商品分类

Pressure Testing Valves: API 598 Shell and Seat Tests Explained

Every valve is pressure-tested before it leaves the factory. Two tests, two different questions: the shell test asks whether the pressure boundary holds, and the seat test asks whether the closure isolates. API 598 is the standard that defines how both are performed for most valves in oil and gas service.

Understanding the difference matters for two practical reasons: it tells you what a test report is actually certifying, and it stops a metal-seated valve being rejected for a "leak" that the standard allows.

The shell test: is the pressure boundary sound?

  • What is tested: the body, bonnet, body joints and the stem seal, with the closure open. The valve is filled and pressurised so that the whole pressure envelope is stressed.
  • Test pressure: normally 1.5 times the valve's rated pressure at 38 °C (the CWP rating), using water or another suitable liquid. Gas testing at this pressure is not a substitute.
  • Duration: set by the standard in a table by valve size - short for small sizes, several minutes for large ones.
  • Allowable leakage: none. No visible leakage through any body joint, wall, or stem seal. Any leakage is a rejection, not a reading to be discussed.
  • Why it exists: a casting or forging defect that leaks under 1.5× rating would fail in service. The shell test is the strongest single non-destructive check a valve gets.

The seat test: does the closure isolate?

  • What is tested: the closure, one side at a time. The valve is closed; pressure is applied to one side and leakage past the seat to the other side is measured.
  • Seat test pressure: normally 1.1 times the rated pressure for the standard closure test. Many specifications also require a low-pressure seat test - typically around 5-7 bar - because a valve that seals at high differential can still weep at low pressure, where the seat load is different.
  • Duration: again set by size in the standard's table.
  • Allowable leakage: this is where the specification matters:
  • Soft-seated valves (PTFE, PEEK, elastomer) must be bubble-tight - zero visible leakage.
  • Metal-seated valves are permitted a defined, small leakage rate, expressed in bubbles or drops per minute, scaled by size. A metal-seated valve that shows the permitted rate is compliant, not defective.
  • Why it exists: the seat test is the acceptance test the plant will live with. It also reveals seat damage from handling and lapping errors.

What a usable test report contains

Field Why it matters
Valve identification - model, size, class, serial Ties the report to the physical valve
Applicable standard, with edition API 598 edition governs durations and limits
Test medium Water, air, nitrogen, kerosene — different media, different meaning
Shell test pressure and duration, result 1.5× CWP, no leakage
Seat test pressure (high and low), duration, measured leakage per seat, per direction The numbers, not "OK"
Inspector and date Who accepted it
Any deviation or concession The part that gets left out and matters most

If a report says "tested per API 598: passed", it is a statement. If it gives pressures, durations and measured leakage in the units the standard uses, it is evidence.

Gas versus liquid testing

Water is the default for the shell test: safe to handle, easy to see leakage, no stored energy. Compressed gas tests are used when the service demands it - oxygen service, some hydrogen duty, or when the specification calls for it - but they carry stored energy and need proper barricading and controls.

Kerosene is used for some seat tests because its low surface tension finds leaks that water misses. If the purchase order says "seat test with kerosene", the report needs to say kerosene.

Where the arguments come from

Claim Reality
"The valve leaks, reject it" If it is metal-seated and within the standard's rate, it is compliant. Reject against the specification, not against intuition
"It passed at the factory so it will seal in the plant" Handling, storage, contamination and seat set all degrade a closure. Fit a strainer upstream, and follow the storage rules
"The test report says passed, so we are covered" Check the pressures and durations match the standard and the order; a report with no numbers proves little
"We do not need a low-pressure test" Many specifications require it precisely because high-pressure sealing hides low-pressure weeping

How this connects to third-party inspection

A production test is performed by the manufacturer and witnessed - or not - depending on the inspection level agreed. Third-party inspection is the mechanism that puts an independent witness at the shell and seat test, which is why the inspection and test plan has to be agreed before manufacture, not after.

Related reading

Need the test requirement written into the order?

Send the valve list and the acceptance criteria the project uses. We will confirm the test sequence, the witness points, and the report format before the order is placed.

Request a quote: contact form or websale@sawye.com

发表回复

登录后才能评论