Double block and bleed and bypass

[engr2GW]

Hello all,
In the attached drawing, there is a control valve with isolation valves on either side of the control valve. A bybass is used to continue to flow gas if the control valve is to be maintained or worked on. each of the isolation valves have bleeds to bleed off pressure before working on the valve while flowing through the bypass. I have not seen any industry code and standard that require the isolation valves on either side to be two each (as opposed to one each as shown).
Is it required to have two valves to isolate?
experience or reference document will be appreciated if available.
Thank you in advance

As much as possible, do it right the first time...

 

[engr2GW]

Additional information: Operating pressure is 1100 psig

As much as possible, do it right the first time...

 

[georgeverghese]

In modern practice, integral body bleeds on isolation valves are not used. Instead, install a separate line bleed.
Single block and bleed is used in ANSI 600lb service in many OpCos in flammable fluid, non toxic service.
What is the service fluid here?

 

[engr2GW]

Thanks, @georgeverghese. This is natural gas, could be sweet or have some H2S.

As much as possible, do it right the first time...

 

[engr2GW]

Okay, and this is CL 600 service, the bleed is on the process piping, separate from the valve

As much as possible, do it right the first time...

 

[georgeverghese]

If H2S content is sufficient to make it toxic, then a DBB is preferred for all piping classes, with bleed routed to LP or LLP flare or vent header. Some LP flare headers may at times have considerable backpressure.
For non toxic service, DBB is typically seen only for ANSI 900lb and higher piping classes.
For class 300lb and 600lb, some operators insist on DBB when the fluid is dirt laden, and single block and bleed still leaks badly.

 

[LittleInch]

Nothing attached.

You need to use the attachment box at the bottom of the text box and follow the instructions.

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

 

[engr2GW]

Tried attachement again, I believe it should work this time. Sorry about that.

As much as possible, do it right the first time...

 

[pierreick]

Hi,
Sounds good to me.
Release to safe location.
You may find pointers in the document attached.
Pierre

 

[pierreick]

Hi,
Norsok standard (end of the document)
Pierre

 

[LittleInch]

A couple of things here:

1) I loath these unregulated bypass valves and when I get an opportunity, either delete them or argue against them in a HAZOP. If you can substitute a manual valve, especially if its a simple isolation valve then you don't need the control valve or yuo are eliminating your primary control mechanism and then relying on alarms and shutdowns to control your plant.

As a designer, you never know how someone will operate the plant in the future. LC, CS etc means nothing. If the valve isn't there, then no one can open it... Control valves don't often fail and if they do, replacing them with a manual valve which in all likelihood isn't being monitored 24 hrs/day, is a backwards step in terms of process safety. IMHO.

2) Anyway your main point about isolation.

"I have not seen any industry code and standard that require the isolation valves on either side to be two each (as opposed to one each as shown)."

That's because there isn't one. Many companies and projects create an isolation philosophy document to guide designers and operators on how to isolate bits of plant. Organisations like the HSE in the UK produce guidance documents like the one attached. there has been a gradual creeping over the last few decades of more valves, i.e. double isolation and bleed and it's quite hard to try and stem the tide. No on has yet told me what happens if the bleed between two valves shows a leak ( of varying amounts). Does the isolation job then stop until the first valve van be repaired or replaced or do you then rely on the second valve to seal? If so why have the expense of the second valve?

"Is it required to have two valves to isolate?" - only if your internal procedures require it. Each situation should be assessed. If you bleed off the pressure and then there is no gas coming out of the vent then go ahead, but if the valve isn't replaced within one shift, then blank it off. If there is gas coming out then it starts to get difficult to show that the risk is low enough to proceed further and you need to actually shut down that line further back.

In the meantime your plant is operating on some manual valve....

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

 

[1503-44]

I'm going to disagree with you about 25% on #1. Control valve maintenance frequency and failure rates depend on service conditions. I've seen some in high pressure drop service fail every six months. In those cases a bypass with throttling capability becomes quite handy. But they should only be used during maintenance activities, not as a replacement for the automatic control valve.

#2 I don't generally believe in the double valve isolation theory. Some companies require them under certain conditions, but when you look across those companies, their requirements are so diverse, you can't even find a common theme that would make sense to implement as an industry standard practice. Some companies require them in similar service conditions for pressure over 20Barg, others over 100 Barg. The different and widely varied implementation criteria isn't close enough to even try to derive a standard practice from. When I can't see any uniformity in application requirements, I doubt the reasoning behind the requirements to use them at all, If a company specifically requires them, I use them. Otherwise, I do not. If it was a toxic fluid in a confined space, that gets my attention.

A double block and bleed is common for metering system and change of ownership of a pipeline amongst different companies applications. When you stop flow, or metering, in those situations you want to ensure nothing is going through.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[LittleInch]

MR 44,

If it's failing that often then you need a true 2 x 100% capability and install a second control valve in parallel. I've often "won" this argument in HAZOPs and whilst I accept it is often seen, I really don't know how much of that is "we did this last time" engineering as opposed to properly considering what is the impact if some operator just goes and opens the valve fully open. I hate them about as much as slopes on pig traps and not including vent lines in the minor barrel.

The DIB philosophy started in the N Sea platforms post Piper Alpha when everyone, quite rightly, became worried about gas explosions on board oil rigs. Nowadays if you try and say it's OTT, you get accused of sending people to their death (that's what someone said to me in a meeting) instead of properly assessing the risks and reality of operating plants. The issue often is that the people writing the procedures and policies have power without responsibility for budgets or operations.

If the isolation valves are only used on clean service and only to isolate, not to stop or start flow, then they should be perfectly capable of isolating the fluid for maintenance. The risk increases where a valve is used in process conditions in unclean service to start and stop flow. Then these are much more likely to pass and need a second valve.

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

 

[1503-44]

Yes. All good points. You notice only disagreed by 25%.
I think where I was going was, it's good to consider on a case by case basis, rather than to try to make a hard fast policy about it. Even considering PiperA, some companies do and others don't. There is no consistency, so what real conclusions can we draw from that. So I don't argue one way or another and just follow the company procedures.


--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."