Relief Headers
Relief Headers
(OP)
I have 2 separate 5,000 gallon reactors each fitted with their own rupture disk (no PSV), but tie into the same relief header. They’re (the RDs) stamped with 82 psig @ 72 F. Doing a quick calculation (assuming only water in the reactor) gives me around a 2-2.5” required area and around 30,000 lb/hr capacity (?)(fire case). Both of these relief lines go 20’ out from the reactor with 2” lines, then connect at a “T” and continue another 100’ in a 2” line before emptying into an open-top tank. How can I go about determining whether or not the “T” is OK, or if I should break it and run 2 separate 2” lines out to the tank?





RE: Relief Headers
Go to the site and on the menu at the left select "Free Articles", "Pressure Relief Desing". You will find my series on rupture disks there.
You treat the rupture disk (and holder) like any other piping component. Start the calculation backwards since we usually know the terminal pressure (atmosphere, some seal pot, etc.) Then calculate the final pressure at the vessel. This must not exceed the design pressure (or MAWP if known) plus allowable accumulation.
I discuss this and other issues in my articles such as how one rupture disk can affect another one tied into the same relief header.
RE: Relief Headers
Just a quick comment. You are sizing the relief based on water in the reactor. Do you actually make only compounds in the reactor that use water as the solvent?
Second comment. All of our reactors have the reliefs sized for run away reactions as these cases are much worse than the fire case FOR US.
If the reactor was at our location the rupture disk size would be in the 10" to 12" range.
Granted all those things are based on what my company does but it may give you a little pause as to how you are calculating your relief capacity.
Good luck this is a difficult area.
StoneCold
RE: Relief Headers
And I’m not worried about a runaway reaction since there’s not a real temperature dependent reaction going on. I’m more worried about a fire starting underneath the reactor and boiling all of the water out of the reactor.
RE: Relief Headers