Experimental design methods and flowsheet synthesis of energy systems

Abstract This work presents a discussion of how to utilise well known methods from the field of experimental design in the flowsheet synthesis of energy systems. The work is based on an earlier work, where a methodology for improving large scale energy systems using a combination of simulation, experimental design and mathematical programming was presented. The methodology is suitable for synthesis of large scale and complex problems with few degrees of freedom. A conceptual bio-fuel indirectly fired microturbine is used to illustrate how a simulation model and experimental design can be used to build an optimisation model of an energy system, and how this model performs compared to a model, where all the units are modelled in detail. The new methodology has a good potential for reducing the optimisation problem and subsequently for solving more complex problems than the traditional methods are able to solve. For the example problem presented in this work, the developed model performs at least as well as the MINLP model with all units modelled in detail. However, as the degrees of freedom of the system increase, the value of the new methodology decreases, since more computational effort is needed to obtain a representative regression model.