background image design
Customer Story

Sartomer charts chemical operations data to match golden batch production

Woman in protective equipment at control panel
  • Challenge
    Small-batch manufacturing with swing production causes deviations affecting batch yield and quality that can bring bigger problems later.
  • Solutions
    PI System technology that allows engineers to achieve real-time visibility into production runs and detect problems as they occur.
  • Benefits
    Increased yield and smoother plant operation, resulting in cost savings

Paints, inks, coatings, flooring, electronics, and fiber optics are just some of the products that use acrylic and methacrylic monomers produced by Sartomer—a subsidiary of global chemical company Arkema. Located in West Chester, Pennsylvania, Sartomer meets the specifications of hundreds of different customers by using PI System™ technology (PI ProcessBook, PI BatchView, and Visual Basic®) that allows it to be proactive with small-batch monitoring. With the PI System, Sartomer has real-time visibility into production runs, and in the control room, any person, new or experienced, can detect off-track batches and minor deviations as they occur.

The operators... can easily see and quickly identify an off-track batch. That leads to more robust problem identification.
Kelsey Duffy
Production engineer

Better visibility means better problem-solving

There are many challenges of small-batch manufacturing with swing production: Sartomer has 11 different (multipurpose) reactors and more than 200 recipes. Each recipe is slightly different, as is each vessel. As a result, it is difficult “to identify small deviations between how your batch is running and an ideal (golden) batch,” even though these deviations can affect batch yield and quality and become major problems later in the production process, said production engineer Kelsey Duffy.

The solution for Sartomer was a PI BatchView display in PI ProcessBook that runs 24/7 untouched and uses Visual Basic code to automatically update the display whenever a batch finishes or a new batch starts. The display charts the current batch alongside a golden batch and the three most recent batches of that same product in that same vessel. The plant has color-coded the display, using green for the current batch, blue for the three most recent previous batches, and yellow for the golden batch. Whenever a new recipe starts, the Visual Basic code identifies the relevant batches using the batch ID, which corresponds to a PI Tag™.

“Engineering has gone through and has identified all the different golden batches that we need and has recorded their batch IDs,” Duffy said. These golden batches have been anchored to a results table so that the golden batch for a particular recipe is always displayed.

With this automated display, Sartomer engineers can compare a current batch to a golden batch and recent batches without any manual interaction or configuration by the operators. An engineer can just walk into the control room to see how the batches are running. “It's very obvious if the current batch is matching the historical batches and also the golden batch that we have for each recipe in each vessel,” Duffy said. “You can actually identify small deviations that are vessel-specific and recipe-specific and have a very good, very precise monitoring of that batch.” Better production visibility means Sartomer can be more proactive in catching problems. Once a problem is detected, plant engineers can take corrective action to save a batch from being off-spec or to adjust the process to improve yield.

Salvaging yield of an off-spec batch

As a use case, Duffy described the batch shown in Figure 1. The current batch in green is easily spotted as off-track from the golden batch and previous batches for this reactor and recipe. Because engineering was notified earlier, corrective action was taken to run the batch longer than normal to keep the yield up. "We were able to salvage a few thousand dollars’ worth of raw materials by not losing this yield," Duffy said.

Chart showing example data

Quicker identification of a temperature spike

In a second use case, shown in Figure 2, there was an obvious spike in the green current batch at about three hours into the batch, but the operator was out of the control room, taking a sample to the lab. When he came back, he noticed the problem. Without this display, the plant wouldn't have detected the problem until four or five hours later in the quality lab. With the display, Duffy said, “we had more flexibility with what we were going to do with that solvent, knowing that it had to have something special done to it."

A 24/7 automated display is a perfect fit for batch plants, “especially ones with swing production vessels like we have,” Duffy said, noting that it’s a good display for anyone with a repeatable unit operation. “You don't even have to be a batch plant. You can call every 24-hour period a batch in your continuous plan. You can still find benefit out of this concept. Ultimately, it allows our plant to run smoother.”

For more information about Arkema and the PI System, watch the full presentation here.

graphics design elements

How can we help you?

We’ve been delivering reliable data for over 40 years. In fact, 95 of our first 100 customers still use our platform. We’d love to help you too.

©2021 OSIsoft, LLC