AI-Based Models of Emissions at the Stack From Industrial Furnaces

  • Alessandro Della Rocca, Tenova S.p.A.,
  • Michele Roveda, Tenova S.p.A.,
  • Nicolò Giso, Tenova S.p.A.

By means of machine-learning algorithms, predictive models of gas emissions from reheating furnaces at the stack are created based on process data: temperatures, air and fuel flowrates, pressures, oxygen content, etc. Those models are not aimed to replicate, even with a very good approximation, measurements of NOx or CO2 performed by the dedicated gas analyzers. The target is to create virtual sensors able to correlate the scheduled production to the pattern of process conditions and to predict the concentration of any gas released, thus allowing to organize the production schedule according to a forecast of the environmental impact.

  • Date:Tuesday, June 29
  • Time:11:00 AM - 11:30 AM
  • Room:207 A
  • Location:207 A
  • Session Type:Presentation Only
  • Is the main contact one of the authors?:Yes
  • Master Session:Environmental: Emissions Control I
  • Session Chairs:David Gilles, Vern Martin
  • Speaker 1 Country:Italy
Alessandro Della Rocca
Tenova S.p.A.
Michele Roveda
Tenova S.p.A.
Nicolò Giso
Tenova S.p.A.