In a previous post (here) I wrote about the value of obtaining data when working to solve any given problem. That post inspired a good friend of mine, Michael Todd, retired project manager, to capture two of his own experiences, which I gladly share below. If you have your own stories to share, please do so! I’ll post them right away.

Story 1

I was a technical area supervisor at one of our plants back in the day where we manufactured a specialty coated film. The coating was applied to a 60-inch wide polyester base and then dried in an arched film dryer at line speeds in the hundreds of feet per minute. In order to expand capacity, we added a festooned dryer section, theoretically allowing significantly increased line speed, increasing the coater/dryer capacity substantially.

The festooned configuration in a perfect world is a bit of a trick to control with multiple “steering rollers” along the sequence working to keep the film aligned down the centerline of the machine. With the sticky coated side down toward the forced-air dryer rolls, control upsets could easily lead to a “crash” and a very time-consuming clean-up.

Well, we could hardly get that beast to run for frequent failures. Time and again, the sticky coated side collapsed and stuck to everything. Extremely frustrating. Of course, we commenced with a massive electro-mechanical troubleshooting effort that elapsed over some number of months trying to resolve the issue. Experts from within and from outside the company were brought to bear. Ultimately to no benefit.

Meanwhile, operations continued to struggle, with continuous “crashes”, lost production, massive frustration. Careful evaluation of nothing more than simple data provided the breakthrough needed. But first, a little background.

“Skew” is one of the properties of polyester film supplied in rolls. In our case, our supplier produced the virgin polyester film in 180-inch-wide sheets which are in part stretched to gain width, then slit and rolled up in “master rolls”. As a result of the stretching, the 60-inch center cut is pretty “straight” while both of the-side cuts possess some degree of “skew”, one being “left hand”, one being “right hand”. The film tends to track through machines accordingly. Normally, machine controls are able to compensate. But the festooned dryer was a different animal, much more finicky.

So, while the massive technical assault of the electro-mechanical control system was unfolding, one of our younger engineers started looking at some data related to the polyester film. By and by, he discovered that the crashes were occurring when lots of incoming polyester film were being switched. Specifically, most acutely, when rolls of polyester with right-hand skew were being fed into the coater/dryer right behind rolls with left-hand skew. And vice versa. The coater/dryer controls couldn’t handle the perturbation, thus the crashes ensued. Eureka!

The young engineer’s discovery led to a very simple modification to our operational protocol regarding how the incoming film rolls were staged for consumption, minimizing the most difficult transitions. Close study of the data revealed the cause of this problem and the solution came soon thereafter. One more administrative responsibility for operations, but anything that would prevent the crashes was welcome. And we lived happily ever after! The controls themselves were never able to compensate for the “left” to “right” transitions.

Oh by the way, this young engineer and I became the closest of friends. And in subsequent years he rose to the position of CEO of one of DuPont’s spin-off businesses in the electronics materials business, DuPont PhotoMask. And then incredibly sadly, he passed away at age 53 due to esophageal cancer. One of the saddest days of my life.

Story #2

One of my great moments, in my third year with DuPont. Stationed at Victoria Texas, representing the Design Division at the site for construction and startup of a massive Adipic Acid expansion. Very big equipment, the area in question built principally of Titanium. Tricky to modify in the field (titanium welding is very sensitive to contaminants, but I digress).

We were trying to get started-up. A very large distillation column in the system, which was basically designed to operate at atmospheric pressure, repeatedly hit a high-pressure interlock at 15psi while heating up, tripped the system out.

After several tries and fails – we did what any enterprising group of engineers do in a crisis – we retired to a conference room, pulled out all of the P&IDs, and the debate ensued. All of these engineers were very senior to me. Many theories of what might be wrong. Personally, having witnessed the situation over the shoulder of the board operator, I had this nagging feeling that the tower just needed to burp, couldn’t expel gas. At some point while the meeting continued, I decided to take a walk. I stood outside the unit, looked up and studied the piping system. Soon concluding that there was no way for the inerts to escape the system. They were sealed in by the liquid level in the condenser/reboiler! We needed a vent line atop the condenser head. The fix wasn’t easy because of the difficulty of field modifying a titanium code vessel, but we added a small nozzle at a strategic point on one of the heads, and voila – she burped. Off we went.

Just as data are essential, closely examining the physical plant (collecting field data as it were) can sometimes reveal things not obvious from the paper we produce!

END