Featured Publication

Stacked deep-learning models predict cogeneration unit gas emissions

This study explores the use of stacked deep-learning algorithms to predict and monitor gas emissions as a more efficient, cost-effective alternative to traditional Continuous Emissions Monitoring Systems (CEMS). Highlighting a case study on Suncor Energy's Firebag Cogeneration Unit 194 in Alberta, the paper details a cloud-based PEMS that utilizes data from physical sensors to accurately forecast NOₓ concentration, mass flow rate, and temperature.