The total capacity of the worst case should be achieved with all the valves. Not by each valve.
Case A: you are designing considering single Jeopardy cases.
What you need to calculate are the relief scenarios. Say that your design case is for 6 million Sm3/d. You split your relief capacity between both devices with 3 million each, Set the second one (the disc) at 105% MAWP, an the valve at some value below MAWP. You provide a difference greater than 5% because the rupture disc are less precise than valves regarding the setpoint.
So, for now: MAWP 94.6 bar
Disc: 99.33 bar
Your disk has a manufacturing range of -5%/0%. This means that it could be marked anywhere between 94.36 and 99.33
That disk has a burst tolerance of +/- 5%. So the minimum burst pressure would be 89.64
Then you set your PRV with some margin below this pressure to be sure that your valve will always open before the disc ruptures.
You are complying with the ASME and API codes since you have one device at or below the MAWP and a supplementary one at a pressure not higher than 105% of MAWP
In this case you have to calculate your relief capacity at 116% of MAWP (109.7 bar). This is the maximum permitted accumulation when multiple devices are used
You can go the other way, starting from the operating pressure, then the PRV, then the burst margin and manufacturing range and ending with the design MAWP
------------------------------
Case B: you are considering the simultaneous occurrence of the worst case overpressure event AND the failure of the PRV (as you mentioned in earlier post)
Your worst case relief requirement is 3 million. You set a PRV at or below MAWP. And you calculate the capacity of 3 million Sm3/d.
You also consider a case where you have the same event where the combined with the failure of the PRV needs a 3 million Sm3/d relief. This is a double jeopardy case and ouside the scope of the code. Acceptance criteria here are the sole responsibility of you or your client. You install a Rupture Disk for 3 million Sm3/d at a pressure decided by you. 105% is a possibility, but not something required by the code.
-------------------------
If you have existing equipment designed with this configuration, take a look at the scenarios that were deemed credible and their quantification.
Also, the numbers you mention does not fit well with the P&ID you showed before, you cannot relieve 3 million cubic meters per day using a 1 x 2 valve.
Regards
Edited by Saml, 31 August 2016 - 06:20 AM.