Very high pressure gas relief - supercritical - Z>2 1

[gelsi]

Hi,
could you help me with the approch to a PSV calculation?
Follow some conditions:
fluid ethylene
critical pressure 50.6 bar
relief pressure = 3000 bar
relief T = 80°C
density about 602 kg/m3
Z by density ~5

It has the density of a liquid but is a supercritical gas, which calculation method can i use?
Tx

 

[don1980]

Supercritical fluids are compressible, like a super-high density vapor. Thus the vapor equation is used.

 

[georgeverghese]

Since T=80degC > Tc of 19degC for C2=, this can only be dense phase gas. BTW, Z can only be a max of 1.0; suspect you mean Z= 1/5 = 0.2

 

[gelsi]

don1980, we are agree, but when Z is outside of range 0.8 -:- 1.1 it is raccumanded not to use vapor equation... do you have experiance about thi kind of item?

georgeverghese, follows my Z calc from density, value of Z is high because of very high pressure:
ro=m/V kg/m3 602
t °C 80
P bar 3000
PM kg/kmol 28,05

R m3 Pa K−1 kmol−1 8314,472
m3 bara K−1 kmol−1 0,08314472
R' m3 bara K−1 kg−1 0,002964161
T K 353,15

Z - 4,8

 

[apetri]

Gelsi,
working above cricoP you should use the integral method but there are also other methods producing similar results, see this thread to compare different methods...

https://www.cheresources.com/invision/topic/16104-psv-discharging-temperature/

just a question, do you really need to calculate the area with a Pin = 3000 Bar ?

 

[gelsi]

operative pressure 2450 bar
set 3000

 

[apetri]

you did not specify the application and the flow...

at 80 C 3000 Bar.a predicted density for ethylene is about 600 Kg/M3
( Prode MBWR 600.3 Prode PRX 591.05 Nist Refprop 601.99 )
For NIST Refprop error should be < 0.5% , for Prode MBWR < 1.0%

if you model the PSV with the suggested integral procedure the estimated (outlet) temperature is about -100 C (isentropic + adiabatic solution)
could be a problem for high flows...

 

[georgeverghese]

Okay, due to the very high pressures here, z could exceed 1.0, I guess. Perry says z=1.2 for the highest listed P=500bar at say T=27degC for ethylene.

Suggest getting specialist assistance for this RV sizing, since it is possible that C2= may isentropically condense at some intermediate pressure which is higher than choke pressure, before finally revaporising at the RV exit.

 

[georgeverghese]

Check if there is a PSV sizing routine on some process simulator that can run this PSV sizing using the integral method. The reference literature in this routine should also state that it can handle supercritical pressure relief cases.
One way to check the accuracy of the z prediction for C2= is to see what the process simulator says for the Perry listed condition of 500bar, 27degC - no where near 3000barg, but better than not checking at all.

 

[rense]

I have used the integral method (HEM model) in Prode for pressures up to 900-1000 Bar, in principle there are no limits but I never found such high values (3000 Bar) as reported for this application..

 

[gelsi]

apetri,
[ul]
[li]discharge flowrate is ~35000 kg/h[/li]
[li]application is reactor effluent (PE plant)[/li]
[li]case is closed outlet[/li]
[/ul]

 

[gelsi]

errata corrige
effluent
feed

 

[apetri]

gelsi,
in my opinion the main problem here is not to solve with the rigorous integral method (as said in previous posts, the suggested thermodynamic library includes the required procedure and thermo models for that) but to make sure that design conditions / calculated values produces the correct PSV selection,
this may be the work for some specialist able to consider all the aspects of the problem... (process, safety etc.)