European project RFQs often carry a single line: “seat test to EN 12266-1, Rate A”. Chinese manufacturers tend to react in one of two wrong ways — either they apply API 598 drops/bubbles and get marked down for using the wrong standard, or they treat EN 12266-1 as “a stricter API 598” and quote every valve at Rate A, distorting both cost and process. The two standards measure the same physical quantities (shell, seat, back seat) but speak different acceptance languages: one uses lettered rates (A–G) multiplied by DN, the other uses drops per minute and bubbles per minute looked up by size. They cannot be translated into each other; each must be executed as referenced.
This article follows the chain: how the two parts divide → the four test references → pressure and medium → hold times → choosing a rate → what Part 2 adds → the PED relationship → reports and bench requirements. Figures follow EN 12266-1:2012 and its Annex A.
1. Two parts: mandatory and supplementary
| Part | Status | Scope |
|---|---|---|
| EN 12266-1:2012 | Mandatory | Pressure tests, procedures and acceptance criteria: shell resistance and tightness, closure resistance, seat tightness |
| EN 12266-2:2012 | Supplementary | Back seat, antistatic, cryogenic, cycling, torque/thrust — when referenced by the product standard or order |
Part 1 covers production testing of metallic valves and may also be used as a type or acceptance test. Safety devices such as safety valves are outside its scope (they follow the ISO 4126 family). That matters: a customer citing EN 12266-1 for a safety valve has cited the wrong standard.
2. The four test references: P10, P11, P20, P12
This is where Chinese-language summaries most often go wrong. Annex A of EN 12266-1 defines:
| Reference | Test | What it proves | Pressure |
|---|---|---|---|
| P10 | Shell resistance | Body, bonnet and joints do not rupture or permanently deform | 1.5 × PS |
| P11 | Shell tightness | No external leakage across the pressure boundary | 1.5 × PS |
| P20 | Closure resistance | Pressure capability of the closure where the valve is the sole means of isolation | 1.5 × PS |
| P12 | Seat tightness | Leakage past the seat in the closed position | ≥ 1.1 × PS |
P11 is not the back-seat test. The back seat sits in Part 2 (P21) and is performed only when a product standard requires it. Treating P11 as a back seat has two consequences: shell tightness is skipped, and the report contains a test reference the standard does not define — an auditor spots it immediately.
3. Pressure and medium: beyond 1.5× and 1.1×
- P10, P11 and P20 are tested at 1.5 × PS (PS = maximum allowable pressure at ambient temperature);
- P12 is tested at no less than 1.1 × PS; with a gas medium, at the lower of 1.1 × PS and (6 ± 1) bar, and that reduction applies only to DN ≤ 80 at all ratings, and DN 80–200 at ratings up to PN 40 / Class 300;
- Shell resistance and tightness should be tested with liquid unless otherwise agreed; air must be vented from the cavity first;
- A gas shell test must follow a successful liquid strength test and needs proper safeguards — stored energy, not paperwork.
Acceptance: P10, P11 and P20 permit no visible leakage with either medium. At P11, leakage from the operating-mechanism seal is tolerated, provided there is no visible leakage when the pressure is reduced to 1.1 × PS — a clause frequently missing from report conclusions.
4. Hold times: production and type tests are different orders of magnitude
| DN / mm | P10, P11 (production) | P12 (production) | Type test (P10, P11, P12) |
|---|---|---|---|
| ≤ 50 | 15 s | 15 s | 10 min |
| 65 – 150 | 60 s | 60 s | 10 min |
| 200 – 300 | 120 s | 120 s | 10 min |
| ≥ 350 | 300 s | 120 s | 10 min |
Two exceptions worth memorising:
- At DN ≥ 350 the shell test runs 300 s but the seat test only 120 s — the two columns do not scale together;
- Where line cycle time is shorter than the specified duration, the standard permits testing to cycle time only with a statistically valid full-duration sampling programme and SPC evidence of batch capability. That second half is routinely dropped in local practice, and it does not survive an audit.
5. Rates A–G: the product standard chooses, not the manufacturer
Table A.5 gives maximum allowable seat leakage per rate in mm³/s, scaled by DN:
| Medium | A | B | C | D | E | F | G |
|---|---|---|---|---|---|---|---|
| Liquid | No visible leakage during the test | 0.01 × DN | 0.03 × DN | 0.1 × DN | 0.3 × DN | 1 × DN | 2 × DN |
| Gas | 0.3 × DN | 3 × DN | 30 × DN | 300 × DN | 3000 × DN | 6000 × DN | Not given for gas |
The selection rule is one sentence: the applicable product standard specifies the rate; an order may tighten it but may not choose a looser one. In practice soft-seated valves are generally called at Rate A (no visible leakage), metal-seated valves at C or D depending on the product standard — gate, ball and butterfly each have their own clauses.
