Gate Valve throttled as a root valve for instrumentation 2

[boofi]

Hi y'all,

Yes, another question about throttling a gate valve. I know for typical applications throttling a gate valve is inadvisable. However, one of the managers at my site constantly requests that gate valves that are root valves to instrumentation (e.g. manual pressure gauges) are maintained at 90% closed. Cracked open, essentially. The reasoning is that if there is a leak at the instrument the valve can be closed quickly, minimizing exposure.

Would gate and seat erosion be a factor here? Is this some kind of industry accepted practice? I imagine the flow at the gate and under the gauge would be mostly stagnant, reducing risk of damage to the valve.

I mean I guess it's risky either way. If the gate is wide open and the threads leak, the operator would have to sit there and spin the handle while being exposed. OR if the valve was cracked open and the gate was damaged, the operator would be there messing with the valve trying to close it to no avail, and still being exposed.

Thanks in advance for the clarification.

 

[TugboatEng]

You should still be using needle valves for your instrumentation. They're even less expensive. Yes it is standard practice to run instrumentation valves only cracked. For us in the Marine industry it prevents massive loss of fluid in the event of a gauge failure in an unobserved space. It also extends the life of the gauge as it allows the operator to throttle until the gauge stops fluttering.

 

[LittleInch]

Tugboat - If you've got "fluttering" then you really should be using snubbers. e.g.

https://www.wika.co.uk/910_12_en_co.WIKA

they would also do the same job in the event of a breakage of the guage

boofi - If you're getting that many leaks from instruments then you're doing something wrong.

The problem with "cracking" valves is the inherent risk that the gauge gets shut off by accident and hence doesn't register or do its job. If any of these items are part of the safety system then they should be fully open IMHO and locked open. Far too many disasters have come from instrumentation designed to save your life being inactive as someone has isolated the guage / switch / transmitter. It's very hard to get something like a gate valve just right and then locked in that position. Use the correct equipment and you don't have to.

Ball valves give you a much faster action to close it off and usually instuments are mounted on DBB blocks with needle valves and small ports. How are yours connected? Any photos?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[gerhardl]

Agree fully with LittleInch, but would have liked more details and sketch of the situation to comment further! Why select gate valves at all for this application?

 

[TugboatEng]

LI, yours is the best answer.

 

[moltenmetal]

Root isolation valves for installation on piping need to be mechanically robust enough to not be damaged in service. They need to be capable of tight shutoff to permit the instrument to be removed/replaced, or calibrated in place, without becoming a hazard to the facility or maintenance workers.

Generally, on large welded pipe, the root valve is a welded block valve from the PIPING spec- generally 1/2", though some insist on it being larger still (1"). What type of block valve- gate, ball, plug etc.- as long as it's in the spec, doesn't matter much, because an instrument like a gauge isn't a flowing service. There is no worry about a partially open gate valve "chattering" due to fluid flow when it's used as an instrument block, because there's no fluid flow THROUGH that valve in service. That said, the proper position for that root block valve in service is open, rather than partially open, as noted above.

Throttling using an almost closed gate valve is a really terrible way to damp pulsations from a pressure gauge. It's ineffective because the valve Cv is too large, and it risks the valve being fully closed and the instrument being "muted" as noted by a previous commentor. The right solutions are either a snubber, liquid gauge filling or preferrably both.

An instrument block and bleed assembly (i.e. an "Anderson Greenwood valve") is sometimes added TO that welded root block valve, but most piping designers are reticent to weld the block/bleed valve assembly directly TO the piping, though that is sometimes done.

If you can thread the instrument block/bleed valve into an o-let or half coupling and skip the piping spec root block valve, that's an easier sell of course- but only for services where a leaking root thread won't cause a serious problem.