Use of Control Valve as ESD valve 2

[1234970]

Can anyone guide me if there exists some standard that requires us not to use control valve as Emergency shutdown valve? I know it would be a good engineering practice but not sure if legislation requires it. Thanks for your help.

 

[jsummerfield]

What is your definition of an Emergency shutdown valve?

What are you calling a control valve?

What industry is this for?

Typically control valve seat leakage is designated by the Fluid Controls Institute standard FCI 70-2 or IEC 534-4. A cage style globe valve complies with Class II or perhaps class III. Leakage Class II leaks like a sieve. Classes IV, V and VI are regarded as tight shutoff, because they are tighter than class II and III. Neither Class IV, V nor VI are bubble tight. FCI 70-2 does not address fire safe issues.

I have an automated 16-inch butterfly valve out for inquiry. I do not call this a shutdown valve but it is to shutoff cooling water. I specified bubble-tight shuttof in the normal direction and Class VI in the reverse direction. I also specified fire safe per API 607 as required for a butterfly valve per the line specification valves.

Generally, an emergency shutdown valve would be a separate item than the control valve. It would also be operated by a different control system. I prefer clear definitions of shutdown applications. Typically we have equipment shutdowns, process shutdowns and perhaps a safety shutdown or an emergency shutdown. My definition of an emergency shutdown is a fire on an offshore platform. The control system would typically shut everything down, perhaps with exclusions of things like firewater pumps, emergency generators, etc. A process shutdown may stop the inflow to a vessel with a high level or stop the pumps on a low level.

I prefer not to consider a control valve as any type shutdown valve. Please clarify your details for a more direct response.

John

 

[Elvie]

Any seriuos Engineering applies a clear split between "CONTROL" and "SHUT DOWN" functions.Then the design for the one function excludes a proper work for the other.
In these equations the Actuators design is part of segregation as each function implies different parameters especially the closing time.

 

[reliablesam]

Hi You have to check IEC 61508 which outlines how any safety insttrumented system (SIS) shall be designed. An effective emergency shutdown system shall be completely independent from any control system otherwise it can be affected adversly by any failure in the control system. I agree also with jsummerfield regarding requirement for leakage rate and that ESD valves undergo fire test.

sam

 

[zdas04]

I once did a fire/personal injury investigation where a throttling valve was used as a spec break. On the occasion of the fire, an upset caused the valve to go shut. 36 hours later, the upstream pressure (it was a gas well) had built to 1,200 psig. Because of leakage past the actuated globe valve and a relief device with grossly inadequate capacity, a 40 psig control-gas system saw most of the 1,200 psig and a flange failed. A worker entered the space and did something stupid and was badly burned.

Since that incident, I've been very careful never to allow a throttling device to be counted as a spec break.





David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

 

[TD2K]

I've taken spec breaks across throttling valves many times but have always asked myself the question 'how is the downstream piping and equipment protected' especially for something as straight forward as a control valve going fully open.

 

[zdas04]

TD2k,
An Engineer doing a conscience risk-analysis of a valve decision is always appropriate. If you can satisfy yourself that the answer to the question you always ask yourself is "adequately" then there is no issue.

On the fire I investigated, no one had ever asked that question. The leakage past the throttling ESD on the first spec break was not included in the list of credible scenarios in the sizing of the relief for the piping in the control-gas system because that would be double jeopardy and therefore not credible (in their opinion). In real life, throttling valves always leak in no-flow situations and it should have been considered. They could have considered it by either putting an on-off control valve in front of the control-gas system or by putting a bigger PSV on volume pot. The did neither and a worker was badly burned, the building burned down, and the well was off production for over 6 months while everyone investigated.

David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

 

[TD2K]

I hear you. I looked at a fuel gas system once where there was a spec change across the control valve. Credit was taken for the 'normal' flow across the valve in sizing the PSV.

Bad, bad idea as the 'normal' flow for the valve was with the downstream users consuming fuel gas to meet winter heating requirements. What happens during the summer when they didn't use much fuel gas? Overpressured fuel gas piping if that valve ever failed open. Oh yes, and a later plant revamp later removed much of the duty from these boilers but of course, the PSV audit file still was taking credit for the 'normal' flow on the data sheet.

 

[jsummerfield]

There are many reasons for a specification break besides changing the ASME flange class. Examples include stainless steel downstream from a valve refrigerates below carbon steel limits, metalurgy changes do to the velocity, etc.

BTW, many line class block valves are not bubble tight.

John

 

[zdas04]

jsummerfield,
Probably millions. Just wondering, does that make it right?

My preference is to specify trunion ball valves for spec breaks that are due to changes in pressure ratings. That is the only reason I've ever made a spec break (generally I have the same fluids upstream of a spec break as downstream, but I wouldn't be shocked if others have seen other reasons).

David

 

[sdl]

It is strongly recommended that the ESD and BPCS be separate and independent. I have seen where an ESD valve is backed up by a control valve (the solenoid controlled by the SIS and the throttling done by the BPCS). Depending on the failure data, this design can meet the applicable SIL.
If the control valve duty is used as a spec break, that particular loop is performing a safety control function (continuous mode) and is calculated a little differently from the normal ESD valves (demand mode) and should be incorporated into the SIS.

jsummerfield makes some good points. It is very important that the proper equipment be used in the applications. There are a lot of processes out there that are inherently dangerous and while cutting costs is always a consideration, safety must be first and foremost.

Just my two cents.