3 replies to this topic

[eastorca]

Hi all,

I found some interesting topics about supercritical PSV in this forum; however, I can not reply so I want to start a new one and raise my concerns. Hope someone could help me to sort them out.

"Rigorously size relief valves for supercritical fluids by Ryan Ouderkirk" provided a better sizing method instead of the reluctantly using equation 8, 9 and 12, section 5.15.2.2 of API 521 which are based on the phycal properties of air and perfect gas law. Article comes with an example for single component (n-butane). If we spread this out. In case vessel contains oil (multi-component, different boiling range), dense phase will apprear if fluid is heated to supercritical condition. Dense phase could be a homogeneous mixture. Question:
1. With assumption that fire occurs during long term, this leads to temperature increase step by step although fluid is relieved by PSV to maintain vessel pressure. Is that reasonable if we consider molecular weight of fluid inside vessle is constant value?
2. Article present equation to calculate the isentropic-nozzle mass flux: G = (2*(H0 - Hb)^0.5 / Vb x 3955.77 (reference 21). Could anyone tell me the basic of this equation? where does it come from? and how to consolidate/prove mathematically.

Thanks,
Eastorca

Attached Files

•  18. Rigorously Size Relief Valves for Supercritical Fluid.pdf   95.32KB   303 downloads

[staffel]

eastorca,
PaoloPemi has provided very good information and a Excel page for solving (numerical integration) the isentropic nozzle and comparing resulyts against other methods in this thread

http://www.cheresour...ng-temperature/

which I would recommend to read,
the basis of the numerical procedure are discussed in API and several other sources,
see the paragraph "Sizing by direct integration of the isentropic nozzle flow calculation" in API,
the method solves the energy balance (isentropic flash) and the mass flux calculation,
in the paper you mentioned the.constant 3955.77 is included for unit conversion,
hoping this helps

Edited by staffel, 12 September 2012 - 03:28 AM.

[eastorca]

Dear staffel,

That's interesting reference. How about my question number 1. Is that reasonable if we assume molecular weight of dense phase (mixture) is unchaged during PSV operating?

Thanks

[marchem]

1) for a mixture above cricondembar (or dew/bubble lines in dense phase) there is no change of phase, hence the molecular weight is constant (feed composition),
please note that within this condition (no latent heat absorbed) final temperatures are much higher (than vaporing fluid)
2) consider the energy balance across the orifice (in = inlet, o = orifice)
hin+1/2*vin^2 = ho+1/2*vo^2
where vo is the speed of sound for choked flow
assuming vin is negligible
hin-ho = 1/2*vo^2

hoping this helps.