Which upstream pressure to use for tube rupture flow rate?

[wolf359]

I have an exchanger where HP gas is cooled against LP cooling water. The LP side (and local pipework up until the block valves) has had its design pressure increased according to the 10/13 rule, therefore there is no need for a local relief valve in the event of a tube rupture.

However, in such an event it will still be necessary to relieve the HP gas via the cooling water network, and therefore the relief valve on the cooling water expansion vessel must be sized for this service, and a check must be made to verify that the return header piping is sufficiently rated for the back pressure.

My question is, what do I use as the pressure upstream of the rupture, in order to calculate the flow? Max operating? Design Pressure? Design Pressure + Accumulation? My own gut feel is to go conservative but I am under pressure to relax my assumptions.

Can anybody point me in the direction of any standards which I can use to justify a relaxation of my assumptions? I specifically need references; just stating that some allegedly experienced engineer on an internet forum told me to use my own judgement will not wash with this client..

 

[don1980]

See API 521, section 5.19. It provides extensive guidance on tube failures, but it doesn't answer your specific question (what upstream pressure to use). Nor am I aware of any standard that will answer that question. Standards intentionally leave that type decision to the user because each case is different (different risks and consequences) and each case needs to be evaluated separately. I would use the upper end of the normal range of operating pressures. To me, it's double-jeopardy to use a value higher than that. To do so would mean you're basing the scenario on two independent failures occuring at the same time (tube failure, and loss of process control).

Every user is free to be extra conservative if they want to, and in some cases that's justified (e.g. highly toxic fluid). but that's a special circumstance. Using the max normal pressure is perfectly acceptable, and in my experience, that's what most engineers use.