Chemical reactor fouling is a widely recognized challenge for maintaining production rates and preventing unplanned downtime in industrial plants. While the extent of fouling is difficult to measure directly, a long-term effect of fouling on the process may be evident by incorporating process knowledge with indirect measurements. A primary shortcoming of this heuristic fouling indicator is correlation between indirect measurements and continually changing process and ambient conditions.

This talk presents a moving horizon estimation (MHE) approach to estimating fouling parameters. The estimation approach utilizes first principles and empirical elements to define the process model. Preliminary results indicate the quantitative fouling indicator tracks with heuristic process observations and provides a consistent measure of fouling. This indicator enables production decision makers to quantify the effects of process improvements in order to reduce fouling rates over extended periods of time under varying conditions.