Just to put things into perspective here as well in regard to the " i am at 99% capacity concern"
1) if the relief case is a full flow blocked outlet then being at capacity means that if there is a possibility of the "full flow" increasing in the future ( due to an increase in throughput of the plant) this PSV will be too small and will need upgrading. so that could be areason to put in a larger valve now.
2) if this is a fire case then unless the composition of the contents change dramatically in the future the relief load is stable. So the valve is OK.
3) When you model fire cases using dynamic simulation , say Hysys or VMg et al you size the PSV based on the peak pressure
and calculate say a 32.55mm diam port required . Now a J is 32.50 , so thats too small according to the code!
next up is a K orifice at 38.80
if we use a J orifice what happens ? the valve cannot pass the required load at the relief pressure and the vessel pressure rises a little above the 121% MAAP, to 121.5% maybe just for a few seconds. Is that unsafe ?
Don't get me wrong here guys I am not advocating that engineers regularly recommend the installation of PSV's that are too small !!
Its just that in this day and age of computers and 8 decimal places and, I feel, a lack of practical thinking engineers adhere to the recommended values too much.
recommended BP on a conventional PSV valve 10% i calculate 10.6% at peak flow for 5 seconds is that a problem?
the depressuring takes 16 minutes , API says 15 mins , we need to redesign... do we
as far as i am aware the 15 mins is an arbitrary number that was originally a guideline by API to use.
It now seems to be a standard number to stick to or perish..
ramblings of an old engineer maybe but a concern i think that the old " ..thats near enough to 1 decimal place" philosophy is dying
critical thoughts from others welcome :-)