Process Integration techniques have been extensively employed in the past 40 years to enhance the management of industrial chemical processes, with Heat Integration becoming one of the chief strategies to foster a more efficient use of energy. However, despite the overwhelming amount of studies in this time frame, some key aspects still remain overlooked, and therefore require a deeper investigation. In particular, scant attention has been placed on how to handle unclassified process streams‒whose thermal identity cannot be defined prior to optimization‒and on how to efficiently deal with the total network area targeting for the assessment of the capital expenses. The present thesis will therefore focus on the development of a comprehensive model suitable for the simultaneous optimizati...[ Read more ]

Process Integration techniques have been extensively employed in the past 40 years to enhance the management of industrial chemical processes, with Heat Integration becoming one of the chief strategies to foster a more efficient use of energy. However, despite the overwhelming amount of studies in this time frame, some key aspects still remain overlooked, and therefore require a deeper investigation. In particular, scant attention has been placed on how to handle unclassified process streams‒whose thermal identity cannot be defined prior to optimization‒and on how to efficiently deal with the total network area targeting for the assessment of the capital expenses. The present thesis will therefore focus on the development of a comprehensive model suitable for the simultaneous optimization with energy integration of the process flowsheet, with the emphasis laid on achieving an economic tradeoff between energy and capital costs. Specifically, building on two already existing formulations for heat integration, a new disjunctive method is developed for the modelling of unclassified streams, and several examples are undertaken to prove its effectiveness. Then, a completely innovative shortcut approach for the area targeting of heat integration problems with variable stream data is going to be discussed. This novel technique avoids the use of disjunctive formulation, greatly reduces the number of binary variables and therefore remarkably shortens the solution time when compared to conventional rigorous methods. Finally, the overall model is successfully and efficiently extended to the optimization of MHEXs, which are well-known to be of particular significance for the energy recovery in several cryogenic applications. Numerous relevant case studies are proposed all throughout this thesis, with the results clearly demonstrating the advantages of the proposed framework, as the number of constraints, variables and binary variables are all significantly reduced, thus greatly benefitting the models' numerical performance and computational efficiency.