Home › Guides › API 598 vs EN 12266-1 vs ISO 5208

API 598 vs EN 12266-1 vs ISO 5208: which valve pressure test standard is in your PO

How the three valve pressure test standards differ, what API 6D changes for pipeline valves, what to write into your purchase order, and what a third-party witness checks at the test bench.

Standards Comparisons2026-11-1310 min read

API 598EN 12266-1ISO 5208API 6D

Short answer: API 598 is the inspection-and-test standard most US-style purchase orders name for gate, globe, plug, ball, check and butterfly valves. EN 12266-1 is the European standard for production pressure tests of metallic industrial valves, and ISO 5208 is the international one. All three cover a shell test and a seat (closure) test; they differ in scope, test options and how seat leakage is accepted. Test to the standard — and the edition — named in your PO.

By Kevin Lyu (吕中国), ValveWitness Inspection, Ningbo · Published 2026-11-13 · Last updated 2026-11-14

In our sample report: Section 9 is a pressure-test witness log written to API 598, with an example basis of ASME Class 150, NPS 2, Group 1.1 (WCB). The same section notes that pressures and minimum durations must come from the edition named in the PO — API 598, API 6D, ISO 5208 or EN 12266-1. See the sample report page and the PDF.

API 598, EN 12266-1 and ISO 5208 compared

API 598EN 12266-1ISO 5208
PublisherAmerican Petroleum Institute (API), USCEN (European standard; sold nationally as BS EN, DIN EN, etc.)ISO (international)
Edition at time of writing20232012 (EN 12266-1:2012)2015 (ISO 5208:2015, 4th edition)
ScopeInspection and pressure testing of gate, globe, plug, ball, check and butterfly valves — resilient-seated, non-metallic-seated and metal-seated. Supplements the API standards that reference it.Production pressure tests of industrial valves made of metallic materials. Not for safety devices. A valve product standard takes precedence where it differs.Pressure testing of industrial metallic valves, used together with the valve product standard, which takes precedence where it differs.
Tests coveredShell test; backseat test; low-pressure closure test; high-pressure closure test; double block-and-bleed closure test; visual examination of castings; optional high-pressure pneumatic shell testShell strength test (P10), shell tightness test (P11), seat tightness test (P12, for isolating and check valves); control-valve seat tightness goes to EN 1349Shell test, backseat test where applicable, closure (seat) test
Which closure tests applySet by valve type, size and pressure class in the standard, plus any optional tests the buyer specifiesSet by the standard and the product standard; optional tests are agreed separatelySet by the standard and the product standard
How seat leakage is acceptedMaximum allowable leakage depends on valve size and seat type (resilient or metal seat)Lettered leakage rates in Annex A; the rate comes from the product standard or the orderTen lettered leakage rates (A, AA, B, C, CC, D, E, EE, F, G). Rate A = no visually detectable leakage; the others scale with valve size (DN)
What your PO should stateStandard and edition; any optional tests (e.g. high-pressure closure where optional)Standard and edition; the leakage rate; any additional testsStandard and edition with the leakage rate, e.g. “ISO 5208:2015, Leakage rate A”

This table explains how the standards are organised. Test pressures, durations and allowable leakage values must be read from the licensed edition named in your PO; we do not reproduce the standards’ tables here.

Why the edition matters

  • POs often say “API 598” with no year. Ask which edition the factory’s test procedure follows, and make it match your PO.
  • EN 12266-1 and ISO 5208 do not fix one leakage rate for every valve. If the product standard does not set it and your PO does not state it, nobody can say whether a small seat leak is acceptable.
  • Write four things into the PO: the standard and edition, the leakage rate or class, the optional tests you want, and which tests are hold or witness points for your inspector.

How API 6D differs for pipeline valves

API 6D is a product specification for pipeline and piping valves (typically ball, gate, plug and check valves used on pipelines). It includes its own pressure-test requirements. API 598 is a test standard used across many general-purpose valve types, including those built to product standards such as API 600.

If your PO or ITP names API 6D, the valves are tested to API 6D, not API 598. A shop procedure written only around API 598 is not automatically acceptable. In practice, the differences buyers should watch for are:

  • Hold times. API 6D generally requires longer minimum pressure-hold durations than API 598 for a comparable valve size. Do not assume the short API 598 durations (for example the 15-second example used for NPS 2 in our sample report) apply to a 6D valve.
  • Extra tests for pipeline features. Where the valve is specified as double block and bleed (DBB) or double isolation and bleed (DIB), API 6D requires the corresponding cavity / isolation tests. Those tests are not a substitute for, and are not always present in, a generic API 598 sequence.
  • Leakage acceptance. Soft-seated pipeline valves are typically required to show no visually detectable leakage under the specified seat test. For metal seats, API 6D commonly points to lettered ISO 5208 leakage rates rather than the API 598 size-and-seat table. The PO should name the rate.
  • Gas and liquid seat tests. Pipeline specifications often require both a hydrostatic seat test and a low-pressure gas seat test. Confirm which fluids and pressures the edition and your data sheet require, rather than assuming one closure test is enough.

None of this is a licence to mix the standards after the fact. A PASS recorded against API 598 does not, by itself, demonstrate conformity to API 6D.

