Reactor Pressure Relief

[RPG]

Anyone have any thoughts on the appropriate relief of a reactor that has two phases of operation. Phase 1 (Mixing) , small rupture disk at a high pressure (below rating of vessel). Phase 2 (Reaction) is where I need to relieve the reactor at a lower pressure to keep the size of the rupture disc small. Does anyone have experience with switching out rupture discs on a reactor based on the phase of a batch.

 

[TrevorP]

I would thoroughly recommend you do not do this. It is frought with hazards and danger.
What if the wrong disc is put in at the wrong stage?
What if someone forgets to change a disc over?
What about hazards due to exposure when changing the disc over?
Size your relief vent for the worst case scenario, and set the set pressure to that required (Usually design pressure of vessel) and leave it at that. If your vent is too small for the worst case, make it bigger. Anything else is unsafe.

 

[unclesyd]

Adding to TrevorP's excellent response especially the last statement.
A point not mentioned is that rupture disks don't like to be handled. One small scratch or dent can seriously affect the bursting pressure, that's why they come in nice boxes.

We run a highly reactive continuous process where we have "dump/vent valves" that are triggered by different process anomalies that will vent/dump the units contents to hopefully quench the process way before the PSV's come into play.
If there is an process excursion that would require the PSV's to operate the dump valves will come into play as the pressure increases to relief pressure.

 

[RPG]

Using a "dump valve" is an option, the problem is that during the non-reactive phase the pressure goes higher than the pressure that I would want to dump at during the reactive phase. A safety system would not have a way of knowing what phase the reactor is in, any suggestions?

 

[StoneCold]

RPG
Explain to us why you need to dump the reaction phase at a lower pressure? Does it have to do with the design temperature of the vessel?

StoneCold

 

[RPG]

During the reaction phase, the system operates at a relatively low pressure. In the event of a malfunction (runaway reaction) the sooner you can dump the reactants the less likely that you will have an excursion that will lead to a vessel overpressure. If you wait until the pressure is near the MAWP of the vessel, this will require a much larger dump valve and line to empty the reactor contents quickly.

 

[MortenA]

A PAHH and a blow down valve could do this - i do however not think it would allow you to "undersize" the bursting disc and vent header connection.

The primary sysytem (PAHH) could fail. Remember to put a PAHH on a seperate instrumten (not you nomal pressure indicatior) since the reason for the process getting out of hand could be a false reading from this instrument!

Best regards

Morten

 

[StoneCold]

RPG
I am not following you about a higher reactor pressure requiring a larger dump valve. Increasing the reactor pressure should decrease the dump valve size. Unless you are worried about the rate of reaction being faster so you think you should dump the reactor faster.

We don't have dump valves but we do have batch reactors with rupture disks and we do everything we can to avert a relieving event. Having a higher rupture disk pressure increases our time to react, and in most instances allows enough time for the limiting reagent to run out or cooling to bring the reactor back under control. If it was my unit I would set the rupture disk as high as possible (less than or equal to MAWP) and leave it there. I would set the dump valve at a lower pressure but still comforably above the highest normal operating pressure. Then you are always covered and you are not relying on changing the rupture disk to save you. Because unless you have very stringent documentation, and signoffs by people who are willing to take responsibility you are going to have an accident.

 

[JStephen]

Put two nozzles on the thing. One has the high-pressure rupture disk sized to protect the vessel. The other has a manually operated valve, then the low pressure rupture disk for the process. And you'd have to open and close that valve each time. Would that work? Just an idea.

 

[phex]

i would go another way in both cases. size a rupture disk/PSV for the high pressure safety. For the runaway reaction case, install another mode of trip, for example strong temperature rise and do this with a FO purge valve to flare/slop. that way you have to completely independent systems and no need for human intervention to have the right safety system in the right moment.

just my two cents,
chris

 

[ProcsssDr]

I second TrevorP's comment:
DO NOT design a safety system that requires human intervention for it to operate properly, i.e. switching rupture discs depending on mode of operation. Design the pressure relief system for the worst case scenario. Then use pressure controls or dump valves to limit maximum pressure for each operating mode. If the controls fail, the pressure relief will save you no matter what operation mode you are in. Trying to save pennies on a safety system, can cost you big bucks in court.

Just my humble opinion!
ProcessDr

 

[unclesyd]

You might want to checkout the two web addresses.
This is the same concept (SIS) that is used at our site. The original system was controlled by Provox but more of the control is being offloaded to local DSP’s with over site by Provox.

SIS (Safety Instrumented Systems).

Check out the FieldView pdf

http://www.emersonprocess.com/fisher/products/documentation/Advertisements/fishersafetyintrsys.html

http://www.emersonprocess.com/sis/index.html