API 598:2023 is the eleventh edition of the valve inspection and testing standard. It covers manufacturer inspection and pressure-test requirements for resilient-seated, nonmetallic-seated and metal-to-metal-seated gate, globe, plug, ball, check and butterfly valves.
An API 598 reference in a purchase order does not establish the valve design, material selection, pressure-temperature rating or suitability for the intended service. It defines which shell, backseat and closure tests apply, how test pressure and minimum duration are determined, and what leakage may be accepted for the valve type and seat construction.
Purchase and inspection reviews must connect the valve type, DN or NPS, ASME class, seat design, test medium and witness requirements before the inspection and test plan is approved. This guide applies those inputs to the 2023 test matrix, pressure rules, hold times, leakage criteria and final test report, while the licensed API standard remains the contractual reference.
API 598 latest edition and official PDF access
The current edition is API Standard 598, Eleventh Edition, published in February 2023. The document contains 24 pages and replaces the tenth edition issued in October 2016.
The edition should be stated in the purchase order, inspection and test plan, and final test report. Test requirements, pressure bases, minimum durations and leakage allowances are edition-sensitive, so data taken from a 2016 or earlier PDF should not be applied to a 2023 contract without verification.
Where to obtain the official API 598 PDF
The licensed PDF and printed edition are available through the official API Std 598 listing. Access conditions depend on the selected licence and delivery format.
API 598 is a copyrighted standard, so this page does not provide a free copy or reproduce the complete document. Use the licensed edition for contractual acceptance, then use the technical sections below to check how its test matrix, pressure rules, hold times and leakage criteria apply to the ordered valve.
What API 598 covers
API 598 controls inspection, examination and pressure testing performed by the valve manufacturer. Its seat scope includes resilient seats, nonmetallic seats such as ceramic, and metal-to-metal seats.
The standard specifies and addresses:
- Shell tests.
- Backseat tests.
- Low-pressure closure tests.
- High-pressure closure tests.
- Double block and bleed high-pressure closure testing.
- Visual examination of castings.
- An optional high-pressure pneumatic shell test.
The required test package depends on the valve type, DN or NPS, ASME class and seat construction. Some tests are mandatory, some are optional when specified by the purchaser, and a backseat test applies only to valves with a backseat feature, except bellows-seal valves. Applying API 598 to a valve type outside its stated scope requires agreement between the purchaser and manufacturer.
Supplementary examinations such as magnetic particle, radiographic, liquid penetrant or ultrasonic examination are not automatic API 598 requirements. The purchase order must state which additional examinations, inspection access and witness points are required.
Passing an API 598 test confirms acceptance under the specified test conditions. It does not establish valve design, material compatibility, pressure-temperature rating, fire-safe performance, fugitive-emission performance or suitability for the operating medium. Those requirements must be defined through the applicable design standard, project specification and other relevant valve standards and documents.
Which API 598 tests apply to each valve type
Every valve covered by API 598 requires a shell test, but backseat and closure-test requirements change with valve construction, size and ASME class. A report marked only “API 598 tested” does not confirm that the correct test combination was performed.
Backseat-test applicability
A backseat test is required for gate valves, globe valves and parallel slide gate valves fitted with a backseat feature. Bellows-seal valves are excluded. Plug, ball, butterfly and check valves do not receive this test under the API 598 matrix.
Size and class dependent closure tests
Gate valves, plug valves, API 609 Category B butterfly valves and trunnion-mounted ball valves use two size and class bands. Low-pressure closure testing is required for valves DN 100 (NPS 4) and smaller through Class 1500, and for valves above DN 100 through Class 600. It becomes optional above those class limits.
High-pressure closure testing follows the opposite pattern for these valve groups. It is optional in the lower bands and required above Class 1500 for DN 100 and smaller, or above Class 600 for sizes greater than DN 100.
Valve-specific closure requirements
- Globe valves and parallel slide gate valves: the low-pressure closure test is optional, while the high-pressure closure test is required.
- Floating ball valves: the low-pressure closure test is required, while the high-pressure closure test is optional.
- API 609 Category A butterfly valves and check valves: the low-pressure closure test is optional, while the high-pressure closure test is required.
- Lubricated plug valves: the low-pressure closure test is optional, while the high-pressure closure test is required.
When both high- and low-pressure closure tests are performed, API 598 recommends completing the high-pressure test first. High-pressure testing can affect the subsequent low-pressure sealing performance of a resilient seat.
