I found the following definition from the semiconductor industry:

Definition of shutoff valve:  A valve designed for and capable of positive closure to prevent flow within a piping system. Typical shutoff valves include, but are not limited to, manually-actuated, power-actuated, or spring-actuated fail-safe shutoff valves. Usually excluded are self-actuated valves, such as check valves, pressure regulators, flow controllers, and other devices that are not intended to provide positive shutoff for safety isolation. [SEMI F1-90]

Is there a similar definition in a ANSI, ASME, or BPV code?

I am working with our plant operators to ensure our LOTO program meets OSHA standards, and I keep getting asked the question why certain types of valves can't be used, specifically check valves.  I have been able to find plenty of statements to not use them, but no engineering reference is provided.  The closest document I found is "The Safe Isolation of Plant and Equipment" which I learned about in a post on this website.  However, it doesn't reference an engineering spec either.

I am afraid that my answer is only reinforcing what you have already stated, but do you actually need any more?

From common practice all over the world a 'positive shutoff' is a shutoff where the closing element is mecanically secured against accidental opening, eg. valves closing directly with mechanical devices and mechanical force and securly locked in the closed position, before mechanically operated to open position again.

And again by one sided closing pressure from springs or helping one-sided closing pressure from fluid, a positive closure can not be said to be given.

Exception on above list of constructions is globe-check valve with spindle (and handwheel) for shut off, and regulating valves constructed with (often protected or retarded, away from cavitation exposed areas) positive sealings with spindle and construction for drop-thigt closure.

Even those, especially the regulating valves, will in addition normally have mounted revision valves (positive closure valves) upstream.

If you work in a nuclear plant in the US, you can look at the Nuclear Regulatory Commission's Information Notices.  There are more than a few that discuss check valves leaking.  I think the worst case detailed was a plant where they had something like 9 different check valves in series leaked by (and caused problems).  Look under the subject of "air entrainment in emergency core cooling piping."

MORE NEWS