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

2

Allowable Total Backpressure For Conventional Psv


12 replies to this topic
Share this topic:
| More

#1 Butterfly

Butterfly

    Gold Member

  • Members
  • 87 posts

Posted 21 January 2025 - 12:43 PM

Hello

 

I come back to this topic, there was a similar query years ago, but I wanto to go a little bit more in detail to be sure that I fully unserstand. Please, correct me if anything I state is wrong.

 

Let's assume that I want to specify a new conventional PSV (even though for this example I would have chosen a balanced one).

 

I attach a diagram with backpressure values for different scenarios (no discharge and multidischarge).

 

PSV set pressure is 23.3 barg. 

 

I have a variable superimposed pressure (0.3 to 1.37 barg). In GEPF, built up backpressure is 2.73-1.37= 1.36 bar.

 

In a conventional PSV, as per API 520, if allowable overpressure is 10% of Pset, then, allowable built up backpressure is 10% Pset. For this particular PSV that would be 2.33 bar

 

If in the process datasheet I specify maximum superimposed backpressure as 1.37 barg, then, I could specify maximum total backpressure as 1.37 + 2.33 = 3.7 barg. Is this correct?

 

Then, for my case, since maximum total backpressure is 2.73 barg, PSV would work (built up, which is 1.36 bar, is less than 2.33 bar).

 

Now, in reality, would it be ok to specify a conventional PSV like this, giving the range of variable superimposed backpressure and the value for maximum built up backpressure of 2.33 bar? Would the PSV open as long as total backpressure is less than 3.7 barg? Or is it wrong to specify a conventional PSV when superimposed pressure is that variable?

 

 

 

In reality, what I am doing is checking an existing system with flarenet (because some new discharges have been added). I came across this conventional PSV (and it discharges to flare system, where superimposed pressure is variable). With flarenet I can get the backpressure values for the different scenarios, and I wanted to check if the PSV is still valid for multi-relieving scenario. 

 

Thank you for your help


Edited by Butterfly, 21 January 2025 - 12:47 PM.


#2 Butterfly

Butterfly

    Gold Member

  • Members
  • 87 posts

Posted 21 January 2025 - 12:52 PM

Sorry, here is the attached file

Attached Files


Edited by Butterfly, 22 January 2025 - 10:50 AM.


#3 latexman

latexman

    Gold Member

  • Admin
  • 1,793 posts

Posted 21 January 2025 - 01:43 PM

Is there a reason there is not a scenario for this PSV being open and all other PSVs being open at same time?

#4 Butterfly

Butterfly

    Gold Member

  • Members
  • 87 posts

Posted 21 January 2025 - 05:03 PM

In General Electric Power Failure all other PSVs are opened (multi-relieving scenario). With drawings second and third, I wanted to show backpressure contibution (which backpressure is due to other PSVs and which backpressure is due to my particular PSV).

 

Regards



#5 latexman

latexman

    Gold Member

  • Admin
  • 1,793 posts

Posted 21 January 2025 - 05:26 PM

Sorry, I was not clear. Is it credible or non-credible for ALL PSV’s (General Electric Power Failure + your particular PSV) to be open simultaneously?

#6 Butterfly

Butterfly

    Gold Member

  • Members
  • 87 posts

Posted 21 January 2025 - 06:55 PM

Yes, it is credible. In fact, it is an existing system I am looking at, with this conventional PSV. My PSV has a discharge in power failure. In total power failure, there are other PSVs which also discharge, and all together, they give a total backpressure in this PSV of 2.73 barg. I was confused because this PSV is a conventional one (I would have specified a balanced one). I didn't find the vendor datasheet, only process datasheet. 

 

With flarenet, ignoring the PSV I am looking at (but with the other PSVs that are open in general electric power failure being active), I was able to get the value of 1.37 barg of backpressure.

 

My first "impulse" was to say: this PSV is not valid for a general power failure scenario, because total backpressure is 2.73 barg, greater than 10% of set pressure (10% of 23.3 barg). 

