Overpressure Protection with a Closed Valve?

[Bambie]

Design rules for valves in ASME B16.34, B31.1, Sections VIII, III etc. require analysis of stem retaining structures to ensure that failure does not result in stem ejection from the valve body.

Analysis is not required to ensure suitable margins exist under pressure and seating preload to prevent the valve failing open.

Without this analysis, am I permitted to rely on a single closed valve as the only overpressure protection provided for a low pressure system connected to a high pressure system in accordance with all requirements in ASME VIII Division 1 Appendix M?

Would I be required to block the stem closed with an external, fully analyzed device if the manufacturer was not motivated to provide the factor of safety on the yoke sleeve or stem nut?

 

[Duwe6]

NOPE. Code overprotection requires Code-recognized overprotection devices. We run into this problem all the time at Pressure Reducing Valves: if the downstream side can see high pressure [it can, as most Reducing Valves fail 'through'], then the low-pressure side requires protection. If the low-pressure side can be calculated to a higher MAWP than normal operating pressure, the added PRV can be fairly small. If the low-side MAWP calcs to equal the PRV setting on the high-pressure side, you do need an additional PRV.

"stem retaining structures" are not Pressure Relieving Devices.

 

[Bambie]

Duwe6,

Most pressurized systems have drains to vented tanks that are isolated by single valves during normal operation. The vents are sized for atmospheric fill rates, not high pressure fill rates. Appendix M permits this arrangement so long as appropriate administrative controls are in effect (locks and blocks).

My question is whether the stems should be blocked if valve failing open is possible.

 

[Duwe6]

Answer is still NOPE. I think you are describing a Spec-Break, transitioning from a high-pressure system to a low-pressure drain tank/knock-out pot. If that is the case, the spec break will be on the downstream flange of the block valve. So that intervening block valve has to meet the spec of the HP system. The attaching LP pipe is installed with a flange that matches the HP valve. Thus you occasionally see Class 900 or 1500# flanges on Sch40 pipe. Just to tie into the spec-break valve.

For safety, think about installing a spectacle-blind on the LP side of that valve.

 

[Bambie]

Duwe6,

Safety is exactly the reason for this post...

I'm not sure which question your NOPE answers.

Perhaps you haven't looked closely at the internals of a manual globe or gate valve.

They all have a yoke sleeve or similar stem retention structure to apply seating load via hand torque.

The yoke sleeve is NOT required by code NOT to fail under hand torque and system pressure (believe it or not).

ASME VIII Appendix M does NOT require this assurance for 'stop valves' used for overpressure protection via isolation (no matter what the valve's pressure rating).

There are small diameter 1500lb valves in the market place that will pop open if an operator applies 25 ftlb of torque to the handle.

I think this is a loop-hole in the code - don't you?

Say anything but NOPE!

 

[Duwe6]

Wasn't aware of this loop-hole. If it was me, I'd use a 1500# [or whatever class is appropriate] Globe valve. Sounds counterintuitive, but globes are the easiest to lap to full contact. Would have that valve lapped in until there was a nice, wide contact band when blue-checked. Now you have a valve that will close tight, and one that is rated for the full design pressure, open, closed, or throttled. I'd still put a spectacle blind beyond it, for a "suspenders + belt" insurance.

 

[Bambie]

Duwe6,

I agree with the blind flange idea but the reason would be compliance with the intent of ASME VIII DIV 1 APPENDIX M for overpressure protection using an engineered pressure retaining device.

I believe an 'engineered' external stem compression device or block installed on the valve would provide the same assurance and code compliance.

 

[Bambie]

Duwe6,

AThe ESCD (External Stem Compression Device) is attached.

 

[Duwe6]

If the valve meets ANSI/ASME B16.34 Class 1500, you do not need the added yoke & stem retainer. The valve is indeed rated for the uplift on the stem. That is why the stem has an Acme thread-pattern, and the yoke is robust. Typically, the handwheels are of the 'hammer' type, to allow extra seating force. Buy a good brand, from a 1st-World country and you can stop worrying.

 

[Bambie]

Is there anyone else out there who actually understands this issue?

 

[BigInch]

It is not uncommon to include a supplemental specification requiring that stems be blow out proof, that shafts withstand the torque needed to close the valve under full rated differential pressure and to provide a shear key in the positioning train that will fail before stem can be damaged.

 

[Bambie]

BigInch (my saviour),

If it is not 'uncommon' to provide this additional assurance in industry, isn't that an indication of a loop hole and should it not be a Code requirement following the normal laws of evolution? I have submitted a written inquiry to the ASME Code Committee (attached), but have not received a reply.

 

[BigInch]

I would agree that it could be a "deficiency".
Shaft failures under torque levels needed to turn those sometimes frozen valves have long been a problem, as have installing motors with larger torque capacities than what the valve shafts can withstand.

 

[Bambie]

BigInch,

We have had several yoke sleeve failures in our plant coincident with overtorquing and pressure transients. The valves were 3/4"nps 900lb bellow sealed plug valves with aluminum bronze yoke sleeves shaped like a top hat with a very small radius at the 90 degree shoulder, which failed in bending. The Manufacturer actually claimed that the yoke sleeve was 'designed' to fail before disc and seat damage could occur from over torquing. I can't imagine that industry requested this feature.

 

[BigInch]

More likely the industry is putting up with the deficiency, or even more likely, oversizing the motor. In any case, the smaller the valve, it seems to increase the likelihood of breaking it.