| Main Page |  |
Publications |
Optimal operations of extractive distillation for regular and middle-vessel batch columns are presented based on a profit function. Detailed models are used for the rigorous dynamic optimization considering all operational decision variables, including reflux ratio, solvent feed rate, heat duties, and possible product withdrawals during the process. Optimal feed distribution and stream configuration at the middle section of the middle-vessel column are investigated. Separation of a minimum boiling azeotropic binary mixture (acetone and methanol using water as solvent) involving different separation duties and feed compositions is presented as a case study. The performance of the middle-vessel column is significantly influenced by the middle-section stream configuration, with the best profit when the stream configuration is allowed to vary during the operation. The optimal operating policy for the middle-vessel column involved the feed being charged mainly to the reboiler still with low holdup in the middle vessel during the operation.
The multivessel system is a novel batch distillation configuration that offers improved separation performance compared to the conventional single-column batch rectifier. The optimal design of this system based on an economics performance index is solved for the first time here by adopting an evolutionary adaptive search technique. The utilization of an overall profitability index, and simultaneous consideration of key design and operation decision variables, allow the true optimum of the system to be obtained. The optimal system configuration is dependent on the separation duty and can contribute significantly to achieving a higher separation performance. The combined energy efficiency and production rate of the multivessel system is found to be greater than that of the regular column, and the benefit is more prominent when separating mixtures with more components. Further theoretical case studies are also presented to highlight the effect of feed composition, relative volatility, and product specification on the optimal column section configuration and feed distribution of the multivessel system.
This work presents a robust method for the integrated design and operation of batch distillation whereby optimal column sizing, process flexibility and operating policies are obtained simultaneously based on the complex economic trade-offs between capital investment, production revenue and utility costs. The proposed stochastic framework, which utilises a Genetic Algorithm and a penalty function strategy, is found to be successful in obtaining profitable and feasible column designs for many design scenarios including binary and multicomponent mixtures, single duty and multipurpose columns, as well as for regular and complex column configurations. The method can also be used with column models of different complexity. Given a set of design specifications and separation requirements, the optimal number of stages, reboiler duty, reflux profiles, product recoveries, time interval of each distillation tasks, process allocation and number of batches can be obtained. Several design case studies are presented and a comparison of optimal designs for various design scenarios, such as different p roduction time, capital costs, process allocation and mixture characteristics, are discussed.
A methodology for simultaneous determination of optimal batch distillation configuration, design and operation is presented. The configuration design methodology utilises a mixed integer dynamic optimisation (MIDO) formulation approach where the optimal batch distillation system is obtained automatically based on its maximum overall profitability for a given separation duty. Using rigorous models, the MIDO problem is solved using a practical stochastic solution approach of genetic algorithm and penalty function. The feasibility of this methodology is demonstrated for both binary and multicomponent separation scenarios. In the binary separation case study, the effect of feed composition for different binary mixtures on the optimal configurations, i.e. regular versus inverted columns, is investigated and discussed. The regular column was found to be more profitable for feeds with a high fraction of the light component whilst the inverted column is optimal for heavier feeds. The optimality of a particular configuration over another is, however, case study specific, depending on, for example, how easy the mixture is to separate. In the multicomponent separation case study, the results obtained highlight the superiority of the multivessel configuration over the regular and inverted configurations.
For more info please contact me.
|