Valve testing is the final line of defence between a manufacturing process that meets the specification and a product that will perform reliably in service. Yet valve testing — hydrostatic shell testing, seat leakage testing, functional testing of actuators and trim — is frequently treated as a box-ticking exercise rather than the meaningful quality verification it should be. This guide covers the major valve testing standards in depth: what each test actually verifies, where the standards differ in important ways, and how to write a test specification that catches the problems that standard testing alone would miss.
- API 598, EN 12266-1, and BS 6755 are the primary valve testing standards for industrial applications, with significantly different acceptance criteria for seat leakage — a valve that passes one standard may fail another for the same leakage rate
- Shell hydrostatic testing at 1.5× rated pressure verifies body and bonnet integrity, but does not test the seats or the valve’s ability to perform under actual operating conditions including thermal cycling and pressure cycling
- Factory Acceptance Testing (FAT) for critical valves should extend well beyond the mandatory minimum tests required by the purchase standard to include functional cycling, torque/thrust verification, and any service-specific tests relevant to the application
The Purpose of Valve Testing
Valve testing serves multiple purposes that are sometimes confused. Manufacturing tests (hydrostatic shell test, seat leakage test, functional test) verify that the individual valve has been manufactured correctly to its design specification. Type approval tests verify that a valve design meets the requirements of a standard, regulation, or project specification — these are usually performed once on representative samples and the results applied to all valves of that design. Performance qualification tests (fire testing, fugitive emissions testing, low-temperature testing) qualify the valve design for specific service conditions and are also typically performed once per design qualification.
Understanding which category a test falls into is important: when a project specification requires “fire testing to API 607” on every valve, the intent is almost certainly that valves of a type-approved fire-safe design should be supplied, not that every individual valve should be fire tested. Clarifying this distinction prevents unnecessary costs and delivery delays while ensuring the actual intent of the requirement is met.
API 598: Valve Inspection and Testing
API 598 is the most widely referenced valve testing standard in oil and gas and petrochemical applications worldwide. It specifies the minimum testing requirements and acceptance criteria for production testing of all types of industrial valves. Understanding its structure is fundamental.
API 598 requires three standard production tests for every valve: the shell (hydrostatic) test, the high-pressure closure (seat leakage) test, and the low-pressure closure (seat leakage) test. A fourth test — the backseat test — is required for valves with backseats (typically gate and globe valves) to verify the integrity of the stuffing box isolation feature.
Shell test (hydrostatic): The valve body is pressurised to 1.5× the rated working pressure (at the temperature rating for the applicable pressure class) with water. Test pressure is held for a minimum duration that varies with valve size — from 15 seconds for DN25 to 120 seconds for DN450 and above. Acceptance criterion: no visible leakage through the body wall or bonnet. Note that leakage at stem seals and seats is not being tested here — the closure element is typically in the open position during the shell test to eliminate the seats as a leak path.
High-pressure closure test (seat leakage): Test pressure is the same as the shell test (1.5× rated). The valve is closed and seat leakage is measured. Acceptance criteria are defined by valve type and seat material. For metal-seated valves, the permissible leakage is expressed in millilitres per minute per inch of nominal pipe size (a larger valve is allowed more leakage in absolute terms). For soft-seated valves, the acceptance criterion is zero visible leakage (zero bubbles in a submerged test, zero drops in a measuring cylinder).
Low-pressure closure test: Conducted at 5-7 bar (80-100 psi) air or nitrogen. This lower-pressure test catches seat leakage that might be masked by the deformation of soft seats under high hydrostatic test pressure — a soft seat can appear to pass the hydrostatic test because the high pressure deforms the soft material into minor surface imperfections, then opens those leakage paths when pressure is reduced. The low-pressure test uses air or nitrogen (not water) and leakage is measured by counting bubbles from the outlet immersed in water, or by measuring downstream pressure rise in a closed system.
API 598 Leakage Rates: Metal Seats vs Soft Seats
The leakage criteria in API 598 differ significantly for metal-seated and soft-seated valves, and the criteria are size-dependent. For metal-seated gate, globe, plug, and check valves, the allowed leakage at high-pressure closure test is given in Table 1 of API 598 in cc/min per inch of nominal pipe diameter. For a DN150 (6″) metal-seated gate valve, the permitted leakage rate is 9 cc/min — a number that surprises many engineers who assume “tested to API 598” means “tight shutoff.”
