In 2026 we still receive test reports headed “EN 13828:2003”. That alone tells the story: BS EN 13828:2025 was published on 28 February 2025, and the 2003 edition was withdrawn the same day. CEN gave member bodies until August 2025 to adopt it. A report issued in 2026 citing the 2003 text cites a withdrawn document — and if it also claims “PN16”, it contradicts itself, because the 2003 scope stopped at PN 10.
EN 13828 covers manually operated copper alloy and stainless steel ball valves for potable water supply in buildings. It is a product standard, not a test-method standard. What separates it from industrial ball-valve standards is two things: an acoustic class, and dezincification promoted to a material gate. This article covers: what the 2025 edition changed, the scope boundary, the thirteen checks, dezincification, endurance and the 90 °C test, acoustics, report boundaries, and the Chinese mapping.
1. What the 2025 edition changed
| Item | EN 13828:2003 | EN 13828:2025 |
|---|---|---|
| Size range | DN 8 – DN 100 | DN 6 – DN 100 |
| Pressure class | PN 10 | PN 10 or PN 16 — two declared classes |
| Materials | Title centred on copper alloy | Title names copper alloy and stainless steel |
| Temperature | 65 °C continuous, 90 °C for up to 1 h | Unchanged |
| Technical content | — | Tables 1, 2 and 4 revised; end connections extended; full-bore and reduced-bore sizes adjusted; design rules, test methods and marking updated |
One hard procurement rule follows: a datasheet claiming “PN 25” or “PN 40” while citing EN 13828 has not read the standard. The ceiling is PN 16 and 65 °C. Gas valves belong to EN 331; heating and industrial valves are outside this scope too.
2. Thirteen checks — miss one and the report is partial
The 2025 edition sets a test sequence in Clause 6 and then groups the checks. Regrouped the way a buyer reads them, there are thirteen:
| Group | Clause | Check | Typical failure mode |
|---|---|---|---|
| Mechanical | 6.3 | Operating torque | Fine when new, seized in six months |
| 6.4 | Resistance to torque | Wrench over-torque during installation cracks the body | |
| 6.5 | Resistance to bending | Valve used as pipe support, neck fractures | |
| 6.6 | Torque on captive rotating nuts and tail pieces | Joint weeps; threads strip after one re-make | |
| 6.7 | Stop resistance | Handle forced past the stop, ball travels beyond the seat | |
| Hydraulic | 6.8 | Leak-tightness (seat when closed, stem to atmosphere) | Will not shut off; stem seepage |
| 6.9 | Hydraulic strength | No visible leakage, no permanent deformation | |
| 6.10 | Replenishment of water held between ball and body | Stagnant pocket, water-quality complaints | |
| 6.11 | Resistance to pressure jumps | Seat deformation after water hammer | |
| 6.12 | Angular seal | Angle-pattern valve weeps when installed off-axis | |
| Acoustic | 7 | Noise class by the EN ISO 3822 laboratory method | Night-time flow noise complaints in dwellings |
| Endurance | 8.1 | Resistance to 90 °C water | Seal ageing in hot-water duty |
| 8.2 | Mechanical endurance of the control device | Torque doubles after a few thousand operations |
Plenty of reports carry two lines — “leak-tight: pass / torque: pass”. That is not a summary, it means only two checks were run. The acoustic and endurance groups are the ones most often skipped, because they need a laboratory and a cycling rig, not a pressure pump.
3. Dezincification: the real gate for brass valves
The 2025 edition makes dezincification resistance an explicit requirement, tested to EN ISO 6509-1, with limits taken from the material standards (EN 12164 / EN 12165):
| Alloy | Maximum dezincification depth |
|---|---|
| CW602N (arsenical brass, DR grade) | ≤ 100 μm |
| CW625N / CW626N | ≤ 200 μm |
Why it matters: dezincification is selective corrosion that eats inward. The body looks sound while the wall has turned spongy, and it perforates under pressure. The UK requires DR brass for concealed fittings; Scandinavia and Australia push the same way. Conversely, in soft, low-chloride, ambient-temperature water, ordinary brass serves for decades — so the correct approach is decide by market rules and water conditions, then report the measured depth rather than the word “compliant”. Keep one distinction clear: whether a valve may contact drinking water is decided by hygiene regulation (metal release assessed by long-term rig tests such as EN 15664-1); EN 13828 does not replace that assessment.
