Energy integration, motivated by the high cost of energy and the rapidly diminishing sources of energy, is a rule rather than an exception in modern process industries. Energy integrated networks offer significant cost benefits, however are quite difficult to operate and control, especially in the context of transitions between different steady states, owing to the strong interactions among the individual units and slowly evolving network-level dynamics. Controlling effectively these integrated plants is a critical link to the economic viability, and the energy and environmental sustainability of the chemical and energy supply chains.;In this thesis, a generic analysis framework for networks with tight energy integration is developed. Initially, two classes of simple networks characterized by the presence of significant energy flows (either arising from energy recycling or of external origin) are identified and shown to exhibit similar (though not identical) two-time scale dynamic behavior. For each case, a model reduction framework based on singular perturbations is developed and subsequently, a hierarchical control strategy, with each tier addressing control/optimization objectives in each time scale, is formulated. The effectiveness of the model reduction and controller design framework is illustrated via simulations of representative systems from the process and energy industries.;The latter part of the thesis analyzes complex networks, which are shown to be composed of interconnections of the simple prototypes identified earlier. A graph-theoretic approach is incorporated in this case, which systematically accounts for these connections and algorithms are developed to parallel the singular perturbation-based model reduction procedure. These algorithms are generic, efficient and scalable to large scale networks. A class of networks with material integration, featuring a large recycle of solvent is also analyzed. These networks are shown to exhibit similar dynamic characteristics and be amenable to a hierarchical control strategy.;The results of this thesis show how tight integration can actually facilitate control and operation of the integrated processes, and lay the foundation for the development of computational tools that will enable dynamic analysis and controller design for complex integrated networks in an automated and scalable fashion.
展开▼
