Testing to API 598, API 6D and BS EN 12266 – what the standards actually require

But each standard evolved in different environments, shaped around the different ways a valve can fail. With over 30 years spent working to these standards, here’s what each one actually requires.

API 598 – the industrial workhorse

API 598 is the standard most people imagine when they think of valve testing, it’s the one that turns up most often across refineries, chemical plants and power stations. It covers gate, globe, check, ball, plug and butterfly valves, and sets out a shell test, a seat test and, where relevant, a backseat test for gate and globe valves.

The shell test is unforgiving, no visually detectable leakage through the body, the body liner or the joints. The seat test allows more nuance, because it recognises that not every seat is built the same way. A resilient, soft-seated valve is expected to seal completely, with no leakage. A metal-seated valve is allowed a small, size-scaled leakage rate, because metal-to-metal contact can never be quite as absolute as an elastomer pressed against steel. It’s a pragmatic standard, and that’s exactly why it’s the default across general applications.

API 6D – built for the pipeline

API 6D covers similar valve types – ball, gate, check and plug – but it was written for a different environment such as a valve buried in the ground, or sitting in a remote station, doing its job for decades and sometimes not easily accessible. 

Pipeline valves often need to demonstrate fire-safe performance, because a pipeline running through populated or environmentally sensitive areas can’t afford to leak under fire conditions. Sour service compliance frequently comes into it too, where the product being carried makes material selection as much a part of the test as the seal itself. 

BS EN 12266 – the European route

BS EN 12266 approaches the same problem from a different angle. Rather than one document covering pressure and functional testing together, it’s split in two: 12266-1 handles pressure and leak-tightness testing, while 12266-2 covers functional requirements like fire-safe design and emissions.

Where API tests lean on its own leakage classes, EN 12266 uses a set of leak rate classes running from A through to G, aligned with ISO 5208, giving a specifier a genuinely granular way to say exactly how tight ‘tight’ needs to be. Test durations scale with valve size too, from a short hold on a small-bore valve to several minutes at the large-bore end. For projects moving through Europe, or working to client specifications built around ISO and EN frameworks, BS EN 12266 is often the standard.

Why it actually matters which one you ask for

Most valves would pass more than one of these standards without a problem. But this isn’t the same as having paperwork that says ‘was tested and certified to’. Knowing which standard your specification calls for, and why, means you get a valve tested against the failure modes that are actually relevant to where it’s going.