4. Endurance and the 90 °C test: the two most frequently omitted
Mechanical endurance is banded by size: 5 000 cycles at DN ≤ 15, stepping down to 500 at DN 80 – 100, and three times those numbers for power-driven valves. Each cycle is open — hold 5 s — close and hold torque 5 s, with half the cycles on cold water and half at 65 °C. After the full run the samples are stored for one week, then checked for tightness and torque. That cold/hot split plus the storage week is designed to expose rebound loss in the seats and scale-induced sticking — exactly what a new-valve test cannot reveal.
The 90 °C water test: flow for 60 min in water at (90 ± 2) °C, then cool and check tightness and operating torque. Flow rates are set by size — DN 6 – 15 covers 0,05 – 0,15 L/s. This corresponds to short-term over-temperature in domestic hot water (a heat exchanger excursion), not continuous duty.
5. Acoustics: the requirement industrial standards do not have
Valves in buildings are heard by occupants. EN 13828 assigns a noise class using the EN ISO 3822 laboratory method, with full-bore ball valves defaulting to group I and flow/class values given in Table 11. This requirement is essentially absent from the Chinese system — GB/T 15185 has no noise classification and GB/T 13927 does not address it. For European residential, hotel and hospital projects the acoustic class is frequently a tender item, and a report without it is technically incomplete.
6. A report covers only the sizes and pressure classes tested
This is the most expensive trap in procurement: a report is valid only for the DN and PN class that were tested. Using a DN 15 PN10 report to cover a DN 50 PN16 order is out of scope. Equally, data generated against the 2023 prEN draft cannot simply be presented as evidence against the published 2025 text — the check names may match, but clause numbers and criteria moved.
7. How Chinese standards map onto it
| Dimension | Chinese counterpart | Gap |
|---|---|---|
| Product | GB/T 15185-2016, flanged iron and copper ball valves, DN 50 – 300, copper alloy to PN 25 and −40 to 180 °C | Flanged range only; small threaded valves not covered |
| Pressure testing | GB/T 13927-2022 | Shell and seat procedures only; no torque, endurance or acoustics |
| Hygiene | GB/T 17219 (safety assessment of equipment and protective materials for drinking water) | Material hygiene, a separate line from mechanical performance |
| Acoustics and dezincification | — | No mandatory noise class; dezincification usually called out by material standard or customer spec |
For export work the practical answer is a two-layer report: mechanical and pressure items recorded internally to GB/T 13927 / GB/T 15185, and conformity evidenced by a third-party report to EN 13828. The data can be reused, but a clause mapping table must be attached — otherwise the reviewer asks “which clause is this?” and the file stops there.
8. Configuring the bench
- Combined torque and endurance station: dual cold/hot water circuits, able to switch between cold and 65 °C, with timing accurate to “hold 5 s, hold torque 5 s”
- 90 °C loop: held at 90 ± 2 °C, flow set by size (0,05 – 0,15 L/s for DN 6 – 15), with a cooling position for the re-check
- Angular seal fixture: angle-pattern valves must be clamped in their off-axis attitude, not tested as straight-through
- Acoustics: needs an EN ISO 3822 laboratory — normally outsourced
- Dezincification and salt spray: ISO 6509-1 and EN ISO 9227, normally sent out
For the basics of pressure and tightness judgement see choosing a tightness test method; for sharing one frame across several test types see the valve comprehensive test bench.
FAQ
Is a 2025-edition report necessarily harder to obtain than a 2003 one?
The item count barely changed. What changed is that PN16 is a new class — a 2003 report cannot support a PN16 claim at all — and the test methods and several tables were revised, so clause numbers moved and old records need remapping. In practice: run new projects to the 2025 text; legacy certificates remain usable while valid, but a PN16 requirement forces a re-test or an extension.
Can an EN 13828 valve be used as an EN 331 gas valve?
No. EN 13828 is potable water, EN 331 is building gas; medium compatibility, tightness criteria, temperature envelope and endurance counts all differ. The only thing they share is roughly the same size range. An EN 13828 certificate on a gas branch is a wrong-standard situation.
Can a PN16 valve ship on a 2003-edition report?
No. The 2003 scope ends at PN 10, so “PN16 to EN 13828:2003” is logically impossible. Either re-test to the 2025 edition, or declare PN10 and obtain written customer agreement.
Does skipping the acoustic group block delivery?
Depends on the project. For European residential, hotel and hospital work the acoustic class is often a tender or acceptance item; industrial and non-residential commercial use rarely mandates it. But in terms of report completeness, missing the acoustic group means one of thirteen checks is absent, and technical review will flag it.
Specific clauses follow the official EN 13828:2025 text; the tables above are for quick orientation in selection, quoting and test discussions. For the gas counterpart see EN 331; for European pressure-test references and leak rates see EN 12266-1; for EU classification and modules see PED and CE marking; and the full 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 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 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