European gas enquiries often carry a line like “ball valve to EN 331, MOP 0,5 bar, DN 25”. Two wrong reactions are common on the supplier side: treating it as an ordinary plumbing ball valve and shipping it after a 1.5× hydrostatic test to GB/T 13927, or treating EN 331 as “a stricter EN 12266” and claiming EN 331 compliance for every ball valve in the catalogue. Neither holds — EN 331 is a product standard for manually operated shut-off valves in building gas installations. Its pressure classes run in five MOP steps, its endurance cycles are banded by size, the tightness medium is air rather than water, and the mechanical loads are applied in the least favourable declared mounting position.
This article runs through: what EN 331 actually covers, the eight performance tests, the endurance table, the torque and bending tables, the test conditions, four widespread misstatements, the boundary against EN 13828, and how Chinese standards map onto it. Figures follow EN 331:2015 and its Tables 8 and 9; the contract and the customer specification always govern.
1. First, confirm what EN 331 covers
| Item | EN 331:2015 |
|---|---|
| Product | Manually operated ball valves and closed bottom taper plug valves (metallic bodies) |
| Nominal sizes | DN 6, 8, 10, 12, 15, 20, 25, 32, 40, 50, 65, 80, 100 |
| Maximum operating pressure | 0,2×10⁵ Pa, 0,5×10⁵ Pa, 1×10⁵ Pa, 5×10⁵ Pa, 20×10⁵ Pa — i.e. 20 mbar, 50 mbar, 100 mbar, 500 mbar, 2 bar |
| Temperature | −5 °C or −20 °C up to +60 °C, as declared |
| Gas families | First, second and third family per EN 437 |
| Installation | Domestic and commercial installations, indoor or outdoor; not directly buried — underground valves need a protected encasement |
| Committee | CEN/TC 236 |
The pressure classes are where most errors start. The top class is 20×10⁵ Pa, which is 2 bar. Marketing sheets that claim “EN 331 approved, rated 5 bar or 16 bar” have imported the ratings of industrial ball valves (EN 12266, ISO 7121 territory). A purchase specification asking for MOP 5 bar while citing EN 331 is itself a contradiction that needs clarifying before quoting.
2. Eight performance tests — miss one and the report is incomplete
| Test | What it measures | Why it matters on site |
|---|---|---|
| Internal tightness (seat) | Gas passing the seat in the closed position | A valve that does not shut is not a valve |
| External tightness (body, stem) | Leakage to atmosphere at body, bonnet and stem seal | The direct risk point indoors |
| Flow coefficient Kv | Through-flow in the open position | Pressure loss on small sizes shows up as low burner pressure |
| Operating torque | Torque over full travel, both directions | Basic hand-feel and the input for actuator sizing |
| Torsion resistance (MT1, MT2) | Wrench over-torque during installation, in-service torsion | Cracked bodies, stripped threads, deformed stems |
| Bending resistance (MF1, MF2) | Pipe misalignment, valve used as a pipe support | Neck fracture — the most common field failure |
| Endurance cycling | Tightness and torque after repeated operation | Separates “leak-free when new” from “leak-free in two years” |
| Temperature and salt spray | Low temperature (−5 or −20 °C), +60 °C, EN ISO 9227 salt spray | Northern winters, proximity to appliances, coastal and outdoor meter boxes |
Note what is not in that list: there is no separate fire-resistance test. Claims that “EN 331 requires a fire test, followed by 1.1× hydrostatic after cooling” borrow requirements from building gas pipework rules (the EN 1775 family, national gas regulations) or from industrial valve standards. If a customer genuinely needs fire qualification, cite ISO 10497 or API 607 separately and report it as a separate item — do not fold it into “EN 331 compliant”.
3. Endurance is banded by size — 5,000 cycles on DN 15 proves nothing about DN 100
| Nominal size DN | Number of operations |
|---|---|
| ≤ 15 | 5 000 |
| 20 – 25 | 2 500 |
| 32 – 50 | 1 000 |
| 65 – 100 | 500 |
Passing endurance is not “it still turns”. Clause 4.10.2.1 requires three checks afterwards: external and internal tightness at ambient, external and internal tightness at (60 ± 5) °C, and operating torque still within limits. The elevated-temperature pass is the one that catches problems — plenty of PTFE seats still seal at ambient after cycling and only start weeping once the seat has been held at 60 °C and lost some of its rebound. A report with ambient-only endurance has a missing item.
4. Torque and bending: two tables, split by MOP band
Torsion (MT1, MT2) and bending (MF1, MF2) are tabulated against MOP. Excerpt below, in N·m, enough to size fixtures and sanity-check quotations:
| DN | MF1 | MF2 | MT1 | MT2 |
|---|---|---|---|---|
| Higher pressure band (MOP 1×10⁵ Pa and above) | ||||
| 15 | 105 | 53 | 75 | 40 |
| 25 | 340 | 170 | 125 | 100 |
| 50 | 1 100 | 550 | 250 | 200 |
| 100 | 2 500 | 1 250 | 465 | 370 |
| MOP 0,2 – 0,5×10⁵ Pa band | ||||
| 15 | 85 | 30 | 75 | 30 |
| 25 | 200 | 50 | 125 | 50 |
| 50 | 450 | 100 | 250 | 100 |
Two details get missed. First, how the loads are applied: valves with threaded or flanged ends take both MT and MF; valves intended only for flexible appliance connectors take MT1 and MF1 alone. Second, stop resistance: clause 4.8 requires the open and closed stops to withstand at least 1,5 times the torque of Table 7. Without that margin, a firm twist collapses the stop and the user believes the valve is shut when the ball has already travelled past the seat.