Shell, backseat and closure tests
Shell, backseat and closure tests examine different pressure boundaries and leakage paths. A valve can pass the shell test but still fail at the backseat or through the closed seats, so each result must be recorded separately.
Shell test
The shell test checks the strength and soundness of the assembled pressure-containing structure. The valve ends are closed, the closure member is partly open, and test pressure is applied inside the body. The body, bonnet, cover, fixed joints and other pressure boundaries are then examined for visible leakage.
The packing gland is tightened enough to maintain the required pressure, which also exposes the stuffing box to the test except on bellows-seal valves. Liquid-filled valves must be essentially free of trapped air before testing because compressed gas can affect pressure stability and increase stored-energy risk.
Backseat test
The backseat test verifies leakage past the stem or shaft-to-bonnet backseat seal. It applies to valves fitted with a backseat feature, except bellows-seal valves. The assembled valve is fully opened, its ends are closed, and pressure is applied with the packing gland loose or the packing not installed.
For valves DN 100 (NPS 4) and smaller, the backseat test may be combined with the shell test when leakage is monitored with an appropriate volumetric and visual detection method. Passing this test does not authorise packing adjustment or replacement while the valve is pressurised.
Low- and high-pressure closure tests
A closure test measures leakage past or through the closed seat interface. Low-pressure testing uses air, nitrogen or inert gas, with leakage detected by bubbles or a calibrated volumetric device. High-pressure testing follows the same basic closure procedure, but liquid leakage is observed as drops when a liquid test medium is used.
Test direction must match the valve design:
- Bidirectional gate valves: pressurise successively from each side, with the opposite side open to atmosphere.
- Globe valves: test with pressure under the disc.
- Check valves: test from the downstream side.
- Unidirectional valves: test from the marked pressure side.
- API 609 Category A resilient-lined butterfly valves: test in one direction.
- API 609 Category B resilient-seated butterfly valves: test in both directions.
Where a preferred flow direction is specified, the nonpreferred-direction test must use the reduced differential-pressure rating applicable in that direction.
The seat interface must be clean and free from oil, grease and sealant before low-pressure testing, except where lubricant forms part of the primary seal, as with a lubricated plug valve. Excessive closing force or an unapproved leverage device must not be used to obtain a passing leakage result.
How API 598 test pressure is determined
API 598 test pressure is not calculated from the valve’s normal operating pressure or from one universal multiplier. The correct value depends on the test type, valve material, pressure class, end connection, valve design and applicable purchase specification.
Shell test pressure
- Steel and nonferrous-alloy valves: the hydrostatic shell test pressure is generally 1.5 times the pressure rating at 38 °C (100 °F), rounded up to the next 25 psig. This basis applies to the class and material rating, not directly to the project operating pressure.
- Class 800 valves: use the same 1.5-times relationship with the API 602 pressure-temperature rating.
- API 609 Category A butterfly valves: use 1.5 times the valve’s maximum cold working pressure.
- Ductile-iron and grey-iron valves: have fixed minimum pressures in API 598 according to class and size, so the steel-valve multiplier should not be substituted.
Hubs, flanges or other end connections with a lower ambient-pressure rating can limit the test pressure below that of the primary valve assembly. The complete pressure boundary must therefore be checked before selecting the shell-test value.
High-pressure closure and backseat pressure
- Valves other than butterfly and check valves: high-pressure closure and backseat testing generally uses 110% of the maximum allowable pressure at 38 °C (100 °F) under the applicable purchase specification.
- Carbon-steel, alloy-steel, stainless-steel and special-alloy check valves: use the same basis for high-pressure closure testing.
- Butterfly valves: high-pressure closure testing uses 110% of the design differential pressure at 38 °C.
- Power-operated and manually operated gear-actuated globe valves: use 110% of the design differential pressure used to size the operator.
These cases cannot be calculated safely from the body pressure class alone.
Low-pressure closure and backseat pressure
The specified low-pressure closure and backseat range is 5.5 ± 1.5 bar gauge, equivalent to 80 ± 20 psig. The tolerance defines both the minimum and maximum pressure, so a test outside 4.0 to 7.0 bar gauge does not match this stated range.
Low-pressure testing uses air, nitrogen or an inert gas. The test equipment and leakage-detection method must be suitable for the selected pressure and capable of resolving the allowable leakage rate.
