Closing these gaps will require coordinated action rather than reliance on either standard alone. First, purchasing specifications should stop treating“ DVT” as a single, interchangeable requirement. Procurement documents should name the standard, edition, and applicable annex or clause, and state which protocol’ s acceptance criteria and cycle counts govern the qualification. Second, API and ISO should pursue closer alignment on scaling and re-qualification π www. valve-world. net Valve World November 2026 15
format of the equivalent, longer-established Annex F in API 6A, and covers the full range of valve types within the API 6D scope: ball, gate, plug, check, and axial valves. ISO 23632:2021 is a different kind of document. Published in May 2021 by ISO Technical Committee TC 153, it is a full, standalone International Standard rather than an annex to a broader specification. It was developed for use across industrial applications generally, not oil and gas alone, drawing inspiration from Shell’ s longestablished MESC SPE 77 / 300 specification and following a test protocol structured similarly to ISO 15848-1( fugitive emissions testing). Its scope, however, is narrower by valve type: it applies only to metallic ball valves and butterfly valves.
Where the standards diverge
Cycle distribution count and test medium. This is arguably the most consequential technical difference. API 6D Annex F specifies 200 cycles in total: 160 ambient cycles run in liquid( hydrostatic) service under Clause F. 19, followed by 40 gas cycles split between maximum and minimum temperature( Clause F. 20). ISO 23632 specifies roughly 205 cycles, but runs them entirely in gas— about 50 at initial ambient temperature, 50 at maximum temperature, 50 at mid-ambient temperature, 50 at minimum temperature, and 5 at final ambient temperature. Because gas is more demanding for leak detection than liquid, ISO 23632’ s all-gas approach is generally the tighter test for seat integrity. API 6D also includes a body shell pressure-andtemperature cycle( Clause F. 24) with no equivalent in ISO 23632. Test sequence. API 6D Annex F runs an 18-step sequence( Table F. 1) covering hydrostatic body and seat tests, force / torque measurement, ambient and high / low-temperature cycling, dedicated bodyshell cycling, and separate high and low-pressure gas seat tests, concluding with a final torque check and post-test examination. ISO 23632 uses a more compact, integrated sequence that pairs a closure test with a torque measurement at each of cycles( initial ambient, maximum, mid-ambient, minimum and final ambient), followed by a Maximum Allowable Stem Torque( MAST) check and a final post-test examination. Temperature control. API 6D bases stabilisation on the rate of change( less than 0.5 ° C / min) and allows temperature to overshoot the target extreme by up to 11 ° C( 20 ° F). It requires temperature monitoring at a minimum of two locations: the seat area and the bonnet or stem-seal region. ISO 23632 instead requires the temperature to settle within ± 5 % or ± 15 ° C of target( whichever is smaller), to hold for at least 10 minutes before leakage is measured, and prohibits any overshoot beyond the set tolerance. It also mandates three monitoring points: the internal flow path, the body near the seat, and the stuffing box plus an optional fourth, and requires that any temperature-dependent flow measurement be corrected using the ideal gas law. Torque monitoring. API 6D measures breakaway and running torque( F. 18) prior to starting dynamic open / close cycling tests, as well as final torque at the end of testing before post-test examination. In contrast, ISO 23632 measures unseating torque at each of its eight measurement steps, permits a seating-torque variation of up to ± 10 % during cycling, and explicitly requires both Maximum Allowable Stem Torque( MAST) verification and torqueinstrument accuracy within 5 % of the measured maximum— none of which are explicitly specified in API 6D Annex F. Leakage acceptance criteria. API 6D Annex F is prescriptive, assigning specific acceptance criteria for high- and low-pressure gas seats based on ISO 5208 leakage rates: Rate B for soft-seated valves, Rate A for lubricated plug valves, Rate D for metal-seated valves, and Rate F for metal-seated check valves. In contrast, ISO 23632 does not prescribe a fixed rate within the standard itself; instead, it requires the manufacturer to declare the allowable seat leakage rate( also referenced to ISO 5208) prior to testing as part of a formal declaration of conformity. Scaling of qualification. For pressure class, API 6D treats Class 900 and Class 2500 as qualifying only themselves, while other classes extend downward( e. g., a Class 600 test also qualifies Class 300 and 150). ISO 23632 uses a single rule: any class below 900( or PN 160) qualifies the same and lower classes, and the same logic applies above that threshold. For valve size, the two standards are closer in philosophy: both qualify the full small-size range up to a threshold( NPS 2 / DN 50), and both use a“ three sizes smaller, two sizes larger” rule for mid-range sizes, though the exact size bands and the large-size extrapolation rules differ slightly. Documentation and certification. ISO 23632 is markedly more formal: it requires a manufacturer’ s declaration of conformity, a signed qualification certificate that explicitly cites the standard and edition, a standardised test report template( Annex A), calibrated and traceable measuring
END USER ARTICLE
instruments, and documented personnel qualification records. API 6D Annex F leaves certification, personnel qualification, and reporting format largely to the manufacturer’ s internal procedures, with calibration governed only by a general reference to the manufacturer’ s own calibration system.
Identifying the gap
Taken together, the gap between the two standards falls into three practical categories. The first is a scope gap: API 6D Annex F covers a wider set of valve types but is optional and non-binding, while ISO 23632 is mandatory in nature once adopted but is restricted to metallic ball and butterfly valves— leaving gate, plug, check, and axial valves without an equivalent formal international standard. The second is a test stringency gap: ISO 23632’ s all-gas cycling, tighter temperature tolerances, continuous torque monitoring, and mandatory MAST verification make it the technically more demanding protocol, even though API 6D Annex F runs more total test steps and uniquely includes body-shell pressure / temperature cycling. A purchaser specifying“ DVT per API 6D” and one specifying“ DVT per ISO 23632” are not asking for equivalent stringency, yet both are frequently treated as interchangeable in commercial specifications. The third, and the one flagged directly in the source material, is a scaling and re-qualification gap: while both standards apply broadly similar logic for extending qualification across sizes, neither standard is fully explicit— nor mutually consistent— about exactly when a design change, material substitution, or manufacturing process change is significant enough to trigger a new validation test. This ambiguity leaves manufacturers and purchasers to negotiate re-qualification triggers case by case, undermining the very standardisation both documents set out to achieve.
Recommendations for the industry
Closing these gaps will require coordinated action rather than reliance on either standard alone. First, purchasing specifications should stop treating“ DVT” as a single, interchangeable requirement. Procurement documents should name the standard, edition, and applicable annex or clause, and state which protocol’ s acceptance criteria and cycle counts govern the qualification. Second, API and ISO should pursue closer alignment on scaling and re-qualification π www. valve-world. net Valve World November 2026 15