5. Test conditions and sample count
Clause 5.1.1 sets testing at (20 ± 5) °C with air as the medium, values corrected to 15 °C and 1013,25×10² Pa absolute, in the mounting position declared by the manufacturer — and when several positions are declared, the least favourable one governs. Clause 5.1.2 requires three samples per size, run through the sequence of Clause 6 (AVCP).
Those three clauses dictate the bench design: air as the medium rules out a water pressure rig; the 60 °C re-check demands an environmental chamber on the cycling station; and the least-favourable-position rule means fixtures must be re-orientable, not welded into one horizontal attitude.
6. EN 331 and EN 13828 are not interchangeable
| EN 331:2015 | EN 13828:2025 | |
|---|---|---|
| Medium | Fuel gas (families 1, 2, 3) | Potable water in buildings |
| Size | DN 6 – 100 | DN 6 – 100 |
| Pressure | MOP 20 mbar – 2 bar, five classes | PN 10 or PN 16, two classes |
| Temperature | −5 or −20 °C to +60 °C | 65 °C continuous, 90 °C for up to 1 h |
| Distinctive | Tightness, torque, torsion/bending, endurance, salt spray | Adds acoustic class and dezincification |
The certificates do not swap. An EN 13828 potable-water valve on a gas branch may have the pressure rating, but material compatibility, tightness criteria and the temperature envelope all differ; an EN 331 gas valve in a drinking-water line runs into a separate hygiene assessment (GB/T 17219 in China).
7. How Chinese standards map onto it
There is no single Chinese national standard equivalent to EN 331; coverage is split:
- GB/T 15185-2016, Flanged iron and copper ball valves: DN 50 – 300, copper alloy up to PN 25 and −40 to 180 °C. It covers the flanged size range; small threaded gas valves are outside it
- GB/T 13927-2022: pressure test procedures and acceptance — useful for shell and seat, silent on gas-specific endurance and torque
- Installation and acceptance: indoor gas installations follow CJJ 94-type codes for piping, joints, pressure testing and purging
Practical conclusion: for European projects, cite EN 331 directly. For domestic projects, combine GB/T 15185 and GB/T 13927 and layer the gas industry codes on top. Do not write “equivalent to EN 331” — the auditor will ask for the standard number, and there isn’t one.
8. Configuring the bench
Working backwards from the eight tests, a complete EN 331 verification needs four capability blocks:
- Tightness: pressure-decay or flow method with air, covering both the internal (seat) and external (body, stem) circuits, repeated at ambient and at 60 °C
- Torque: full-travel torque curve with peak and average, feeding the 1,5× stop criterion
- Endurance: 500 to 5 000 cycles by size, with a chamber able to switch samples between ambient and 60 °C
- Torsion and bending: MT1/MT2 torque plus bending on the corresponding lever arm; fixtures rated for 2 500 N·m at DN 100
For choosing between pressure-decay, flow, tracer-gas and vacuum methods, see how to choose a tightness test method. For running pressure, tightness, torque and life cycling on one frame, see the valve comprehensive test bench.
FAQ
How do EN 331 and EN 12266 relate?
EN 12266 is a pressure test method standard — shell, seat and back seat procedures and leak criteria. EN 331 is a product standard that adds torque, torsion and bending, endurance, temperature and salt spray on top of pressure testing. During EN 331 certification the pressure tests may follow EN 12266 methods, but the report must add the EN 331-specific items; a set of EN 12266 pressure records is not a substitute.
Can an MOP 0,5 bar valve go on a 2 bar line?
No. MOP is the valve’s own maximum, not the line’s. EN 331 bands at 20 mbar, 50 mbar, 100 mbar, 500 mbar and 2 bar, and the torque and bending limits differ between bands. A low-band valve on a high-band line is outside its pressure envelope and may also fail the bending requirement. Select by the line’s maximum operating pressure, not its normal running pressure.
Are the endurance cycles run under pressure?
EN 331 endurance is a cycling test of the operating mechanism and sealing system; what counts is the tightness and torque re-check afterwards, not pressurisation during cycling. In practice, follow the standard and the notified body’s protocol — but do not skip the 60 °C re-check, which is where technical reviews send reports back.
Are there limits on copper alloy bodies for gas?
They are permitted, with conditions: size and pressure band are restricted, and dezincification sensitivity must be assessed to EN ISO 6509-1. High-zinc brass in condensate-bearing or sulphur-bearing gas dezincifies from the inside; the body looks intact until it perforates. Where dezincification resistance is required, use a CW602N-type alloy and report the measured depth.
Specific clauses follow the official EN 331:2015 text; the tables above are for quick orientation in quoting and test discussions. For the European pressure-test reference numbers and leak rates see EN 12266-1; for EU market classification and modules see PED and CE marking; for the potable-water counterpart see EN 13828; and the full standards map is in 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 life-cycle test bench — pressurised cycling with automatic counting and stop-on-failure
- Valve torque test machine — break-to-open and seating torque curves for actuator sizing
- Valve pressure test machine — shell at 1.5x and seat at 1.1x, with a report per valve
- Valve testing equipment range — browse all models by valve type and test item