separator calculation (Ki ratios)

[Malimedo]

Hi, I'll be very helpfull to see someones separator calculation in order to obtain liquid and gas composition at desired separator pressure and temperature.
I'm getting sum of mole fractions of gas phase=1, but for liquid fraction I get numbers different than 1.
I attached my example.
Also I used Kessler-Lee correlation for C7+ critical temperature and pressure, Edmister correlation for C7+ accentric factor, and Wilsons correlation for Ki.

 

[MortenA]

It look like 1 to me?

Best regards

Morten

 

[Malimedo]

sum of xi is 0.8239
I'm not sure if it's correct.

 

[MortenA]

Now i see it - whats the vapour fraction?

If its 1 then it may be rounding?

Im assuming that its not something like "dry basis" where in reality water is present?

Best regards

Morten

 

[Malimedo]

I appologize for not explaining the symbols.
xi is the composition of the liquid phase
yi is the composition of the vapour phase

zi is composition before separation
As far as I know, this should be satisfied: sum zi = sum yi = sum xi
nL is quantity of the liquid phase, ng is quantity of the gas phase. I guessed nL by newton-rhapson iterration method in order to satisfy sum zi=1, calculated from nL and corresponding Ki ratio.
omega is accentric factor for the component - from book, and omega for C7+ fraction is calculated from correlations (MC7+ -> pseudocritical T and p -> omega C7+)
nL + ng should equal one.

 

[MortenA]

As i suspected the nL is very low.

I think it may be a round error that gets accepted because the overall error is sufficient low.

If this does not make you "happy" then you must call the support line

Best regards

Morten

 

[Malimedo]

Thanks. So you think such big error occurs with condensates that vaporize very quickly?

 

[MortenA]

No - I would assume that your simulator is steady state - and its always at equilibrium.

Im talking about a numerical error that is accepted (by the software) because the liquid part is small compared to the overall phase.

Best regards

Morten

 

[Malimedo]

I don't use simulator, unfortunatelly my antivirus detects a virus in registration tool ?.

I just made spreadsheet calculation and then guessed nL using solver, ie. I guessed nL until sum of zi was zi=1
I think this method should work, but anyway - the error
repeatedly occurs (I have only one analysis, so I just change separator pressure)

 

[MortenA]

well my guess is the only one i will still cling to (except then maybe an error en your homemade sheet): The overall error is acceptabel since your liquid fraction is so small.

Best regards

Morten