TY - JOUR
T1 - Smart solutions for clean air
T2 - An AI-guided approach to sustainable industrial pollution control in coal-fired power plant
AU - Lim, Juin Yau
AU - Teng, Sin Yong
AU - How, Bing Shen
AU - Loy, Adrian Chun Minh
AU - Heo, SungKu
AU - Jansen, Jeroen
AU - Show, Pau Loke
AU - Yoo, Chang Kyoo
PY - 2023/10/15
Y1 - 2023/10/15
N2 - Conventional fossil fuels are relied on heavily to meet the ever-increasing demand for energy required by human activities. However, their usage generates significant air pollutant emissions, such as NO , SO , and particulate matter. As a result, a complete air pollutant control system is necessary. However, the intensive operation of such systems is expected to cause deterioration and reduce their efficiency. Therefore, this study evaluates the current air pollutant control configuration of a coal-powered plant and proposes an upgraded system. Using a year-long dataset of air pollutants collected at 30-min intervals from the plant's telemonitoring system, untreated flue gas was reconstructed with a variational autoencoder. Subsequently, a superstructure model with various technology options for treating NO , SO , and particulate matter was developed. The most sustainable configuration, which included reburning, desulfurization with seawater, and dry electrostatic precipitator, was identified using an artificial intelligence (AI) model to meet economic, environmental, and reliability targets. Finally, the proposed system was evaluated using a Monte Carlo simulation to assess various scenarios with tightened discharge limits. The untreated flue gas was then evaluated using the most sustainable air pollutant control configuration, which demonstrated a total annual cost, environmental quality index, and reliability indices of 44.1 × 10 USD/year, 0.67, and 0.87, respectively.
AB - Conventional fossil fuels are relied on heavily to meet the ever-increasing demand for energy required by human activities. However, their usage generates significant air pollutant emissions, such as NO , SO , and particulate matter. As a result, a complete air pollutant control system is necessary. However, the intensive operation of such systems is expected to cause deterioration and reduce their efficiency. Therefore, this study evaluates the current air pollutant control configuration of a coal-powered plant and proposes an upgraded system. Using a year-long dataset of air pollutants collected at 30-min intervals from the plant's telemonitoring system, untreated flue gas was reconstructed with a variational autoencoder. Subsequently, a superstructure model with various technology options for treating NO , SO , and particulate matter was developed. The most sustainable configuration, which included reburning, desulfurization with seawater, and dry electrostatic precipitator, was identified using an artificial intelligence (AI) model to meet economic, environmental, and reliability targets. Finally, the proposed system was evaluated using a Monte Carlo simulation to assess various scenarios with tightened discharge limits. The untreated flue gas was then evaluated using the most sustainable air pollutant control configuration, which demonstrated a total annual cost, environmental quality index, and reliability indices of 44.1 × 10 USD/year, 0.67, and 0.87, respectively.
KW - Air pollutant control
KW - Data augmentation
KW - Monte-carlo simulation
KW - P-Graph
KW - Sustainability enhancement
U2 - 10.1016/j.envpol.2023.122335
DO - 10.1016/j.envpol.2023.122335
M3 - Article
SN - 1873-6424
VL - 335
JO - Environmental pollution (Barking, Essex : 1987)
JF - Environmental pollution (Barking, Essex : 1987)
M1 - 122335
ER -