HC Gas Properties above critical point 2

[jeyx]

Dear All,

can you please tell me

1) Why in Hysys, if we have HC gas well above its critical point (by temp) we still can turn it into the liquid phase by increasing presure?

2) How do you understand the critical point and especially the critical pressure/temperature?

These seem to be simple issues which are often difficult to understand/explain

I asked my colleague - the principal process engineer and he could not give me the clear answer

thank you,
Jeyx

 

[25362]

If you are speaking of hydrocarbon mixtures visit thread798-70867.

 

[joerd]

There is also a numerical issue here - HYSYS decides at some point that the dense phase is more like a liquid than like a vapor. From their solutions database I quote:

HYSYS reports a vapor fraction for a stream under supercritical conditions as zero or one. Theoretically, this value does not have a specific physical meaning since there is no distinct liquid and vapor phases in a supercritical region. However, for the purpose of flash calculations, HYSYS will determine the vapor fraction of a dense-phase stream based on the following criteria:

If the compressibility factor (Z) is less than 0.3 and the isothermal compressibility factor (beta) is less than 0.1/P, a vapor fraction of zero (Vf=0) is assigned to the stream. Otherwise, the vapor fraction would be one (Vf=1). For a supercritical fluid stream, if its vapor fraction equals one (Vf=1), vapor correlations are used for the physical property calculations, otherwise, liquid correlations would be used.

The bottom line is that the vapor fraction for a dense-phase stream is not an indication of its phase state (liquid/vapor) anymore, but only an indication of which correlations (liquid or vapor) are used for the stream. For the same reason, any attempt to do bubble or dew point calculations by specifying Vf=0 or Vf=1 should be avoided.

Cheers,
Joerd

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[UmeshMathur]

(1) PURE COMPONENTS
A pure component above its critical temperature can NEVER be liquified, no matter what the pressure.

(2) MULTICOMPONENT MIXTURES
To understand the issue you have raised, it is instructive first to study the concept of the phase envelope. This is a curve showing the locus of points on the two phase boundary on a P-T plot.

For many wide-boiling mixtures, the critical point is located BELOW the so-called cricondenbar pressure. In fact, it is this phenomenon that gives rise to the problem of "retrograde condensation", where reducing pressure for a supercritical mixture at constant temperature (starting above the 2-phase boundary) can result in liquid separation at temperatures above the true critical temperature of the mixture. This can, for example, have important safety implications for gas pipeline networks.

For a decent explanation of the relevant theory, please consult Professor S. M. Walas’ excellent book, "Phase Equilibria in Chemical Engineering", (Butterworth, 1985, pp 94-96). Walas also has excellent discussions of the critical state and critical properties at pages 87-94. This book contains an encyclopedic coverage of the subject and has numerous diagrams to aid comprehension. Two other good books on this subject are:
Professor T.F. Daubert, “Chemical Engineering Thermodynamics” (McGraw-Hill, 1985), and
Professors J.M. Smith, H.C. van Ness, and M.M. Abbott, “Chemical Engineering Thermodynamics”, (6th Edition, McGraw-Hill, 2001).

By the way, joerd's post appears to me to refer only to how Hysys handles computational issues related to flash calculations for a supercritical fluid, and not to the underlying phenomena raised by your original question.

 

[Samiran]

It is extremely important to understand that thermo properties predicted by Hysis may not be accurate at that region. So if you are designing a column operating at pressures close to critical be very careful. The liquid in the trays may not be liquids any more at the top and may run into offspec product. This is where heritage data comes very handy.

Samiran