1 reply to this topic

[DivineFist001]

Simulation background:

I have to separate mixture of components via distillation, I have already performed Radfrac distillation and got desired results. 

Property method used: Peng-Robinson.

 

To optimize it further and design as per requirement, i have to simulate the given problem in BatchDistillation method for Aspen Plus.

 

 

Issue Faced :

While Performing Batch distillation simulation, I have gotten below severe error:

"ERROR T-LOOP NOT CONVERGED IN 37 ITERATIONS, FLASH CALCULATION CONVERGED BUT RESULTS MIGHT NOT BE RELIABLE".

Property method used: Peng-Robison

Flash Convergence calculation method used: Gibbs

Flash convergence iterations : 1000

Batch Distillation iterations : 1000

The only tear-stream present in the simulation is Reflux.

Initial Condition : Total reflux

Condenser type : Total

 

NRTL method was initially tried, but was not able to predict known binary azeotrope of some selected components. 

[SilverShaded]

Dont know why it didn't converge but as far as NRTL goes, both UNIQUAC/NRTL methods should be tuned to real data before use.  Doing that for multi-component mixtures is not simple.