PSV Sizing for Shell and Tube Heat Exchanger

[khatarnak]

Hi All,
I would really appreciate any help on this matter. Typically, I have sized PSVs for tube ruptures in a heat exchanger since the tubes were always at higher pressure. I ahd done these using API 520 and formulas in CRANE book.

What do I need to do and what formulas to use if the shell side is actually at high pressure ?

Please advise.

Thanks,

 

[dcasto]

the same and then have the PSV somewhere on the tube side. You assume a tube failure and go.

 

[khatarnak]

dcasto,
Thanks for your reply. When you say 'PSV somewhere on the tube side' what exactly you mean by that.
In the scenario I mentioned, the high pressure gas from the shell side will enter the tube side.
What flow rate do I need to use to size the PSV . Also, where exactly this PSV should be installed.

Thanks.

 

[maddocks]

You should install the PSV as close as possible to the exchanger; however, if you have the chance, it can actually be cheaper to get the same design pressure on both sides sometimes. ALternately, you can design the tube to 2/3 of the shell and potentially justify that the possible max pressure it could see would be the same as the original hydrotest pressure.

 

[lizking]

Note that the 2/3's rule is now obsolete due to new Code hydrotest pressure factors. If MAWP of high pressure side is less than the corrected hydrotest pressure of the low pressure side, tube rupture scenario can be negated.

 

[maddocks]

Good point - thanks for the correction. I'd forgotten about that.

 

[EGT01]

One other point to remember, it is not just the rating of the low pressure side of the heat exchanger that needs to be considered but the evaluation should be made for any other piping and equipment that is associated with the low pressure side of the exchanger.

 

[shahyar]

2/3 is now 10/13.

 

[khatarnak]

Thanks all for all your comments.
Will it be possible to find out the formulas to calculate the Q ( required flow rate ) I need to use for sizing the PSV.

Thanks in adavance.

 

[maddocks]

just use the eng-tips search feature - we've discussed the tube rupture formula's in here lots of times...

 

[roydm]

In the HAZOPs (Petrochemical) I have attended the valve has been for locked in thermal relief, sizing for a rupture has never been an issue. By specifying one of the valves to be "Locked Open" can't you avoid the relief valve altogether to minimize the possibility of leaking hazardous fluids?

 

[maddocks]

Only if the low pressure side has enough PSV capability to relieve a tube rupture.