Phase equilibrium dominant parameter matrix of multicomponent distillation trains

Plant−Model Mismatch Analysis in Deethyleniser Simulation. 2. Application to Data-Model Selection and Tuning

“…Application of the criteria proposed in the work of Gros et al, 18 viz., a strict key component specification and low relative volatility, leads to a pinch zone in the rectifying section of the deethyleniser. The key thermodynamic data-model parameters for a high-purity ethylene column operating very much like a ethylene/ ethane splitter in the top of the column at least are, according to Gros et al, 18 those required for accurate prediction of the dominant separation variables. These are the equilibrium ratios (K i ) y i /x i ) of the key component being removed in each sensitive zone or column section.…”

“…It is envisaged that further research could be carried out with the objective of determining whether independent k ij (ethylene/ethane) tuning in both rectifying and stripping sections according to the criteria given by Gros et al 18 above, concurrent with the tuning of independent rectifying and stripping section efficiencies, is warranted.…”

“…Although the data-model parameters according to Gros et al should ideally be fitted to each sensitive zone temperature and at low concentrations of the key component, the approach in this work has been to fit k ij (ethylene/ethane) over the whole range of data available in work by Barclay et al This value is then used in the simulation model tuning procedure using effective average equilibrium tray efficiencies as semiempirical fitting parameters that adjust for any inadequacies in the data-model tuning procedure.…”