 

But then I read API 520 carefully and realize it refers to built up backpressure 

 

Quote: in a conventional PRV application, the allowable built up backpressure is equal to the allowabel overpressure

Unquote

 

In GEPF, built up backpressure is 2.73-1.37 barg (is the contribution from my single PSV). Less than 10% set pressure. 

 

But I was used to see conventional PSVs discharging to atm, so that total backpressure = built up

 

Maybe, even though built up backpressure in multi relieving scenario is less than 10% of Pset, the conventional PSV is not valid because superimposed backpressure is not constant. Or mayne the conventional PSV is, in fact, perfectly ok. I would like to clarify and understand this.

 

I hope I made myself clearer now.

 

Thanks and regards.



#7 shvet1

shvet1

    Gold Member

  • Members
  • 469 posts

Posted 21 January 2025 - 10:57 PM

What do you mean 'specify'? Do you want to replace this PSV with an another one? Or modify flare header?

From my point, 'to specify' means to prepare a datasheet for purchasing. Do you want to purchase a new one PSV?

 

PSV superimposed backpressure is variable during a common relief case. And what is a porblem? Some PSV will open at the upstream pressure fewer than the set pressure. Will it be a problem? Will the accumulation exceed the allowable one and a vessel rupture?

From here it seems as there is no difference in the scenario result.

 

Why do you want to replace a conventional PSV with a balanced bellows one? Conventional design is more robust, reliable, and - what is more important - safe as the flare gas won't break through in case of a bellows failure.

 

Identify the problem you are solving. What will you improve if there is no a problem? Try to shift your mind and think about how it works instead of how to comply with codes or practices.


Edited by shvet1, 22 January 2025 - 09:15 AM.


#8 latexman

latexman

    Gold Member

  • Admin
  • 1,793 posts

Posted 22 January 2025 - 09:21 AM

It appears the attached diagram does not have the scenario and backpressure data for ALL PSV’s (General Electric Power Failure + your particular PSV) to be open simultaneously, correct?

The third sketch does not mention the status of “the other PSVs”.

#9 Butterfly

Butterfly

    Gold Member

  • Members
  • 87 posts

Posted 22 January 2025 - 10:55 AM

I tried to make the drawing (attached, edited) clearer now.

 

I included a new scenario, single valve relieving (sketch 4).

 

I think I was confused about two concepts:

 

  • API-520 mentioning allowable built up backpressure for a conventional PSV
  • Potential problem with conventional valve if backpressure is too high

 

Let’s say this particular PSV-A (I named it so in sketch) has several relieving cases. And superimposed backpressure is variable (because the valve discharge goes to flare system).

 

  • Single relief case (sketch 4). Just before valve opening, superimposed backpressure is (0.1 x 100)/23.3 = 0.43 % of set pressure

Built up backpressure %, I would say is (1.34-0.1)x100 / 23.3 = 5.32 % of set pressure.

 

  • GEPF case (sketch 3). Just before valve opening, superimposed backpressure is (1.37 x 100) / 23.3 = 5.87 % of set pressure

 

 

With this variation of superimposed backpressure (0.43%-5.87%) I understand that, for the vessel being protected by this PSV, depending on the scenario, set pressure at which the PSV will start to open is going to be different depending on the scenario (because the PSV sees a differential pressure).

 

 

Thanks for reading



#10 snickster

snickster

    Gold Member

  • Members
  • 81 posts

Posted 22 January 2025 - 04:59 PM

I believe when you are looking at superimposed back pressure in a system that is not compensated for in the set pressure of the conventional relief valve, you must add the superimposed back pressure to the built up back pressure due to flow and the total of this sum must be below 10% of the set pressure.  This is how I have always designed relief systems. 

