ANSI/FCI 70-2 defines six seat leakage classes for valves. Class IV (0.01% of rated capacity) is standard metal-seated quality, Class V is a much tighter liquid-tested metal seat, and Class VI is the bubble-tight soft-seat grade measured in bubbles per minute of air. The class must be named on the purchase order to be enforceable.
FCI 70-2 was written for control valves, but its class numbers have become the universal shorthand for seat tightness across isolation valves too. The ladder runs from Class I (no test, by agreement) through II and III (0.5% and 0.1% of rated capacity — commodity dampers and old designs), to the three classes that appear on real datasheets: Class IV, the default for metal-seated control and isolation trim; Class V, a high-pressure water test allowing only drops per minute per inch of seat — the severe-service metal seat grade; and Class VI, the soft-seat air test counted in bubbles per minute.
Two traps catch buyers. First, the classes use different test fluids and pressures — a Class V valve is not "one grade below" Class VI on the same scale, it is a different test entirely. Second, quoting "tight shutoff" without a class number is unenforceable; the supplier will test to whatever their default is.
For isolation valves ordered to API 598 or ISO 5208, map your intent: soft seats → bubble-tight (Class VI equivalent), metal seats → state Class IV or V explicitly. VTXFLOW tests to the class stated on the order and documents the result on the inspection certificate.