Optional high-pressure pneumatic shell test
A high-pressure pneumatic shell test is performed only when specified in the purchase order. It follows the hydrostatic shell test and uses 110% of the maximum allowable pressure at 38 °C, unless the purchase order states another value.
Compressed gas stores substantially more energy than a liquid at the same pressure. The test therefore requires an approved procedure, suitable barriers and appropriate safety precautions, with no visible leakage permitted.
Minimum API 598 test duration by valve size
The minimum test duration begins after the valve is fully prepared and the specified pressure has been reached. Time spent filling the valve, removing trapped air, increasing pressure or stabilising the test system does not count towards the inspection period.
| DN | NPS | Shell test | Backseat test | API 594 check-valve closure | Other-valve closure |
|---|---|---|---|---|---|
| ≤ 50 | ≤ 2 | 15 seconds | 15 seconds | 60 seconds | 15 seconds |
| 65 to 150 | 2.5 to 6 | 60 seconds | 60 seconds | 60 seconds | 60 seconds |
| 200 to 300 | 8 to 12 | 120 seconds | 60 seconds | 120 seconds | 120 seconds |
| 350 to 600 | 14 to 24 | 300 seconds | 60 seconds | 120 seconds | 120 seconds |
| > 600 | > 24 | 600 seconds | 120 seconds | 240 seconds | 240 seconds |
The closure-test duration for other valves applies to each seat-sealing direction. A bidirectional valve tested successively from both sides must therefore complete the specified duration for each direction rather than dividing the time between the two tests.
For a successive double block and bleed high-pressure closure test, each seat receives at least the applicable closure duration. When both seats are pressurised simultaneously, the test duration must be at least twice the value listed for other-valve closure testing.
API 598 leakage acceptance for resilient, metal and check-valve seats
API 598 does not apply one leakage limit to every part of a valve. Pressure-boundary leakage, stem or backseat leakage, and leakage through the closed seat interface have separate acceptance rules.
Shell, backseat and stem-seal leakage
Visible leakage through pressure-boundary walls or fixed body joints is not permitted during the shell test. A liquid test must show no drops or wetting on the external surfaces. An air, nitrogen or inert-gas test must show no leakage through the established detection method.
Visible leakage is also prohibited during the backseat test. Leakage from an adjustable stem seal during the shell test does not automatically reject the valve, but the manufacturer must demonstrate that the seal can retain the valve’s rated pressure at 38 °C (100 °F) without visible leakage. Nonadjustable stem seals, including fixed O-rings, must show no visible leakage during the shell test.
Closure-test leakage rates
Resilient-seated valves require no seat leakage during the applicable minimum test duration. Metal-seated valves may have a measurable allowance that increases with size. Metal-seated check valves use a separate volumetric basis and must not be judged by the manual-valve drop or bubble limits.
| Seat category and size | Liquid closure test | Gas closure test |
|---|---|---|
| All resilient-seated valves | 0 drops during minimum duration | Less than 1 bubble during minimum duration |
| Metal-seated, non-check, NPS ≤ 2 | 0 drops | 0 bubbles |
| Metal-seated, non-check, NPS 6 | 12 drops/min or 0.75 mL/min | 24 bubbles/min or 0.24 mL/min |
| Metal-seated, non-check, NPS 12 | 24 drops/min or 1.50 mL/min | 48 bubbles/min or 0.48 mL/min |
| Metal-seated check valve, NPS ≤ 2 | 6 mL/min | 0.08 m³/h or 3 ft³/h |
| Metal-seated check valve, NPS 6 | 18 mL/min | 0.25 m³/h or 9 ft³/h |
| Metal-seated check valve, NPS 12 | 36 mL/min | 0.50 m³/h or 18 ft³/h |
The table shows representative values rather than every API 598 size row. For liquid testing, the standard treats 1 mL as 16 drops. For gas testing, 1 mL is treated as 100 bubbles. Nonmetallic seats, such as ceramic seats, use the allowance specified for a metal-seated valve of equivalent size and type.
For sizes above DN 1200 (NPS 48), the standard uses formulas. Metal-seated valves other than check valves allow 2 times NPS in drops per minute for liquid and 4 times NPS in bubbles per minute for gas. Metal-seated check valves allow 3 times NPS in cubic centimetres per minute for liquid, or 0.042 times NPS in cubic metres per hour for gas.