 

I have a copy of API 520-2000 which states the following:

 

3.3.2.2 Balanced spring-loaded or pilot-operated pressure relief valves should be considered if the superimposed back pressure is variable. However, if the amount of variable superimposed back pressure is small, a conventional valve could be used provided:

a. The set pressure has been compensated for any superimposed back pressure normally present; and

b. The maximum pressure during relief does not exceed the Code-allowed limits for accumulation in the equipment being protected.

 

3.3.2.3 For example, conventional valves are often used when the outlet is piped into a relief header without compensating the set pressures for the superimposed back pressure caused by other relieving devices. This approach can be used provided the allowable accumulation is not exceeded during the release.

 

I interpret this to be saying that for superimposed backpressure created by the release of other valves, it is allowed to install the subject conventional valve without compensation for the set point (by adjusting the spring compression) as long as the allowable accumulation of the vessel is not exceeded.  

 

The conventional relief valve capacity is reduced when discharge pressure reaches 10% of set pressure for a relief valve sized based on 10% overpressure without spring compensation.  This is because this is the point where forces on the back side of the relief valve disc trying to close it, just equal the forces on the upstream side trying to keep it open.

 

For instance, for a valve set at 100 psig with no spring compensation the valve will open at 100 psig and reach full capacity at 110 psig (10% overpressure).  The spring load will exert a constant load equivalent to 100 psig (set pressure) so when the backpressure reaches 10 psi (10% of set point) the loads on the upstream side and downstream side are just equal, any more back pressure will tend to close the valve.

 

The valve doesn't know whether the backpressure at the valve discharge is due to superimposed back pressure or built up back pressure as far as the capacity is concerned.  So full capacity will be reached as long as the total backpressure at the discharge is maintained below 10% (so valve stays fully open) which is the sum of superimposed plus built-up backpressure.  If full design capacity is achieved, then vessel allowable vessel accumulation will not be exceeded.

 

However, the valve will pop at a higher vessel pressure equal to the set pressure plus superimposed back pressure, but this seems to be allowed per API-520 as it doesn't really matter as long as the allowed vessel accumulation is not exceeded.

 

In your case 3 you are exceeding 10% of set pressure based on the sum of superimposed and built up back pressure.

So you will need to upsize your piping to get it equal to or less than 10%.


Edited by snickster, 22 January 2025 - 05:18 PM.


#11 latexman

latexman

    Gold Member

  • Admin
  • 1,793 posts

Posted 22 January 2025 - 10:19 PM

Butterfly,

Thanks for clarifying the diagram and scenarios. It’s much clearer to me now. Snickster has given you some good advice.

Good luck,
Latexman

#12 shvet1

shvet1

    Gold Member

  • Members
  • 469 posts

Posted 22 January 2025 - 11:57 PM

With this variation of superimposed backpressure (0.43%-5.87%) I understand that, for the vessel being protected by this PSV, depending on the scenario, set pressure at which the PSV will start to open is going to be different depending on the scenario (because the PSV sees a differential pressure).

 

Different, but how different? Any backpressure will reduce pressure differenece accross the PSVs' seat and make those to pop up at upstream pressure fewer that set pressure. Fewer, not higher!

 

Backpressure is not a problem. The problem is a vessel rupture. Can you describe a logic how the vessel's accumulation will be exceeded? Not a complyance with a code/practice.

 

Assuming that vessel's MAWP = PSV's set pressure and allowable accumulation  is 10%, the vessel will rupture when the pressure inside exceeds

 

23.3 * 1.1 = 25.63 barg

 

Can you describe how the pressure inside of a vessel will exceed 25.63 barg? Not all those staff about the backpressure, practices and so one.


Edited by shvet1, 23 January 2025 - 12:00 AM.


#13 Butterfly

Butterfly

    Gold Member

  • Members
  • 87 posts

Posted 23 January 2025 - 04:01 AM

Thank you snickster for the clear response (and the note for the valve with no spring compensation, so that it does not care whether backpressure is superimposed or buit up) .

 

Thanks also to shvet1, I have to worry about my vessel and if accumulation limits are going to be exceeded.






Similar Topics