test at all" (Class I) to "bubble-tight by measurement" (Class VI). Class II is a generic industrial promise that the valve closes; Class III tightens the metal seating; Class IV is the workhorse for metal-seated control valves; Class V is a precision water test at real shutoff pressure; and Class VI is the soft-seat bubble test. The classes are not a simple decimal scale—Class V and VI use completely different units and test pressures, so Class VI is not automatically tighter than Class V at every size.

Soft Seat vs Metal Seat — Which Class?

The seat material is what actually determines which leakage class your valve can reach. A metal seat leakage class (Classes II through V) relies on precisely machined and often lapped metal-to-metal surfaces. The plug and seat are both metal, so the valve survives high temperature, abrasive media, and fire-safe duty—but it can only approach, never quite reach, a perfect seal. Class IV is the normal result for a well-made metal-seated control valve, while Class V demands extra lapping, higher seat load, and a strong actuator to hold the surfaces together at full differential pressure.

A soft seat leakage class, by contrast, uses a resilient insert—PTFE, EPDM, FKM, or PEEK—that deforms slightly under closing force and conforms to the metal counterface. That compliance closes the micro-gaps a hard metal pair cannot, which is why soft seat leakage class performance is what earns a valve its Class VI bubble-tight rating. The trade is temperature and chemical limits: a polymer seat loses its shape above roughly 200 °C (lower for some elastomers), and aggressive media can swell or embrittle it. If you want to compare seat materials across a specific product family, our breakdown of metal-seated ball valves shows where a lapped metal face beats a soft polymer and where it does not.

In practice the choice is rarely "which is better" and almost always "which fits the duty." For hot, dirty, or fire-risk service you accept a small metal-seat seepage and specify Class IV or V. For clean, moderate-temperature service where tight shutoff matters, a soft-seated ball valve or soft-seated control valve delivers Class VI at a lower cost than chasing the same result in metal. Your business should pick the seat for the fluid first, then read the resulting class off the standard.

When Do You Need Class VI Bubble-Tight?

You reach for a leakage class VI rating when even a whisper of pass-through is unacceptable. The classic cases are gas isolation, hazardous or toxic fluids, and valuable products you cannot afford to lose through the seat. A single-seat globe control valve with a PTFE insert, or a soft-seated ball or butterfly valve, is the textbook way to get there: the resilient seat fills the gaps, and the factory bubble test confirms it against the FCI 70-2 table. Clean water, HVAC, and fire-protection loops also lean on Class VI because the fluid is clean enough that the soft seat will not erode.

Two cautions protect your specification from an expensive mistake. First, Class VI is measured, not absolute: a 6-inch valve may legally pass 27 bubbles per minute and an 8-inch valve 45. That is invisible on a roaring line but is not zero, so do not write "zero leakage" on the purchase order. Second, the Class VI test pressure is only 50 psig (or the rated differential, whichever is lower). If your line runs at 500 psig, a valve that passed bubble-tight at 50 psig can still weep in service—because the soft seat may not maintain contact at the higher load. For genuinely high-pressure tight shutoff, a metal-seated Class V design tested at the real ΔP is often the safer call.

When Is Class IV Enough?

For most process plants, a leakage class IV metal-seated valve is more than enough—and asking for more would only add cost and fragility. Class IV allows 0.01% of rated capacity at a modest 45–60 psig test, which a standard metal-seated globe or rotary control valve achieves reliably on the production line. Steam pressure control is the textbook example: a globe control valve in steam service routinely meets Class IV, holds up to temperature, and shrugs off the cycling that would chew up a soft seat. Our guide to high-pressure steam control valve specifications walks through how Class IV trim is selected for that duty.

Class IV is also the right answer whenever the valve sits in a line that already has separate block valves upstream and downstream. In that arrangement the control valve modulates flow most of its life and only needs to "roughly close"; positive isolation is handled by the block valves during maintenance or emergency. Specifying Class VI there buys you nothing but a softer, temperature-limited seat. Reserve the tighter classes for the moments when the closed control valve itself is the only barrier between your process and a leak.