A bubbler or other volumetric device must be calibrated for the required units. The timed test begins only after flow through the measuring tube has been established and stabilised. Leakage acceptance is based on the nominal DN or NPS even when the valve has a reduced bore.
What an API 598 test report and ITP should show
A test report that states only “tested OK” cannot confirm whether the correct API 598 test, pressure, duration, direction or leakage limit was applied. The inspection and test plan should connect each test result to an identifiable valve and the applicable purchase-order requirement.
Set inspection and witness points before testing
API 598 does not automatically require the purchaser to witness every test. The purchase order must state whether the purchaser or third-party inspector will inspect assembly, witness pressure tests, review records or observe supplementary examinations.
Witnessing may be performed remotely by video when properly arranged. The order should also define the required notice period, inspection location and contact details, including any pressure-containing parts manufactured outside the main valve plant.
Magnetic particle, radiographic, liquid penetrant and ultrasonic examinations are supplementary requirements. Their extent, procedure and acceptance criteria must be specified rather than assumed from a general API 598 reference.
Record enough data to reproduce the acceptance decision
An auditable API 598 test record should identify the valve and document the conditions used to reach the pass or fail decision:
- Valve tag, serial number, model, quantity and manufacturing batch.
- Valve type, DN or NPS, ASME class, end connection and bore configuration.
- Body, bonnet, closure member, trim and seat materials where relevant to acceptance.
- API 598 edition and the applicable valve design or purchase specification.
- Shell, backseat, low-pressure closure, high-pressure closure or DBB test performed.
- Test medium, medium temperature, test pressure and pressure unit.
- Valve position, pressure side, seat-sealing direction and observation point.
- Minimum test duration and the actual timed inspection period.
- Measured leakage, measurement unit, allowable limit and final result.
- Gauge or measuring-device identification, calibration status, test date, operator and witness.
For metal-seated valves, the report should record the measured value and unit rather than replacing it with a general statement such as “minor leakage accepted.” For resilient-seated valves, the result should confirm that no leakage was detected during the applicable minimum duration.
Certificates, MTRs and retesting
A certificate of compliance is supplied when the purchaser specifies it in the purchase order. Material test reports are also purchaser-specified documents and should identify the components for which traceability is required.
When a manufacturer transcribes material data produced by another organisation, the manufacturer remains responsible for its accuracy and authenticity. The MTR should identify the data source, while the originating report remains controlled in the manufacturer’s records.
A completed valve does not require retesting unless purchaser inspection is specified. The purchaser may waive retesting after receiving written manufacturer certification that the valve was inspected, examined and tested in accordance with the stated requirements.
API 598 vs API 6D and ISO 5208
API 598, API 6D and ISO 5208 can all appear in a valve specification, but they do not control the same engineering decisions. Their test pressures, procedures, durations and leakage criteria should not be substituted without checking the contract and applicable product standard.
API 598
API 598 controls inspection, examination and pressure testing for the valve types within its scope. It normally supplements a product standard rather than replacing the design, material and pressure-temperature requirements in that document.
API 6D
API 6D is a pipeline-valve specification covering design, manufacture, testing, marking and documentation for pipeline ball, gate, plug and check valves. Requiring API 6D and API 598 together can create conflicting test details unless the purchase order states which requirement governs.
The dedicated API 6D vs API 598 comparison explains how to separate pipeline-valve requirements from general API 598 acceptance checks.
ISO 5208
ISO 5208 specifies pressure testing for industrial metallic valves and uses its own leakage-rate designations. A project using ISO 5208 should state the required rate and test basis instead of assuming direct equivalence with API 598.
A purchase order may combine API 600 or API 602 for valve design, ASME B16.34 for pressure-temperature ratings, and API 598 for final inspection and pressure testing. The broader API valve standards guide separates these design, test, fire-safe and documentation roles.
Confirm the API 598 test basis before production
A purchase order that states only “test to API 598” leaves the edition, optional tests, test directions, purchaser witness points and document package unresolved. Those items can change the inspection plan and final acceptance record even when the valve size and class stay the same.
Before approving the ITP, confirm the API 598 edition, valve type, DN or NPS, ASME class, body and seat materials, test medium, closure direction, leakage basis, supplementary examinations, witness points, certificate of compliance and MTR requirements. Resolve any deviation in the purchase order before manufacturing and testing begin.
For an order-specific review, send these details when you confirm the API 598 inspection requirements with MacoTango.