Intelligent performance enhancement of flue gas waste heat recovery in combined low-temperature economizer – Air heater systems

To enhance flue gas waste heat recovery (FGWHR) and address the significant time delay in traditional Proportional-Integral-Derivative control of the low-temperature economizer (LLTE) and air heater (AR) system, a capacity expansion retrofit integrating an air heater (AR) with the LLTE in a 1000 MW...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Process safety and environmental protection Jg. 202; S. 107812
Hauptverfasser: Li, Huaan, Wei, Dongliang, Wu, Yajie, Zhou, Tianxing, Zhou, Hao
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier Ltd 01.10.2025
Schlagworte:
Algorithm Comparison
Capacity Expansion Retrofit
Flue Gas Parameter Prediction
Flue Gas Waste Heat Recovery
Optimization Analysis
Flue Gas Parameter Prediction
Capacity Expansion Retrofit
Optimization Analysis
Flue Gas Waste Heat Recovery
Algorithm Comparison
ISSN:0957-5820
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:To enhance flue gas waste heat recovery (FGWHR) and address the significant time delay in traditional Proportional-Integral-Derivative control of the low-temperature economizer (LLTE) and air heater (AR) system, a capacity expansion retrofit integrating an air heater (AR) with the LLTE in a 1000 MW coal-fired unit was investigated. The retrofit’s effects on key performance indicators, such as thermal efficiency, power generation, and coal consumption, were evaluated. The predictive performances of Backpropagation, Long Short-Term Memory, Temporal Convolutional Network–Transformer, and Least Squares Support Vector Machine models (LSSVM) for flue gas parameters were compared, with the Constrained Particle Optimization (CPO)- LSSVM model achieving the highest accuracy and generalization. By utilizing the predicted flue gas parameters, real-time optimization was carried out through CPO-HYSYS. The thermal efficiency and environmental benefits were both enhanced. It is demonstrated that intelligent algorithms for real-time flue gas prediction and optimization provide an effective solution to overcome lag in FGWHR systems. [Display omitted] •Practical application of LLTE combined AR on 1000 MW unit is tested.•Compared BP, LSTM, TCN-Transformer, and LSSVM prediction performance for flue gas.•Real-time optimization analysis conducted using CPO- HYSYS.•RMSE for flue gas temperature and flow rate prediction are 0.129 and 15.67.•Coal consumption decreased by 0.26 g/kWh, flue gas WHR increased by 2.51 MW.
ISSN:0957-5820
DOI:10.1016/j.psep.2025.107812