13 replies to this topic

[wenchop]

API 521 states an equation to calculate the orifice size of a PRD, the discharge areas for PRDs on vessels containing supercritical fluids, gases or vapors exposed to open fires.

 

The question is, how to size the tailpipe or discharge piping, given that with this method, there is no load calculation (mass flow)?

 

 

 

[Bobby Strain]

You should tell us what a PRD is. You use strange terminology. Assuming that it is a relief valve, when you have the orifice area, you can calculate the flow rate. But you probably already knew that.

 

Bobby

[wenchop]

Strange terminology?

 

API 521 6TH EDITION quotes:

"The discharge areas for PRDs on vessels containing supercritical fluids, gases or vapors exposed to open fires can

be estimated using Equation"

 

Page 11 acronyms and abbreviations of API 521:

PRD pressure-relief device

Edited by magonz, 28 May 2014 - 06:51 AM.

[fallah]

API 521 states an equation to calculate the orifice size of a PRD, the discharge areas for PRDs on vessels containing supercritical fluids, gases or vapors exposed to open fires.

 

The question is, how to size the tailpipe or discharge piping, given that with this method, there is no load calculation (mass flow)?

 

 

magonz,

 

The fire relief load should has been calculated based on the procedure included in API 521. Then the PRD sizing would be performed using "Flow Resistance Method" included in API 520 part I (as attached) which appears to be same as what you quoted from API 521 6TH Ed. 

Attached Files

•  Pages from API 520.pdf   41.4KB   17 downloads

[wenchop]

Exactly which procedure included in API 521, what page?

[fallah]

magonz,

 

At the moment i haven't API 521 in hand, but for one who is a little bit familiar with this standard it isn't so hard finding "Relief Load Determination" for pool fire...

[wenchop]

If it is so easy, I will gladly await for the page number..

[fallah]

magonz,

 

Page 44 of the Fifth Ed. as attached.

 

 

Attached Files

•  Fire Relief Load.docx   183.16KB   13 downloads

[wenchop]

As shown in your attachment, it goes as far to suggest to estimate the thermal expansion coefficient of the supercritical fluid, but to use this method you need to calculate a fire heat flow, and that is not clear how to do for a supercritical fluid, as there is no "wetted" area.

 

Also, page 43 of 6th edition gives an equation for relief load, but the assumption is ideal gas equations, and in my case the relief pressure is too high to apply ideal gas equations.

 

There is an alternative method in page 44, but it is rigorous, not good for preliminary calculation.   

 

And I still haven't found what I am looking for...

[Bobby Strain]

Magonz,

     Your arrogance won't win friends. Or help, either.

 

Bobby

[breizh]

Magonz,

Consider this paper , you may find what you are looking for .

Don't expect too much from us , just a support .

Regards

Breizh

Edited by breizh, 29 May 2014 - 06:49 AM.

[fallah]

magonz,

 

My comments on your statements are in red color as follows:

 

As shown in your attachment, it goes as far to suggest to estimate the thermal expansion coefficient of the supercritical fluid, but to use this method you need to calculate a fire heat flow, and that is not clear how to do for a supercritical fluid, as there is no "wetted" area.

Yes, for supercritical fluid there isn't true latent heat, but you can treat the vessel containing such fluid as it is very high density gas filled because it makes sense. See also the statement from API 521 in the red box in attached.

Also, page 43 of 6th edition gives an equation for relief load, but the assumption is ideal gas equations, and in my case the relief pressure is too high to apply ideal gas equations.

Yes, the assumptions are based on ideal gas law and might not be valid for high pressure gases but it can be used for estimation as mentioned in API itself. If you wanted to use a PSV for only fire case without another credible scenario, no matter how much does the estimated relief load deviate from the real value but appears by using PRD you are trying to protect the vessel from failure in fire case. Then you can modify the calculated relief load , say, by using multiplyer of density ratio (critical density/maximum perfect gas density) and use as final relief load.

 

Attached Files

•  Supercritical.docx   255.45KB   11 downloads

[wenchop]

Magonz,

     Your arrogance won't win friends. Or help, either.

 

Bobby

 

Your condescension or patronizing attitude won't win friends.

Edited by magonz, 29 May 2014 - 07:01 AM.

[wenchop]

Fallah:

 

Sounds very reasonable, but I cannot argue in front of peers that I used ideal gas equations at high pressures.

 

I also do not want to calculate a wetted area and expansion coefficient for a supercritical fluid. API 521: 

"Unwetted wall vessels are those that have no liquid in contact with the internal vessel walls (e.g. internal walls are

exposed only to a gas, vapor, or supercritical fluid or they are internally insulated regardless of the contained fluids). These include vessels that contain separate liquid and vapor phases under normal conditions but become single phase (above the critical) at relieving conditions."

 

 

What I will do is do the rigorous method, as there doesn't seem to be a simple approach.

 

But thank you anyway. 

 

Regards,

Edited by magonz, 02 June 2014 - 09:52 AM.