Jump to content



Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

Check Valve Failure


This topic has been archived. This means that you cannot reply to this topic.
10 replies to this topic
Share this topic:
| More

#1 harini

harini

    Junior Member

  • Members
  • 12 posts

Posted 22 April 2009 - 03:55 AM

Dear All,

I would like to have some views and comments regarding check valve failure as a credible overpressure scenario while evaluating a PSV's scenario.
API 521 talks about Check valve leakage or failure and the topic addressed there is the leakage that causes reverse flow and pressurise the lower design pressure side.
I would like to know if we can consider a blocked outlet due to a check valve stuck close. I tried reasoning the same with some engineers at work and did not get a clear conclusion.
I tried checking the failure modes for the check valve and the database lists just only common mode of failure and the probability for the failure is 0.21, which will lead to a high risk level.
An example for the arrangement could be a overhead gas line from a contactor to suction drum. I am attaching a simple skrtch for the same.
I would appreciate all your suggestions.

Thanks
Harini

Attached Files



#2 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 22 April 2009 - 06:27 AM


Harini,

Check valve stuck closed is a very realistic scenario in the process industry. If you will notice, for pumped liquid service when using centrifugal pumps, the take-off or branch connection for the minimum safe continuous flow is always provided upstream of the check valve as a good engineering practice for precisely the reason of pump running dead-headed in case the check valve is stuck closed.

I am not conversant of such a scenario in gas service since I can't think of the pipe upstream of the check valve having a lower design pressure than the downstream pipe. Thus the pipe upstream of the check valve is fully rated for the gas shut-off pressure. On the contrary, if a pipe specification break is provided, then it is generally the downstream side of the check valve which has a lower design pressure rating. However, I personally do not recommend providing a pipe specification break i.e. lower pipe rating after the check valve due to the notorious unreliability of check valves and specially if your fluid has some kind of solids which can aggravate check valve failure.

As per the description I have provided in the above paragraph, it appears that a blocked outlet case in gas service due to check valve failure is not realistic.

Hope this helps.

Regards,
Ankur.

#3 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,780 posts

Posted 22 April 2009 - 08:21 AM


Harini:

I believe you are trying to protect the vessel immediately upstream of the stated check valve, although you don't state it outright.

I have taken the liberty of making my comments on your workbook, of which I attach a Rev1 copy for your review.

I believe that you have to consider the blocked flow case for the mentioned vessel - regardless of the check valve's operation. This, then, would ensure that the failed-close case for the check valve is taken care of.

Please correct me if my interpretaion is wrong.
Attached File  PSV_Blocked_Case_Sketch_Rev1.xls   77KB   563 downloads


#4 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 22 April 2009 - 12:41 PM


Dear Art,

The gas blow-by due to a process upset at the suction drum seems a more realistic and governing scenario for a relief valve rather than anything else. I probably missed out the design pressure of the downstream suction drum. Another reason for over-pressurization of the drum could be the source side with a blocked outlet rather than the check valve failure as you have mentioned.

One way to mitigate a gas blow-by from the downstream would be to provide two dissimilar check valves with different operational principles.

Thanks for pointing out the gas blow-by scenario.

Regards,
Ankur.

#5 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,780 posts

Posted 22 April 2009 - 01:00 PM


Ankur:

Yes, you have reflected what I feel about the scenario - as I interpret it.

If the downstream process is subject to a pressure upset (as the downstream design pressure seems to indicate), then the smart thing to do is precisely what you suggest: install two check valves in series, each of a generic different design in order to mitigate any simulaneous failure of both at the same time.

I am curious to hear from Harini, who supplied us with very good input about the application in the form of a simple sketch. Once again, a picture tells all.



#6 fallah

fallah

    Gold Member

  • ChE Plus Subscriber
  • 4,930 posts

Posted 23 April 2009 - 01:06 AM


Dear Art,

Reffering to your comments submitted on the scketch,as far as i know there is no predefined failure position (FC/FO) for check valves similar to what is defined for control valves.Would you please clear this matter,briefly.

#7 harini

harini

    Junior Member

  • Members
  • 12 posts

Posted 23 April 2009 - 09:35 AM

Dear Art, Ankur and Fallah,

I would like to first thank every one for the valuable input. As far as the block valve downstream of the check valve is lock-open valve. I understand that we can give credit for LO status of the valve - please correct me if i am wrong. This is the reason i was worried for the block discharge due to the check valve.
And yes i am trying to protect the contactor with the PSV set at 52 Bar g.
Art - I guess it was a typing error for the design pressure of the suction drum. It is same as the contactor vessel 52 bar g. I am sorry for the wrong data. kindly excuse.

Thanks
Harini

#8 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 23 April 2009 - 10:14 AM


Harini,

If as you say the suction drum DP is the same as the contactor and the block valve is LO then I don't think the case of check valve stuck closed is a realistic scenario.

In fact if there is no other source of flow to the suction drum which under all conditions would be at a lower pressure then the contactor then I don't even see the point of having a check valve in the line. Check valves are meant to prevent backflow from the downstream system and if you don't have any other source connected to the suction drum then the check valve would be redundant and thus serve no purpose. Having a check valve implies that there is a possibility of back flow which would only happen if the downstream system sees a higher pressure due to some process upset and is connected to another source, other than the contactor.

Analyze the system properly, and find out if there is a possibility that the suction drum can see a higher pressure which can cause reverse flow to the contactor. If such is the case then the backflow would govern the protection of the contactor. It means that the gas blow-by would be your governing case for the relief valve on the contactor and not the check valve stuck closed.

As indicated in my above post, a gas blow-by (reverse flow) scenario can be mitigated by providing two dissimilar check valves with different design principle.

Hope I have been able to help.

Regards,
Ankur.

#9 harini

harini

    Junior Member

  • Members
  • 12 posts

Posted 24 April 2009 - 01:35 AM

Thank you Ankur for your input. Yes I am going to do a through analysis and then will share the conclusions in this forum.

Regarding the series of two check valve for increases reliability - API says that even if there is a series of check valves, we will have to assume that the valves will fail and have to consider gas blow-by case. The percentage of orifice opening will depend on the company standard or if no standard is available, we will have to assume 10% of the valve diameter as orifice diameter.

In a previous project we had the similiar arrangement with 2 check valves in series and they were classified as safety critical. But inspite of this, i had calculated Gas-blow by for flow through the orifice of 10% of valve size for the low pressureside vessel. Will this be Over design??

#10 fallah

fallah

    Gold Member

  • ChE Plus Subscriber
  • 4,930 posts

Posted 25 April 2009 - 12:06 AM

QUOTE (harini @ Apr 24 2009, 01:35 AM) <{POST_SNAPBACK}>
But inspite of this, i had calculated Gas-blow by for flow through the orifice of 10% of valve size for the low pressureside vessel. Will this be Over design??


I think as per 4.3.4.4 Pressure consideration for series back-flow prevention from API 521,no overdesign has been considered.


#11 harini

harini

    Junior Member

  • Members
  • 12 posts

Posted 09 July 2009 - 04:32 AM

Dear All,

The final design of the system is finalised at last and Yes Check valve failure was the governing case for the PSV instead of blocked discharge. With one check valve, the flowrate was quite higher and it had a lot of impact on the Flare network. So finally itw as decided to put two check valve in series of different type and calculate the flow for 10% of the size of the orifice. I have named the check valve as safety critical in the P&ID.

Thank you all for the best input and guidance.

Kind Regards
Harini




Similar Topics