PSV Inlet Line sizing - process conditions 2

[zamakaze]

i have a gas vessel operating ayt 25 psig and psv set pressure at 55 psig.
during psv inlet line sizing for blocked flow, should i take properties of gas at 55 psig? or 25 psig? for line sizing?
what should be the temperature for line sizing ...will it be operating temp / ambient temp?

 

[RVAmeche]

You should be calculating fluid properties at the relief conditions (pressure & temperature). If you have multiple relief scenarios, these values may vary.

 

[zamakaze]

hi RVAmeche, for a simple block flow case, what should be relief temperature? should it be the max operating or should some other calculation be done for temp.?

 

[Latexman]

To me the blocked flow case means the gas in the vessel is being compressed. Is it happening so slowly the heat of compression can dissipate, or is it happening so quickly the heat losses are negligible? We have no data on that.

This scenario falls under a process upset classification so sizing pressure is set pressure + 10% over pressure, or 60.5 psig, if set pressure = vessel MAWP.

Good Luck,
Latexman

 

[zamakaze]

Dear Latexman, Thanks for your reply.
DO you think that taking the vessel operating temperature would be a reasonable assumption for psv inlet line sizing?

 

[Latexman]

As I said, we have no data on that. We know the compression ratio is about 1.9, but we do not know the composition of the gas, it’s properties, the speed of compression, whether the vessel is insulated or bare, or what else may affect the outcome or risk. You should know all this and make a wiser decision than SGOTI.

If this information is not readily available, then I would look at what it would cost to go conservative in the calculations. If going conservative is essentially free (same PSV), then that’s what I would document in the official PRD record, with an explanation of why I went conservative (same PSV).

Good Luck,
Latexman

 

[zamakaze]

dear latexman, for the relieving temperature, if we assume that gas is getting compressed, would isoentropic compression be the way to estimate the temperature at relieving pressure i.e. T2/T1 = (p2/p1)^(k-1/k). thankyou