| Mass and heat exchange network (HEN/MEN) technologies propose to generate process networks that reduce the cost and increase the efficiency of mass and heat transfer operations. The goal of this work is to address several outstanding issues in this field. MEN design where more than one transferrable component is present has long been known to be of an inherently different dimensionality than single component design. Both the different phenomena present, and their implications for design have been poorly understood, hampering the development of new methods. A single mass exchange unit operation, with the goal of minimum utility cost (MUC), is rigorously examined and leads to the discovery and quantification of a new phenomena. Another outstanding issue was the inclusion of sets of streams that are allowed to mix in HEN/MEN design. Previous technologies were not able to use the possibility of mixing, and yielded non-optimal designs. A linear programming formulation for this problem is developed and proven to generate optimal designs. The insights gained in this development help lead to a new framework for process design, the Infinite DimEnsionAl State Space (IDEAS). This new framework provides a generalized process description, allowing all possible configurations of equipment, while generating a linear constraint set for optimization. The power of IDEAS is shown through two example applications. First, MUC MEN design with multiple transferring components is approached. The resulting networks are shown to be cost optimal. Second, total annualized cost (TAC) for HEN design is considered. This problem is shown to be NP-hard, thus removing the possibility of a convex optimization formulation. By use of IDEAS, the non-convexity is limited to the objective function. This allows the successful solution to a larger class of TAC HEN design problems than previously possible. Two examples demonstrate the capabilities of this new method. |
