Perstorp
            

Perstorp is one of the world’s leading specialty chemicals innovators. The company makes essential properties, such as resins and coatings, used in products found in our pockets, homes, offices and even farms. Founded 140 years ago as a family-owned business in Sweden, Perstorp now spans seven production sites in Europe, the U.S. and China. Annual revenue is about $124 million.

“If it matters, make it better!”

That’s the Perstorp mantra. The chemical manufacturer became an industry leader by making the world a better place, one molecule at a time. Its innovations make phone screens scratch-free, buildings more fire resistant and air conditioning systems more environmentally sound, to name a few examples.

“Using nature’s tiniest building blocks, we can help advance everyday life,” says the Perstorp website. “We’re always ready to take things to the next level, saying: ‘Wow, we did it! Now how can we make it better?’”

Improving processes at Perstorp is perpetual in order to always make the highest-quality, most useful products at the most competitive cost.

A few years ago, uncovering improvement possibilities meant relying on data scientists to analyze historical data using tools few knew, like MATLAB® or Python®. Or engineers would import data into Microsoft® Excel® and do analysis there. Both options took too long for this industry leader, eyeing to be more agile and resilient.

That’s why Perstorp turned to TrendMiner in 2020. If Perstorp had left analytics only to the data scientists and to engineers using only Excel, some analyses wouldn’t get done or at least not as often, explained Krister Forsman, Perstorp Corporate Specialist and TrendMiner champion.

Since using TrendMiner, Perstorp engineers are empowered to analyze time-series data quickly and easily on their own. They can make smarter decisions about processes, faster, using a self-service industrial analytics tool.

To excel or not to excel

Democratizing data analytics is critical to Perstorp’s continuous optimization journey. And that requires easier access to efficient data analysis tools. “In theory, people could extract the data and use complex languages to look at the numbers,” Forsman explained. “But that's not democracy … where it's very easy for everyone.”

A lot can also be accomplished in spreadsheets. But they’re not ideal for time-series-specific analysis or when data volume becomes a bottleneck.

“It’s a unanimous opinion,” Forsman said. “It’s more efficient to do calculations in TrendMiner than in Excel. There’s so much time to be saved. I think in many cases, using TrendMiner is much more than ten times faster.”

Analytics and answers—fast

A friendly UX was a key reason Perstorp took to TrendMiner after evaluating several self-service analytics tools.

“When we checked the requirements on our list, TrendMiner was the only one that satisfied them all,” Forsman said, noting his favorite features are the compare table, search and filtering, the tag builder and the “chat with us” button. All simplify getting answers about data, and the product itself, really fast.

“Every time I have a question, I get an answer within minutes,” he said of the “chat” button. “There's a competent person to help, no matter what the time of day.”

Customer Success Manager Tommy Peters added: “And it’s a conversation with one of our data analytics engineers. They are all experts in TrendMiner.”

The smart things Perstorp people do

Five AspenTech Enterprise IP.21 Historian™ systems, which collect and store continuous time-series data, are connected to TrendMiner, so data is ready for analysis, anytime. Indexed data goes back to 2015 for Perstorp with thousands of quickly accessible, calculated variables (or plant measurements, like for governing pH).

The young engineers, in particular, like that TrendMiner helps them rapidly uncover opportunities for process improvement. So TrendMiner’s proving to be valuable to help attract and retain talent.

Here are some smart things Perstorp people do with TrendMiner:

Batch optimization – Engineers extract reactor batches, overlay statistical analysis, calculate raw material yield for each batch and then extract subphases of batches and features, such as pressure average during a certain phase in the reaction. Then they correlate features with yield to make process optimizations. “If you look at this whole work process, it would be extremely difficult to do in some other tool unless you use really advanced tools like Python and MATLAB,” Forsman explained. “But only a few engineers are experienced in those.”

Yield monitoring - Raw material yield is the highest Perstorp variable cost. So Perstorp uses TrendMiner to maximize yield, reduce catalyst deactivation and, therefore, control costs. Two engineers created the solution, which presents the current yield estimate to the control room and monitors catalyst deactivation to determine the precise best time for catalyst change. It’s a fine line: too frequent or infrequent a change results in higher cost. The process engineer commented: “This would not have been possible in Excel.” Added Forsman: “ The potential savings from this solution could be significant!”

Dashboarding - Without any training on TrendMiner dashboard utilities, an engineer created a solution that provides visual KPIs to continuously optimize operations. The engineer first finds critical process variables that can be used as KPIs, then extracts statistics, determines control limits for key parameters, and visualizes KPIs together with control limits on dashboards.

Environmental reporting - Traditionally environmental variables are “quite complex to extract,” Forsman said. Using TrendMiner, reporting on them is far easier. Said one process engineer who needed to report on water usage: “Using calculated variables, value-based searches and filtering in TrendMiner saved a lot of time in this work, compared to doing it by exporting data into Excel and then doing the analysis there.”

Equipment checks – Perstorp engineers also use TrendMiner to assure equipment performs as specified to meet guarantees. In one case, Perstorp did a value-based search on the seven different conditions, including temperatures and pressures, to find a gap in expected performance. With this analysis, Perstorp can work proactively with the supplier to find a solution.

“I’d like to see us use TrendMiner more,” Forsman said. He looks forward to what the future brings with TrendMiner, since using the solution has become a growing trend.

Ready for a demo?

Engineers can continuously optimize operations to increase yield, lower costs and reduce energy consumption by using TrendMiner. See how easy industrial analytics can be. Register today for a TrendMiner demo.

Sustainability & Perstorp

Sustainability is important to Software AG, so we like to highlight our customers who have the same commitment to and investment in sustainability.

Perstorp’s mission is to be: “the sustainable solutions provider, focusing on the global Resins & Coatings, Engineered Fluids and Feed Additives markets.”

Its landmark sustainability initiative is Project Air, which will reduce carbon dioxide emissions by 500,000 tons per year, corresponding to the annual emissions of around 340,000 new cars running on fossil fuel.

Learn more about Project Air slated for completion in 2025.

Discover Software AG Corporate Social Responsibility projects.