(117c) Batch Distillation Robust Design and Operation in the Face of Feed Variability and Thermodynamic Uncertainties Using Stochastic Modeling in Multibatchds™

Separations using batch distillation often encounter the problem of feed variability. One of the significant problems in optimizing or simulating batch distillation operation for a particular product specification in specialty chemical industries is that the number of components whose thermodynamic properties are known is less than the total number of chemicals in the mixture. In pharmaceutical industries, new drugs bring in new components whose thermodynamic data is restricted to the structure of the molecule. In such conditions, stochastic modeling can be used to study the effect of these uncertainties on product specification and find robust design and operating conditions.

The stochastic modeling involves:

1. Specifying the uncertainties in key input parameters in terms of probability distributions;
2. Establishing the correlation structure of any interdependent parameters;
3. Sampling the distributions of the selected parameters iteratively;
4. Propagating the effects of uncertainties through batch distillation simulation model; and
5. Applying statistical techniques to analyze the effects of uncertainties on the product yield and specification.

Multibatch distillation systems (MultiBatchDSâ„¢) is a software package that captures the essence of these multiple configurations, multiple operating modes, and various fractions of batch distillation by providing the capability to simulate, design, optimize, and control such columns with multiple levels of models. This paper presents two real-world case studies of feed variability and/or thermodynamic uncertainties in batch distillation using the novel stochastic modeling framework in MultiBatchDS.â„¢