(13c) Predictive Maintenance the First Digitization Success for the Chemical Industry

Over the past decades, maintenance strategies have evolved in the chemical industries to minimize unexpected equipment failures along with the associated cost and losses. Preventative maintenance plans are in place for many pumps, compressors and other important equipment, and for critical units condition-based monitoring programs are being implemented. In the latter case, not only time and utilization become factors in the maintenance decision, but actual process conditions and consequential performance metrics are integrated in metrics that trigger maintenance and repair action.

For the chemical industry, advancements in digitization and data utilization along with smart devices and sensors are ushering in a new area, often referred to as Industry 4.0. While many of these concepts, ideas and developments are still in their early stages, equipment maintenance processes are already harvesting successes from advanced data utilization and predictive modeling. Data mining, increased computing power, and user-friendly implementations of machine learning models enable early detection of impending failures, in some cases with significantly improved lead times compared to condition based monitoring systems.

In this talk we will summarize the current technologies on the market, focusing on the key methodologies and algorithms. While the actual modeling approach is a critical component for success, many factors may impact the ease of implementation and quality of the results. We will discuss key elements of those along with some thoughts on how to identify strong partners in the journey of predictive maintenance application for larger chemical complexes, and share some success stories.