Enhancing sustainability of heating systems in humid regions: A framework for flue system design based on thermal-acoustic-aerodynamic performance

This study investigates the performance of flue systems in a library building located in Rize province, a region characterized by high humidity and variable climatic conditions. The heating system of the building is designed using a solid fuel (coal) boiler with a thermal output of 100,000 kcal/h. The flue system connected to the boiler was modeled using square and rectangular cross-sections in combination with four distinct chimney materials: chamotte form stones, varnished ceramic, ceramic with an internal varnish layer, and concrete form stones. A total of eight alternative design scenarios were developed based on combinations of cross-sectional geometry and construction material. Performance assessments were conducted using the Kesa-aladin software, allowing for precise simulations of thermal, aerodynamic, and acoustic behavior under real-world operating conditions. To further enhance the reliability of the evaluation, a random forest regression model was applied to determine the relative importance of performance parameters. The analysis revealed that maximum negative pressure and flow velocity were the dominant predictors of chimney performance, followed by thermal conditions (inner wall temperature and insulation temperature) and acoustic effects (flow sound). This prioritization underscores the critical role of aerodynamic and thermal stability in flue optimization. Additionally, two multi-attribute decision-approaches, Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR), were employed to rank the eight chimney alternatives. Both methods consistently identified chamotte-based chimneys as the optimal solutions, irrespective of geometry. While TOPSIS ranked concrete alternatives higher than ceramics, VIKOR favored ceramics after chamotte, demonstrating how methodological differences can influence final decision-making. Overall, the integration of simulation, machine learning, and multi-attribute decision-approaches methods provides a comprehensive framework for chimney system evaluation. The findings highlight that optimizing pressure stability, gas flow dynamics, and thermal insulation are key for improving chimney efficiency and sustainability. This research not only provides a replicable framework for the design of energy-efficient flue systems but also offers practical guidance for architects and engineers working in regions with complex climatic challenges.