Whether cutting costs or headcount, offering supply chain insights, or giving advice on “continuous improvement,” facility employees who have endured (survived) consultants over the years are familiar with these programs and often refer to them as “the flavor of the month.” That is how it was when Elmer’s process plant looked to apply “lean manufacturing” to its processes.
Yet another square peg aiming to fill an incompatible hole, “lean” practices were made famous by Toyota, where the concept became highly refined. Its success at lean manufacturing spawned numerous experts and consultants eager to guide new padawans in their discipline. But how can such concepts and practices apply to processing plants?
At Toyota, these concepts became vital to the efficient and error-free assembly of 20,000 parts from disparate supply chains into high-quality automobiles. However, Elmer’s plant had only two feedstocks and four products. In addition, automotive manufacturing employs many individuals and/or robots on vast and complex assembly lines, while process plants such as Elmer’s operate more like airplanes, with only a few people in the cockpit and a few attendants in the cabin. Crew size is dictated more by safety and handling upsets than actual labor required. What if any “lean” concepts apply?
To explain how lean manufacturing works, it’s best to look at the five Japanese words and phrases that all begin with the letter “S” when spelled in a western alphabet, and form “5S”, a foundational tenet of lean manufacturing. These five words and phrases are: sort, set-in-order, shine, standardize and sustain. On an assembly line, a randomly cluttered workstation can slow down or stop the line when a part can’t be found, or an incorrect or incompatible part might be installed in error. Mistakes on the assembly line might not be corrected before a product is sold, leading to costly and embarrassing recalls.
In a continuous process plant such as Elmer’s, instrument maintenance must wait for a plant outage or “turnaround.” When demand is strong, the cost of downtime is extraordinary, so one could make the best use of the outage by planning, prioritizing, and palletizing (sort and set-in-order) all the jobs waiting due to a shutdown—something Elmer’s plant was already doing. But extensive planning and palletizing was derailed by “found work.”
Much of the manufacturing process takes place inside opaque pipes, vessels, compressors and reactors, and limited instrumentation grants a myopic view of what’s going on inside. Even if Elmer’s could correctly surmise, “I think we have some broken filter elements,” it wasn’t possible to know if it was one or 100, or how difficult it would be to make repairs.
Elmer’s digital networks—Foundation Fieldbus and HART multiplexers—provided insights into the opaque control valve bodies and actuators. Modern positioners provide data that might only be available during an outage, increasing the opportunity to be ready with parts and supplies if an outage is scheduled. Elmer’s could monitor certain control valves for alerts or “triggered data”—information gathered after a configured event such as excessive travel deviation. However, interpreting and prioritizing device alerts, control valves, in this case, was complex. The effort to sort and set-in-order”—to prioritize—needed more resources.
The “flavor of the month” can become a nuisance if its roots aren’t in the process industries. Can you use such canned practices to improve utilization of the vast device data available? When you’re “staffed to run,” sorting through pages of device diagnostics is rare, as daily issues emerge to steal your attention. If you feel you’re being bludgeoned with mostly incompatible continuous improvement dogma, see if you can’t use the program to leverage some help with assimilating and digesting all your digital data.