How Is Leakage Tested?

A seat leakage test follows a repeatable routine so that "Class IV" means the same thing in every factory. The valve is placed fully closed, the test medium is connected to the upstream side, and the specified pressure differential is applied and allowed to stabilize. For Classes II, III, and IV the medium is air or water at 45–60 psig, and the leakage is measured downstream over a set interval. Class V switches to water at the actual maximum service differential pressure, which is why it is the honest choice for high-pressure duty. The actuator must deliver its full rated closing thrust during the test—a detail that quietly verifies valve-and-actuator matching.

Class VI uses the famous bubble test. Air or nitrogen at 50 psig is fed through the closed seat, and the escaping gas is collected in a calibrated tube submerged under water; the bubbles per minute are counted against the port-diameter table. The result is reported as a bubble count, not a volume, which is exactly why people call it "bubble-tight." The discipline shares its philosophy with adjacent standards work—if fugitive emissions matter in your plant, our practical guide to fugitive emission test standards explains the related test thinking for stem sealing.

FAQs

What is a valve seat leakage class?

A valve seat leakage class is a standardized rating—Classes I through VI under ANSI/FCI 70-2—that defines the maximum fluid allowed to pass through a fully closed control valve, plus the required test medium and test pressure. It converts the requirement "the valve must shut off" into a measurable, contractually enforceable number that both you and the manufacturer can verify at factory acceptance.

Is Class VI leakage really zero leakage?

No. Class VI is "bubble-tight" by test, not zero. FCI 70-2 permits a specified bubble count per minute that rises with port diameter: 1 bubble/min at 1 inch, 27 at 6 inches, and 45 at 8 inches. For genuinely zero pass-through you need a separate block valve or a special metal-seated design, not a standard Class VI soft seat.

What is the difference between Class IV and Class V?

Both are metal-seated. Class IV allows 0.01% of rated capacity at a 45–60 psig test, while Class V allows 5 × 10ₙ mL/min per inch of port per psi of actual service differential, tested with water at the real shutoff pressure. Class V is therefore the stricter choice for high-pressure, long-closed duty where the valve must hold against a large ΔP for extended periods.

Can a metal-seated valve reach Class VI?

In practice, FCI 70-2 Class VI is defined for resilient (soft) seats such as PTFE, EPDM, FKM, or PEEK. A metal-seated valve can approach bubble-tight only with extensive lapping and special geometry, so for a documented Class VI you normally specify a soft seat. Treat any "metal-seated Class VI" claim with care and ask for the actual test method.

Does FCI 70-2 apply to isolation valves like gate or ball valves?

No. FCI 70-2 covers modulating control valves only. Isolation and block valves are tested to API 598 (US) or ISO 5208 / EN 12266-1 (international), which use different graded or pass/fail rates. Writing a control-valve leakage class on an isolation valve creates a false sense of security because the test pressure is far below line pressure.

Should I write FCI 70-2 or ANSI B16.104 on the data sheet?

They are the same standard: ANSI B16.104 was the original name and was later published as ANSI/FCI 70-2 (current edition ANSI/FCI 70-2-2006). Writing "ANSI/FCI 70-2" is the current, unambiguous choice, and IEC 60534-4 is the equivalent international reference your overseas suppliers will recognize.

Conclusion

Specifying a valve seat leakage class is not about picking the tightest number—it is about matching the shutoff acceptance test to your fluid, pressure, and risk. A metal-seated control valve at Class IV is the workhorse of process plants, while a soft-seated design earns its bubble-tight leakage class VI rating in gas, clean-water, and hazardous-duty service. Keep in mind that Class VI is a measured bubble count rather than zero, and that FCI 70-2 defines control-valve shutoff, not isolation-valve integrity. Choose the class your process can justify, document it on the data sheet, and verify it at factory acceptance.