Relationship to API 598, in three lines:
- Soft seat: effectively equivalent — both mean no visible leakage;
- Metal seat: not convertible. API 598 gives drops/min and bubbles/min by size; EN 12266-1 gives mm³/s × DN. The same valve can pass under one and fail under the other;
- So the RFQ line must read “EN 12266-1, Rate __” or “API 598 + seat construction + allowable value”. A standard number alone is not a leakage specification, and it is the main source of late-stage disputes.
6. What Part 2 adds: back seat, antistatic, cryogenic, cycling, torque
EN 12266-2 is supplementary — not mandatory unless referenced. It covers exactly the items most often added late in project review:
- Back-seat test (P21): verifies the upper seal with packing relieved;
- Antistatic: resistance across stem–ball/disc–body, typically ≤ 10 Ω, essential for flammable service;
- Cryogenic: links to ISO 28921-2 and BS 6364 for cold seat leakage and cycling;
- Cycling / life: cycles, medium and rejection criteria must be written into the agreement — the standard sets no universal number;
- Torque / thrust: break and run torque at maximum differential, for actuator sizing.
None of these are in Part 1, so “complies with EN 12266-1” does not mean “antistatic passed”, nor “cryogenic capable”. Name the Part 2 clauses explicitly in the specification.
7. Relationship to PED/CE: a harmonised standard gives presumption of conformity
EN 12266-1 is a harmonised standard under PED 2014/68/EU. Testing to it gives presumption of conformity with the relevant Essential Safety Requirements of Annex I, but testing alone is not CE marking — CE is a conformity assessment procedure (classification, module, technical file, notified body, declaration of conformity). Conversely, PED affects testing in one practical way: for Category I valves, statistical proof testing may replace unit-by-unit shell testing, an exception written into EN 12266-1 itself. Useful for line planning, provided the statistical scheme can be produced on request.
For classification, modules and the CE marking route, see the PED 2014/68/EU and CE marking article; for the ten-class ISO system, see ISO 5208 leakage rates, and for the US-side comparison API 598 vs GB/T 13927.
8. Reports and benches: the four records European projects ask for first
- Standard and edition: write “EN 12266-1:2012”, not “EN 12266”;
- Actual test pressure and medium for each test, with temperature;
- Hold duration and acceptance rate: P12 must state Rate A/B/C/D — “passed” alone is not a result;
- Traceability: serial or heat number tied to EN 10204 3.1 material certificates, with inspection signed independently of production.
On the equipment side, three things matter: gauge accuracy and calibration interval; fixtures that add no load to the seat; and gas-test venting and safeguards. A 300 s shell hold at DN ≥ 350 demands more pressurisation control and stability than the usual API 598 rhythm, so size pumps and accumulators for the strictest band.
EN 12266-1 governs how to pressurise and how to judge — not whether a valve is fit for a supply network. The product-side requirements sit elsewhere: in Europe under EN 1074-1/-2 (PFA/PMA/PEA, eleven type tests, the 0.5 bar tier), in the US under AWWA C515 and C509 (2.5× shell, 2× gate, 500 pressurised cycles). Both require higher shell pressure than EN 12266-1 P10, so confirm which standard governs before quoting. The drinking-water hygiene gate is covered in NSF/ANSI 61 and 372.
FAQ
Is EN 12266-1 Rate A stricter than API 598?
For soft-seated valves they are effectively equivalent — both mean no visible leakage. For metal-seated valves it depends on size and rate class. Convert both allowable values to mm³/s and compare numbers, not names.
Is P11 the back-seat test?
No. P10 is shell resistance, P11 shell tightness, P12 seat tightness, P20 closure resistance. The back seat is P21 in EN 12266-2 and only applies when a product standard requires it.
Can a gas medium be used for the shell test?
The standard permits it by agreement, but only after a successful liquid strength test, with safeguards. In practice liquid is preferred unless the valve cannot see water, such as oxygen service.
The customer wrote “EN 12266” with no rate — what now?
Ask for clarification. Without a rate the requirement cannot be judged: A and G differ by more than two orders of magnitude, and the quoting and lapping process differ just as much.
Specific clauses follow the official EN 12266-1:2012 and EN 12266-2:2012 texts; the tables above are for quick orientation in quoting and test discussions. For EU market access classification see PED and CE marking, for US pipeline valve testing see API 6D test clauses, Building gas and potable-water valves follow product standards rather than test-method standards — EN 331 for gas and EN 13828 for drinking water — and for the full standards map see the valve test standards index.
Related test equipment
Each test above maps to a bench. Pick by valve type and by the item you need to prove:
- Valve pressure test machine — shell at 1.5x and seat at 1.1x, with a report per valve
- Valve torque test machine — break-to-open and seating torque curves for actuator sizing
- Valve life-cycle test bench — pressurised cycling with automatic counting and stop-on-failure
- Valve testing equipment range — browse all models by valve type and test item
From test requirements to equipment selection
Provide the valve type, size, medium, pressure, applicable standard edition and acceptance items. Equipment test capability is not equivalent to product certification or regulatory conformity.