A Multi-Period MINLP Model for district heating networks design considering production systems architecture optimization
Résumé
The design and optimization of district heating networks is a rather complex task; therefore most of the researchers have been focusing on simulating models allowing performing parametric optimization for proposed architectures. Energy integration techniques combined with thermodynamic second law analysis are a way to perform systematic architecture and parametric optimization of district heating design.This paper proposes a multi-period Mixed Integer Non Linear Programming (MINLP) model for the optimal design of a district heating network. This approach allows considering varying energy demands and supplies, and making discrete decisions such as choosing units for design or turning equipment on and off. The optimal heat production system architecture is obtained by minimizing the total consumed exergy, considering, but not limited to, centralized heat pumps and complementary utilities. The energy conversion technologies are represented by reduced order modelling and their operating conditions (characteristic temperatures and heat loads) are optimized. Numerical results are presented in a case study optimizing the use of a 60°C geothermal heat source connected to a district heating network in northern France. A second case study is analysed, considering the integration of heat pumps. Furthermore, a comparison with results obtained for the same case study using a Mixed Integer Linear Programming (MILP) model is performed. Calculation time and the solution’s accuracy are discussed.