- Seat leakage testing is the final quality gate that determines whether a valve actually seals — and the difference between a valve that passes API 598 Class VI and one that merely meets API 598 Class IV is not just a number on a test certificate, it is the difference between a valve that will hold zero leakage in critical service and one that will allow measurable fluid passage when closed.
- API 598 and IEC 60534-4 define the two principal leakage test frameworks used in industrial valve procurement — one for isolation valves (gate, globe, ball, butterfly, check), the other for control valves — and understanding what each standard requires, how tests are conducted, and what the results mean for valve selection is essential knowledge for specification engineers and procurement inspectors.
- This guide covers both standards in detail: test procedures, leakage classes, acceptance criteria, test fluids, test pressures, and the practical knowledge needed to witness a seat leakage test and evaluate whether a valve meets its claimed performance.
API 598: Valve Inspection and Testing
API 598 is the standard governing inspection and testing of gate, globe, ball, plug, butterfly, and check valves in general oil and gas and petrochemical service. It defines three types of tests that most valves must pass before shipment: the shell (body) hydrostatic test, the seat closure test, and the backseat test (for valves with a backseat feature on the stem). The primary focus here is the seat closure (leakage) test, which determines how much fluid passes through the valve’s primary isolation when the valve is fully closed.
API 598 defines seat test acceptance criteria in terms of the allowable number of drops per minute or millilitres per minute of test fluid through the closed valve bore. The criteria are organised by valve type and seat design: Class I applies to metal-seated valves (gate, globe, check, ball, plug, and butterfly with metal seats), permitting leakage up to a specified drops/minute or ml/minute based on valve size. Class II is a tighter criterion for the same valve types. Class III applies to metal-seated valves in specific configurations. Class IV and Class V apply to valves with higher seating performance requirements. Class VI is the tightest standard class, applicable to soft-seated valves (with elastomeric or PTFE seats), and requires essentially bubble-tight shutoff — the allowable leakage is expressed as a maximum number of bubbles per minute at specific test duration, which for larger valves corresponds to less than 0.15 ml/min.
The test procedure under API 598 for the seat closure test uses water as the standard test fluid (for the normal ambient-temperature hydrostatic test), with air or nitrogen as the optional pneumatic test medium for soft-seated valves where the bubble-counting method is used to verify Class VI performance. The test pressure is the rated working pressure (differential pressure across the closed seat), and the valve is held at test pressure for a specified minimum duration while leakage is measured. For Class VI (bubble test), the test fluid is air or nitrogen at 6 bar (80 psi) or rated working pressure, whichever is lower, and the downstream side of the closed valve is submerged in water while observers count the rate of bubble formation.
The allowable leakage for API 598 Class VI varies by valve size: for NPS ½ to NPS ¾, 0 bubbles per minute (zero leakage); NPS 1 to NPS 2, 0 bubbles per minute; NPS 2½ to NPS 8, 0.1 to 1.5 ml/min (expressed as equivalent bubble counts). These are stringent requirements that soft-seated ball valves and globe valves with PTFE/PEEK seats can routinely achieve, but metal-seated valves typically cannot maintain over extended service life due to surface finish degradation and particle contamination of the seating surfaces.
IEC 60534-4: Control Valve Seat Leakage Classes
IEC 60534-4 (Industrial Process Control Valves — Part 4: Inspection and Test Procedures) defines seat leakage classes specifically for control valves, using a classification system from Class I through Class VI where the classes represent progressively tighter shutoff requirements. The IEC classes differ from API 598 in that they are expressed as a percentage of the rated Cv (or Kv) of the valve, providing a normalised leakage criterion that accounts for valve size and capacity.
IEC 60534-4 Class I specifies no test requirement — Class I valves are not required to demonstrate any shutoff capability, and are appropriate for valves that will never be required to provide isolation (pure throttling service where the valve will always be partially open). Class II allows leakage up to 0.5% of rated Cv. Class III allows 0.1% of rated Cv. Class IV allows 0.01% of rated Cv. Class V is a special test class requiring water at rated differential pressure with a maximum leakage of 5 × 10⁻⁴ ml/min per millimetre of seat diameter per bar of differential pressure — a very tight requirement typically achievable only by valves with special seat designs and excellent surface finish. Class VI requires zero measurable leakage (bubble-tight) in the same test conditions as API 598 Class VI, using air or nitrogen as the test medium.
The most commonly specified IEC 60534-4 class for process control valves is Class IV — “standard” shutoff capability for a good-quality control valve with properly lapped or ground seating surfaces. Class VI is specified when a control valve must provide true isolation — for example, where the valve also functions as a block valve and must prevent any leakage to a downstream system that may be pressure-tested or maintained at a different pressure while the upstream system remains in service. Class VI control valves are more expensive and must be maintained to a higher standard to retain their tight shutoff performance over service life.
Witnessing a Seat Leakage Test: What to Observe
When witnessing a seat leakage test as a purchaser’s inspector or TPIA representative, there are specific things to observe and verify that go beyond simply watching the test operator run through the procedure. Confirm that the test setup matches the test requirements: the test fluid is correct (water for API 598 liquid test, clean dry air or nitrogen for API 598 pneumatic test); the test pressure gauge is calibrated and the gauge range is appropriate for the test pressure; the valve is installed in the correct orientation for the test (some valves must be tested with pressure applied in the defined direction, and installing in reverse can give misleading results).
Verify that the valve is set correctly before pressurisation: for gate and globe valves, confirm that the closure element is fully seated (the handwheel or actuator has been operated to the fully closed position, not just partially closed). For ball valves, confirm the ball is in the closed position with pressure on the correct side (the upstream side per the test standard). Do not allow the test pressure to be increased gradually to mask the initial leakage — the valve should be brought to test pressure and held there for the full specified duration.
For bubble tests (Class VI), observe the entire test duration and count bubbles independently — do not accept the manufacturer’s count without verification. Ensure the downstream connection is properly submerged and that the water is clear enough to see bubbles accurately. If bubbles are observed, confirm the rate is within the allowable limit for the valve size before signing off the test. Any ambiguity should be resolved in favour of repeating the test rather than accepting a marginal result.
Seat leakage testing is one of the most important quality gates in valve manufacturing. A valve that passes the required leakage class at factory test, documented by a witnessed test certificate, has demonstrated its isolation capability under defined conditions. A valve without such documentation, or with a test certificate of uncertain provenance, has not. The discipline of demanding and witnessing proper seat leakage tests — and understanding what they demonstrate — is a core competency for any valve engineer involved in procurement and quality assurance.