Metal-seated ball and butterfly valves have different (generally tighter) criteria. For metal-seated ball valves per API 598, zero leakage is required at the high-pressure test. This is a more demanding requirement than for gate or globe valves because ball valve seats are designed for tight shutoff by their geometry, unlike gate valve seats which are engineered for high-cycle service with accepted nominal leakage.
Soft-seated valves (ball, butterfly, plug with elastomeric or PTFE seats) must achieve zero leakage at both the high-pressure and low-pressure closure tests. “Zero leakage” means zero visible leakage, not zero absolute leakage — practically, this means no measurable flow or bubble formation during the test duration. For Class VI shutoff per ANSI/FCI 70-2, the test methodology is more rigorous and involves specific measurement of bubble count over a defined time period.
EN 12266-1: The European Counterpart
EN 12266-1 is the European standard for industrial valve testing, replacing the earlier BS 6755, DIN 3230, and other national standards it harmonises. Its structure parallels API 598 but with different test pressures and acceptance criteria that are not directly equivalent.
The key differences from API 598:
Shell test pressure: EN 12266-1 specifies the shell test at 1.5× the rated pressure at 20°C, which is the same multiplier as API 598 but the reference temperature differs — API 598 uses the rated pressure at the rated temperature, while EN 12266-1 uses the rated pressure at ambient. For valves rated above ambient temperature, this can result in different absolute test pressures.
Seat leakage rates: EN 12266-1 defines leakage in rates A through G (from rate A meaning no visible leakage to higher rates with defined numerical leakage allowances). The default seat test rate for most valve types in EN 12266-1 is Rate D, which allows a specified volumetric leakage rate that is different from the API 598 allowances. Rate A (zero leakage) is required for soft-seated valves, equivalent to API 598.
The practical implication: a valve specified only as “tested to API 598” may have a different actual leakage performance than a valve tested to EN 12266-1 Rate D, even if both “passed” their respective standards. When specifying valves for tight-shutoff service, specify the maximum permitted leakage rate numerically (e.g., “maximum 1 cc/min” or “zero visible leakage”) rather than relying on the default acceptance criteria of a test standard.
API 607: Fire Testing for Soft-Seated Valves
API 607 defines the fire test procedure for quarter-turn and other valves with non-metallic body or seat seals. The test involves exposing the valve to a hydrocarbon fire at defined temperatures for a defined duration (typically 30 minutes exposure), then testing for both leakage to atmosphere (fugitive emissions during fire) and seat leakage after the fire has been extinguished and the valve cooled.
The rationale for fire testing is straightforward: soft-seated valves (ball, butterfly, plug with PTFE or elastomeric seats) rely on the soft seat material for sealing. If those materials are destroyed by fire — which they will be, since PTFE and elastomers combust or melt — the valve must have a backup sealing mechanism (metal-to-metal seats) that can maintain reasonable sealing even after the non-metallic components have been destroyed.
API 607 acceptance criteria specify maximum leakage to atmosphere during the fire (typically Class IV to Class VI equivalent) and maximum seat leakage after fire. Valves that pass the full API 607 test are described as “fire safe” or “fire tested to API 607.” This designation means the specific valve design, in the tested configuration with the specific materials used in the test, has demonstrated acceptable fire performance — it does not mean any ball valve is automatically fire safe.
Related standards include ISO 10497 (equivalent to API 607 but more widely accepted internationally) and BS 6755 Part 2 (the older UK standard still referenced in some specifications). API 608 covers metal-seated ball valves and includes different acceptance criteria reflecting that metal seats do not burn off in a fire.
Factory Acceptance Testing (FAT): Beyond the Minimum
The mandatory tests in API 598 or EN 12266-1 represent the minimum acceptable quality verification. For critical valves — emergency shutdown valves, safety-critical isolation, high-consequence service — a comprehensive FAT program should extend well beyond these minimums.
Dimensional inspection: Verify face-to-face/end-to-end dimensions, flange bolt hole patterns, bore dimensions (full bore requirement verification), and actuator mounting interface dimensions. For replacement valves, critical dimensions must match the existing spool exactly. Dimensional non-conformances discovered at site are very expensive — discovering them at FAT is far better.
