(372f) Cost-Driven Heat Exchanger Network Synthesis Covering Practical Implementation Issues

Heat exchanger network synthesis is widely investigated in research. It has been a vital topic for decades by now. Implementing heat integration strategies is one main approach to significantly reduce the use of primary energy carriers and thus increasing the energy efficiency and cost effectiveness of a process. Especially against the background of global warming and increasing global competition heat exchanger networks (HENs) positively affect the economic potential of production processes.

The synthesis of HEN is often carried out using simple cost functions which do not represent practical implementation issues like piping, pumping or different fluid properties of the involved streams, different types of heat exchangers and quality ratings of materials. Therefore, the synthesized HEN will likely be a suboptimal solution for the application case. A further downside of an increase in process integration is the higher extend of thermodynamic interaction between the process streams coupled via heat exchangers. As a consequence a disturbance will not only cause local effects, but can propagate throughout an entire network and thus a whole process. The important practical implementation issue of process control is of major importance. Normally, traditional approaches do not incorporate such information. These approaches are mainly dominated by steady-state evaluation. As a consequence, there is a need to assess both, the costs representing practical implementation as well as the impact of heat integration on the process stability and controllability in order to generate optimal HENs.

The solution approach presented covers the complex constraints of practical implementation of HEN solutions as well as steady-state and dynamic assessment of the obtained structures. The utilized cost functions framework is based on the incorporation of different cost function types for each match individually proposed by Rathjens and Fieg [1]. These individual cost factors are used to account for fluid properties, resulting engineering costs and plant layout based factors. The associated MINLP is then solved using a genetic algorithm. Due to the stochastic nature of a genetic algorithm several near cost-optimal solutions are obtained. These solutions are further investigated using different characteristic numbers derived from a linear steady-state model. In order to fulfill the demand of dynamic considerations, simplified mathematical solution models are used. These models are integrated in a modular tool to represent the dynamic behavior as a matter of principle. The control of the HEN is realized by decentralized feedback control. It is accomplished through the adjustment of bypasses on the process to process heat exchangers and the flowrates of the necessary utilities. The combination of different HEN structures and dynamic considerations guarantees much better insight in the production process and thus greatly supports the decision-making [2].

Structurally different solutions for HENs are obtained using the shown methodology. The implementation of detailed cost functions for the different matches can lead to practically more relevant solutions. Obviously, cost-optimal solutions are not necessarily the solutions with the best controllability characteristics and applicability in practice. Therefore, the inclusion of dynamic considerations can be very beneficial when searching for the best solution for practical implementation.

[1] Rathjens, M.; Fieg, G.; Cost-Optimal Heat Exchanger Network Synthesis Based on a Flexible Cost Functions Framework. Energies 2019, 12, 784.

[2] Rathjens, M.; Bohnenstädt, T.; Fieg, G.; Engel, O.; Synthesis of Heat Exchanger Networks Taking into Account Cost and Dynamic Considerations. Procedia Engineering 2016, 157, 341–348.