Datenbestand vom 14. Februar 2018
Tel: 089 / 66060798
Mo - Fr, 9 - 12 Uhr
Fax: 089 / 66060799
aktualisiert am 14. Februar 2018
978-3-8439-3016-1, Reihe Technische Chemie
Simultaneous Optimization of Scheduling, Plant and Process Design of Sequential Multi-Purpose Batch Plants
149 Seiten, Dissertation Technische Universität Dortmund (2016), Softcover, A5
The design of multi-purpose batch plants is a challenging task, because the number of degrees of freedom for optimization is high. Important optimization variables are the process design in form of operating conditions, the plant design in form of size and number of equipment items and the scheduling in form of the batch number and the sequencing of the products. Since all factors mentioned interact, a simultaneous consideration would be beneficial in order to reduce investment and processing costs. But due to the problem size, in most publications, only a part of the factors is regarded simultaneously and the optimization potential is not fully exhausted. Although some holistic approaches exist, they were never applied for complex, realistic problems.
In this work, the task of the simultaneous optimization of all factors is tackled for a new, comprehensive case study: A biochemical plant for protein production. Aspects important for the scheduling of such plants like changeover times and semicontinuous operations are considered as well as comprehensive models of the unit operations to describe the influence of the design parameters on processing times, the mass balance and the conditions of the process streams.
The resulting MINLP model is large and difficult to survey by the modeler. Due to this a model with a clear structure that builds up on smaller, physically meaningful submodels is used for optimization. A MILP and a NLP submodel are used. The MILP submodel is used to obtain a feasible schedule. The NLP submodel consists of comprehensive models of the unit operations that can be used to optimize operating parameters and equipment dimensions.
Optimization runs are performed for three different cases with different product demands. The qualities of the solutions obtained are analyzed in detail, with the result that good and reasonable solutions are obtained in an adequate time.