| Process Intensification (PI) has developed as a novel and revolutionary design approach that provides advantages in size and cost reduction and a closer approach to optimal processing conditions. Despite the numerous advantages PI may provide, the implementation of this concept in process designs has generally been limited to the application of highly efficient equipment developments. The fundamental thinking underpinning PI involves a multiscale process design that consists of designing the process (equipment) to perform at the appropriate lengthscales and timescales to deliver substantial improvements. This procedure invariably involves an understanding of process phenomena and their interrelationships without the spatial constraints of conventional unit operation models. This thesis presents a framework based in functional and systems approaches used to consistently define phenomena-based building blocks. A hybrid qualitative-quantitative methodology is proposed to represent and analyse available knowledge about chemical processes to support the generation of intensive design strategies. Models are based on the physicochemical phenomena arranged into abstract (i.e. equipment-independent) functional, structural and behavioural modules. The qualitative component includes the use of a graphical topology to qualitatively represent and model the process in an equipment-independent approach. The methodology includes the mapping of phenomenological representations into mathematical models. Causal graphs are Introduced to allow the process designer to identify the relationships between variables relevant to the process. The implementation of the phenomena-based models in equation-based software is also used to quantify the net effects of selected design variables under the prevailing process topology and conditions. Process models are hierarchically composed from custom models based on the physicochemical phenomena consistently implemented using the object-oriented features of Aspen Custom Modeler. The potential for more efficient options is increased as the design process is not constrained early to particular unit operations, thus revealing features for intensification and making explicit the associated degrees of freedom. This approach provides increased insights into the processes being modelled and an appreciation of the cause and effect, expanding on existing heuristic rules from the early development of PI. It encourages the implementation of PI principles regarding both the enhancement and the coupling of phenomena via the generation of alternatives derived from model-based reasoning. Within the modelling context, combined advantages are gained in model flexibility, customisation, reusability, complexity and cost. The use of these building blocks at a lower level of aggregation should enable construction and customisation of a large number of processes. In this way, evaluation of novel unconventional and highly efficient processes may be performed with an acceptable level of complexity and cost. Equally, these models are still suitable for the design of traditional plants involving conventional operations. The application of the methodology is illustrated with a case study on aromatic nitration. |
