SKY Member evening: Data-driven design in process industry

Two short presentations and free discussion on the subject.

The link will be sent to SKY members by e-mail.

Machine learning for predicting alkali salt-induced high-temperature corrosion of superheater materials (by Rasmus Fagerlund)

• Short introduction on high-temperature corrosion in black liquor recovery boilers and Åbo Akademi corrosion test method
• Data analysis
• Modeling of data using machine learning methods
• Model performance and conclusions

Modeling of steam usage at a power plant site (by Jere Espo)

• Trials to manage different operational situations with a mathematical model based on process data
• Correct timing of required operator manual work
• Generation of more valid process measurement alarms

Rasmus Fagerlund

• Doctoral student at Åbo Akademi University
• Currently working on modeling of NOx formation for computational fluid dynamics (CFD) models of black liquor recovery boilers (supervisor Docent Markus Engblom)
• Background in chemical engineering
• Interested in AI applications, mathematical modeling and optimization

Jere Espo

Process Engineer

• Team leader of Vantaa Energy Chemistry Team operating at three sites
• Background in analytical chemistry

Roughly 25 years of experience in power plant chemistry including topics like:

• Fuel analysis
• Flue gas cleaning
• Water chemistry
• Chemical safety
• Team leading tasks