Plastics Manufacturer Solves Decades Old Problem in Four Months

The Challenge

The largest manufacturer of acetal resins in the world was having difficulties with uptime in their largest manufacturing facility.

The continuous manufacturing process was designed with three reactor trains, two of which could be operated simultaneously. Due to a phenomenon described as “clogging of the arteries” at the raw material injection points, the reactors had an average run time of less than 30 hours.

The uptime, measured by the percentage of time that two reactors were running, was roughly 60%. An average runtime of 65 hours was required to achieve the uptime of 90% to produce enough resin to supply all their orders on time.

The Strategy

The possibility of finding the solution to a problem that had plagued the plant for more than a decade was met with a high degree of skepticism by both the production team.

An abundance of previously unsuccessful engineering studies, process redesigns and operational changes had placed plenty of doubt in their minds that the problem could be solved. However, management encouraged (some would say forced) them to try once again to solve the problem.

A team of process engineers, production managers and operators used MSLLC designed experimentation techniques to test which process changes could help them achieve the required uptime.

The Experiment

The team identified over 65 changes that could be made to the process to increase the runtime of each reactor. From the 65 brainstormed ideas, the team narrowed the list to eleven ideas that were quick, easy and cheap to change.

Due to the length of time required to execute each experimental trial, the team selected a 12-run design to test the 11 factors. Considering that over 2,000 possible combinations of factors existed, the selected design was very efficient, yet robust enough to provide the insight needed to determine which factors helped, hurt or didn’t make a difference. The factors are shown below:

The Results 
The analysis showed that five of the eleven ideas had a significant effect on the runtime of the reactors. Three factors helped and two factors hurt, while the rest made no difference at all. After verifying the three factors that helped through further experimentation, it was found that two of the factors interacted with one another. After implementing the three factors at their correct level, an average runtime of 125 hours (almost twice the requirement) was achieved. The extra runtime also gave plant mechanics enough time to perform all the required maintenance, a task which could not be done previously. Additional experiments on other parts of the plant that had become the bottleneck since “unclogging the arteries” yielded better operations than had been seen in over 10 years. In total the P&L statement showed a net gain of over $25 million per year in profitability.