Fouling potential prediction and multi-objective optimization of a flue gas heat exchanger using neural networks and genetic algorithms

Abstract

The fouling problem of flue gas heat exchangers is one of the most critical problems to be solved in industrial waste heat recovery applications. In this paper, the fouling potential prediction and efficient design method of a flue gas heat exchanger were proposed. Firstly, the evaluation index of the fouling factor indicating the fouling degree of the heat exchanger was defined. Then, the fouling factors under different coal ash types and structure parameters were numerically predicted, forming the fouling factor database. Finally, the Pareto optimal solution set is obtained by the neural network and genetic algorithm. The results show that fouling factor index FFI effectively reflects the effect of coal ash type and structural parameters on the fouling degree, which can make up for the deficiency of existing fouling evaluation index B/A. The thermal-hydraulic performance of flue gas heat exchanger was optimized with multi-objective optimization method, and the fouling degree of Pareto optimal solutions were compared. Case A (S1/D = 3.00 and S2/D = 1.50) and case B (S1/D = 2.04 and S2/D = 2.19) are applicable to the waste heat recovery of dusty flue gas with and without considering the space size constraint, respectively. Compared with case E (S1/D = 1.50 and S2/D = 3.00) considering only heat transfer performance, thermal-hydraulic and anti-fouling performance of case A and case B are significantly improved.