Long-Term Hybrid AI-Expert Combustion Optimization System for Coal-Fired Electricity Generation NOx Reduction

An ever-increasing concern for the environmental impact of fossil-fuel electricity generation has amplified efforts by academia and industry to identify methods to mitigate this issue [1]–[5]. An attractive option has been the development and installation of optimization systems which utilize machine learning models to represent the combustion process and improve system performance. Consisting of primarily software, these systems can potentially provide significant improvements with little capital expenditure [6]. Numerous potential models and systems have been discussed within the literature, but each of these represent entirely offline or only short duration online performance studies [7], [8]. This work aims to satisfy the need for the long-term evaluation and characterization of the performance and effects of a hybrid AI-expert combustion optimization system (COS) on NOx emission rates as well as the secondary effects on the remainder of the system. The COS uses artificial neural networks to model the combustion process, particle swarm optimization (PSO) to optimize the 80+ manipulated variables, and expert logic to address adverse conditions in real-time. Over a two-year operational period, the hybrid AI-expert system was able to aid in the reduction of the NOx emission rate by more than 22.5% compared to baseline, while improving unit temperature management and realizing high operator acceptance leading to a system service factor greater than 86%.

[1] G. H. Richards, C. Q. Maney, J. L. Marion, R. Lewis, and C. Smith, “Ultra-Low NOx integrated system for coal fired power plants,” Fuel Energy Abstr., vol. 43, no. 3, pp. 218–219, 2003.

[2] J. F. Tuttle and K. M. Powell, “Analysis of a thermal generator’s participation in the Western Energy Imbalance Market and the resulting effects on overall performance and emissions,” Electr. J., vol. 32, no. 5, pp. 38–46, 2019.

[3] F. Wang, S. Ma, H. Wang, Y. Li, Z. Qin, and J. Zhang, “A hybrid model integrating improved flower pollination algorithm-based feature selection and improved random forest for NO X emission estimation of coal-fired power plants,” Meas. J. Int. Meas. Confed., vol. 125, no. 1, pp. 303–312, 2018.

[4] P. Tan et al., “Dynamic modeling of NOX emission in a 660 MW coal-fired boiler with long short-term memory,” Energy, vol. 176, no. X, pp. 429–436, 2019.

[5] J. F. Tuttle, R. Vesel, S. Alagarsamy, L. D. Blackburn, and K. Powell, “Sustainable NOx emission reduction at a coal-fired power station through the use of online neural network modeling and particle swarm optimization,” Control Eng. Pract., vol. 93, no. September, 2019.

[6] D. Hasler and W. Rosenquist, “Coal-Fired Power Plant Heat Rate Reductions,” 2009.

[7] J. Song, C. E. Romero, Z. Yao, and B. He, “A globally enhanced general regression neural network for on-line multiple emissions prediction of utility boiler,” Knowledge-Based Syst., vol. 118, no. X, pp. 4–14, 2017.

[8] J. Smrekar, P. Potočnik, and A. Senegačnik, “Multi-step-ahead prediction of NOx emissions for a coal-based boiler,” Appl. Energy, vol. 106, no. 2013, pp. 89–99, 2013.