Testing an emergency depressurization system
Testing an emergency depressurization system
(OP)
Does anybody have experience or thoughts about the possibility of testing an emergency depressurization system as part of the startup of a plant? I mean to really blow some product to the flare (in this case the depressurization system goes to the flare).
I feel that it would be expensive (loss of product) but at the same time would give some valuable information about the performance of the system. If you expect two phase flow and flashing, the sizing of the orifice plate with reasonable accuracy is not possible (my conclusion from other thread where this topic was discussed). Consequently, during a real emergency depressurization I could get a flow higher than expected, with the possibility of exceed the capacity of the flare (or at least reduce the remaining capacity available for other discharges, risky if we think of a fire situation for instance), or I could get a low flow, not reaching my depressurization pressure in the expected time. Doing some testing would allow me to resize the orifice if required.
Any opinion or suggestions?
I feel that it would be expensive (loss of product) but at the same time would give some valuable information about the performance of the system. If you expect two phase flow and flashing, the sizing of the orifice plate with reasonable accuracy is not possible (my conclusion from other thread where this topic was discussed). Consequently, during a real emergency depressurization I could get a flow higher than expected, with the possibility of exceed the capacity of the flare (or at least reduce the remaining capacity available for other discharges, risky if we think of a fire situation for instance), or I could get a low flow, not reaching my depressurization pressure in the expected time. Doing some testing would allow me to resize the orifice if required.
Any opinion or suggestions?





RE: Testing an emergency depressurization system
RE: Testing an emergency depressurization system
I will recommend that you make use of API 520 & API 521 especially when you have 2-phase flow or liquid flashing across the orifice. the results of your calculation should be tested using HYSYS, which will save you the cost of practically depressurizing the plant and resizing your orifice.
RE: Testing an emergency depressurization system
I've seen a several articles published suggesting the benefits of dynamic simulators to validate relief systems and emergancy systems.
On an high fiedelity simulator project i've worked on we have tested an emergancy depressurisation system. (different industry though)
RE: Testing an emergency depressurization system
you can use Aspentech Flarenet to model this for you. Although it is a steady state analysis it will give you a very good calculation and indicate any potential problems you may have. If you do not have access to this software then get in touch with this specialist flare company, they may be able to assist you with a small study. sales@softbits.co.uk
RE: Testing an emergency depressurization system
We tend to check out blowdown systems prior to a maintenance turnaround when the plant is scheduled to be depressurized and gas freed for inspection purposes. We also take this opportunity to check out the functionality of ESD valves i.e. subject them to a full stroke test.
Regards
Jit
RE: Testing an emergency depressurization system
Personally, I like a lot of what has been said here. Functionally testing is good and I like to have blinds that can be inserted so you don't waste or blow out unexpected liquids.
Then having calculations that show capacities instead of a stop watch or whatever you want to grade the performance is good. I always question why a 30 second rule, the answer is always beacuse thats what Joe told me or something..
RE: Testing an emergency depressurization system
Among depressuring recommendation is the old API RP 521 was to “set the depressured rate to achieve 100 pounds per square inch gauge (690 kilopascals) or 50 percent of the vessel design pressure, whichever is lower, in 15 minutes.