What a third-party witness checks at the test bench

  1. The test procedure, standard and edition on the bench match the PO.
  2. Gauge ID, range and calibration date — and the test pressure falls in a sensible part of the gauge range.
  3. Shell test: test pressure reached, hold timed on a watch or video, no visible leakage through body, bonnet or body joints.
  4. Seat test: the correct side or sides pressurised, gas or liquid as specified, leakage observed and recorded against the required rate.
  5. Backseat test only where the valve has a backseat.
  6. Any cavity, DBB or DIB test required by the product specification or the data sheet.
  7. Stainless steel valves: test water chloride content, if your specification sets a limit.
  8. Serial numbers of the valves witnessed against their nameplates, so the report ties to real units.
  9. Any re-test is recorded with the reason — never “test until it passes” off the record.

Common mismatches buyers run into

  • The PO says EN 12266-1 or ISO 5208 but no leakage rate, so the factory applies the least strict rate the product standard allows.
  • The PO says API 6D but the factory’s test sheet is laid out as API 598, with short holds and no cavity test.
  • A test certificate says “PASS” with no pressures, durations or gauge numbers.
  • Identical results printed for every valve size on the certificate.
  • One closure test is passed but another required test was never run. A valve can be tight on a high-pressure water seat test yet leak on a low-pressure air test, because a low pressure may not push the closure member firmly onto the seat.

What to write into the purchase order

  • The test standard and the edition (API 598, EN 12266-1, ISO 5208 or API 6D — not a mixture unless you have agreed a written deviation).
  • The leakage rate or class, especially for EN 12266-1, ISO 5208 and metal-seated API 6D valves.
  • Which closure tests are mandatory (high-pressure, low-pressure, both sides, gas and/or liquid) and any DBB/DIB cavity tests.
  • The witness or hold-point quantity (for example 100%, or a stated sample while the factory tests 100%).
  • That the factory records pressures, durations, gauge identification and serial numbers — a one-word “PASS” is not a test record.

Questions buyers ask

Is EN 12266-1 stricter than API 598?

It depends on the leakage rate chosen. EN 12266-1 lets the product standard or the buyer select a lettered leakage rate, while API 598 sets its own allowable leakage by valve size and seat type. Compare like with like: same valve type, size, test fluid and leakage rate.

Can a valve tested to API 598 be accepted against an EN 12266-1 or ISO 5208 specification?

Only if the buyer agrees in writing. The test sequence, durations and acceptance criteria are not identical, so a PASS to one standard is not automatically a PASS to another.

What does Leakage Rate A mean?

In ISO 5208 (and the lettered leakage rates used by EN 12266-1), Rate A means no visually detectable leakage for the duration of the test: no visible drops, seepage or gas bubbles. It is a visual criterion under the stated test conditions, not a measured zero.

How is API 6D different from API 598?

API 6D is a product specification for pipeline and piping valves and includes its own pressure-test requirements. API 598 is a general inspection-and-test standard used across many valve types. If the purchase order names API 6D, the valves are tested to API 6D, not API 598. Hold times, additional tests (for example cavity or double-block-and-bleed tests where specified) and leakage acceptance can all differ; read them from the edition named in the PO.

Can pressure tests be witnessed by video?

Yes, if it is agreed in advance. The video should show the gauge, the valve serial number and the full hold, and the report should state who timed the holds.

Who performs the test during a witness inspection?

The factory performs the test on its own bench with its own gauges. The inspector witnesses, times the holds, checks the gauge and records the results.

Sources

  • API 598, Valve Inspection and Testing — API publications store
  • API 6D, Specification for Pipeline and Piping Valves — API publications store
  • EN 12266-1:2012, Industrial valves — Testing of metallic valves — Part 1: Pressure tests, test procedures and acceptance criteria — Mandatory requirements — BSI (BS EN 12266-1:2012)
  • ISO 5208:2015, Industrial valves — Pressure testing of metallic valves — iso.org

Need the test witnessed? See valve pressure test witnessing, the sample report, or who will be at the bench.

We report facts from the samples inspected. Standards named in these guides are public standards we work to, not certifications or accreditations we hold. We do not issue product certificates or accreditation marks. Test pressures, durations, leakage rates and acceptance numbers must be read from the licensed edition named in your purchase order; we do not reproduce copyrighted tables from the standards.

Related

PSI vs DUPRO vs test witness vs loading

What each inspection stage actually covers, why a late PSI can miss casting and dimensional faults under paint and crates, and how to write hold points into your PO.

Related

ASME B16.5 vs EN 1092-1 flanges

Why Class 150 is not PN16 or PN25, which flange and face-to-face dimensions a PSI actually measures, and how API 600 / ASME B16.34 relate to EN 1984.

Related

ISO 2859-1 vs ANSI/ASQ Z1.4

How General Level II and special levels S-3/S-4 set the sample size, how Ac/Re numbers work, and why a DFT deviation is not counted in the visual AQL tally.

Need an inspector at a Zhejiang factory?

Send your PO or specification, the factory city and the date. We confirm scope and price within one working day.

Request an inspectionWhatsApp +86 131 2757 0627