MIND : Optimization method for industrial energy systems

Abstract: The MIND optimization method is a tool for life cycle cost minimization of a flexible range of industrial energy systems. It can be used in analyses of energy systems in response to changes within the system, changes of the boundary conditions and synthesis of energy systems. In analysing industrial energy systems there are a variety of issues to consider in finding the best way of production. Both the energy supply part and the energy demand part is of great significance. The structure of the energy supply part is often decided on economic terms such as fixed costs, fuel prices or energy tariffs but also on availability. The energy demand is depending on the technology employed and the layout of the system. A change to new technology or recondition of old equipment may as well as alterations in the production schedule give considerable overall savings. In order to comprise all aspects in the analysis it is essential that the optimization method can handle: all occurring flows in the energy system, time-dependent components and conditions, non-linearities. A change regarding the production schedule, kind of energy or renewal of process equipment may cause a change of material and energy flows as well as a change of the interaction between them. Since industry as a rule has a production goal to fulfil it is necessary to represent both material and energy flows in the calculations. Time dependency for components indicates that process equipment must be represented in a way that allows different process routes to be chosen within the industrial system. Changes in boundary conditions, such as varying energy rates or climatic conditions, will also have to be represented. This implies that the system has to be represented with a proper time division. It is also necessary to let flows pass between time steps to be able to consider storage of both material and energy.  Optimization of industrial energy systems at the component level involves non-linear relationships, such as energy demand functions and investment cost functions. It is important to use the proper level of accuracy in the representation of equipment units. If non-linear relationships are not included there may be considerable errors involved. The accuracy of representation must be chosen for each industrial system to be optimized. These demands can be met in optimization with mixed integer linear programming. Non-linear relationships can be approximated with step functions and piecewise linear segments giving the opportunity to optimize all levels of energy systems. The objective of the optimization is to minimize the life cycle cost of the studied energy system. The life cycle cost includes both fixed and variable costs. Two applications are presented to show the flexibility of the MIND method, heat treating processes in the engineering industry and milk processing in a dairy.