Chemical absorption and desorption processes are two fundamental operations in the process industry. Due to the rate-controlled nature of these processes, classical equilibrium stage models are usually inadequate for describing the behaviour of chemical absorption and desorption processes. A more effective modelling method is the non-equilibrium rate-based approach, which considers the effects of the various driving forces across the vapour-liquid interface. In this thesis, a new non-equilibrium rate-based model for chemical absorption and desorption is developed and applied to the hot potassium carbonate process CO₂ Removal Trains at the Santos Moomba Processing Facility. The rate-based process models incorporate rigorous thermodynamic and mass transfer relations for the system and detailed hydrodynamic calculations for the column internals. The enhancement factor approach was used to represent the effects of the chemical reactions. The non-equilibrium rate-based CO₂ Removal Train process models were implemented in the Aspen Custom Modeler® simulation environment, which enabled rigorous thermodynamic and physical property calculations via the Aspen Properties® software. Literature data were used to determine the parameters for the Aspen Properties® property models and to develop empirical correlations when the default Aspen Properties® models were inadequate. Preliminary simulations indicated the need for adjustments to the absorber column models, and a sensitivity analysis identified the effective interfacial area as a suitable model parameter for adjustment. Following the application of adjustment factors to the absorber column models, the CO₂ Removal Train process models were successfully validated against steady-state plant data. The success of the Aspen Custom Modeler® process models demonstrated the suitability of the non-equilibrium rate-based approach for modelling the hot potassium carbonate process. Unfortunately, the hot potassium carbonate process could not be modelled as such in HYSYS®, Santos’s preferred simulation environment, due to the absence of electrolyte components and property models and the limitations of the HYSYS® column operations in accommodating chemical reactions and non-equilibrium column behaviour. While importation of the Aspen Custom Modeler® process models into HYSYS® was possible, it was considered impractical due to the significant associated computation time. To overcome this problem, a novel approach involving the HYSYS® column stage efficiencies and hypothetical HYSYS® components was developed. Stage efficiency correlations, relating various operating parameters to the column performance, were derived from parametric studies performed in Aspen Custom Modeler®. Preliminary simulations indicated that the efficiency correlations were only necessary for the absorber columns; the regenerator columns were adequately represented by the default equilibrium stage models. Hypothetical components were created for the hot potassium carbonate system and the standard Peng-Robinson property package model in HYSYS® was modified to include tabular physical property models to accommodate the hot potassium carbonate system. Relevant model parameters were determined from literature data. As for the Aspen Custom Modeler® process models, the HYSYS® CO₂ Removal Train process models were successfully validated against steady-state plant data. To demonstrate a potential application of the HYSYS® process models, dynamic simulations of the two most dissimilarly configured trains, CO₂ Removal Trains #1 and #7, were performed. Simple first-order plus dead time (FOPDT) process transfer function models, relating the key process variables, were derived to develop a diagonal control structure for each CO₂ Removal Train. The FOPDT model is the standard process engineering approximation to higher order systems, and it effectively described most of the process response curves for the two CO₂ Removal Trains. Although a few response curves were distinctly underdamped, the quality of the validating data for the CO₂ Removal Trains did not justify the use of more complex models than the FOPDT model. While diagonal control structures are a well established form of control for multivariable systems, their application to the hot potassium carbonate process has not been documented in literature. Using a number of controllability analysis methods, the two CO₂ Removal Trains were found to share the same optimal diagonal control structure, which suggested that the identified control scheme was independent of the CO₂ Removal Train configurations. The optimal diagonal control structure was tested in dynamic simulations using the MATLAB® numerical computing environment and was found to provide effective control. This finding confirmed the results of the controllability analyses and demonstrated how the HYSYS® process model could be used to facilitate the development of a control strategy for the Moomba CO₂ Removal Trains. In conclusion, this work addressed the development of a new non-equilibrium rate-based model for the hot potassium carbonate process and its application to the Moomba CO₂ Removal Trains. Further work is recommended to extend the model validity over a wider range of operating conditions and to expand the dynamic HYSYS® simulations to incorporate the diagonal control structures and/or more complex control schemes. / http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1350259 / Thesis (Ph.D.) - University of Adelaide, School of Chemical Engineering, 2009
| Identifer | oai:union.ndltd.org:ADTP/282082 |
| Date | January 2009 |
| Creators | Ooi, Su Ming Pamela |
| Source Sets | Australiasian Digital Theses Program |
| Detected Language | English |
Page generated in 0.0029 seconds