Cryogenic liquefied natural gas service is among the most demanding assignments for manual and actuated valves: temperatures near −196 °C, a fluid that flashes violently on release, and thermal cycling every start-up. Valves that perform flawlessly on ambient hydrocarbon duty fail within months at LNG peak-shaving plants and export terminals when three design fundamentals are ignored.
The single most distinctive cryogenic feature is the extended bonnet — a lengthened bonnet column that keeps the stem packing and actuator above the cryogenic zone. The gas column inside the bonnet insulates the packing area, keeping elastomer and graphite seals within their rated temperature window and preventing ice formation on the stem that would tear packing during operation. Bonnet extension length is standardized against temperature class in cryogenic valve specifications; shortening it to save face-to-face length is a false economy that fails at first cooldown.
Standard carbon steels lose impact toughness at cryogenic temperature. Valve bodies and bonnets for LNG duty are typically austenitic stainless steels (304/316) or low-temperature nickel alloy steels, with trim materials — often 316 or nickel alloys — selected for galling resistance at temperature. Ball and gate internals must tolerate differential contraction between seat and body: clearances that seal correctly at ambient will bind or leak after cooldown unless designed for the temperature delta.
Room-temperature shell and seat tests do not validate a cryogenic valve. Cryogenic testing per BS 6364 (or the equivalent LNG project specifications) submerges the valve in a liquid nitrogen or liquid methane bath, cycles it at temperature, and measures seat leakage and operability cold. Extended bonnets are tested assembled. Buyers should witness or review cryogenic test reports for every class of valve intended for cold-box or tank service — ambient test certificates prove nothing about cooldown behavior.
LNG terminals that apply these four checks consistently report dramatically fewer valve interventions in the first operating year — the cryogenic learning curve is expensive to pay for twice.