Advanced power generation processes are challenged by needs for minimization of CO2 emissions, higher energy efficiency and integration of superior processing technologies. These challenges also require the development of more effective models and strategies for large-scale process optimization. This study considers optimization tools and formulations for process synthesis, which highlight equation oriented modelling environments, nonsmooth modelling via complementarity conditions and fast optimization algorithms. We especially focus on novel formulations for process synthesis that take advantage of these state of the art optimization algorithms. All of these are formally represented as bi-level nonlinear programming (NLP) problems and include Gibbs reactor models, vapour-liquid equilibrium with vanishing and reappearing phases, and simultaneous heat integration and flowsheet optimization. These applications are demonstrated on the process synthesis of a coal oxycombustion process for clean electric power generation.