Torque/thrust verification: For actuated valves, measure the actual breakaway torque (for quarter-turn) or thrust (for linear) at the design differential pressure. Compare against the actuator output force with the specified safety factor applied. The actuator must provide adequate margin at minimum supply pressure — verify this with the actual supply pressure minimum, not nominal.
Functional cycling: Cycle the valve a defined number of times (typically 5-10 full open-close cycles at design pressure) and verify: consistent operating torque/thrust, correct stop positions, absence of abnormal noise or vibration, and no leakage development during cycling. For ESD valves, close time verification at minimum supply pressure is critical.
Positioner calibration and stroke check: For control valves with digital positioners, verify: zero and span calibration, response to 4mA and 20mA signals, fail-safe position on signal loss, and any specific features required (partial stroke testing configuration, HART diagnostics, action on instrument air failure).
Fugitive emissions test: For valves in emissions-critical service (ISO 15848 or API 624 compliant valves), verify that the specific valve meets the specified leakage class by testing with helium at the required temperature and after the required number of thermal and mechanical cycles.
Positive Material Identification (PMI) During FAT
PMI — typically performed using X-ray fluorescence (XRF) or optical emission spectroscopy (OES) — verifies that the actual chemical composition of installed materials matches the specified composition. For alloy valves (stainless steel, duplex, Inconel), PMI during FAT catches material mix-ups that could have catastrophic consequences in service.
A common and dangerous error is the installation of carbon steel components (bolts, inserts, body plugs) in otherwise stainless steel valve assemblies due to warehouse or workshop errors. Carbon steel bolts in a 316 stainless valve body will corrode rapidly in seawater or chloride service, potentially releasing the bonnet under pressure. XRF PMI on all pressure-containing and pressure-retaining components catches these mix-ups before the valve leaves the factory.
PMI should test: body, bonnet, stem, seat rings (where accessible), pressure-containing bolting, and welded attachments. For duplex and super duplex valves, PMI should also verify the approximate PREN from the XRF results — while XRF is not as precise as full wet chemistry, it is sufficient to confirm the material is within the correct alloy family and flag obvious composition problems.
Partial Stroke Testing for ESD Valves
Emergency shutdown valves that remain static for extended periods between demands are at risk of becoming stuck due to corrosion, packing creep, or process deposits on the closure element. Partial stroke testing (PST) — moving the valve a defined fraction of its total travel (typically 10-15%) without fully closing it — verifies that the actuator and valve are operable without disrupting the process.
PST is a key element of Safety Instrumented System (SIS) maintenance strategies and is mandated by IEC 61511 for SIL-rated ESD valves in many operating companies’ procedures. The testing can be performed manually, automatically on a schedule (using the smart positioner’s built-in PST capability), or triggered remotely from the DCS/SIS.
PST does not prove that the valve will fully close or fully open — it only proves that the valve can move. A successful PST does not substitute for full stroke testing. The diagnostic coverage credit that can be claimed for PST in SIL calculations is defined in IEC 61511 and is typically 60-90% depending on the PST methodology and the valve type. Verify this with your SIS safety requirement specification and risk assessment.
Documentation and Traceability Requirements
Test documentation is as important as the tests themselves — an undocumented test might as well not have happened in the context of regulatory compliance, insurance requirements, and plant maintenance records. The minimum documentation package for each valve should include the valve data sheet (agreed design parameters), the test report signed by the manufacturer and witnessed by client (or third-party inspector) where required, CMTRs for all pressure-containing materials, and NDE (non-destructive examination) reports where applicable.
For SIL-rated ESD valves, additional documentation is required: the valve’s contribution to the SIL calculation (failure rate data, diagnostic coverage), the proof test procedure, and records of all proof tests performed throughout the valve’s service life. This documentation must be retained for the operational life of the plant and must be accessible for regulatory audit.
Valve testing done properly is an investment in plant reliability, safety, and regulatory compliance. Done poorly — or treated as a bureaucratic formality — it gives false confidence that valves are ready for service when they may not be. Specify your test requirements clearly, witness critical tests personally or through qualified third-party inspection, and keep the documentation. When the valve eventually needs replacing or the plant undergoes a safety audit, you’ll be glad you did.