ISO 5208 grades valve seat tightness into rate classes from Rate A (no visually detectable leakage) through AA, B, C, CC, D, E, EE, F and G, each with a numeric allowance proportional to valve size. The purchaser selects the rate class; European EN valve standards reference ISO 5208 by default.
Where API 598 fixes pass/fail rules per valve type, ISO 5208 hands the decision to the buyer: pick a rate class, and the manufacturer tests to it. Rate A demands no visually detectable leakage for the test duration — the soft-seat grade. Rate B allows a tiny allowance (0.01 mm³ per second per millimetre of nominal size for liquid test), Rate C ten times that, and so on down the alphabet; rates beyond D exist mainly for large or special-purpose valves where drop-tight sealing is neither achievable nor needed.
The standard also defines the test conditions the rates apply to — fluid (water, air or inert gas), pressure (1.1× rated for closure tests), minimum durations by size — so a certificate reading "seat test ISO 5208 Rate A, 60 s, air 6 bar" is complete and auditable. EN 12266, the testing standard cited by EN gate/globe/check specifications, adopts the same leakage rates.
Practical guidance: specify Rate A for resilient-seated valves, Rate B or C for metal-seated depending on service criticality, and always name the class on the PO — see our full API 598 vs ISO 5208 comparison. VTXFLOW's computerised benches test and record to either standard as ordered.