Biller, Nicholas Charles Trinder; (2003) Modelling and control of reactive distillation processes. Doctoral thesis (Ph.D), UCL (University College London).

Text Modelling_and_control_of_react.pdf Download (6MB)

Abstract

Reactive distillation has been applied successfully in industry where large capital and energy savings have been made through the integration of reaction and distillation into one system. Operating in batch mode, in either tray or packed columns, offers the flexibility required by pharmaceutical and fine chemical industries for producing low volume/high value products with varying specifications. However, regular packed or tray columns may not be suitable for high vacuum operations due to the pressure drop across the column section and short path distillation may be more applicable. The objective of this thesis is to investigate the control of reactive distillation in batch columns, tray and packed, and in short path columns. In order to study control fully, it is necessary to develop rigorous dynamic models that accurately capture the process behaviour. The higher the degree of rigour, the more accurately the process conditions and dynamics are captured. However, more rigorous models are more computationally expensive to implement and can be prone to numerical errors, introduced for instance during linearisation. Therefore, in this thesis, the degree of modelling rigour required for both simulation and control purposes is explored in detail for tray and packed batch columns and short-path columns. For batch tray columns, it is demonstrated that to accurately capture the change in process conditions, it is necessary to model pressure dynamics and employ a dynamic energy balance. For packed columns, distributed rate based modelling is compared to lumped equilibrium modelling and it is found that due to the varying conditions within the packing, the efficiency changes, resulting in mismatch between the two methods. The short-path distillation column which has hitherto only been modelled at steady-state, is modelled using a dynamic rate based model, essential for investigating control. Having developed the dynamic models, the control and controllability of these reactive distillation processes are examined. General control properties of reactive batch distillation are discussed and methods are presented for applying linear controllability tools to these non-linear process models. The linear models are then employed to demonstrate the implications for control when adopting one of the three processes.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item