Agile manufacturing is all about achieving rapid changeover between different products with minimum disruption to the process involved. Increasingly, particularly for batch operations, this relies on real-time information.
"Batch production must take into account varying raw material properties and potencies to ensure a quality end product. Access to this sort of information in real-time is very important. Batch systems need to take advantage of this information to dynamically adjust operating quantities. Also, standard batch products need to adjust to variances in ingredient delivery and adjust all remaining feeds to stay within specification without requiring custom engineering," explains Andrew Stump, batch, information and advanced process control segment manager, process systems team, Rockwell Automation, Mayfield Heights, Ohio.
The growing trend of utilizing real-time production data crucially depends on collecting such data within the context of production activities, notes Stump. Two approaches dominate. In the first, real-time batch reporting summarizes production details and process trend data for on-demand review — generally via a fixed format report for a defined purpose. In the second, tools, such as Gantt chart view of activity tied to process conditions at that moment, allow real-time analysis of batch execution — enabling users to shift time horizons, correlate process trend data and look at different parts of batch activity and their impact on production.
"By collecting this data, it enables users to leverage historical event and trend data for real-time golden batch analysis with real-time monitoring software to identify variation in the model," he adds.
However, one size doesn't fit all, Stump says. "There is a need for local standalone skid-based batch control. These systems should offer the flexibility to fit into larger process trains without re-engineering."
An automation upgrade project carried out at the Axis, Ala., plant of Philadelphia-based Arkema, Inc., shows the importance of collecting real-time data. To create impact modifiers for polymers, Arkema's engineers use a batch reactor to produce a liquid latex material that then goes into a spray dryer as part of a continuous process. Each batch takes about 12.5 hours to complete. The plant annually makes about 85 million pounds of 15 different impact modifiers.
Arkema was suffering increasing problems with its legacy distributed control system. (Half the plant used a 28-year-old version and the other half a 15-year-old one.) It was struggling to find both the people and equipment to make necessary repairs. At the same time, a simple recipe change could shut down the plant for ten hours — while failure episodes were costing $42,000 every month.
"More importantly, safety concerns were starting to crop up. The plant experienced a few potentially critical incidents stemming from aging I/O [input/output] hardware and control systems that were overburdened. As the current system lacked condition-monitoring capabilities, plant personnel were not alerted to problems until it was too late," notes John Bryant, engineering and maintenance manager.