Dear All,
I checked some of the projects with different Engineering companies and I found that there are two different configurations for PSV isolation valve on tail pipe:
1. Some of the companies use a flanged to flanged configuration for PSV and isolation valve (usually full bore locked open ball valve) for this the isolation valve is located upstream of expander.
2. And some of the companies use the isolation valve downstream of the expander.
I thought that the first configuration is the best one especially from economical perspective because usually the tail pipe size is two or three times (or more) the psv outlet nozzle and if you use the isolation valve upstream of the expander you can save a lot of money considering that ball valves are expensive and the number of pressure safety valves in a typical plant is great.
But I recently read in a technical article (if you are interested I can share it with you) that in recent years there have been instances in it we had a combination of psv at popping state while the isolation valve in closed position(although it is locked open)! The author explain that this is due to high induced vibration of psv when popping. So one of the recommendations of the author is that use the second configuration to prevent this failure.
What is your recommendation about discussed subject?
Many thanks in advance.

Mojtaba:

With all due respect, your message doesn't make sense.

If a PSV block valve is "locked open", then it should be literally LOCKED and there is no physical way - short of stupidity or sabotage - that allows the block valve to close. You state, "The author explain that this is due to high induced vibration of psv when popping". That is not an "explanation". It is an assertion or claim. A "locked" (or "sealed") valve is either closed or open - until changed by direct, human action.

What your author is saying is that he/she is laying claim to making all Hazops impractical and useless since the "locked" or "sealed" procedure is totally accepted and followed in a Hazop. Either you are not reading/translating corrrectly what the author means, or the author is just "blowing smoke" (fantasizing).

Dear Mojtaba,
Somehow I fear that, either you have mistaken the Relif Valves's tell-tail pipes as their Vents ("tail pipes" you described) or you have grossly mis-understood the concept (please do apologise me for being too-Straight forward).
Actually the very purpose of safeguarding the valuable assets gets Totally lost for

pressure vessels,columns and other highly valuable equipment and pipeline systems

if the system can not be relieved at a much lower(i.e safer) pressure
[/list]within the the limits of its MAWP(Maximum Allowable Working Pressure)
accordingly no blockage/isolation can be permissible on downstream side of a pressure relif,however rupture disks(again safety device) are employed upstream side.
Moreover the concept of Car-seal Open and Car-seal Closed is the specific typical cicumstance;which in no case jepardizes safety of the the system.
As such the vibration,locking etc. referred in your post are beyond understanding(as pointed out rightly by Mr.Art)
Best Regards No Hard feelings
Qalander

Mojtaba,

I have seen both approaches.

"But I recently read in a technical article (if you are interested I can share it with you) that in recent years there have been instances in it we had a combination of psv at popping state while the isolation valve in closed position(although it is locked open)! The author explain that this is due to high induced vibration of psv when popping. So one of the recommendations of the author is that use the second configuration to prevent this failure."

The query is unclear...

Please cite the reference and quote the sentences for better understanding.



JoeWong

Obviously option 2 in initiator post is better, because pressure drop and following that build-up back pressure in tail pipe would be decreased, hence there is lesser effects on PSV relief rate due to build- up BP.Because of safety priority economical aspect is not considered.

Regards

Dear Joe if possible;may you share the referred article for my knowledge updating as I might have been at fault in outright rejection of the concept.
Being human;I am definitely suseptible to errors,ommissions and forgetfulness.
Regards
Qalander

Qalander,
I don't really know which article Mojtaba was referred to.

Mojtaba is in position to response.

Dear Joe, I probably asked your referred article showing both types of system used.as highlighted with green font above
Regards
Qalander

I was referred to previous projects and company (copyrighted) requirement.

Thanks Joe, I stand clarified now from your side.

However I must clarify my position that sentences in your previous post did not imply the same; that's why earlier request was made. In no case I would compell anyone to go for laws breaking and/or copyright infringements for my sake.

Best regards,thanks for continued cooperation.
Qalander

As an engineer, I cannot accept a statement like this. Safety, of course, is the overriding consideration. However, if a safe and effective system can be built at a lower cost, there is no reason to add expense. For a PSV, there are well defined limits for what is/is not an acceptable pressure drop for both inlet and outlet piping. We must respect those requirements. Within those constraints, we should select the most economic solution available. Period. That IS engineering. (Of course, if there are other considerations that exist, they can enter into the selection process, even if they are not quantifiable.)

Dear djack

In a "safety requirement range" you are right, but the conditions of flare network significantly dictates the configuration of outlet piping (of course there are limitations such as superimposed and build-up back pressures of outlet piping specially in API 520).Hence , in my opinion the trade-off between two options in pressure drop and economical point of view determines the better configuration.
Regards

I am glad to hear that...

Dear All,
Thanks a lot for all of your valuable comments.
All of the contents of my first post are extracted from the mentioned article.
To JoeWong: I really confused that which part of my question is dark for you? Let me know! I will send the mentioned article to your mailbox.
To fallah: For both of the configurations the economical aspects are important to. back pressure, isolation valve size and etc are affected by the selected configuration. In my view the first one seems to be cheaper.

CCPS published book " Guidelines for design solution for process equipment failure" Chapter 2.6.1 specifically details a failure of a locked open valve due to " design documentation" error. As a designer it was a wake up call for me.

Mojtaba, Could you forward me the article? You know my e-mail address.

Regards
Rajiv Narang
MAIChE

Dear Rajiv,



Glad to hear this!



Check your mailbax.

Thank you.

Dear Rajiv

How do i prepare that book (or at least mentioned chapter)?

Regards

Dear fallah,

The book is a publication of CCPS, a committe set up by AIChE. You should be able to order the book on AIChE website.

Please write to me at rxnarang@gmail.com, and I will attempt to further clarify this incident, as I understood it.

Regards
Rajiv Narang
MAIChE

Dear All Hello,
Let me enter this extremely useful and interesting discussion of process safety.
If anyone is interested he/she can also find excellent guiding info about 'the problems associated with incorrect locking/misunderstandings by operating personnel' through CSB's site www.csb.gov as a part of their initial animation during investigative course of BP-Amoco Texas Refinery incident of March 2005. Also the final report has a great deal of info as links to many of cited/referenced material. Splendid learning info indeed.[font="Book Antiqua"][/font][color="#008000"][/color]
Best Regards
Qalander

I have read the article that you have send me earlier.

The article indicating that block valve is locked open. Somehow large flow rate passing PSV and block valve induced severe vibration lead to block valve closure (damaged padlock).

I have heard of this type of "concern" before but there is not much action on this. Reason very simple... If we are concern about manual block valve in PSV outlet, should this concern extended to other block valves which are locked open ? If not, why not ? Any criteria to define vibration level or sound power level (SPL) limit lead to this concern ? NO. Have not heard of this with my limited experience. Nevertheless, more studies required...

This type of concern has not appeared in other "reputable" company. So i guess it is still remain an open question.

In many flare or PSV tail pipe design,
i) acoustic induced vibration (AIV) is one the check to be conducted to ensure mechanical integrity of these pipes.
ii) tail pipe will be designed to limit the MA number to an agree limit i.e MA=0.7 even though some company may design for MA =1 for whatsoever reason, like what has been discussed in "Is PSV tail pipe & lateral at CHOKED (Mach no = 1) Accpetable ?"
iii) static and dynamic stress analysis of pipe, support design, etc to ensure piping integrity during major relief
iv) Avoidance of secondary choke event in tail pipe...
v) proper arrangement of block valve (valve handle is on top)
.
.
.

I believe with above "gates", locating block valve just downstream of PSV (upstream of expander) still remain acceptable.

Hope this help.