5 replies to this topic

[Neelakantan]

hi friends,
i have a peculiar problem while using hysys to simulate a TEG regeneration tower. within the column environment the still vent is 104 deg C and all vapour (vapour fraction=1.000). but in the main pfd the same stream shows 127 deg C and 0.9812 vap fraction.

the compositon is same (mole fractions, within the column) and outside. the energy and mol flow same.

i use to get a small difference of 0.1 ~ 0.5 deg C which can be further tuned by the "convergence" parameters.

i thought this is because two phase detection has been disabled in the "advanced" settings. even after turning it on, it is not detecting water in liquid phase in top tray.

any ideas how this happens and how to remove this "error"?

regards
neelakantan

[joerd]

This sometimes happens because of the different flash algorithms in the column environment. Try changing the transfer basis to P-T. This can be found on the Flowsheet tab of the column.

[Zauberberg]

By default, pressure-enthalpy (P-H) transition is used for transfer between the main flowsheet and other sub-flowsheets. You can switch it manually to P-T, as suggested by Joerd, or any other method you find to be more convenient. The difference usually occurs if you have components mapping from one to another fluid package/flowsheet, resulting in different enthalpy content(s) at a given temperature.

[Neelakantan]

Hi J and Z,
thanks for the inputs; yes i understood that P-H are same (because energy flow values were same) and hence i did some fiddling and selected P-T.

but as the basis is same in both environment, i was surprised at this. I finally did locate "whole liquid" detection (or some words to this effect in the solver menu 2/3 phase detction) instead of Liq2 detection which immediately prompted "two phases detected" and convergence failed.

i did adjustments in the reflux rate and the feed tray location and got the still column to converge. with this the P-T and P-H options both give almost same values. (a differecne of 0.6 deg exists, which i believe is within the accuracy level)

regards
neelakantan

[smalawi]

Hi,

This issue happens to me as well, sometimes if the column is complex internally (have exchangers, vessels, pumps...etc) then you get a consistency error as well

The guys gave you some pointers and I can add that it could be you mass and heat balance spec, try to tighten them more use 1e-6 or smaller and see the results. its located in the solver tab

Some times with some packages the stage eff is not 100% then your stream once p-h falshed out of the column to the main PFD will be mixed phase as well. This can not be solved by the mass & heat spec.

cheers,

sm

[Neelakantan]

hi
the problem may be with the peculiarity of the glycol -HC-water system. i used hysys 3.2 with Peng-Robinson-LK package. (as the client used hysys3.2)

when the same case was run in 2006 using the "sepecialised" glycol package, the vapour load and condenser load are different. Also we can use the glycol package to the absorber/exchangers/regen column only. if i included glycol package as default for the inlet scrubber (the dummy water saturator for the feed gas) i am getting lot of HC going along with water. when i use PR for the scrubber it is only water that is removed.

and when the regen column is re-run after adjusting the reflux-ratio specs, i got the top temperature as 97 deg C! i am back to using the PR package.

i was trying the use the existing condenser duty in the run; with the given composition and flow rate, the condenser duty need be increased (from 19 kw to 30 kw). then the column converges easily with about 0.14 as the reflux ratio and top temperature of 103 deg C which appears as the "real world" number as against the 97 from 2006 version.

i have seen in some forums people saying winsim is a better module for teg regeneration. what are your experiences?
regrds
neelakantan