IIoT can change the formula for the chemical industry

Perhaps I am biased (as a chemical process engineer/operations manager in the chemical industry for 25 years) but it appears to me that the chemical industry is the industry that is most challenged to adopt IIoT solutions.  

The industry does indeed face different challenges than other types of manufacturing. Whereas robotics transformed automotive manufacturing, and additive (or 3D printing) is transforming turbine and aircraft engine manufacturing, chemical process optimization isn’t about eliminating manual labor or developing new ways to create complex parts. It is mostly about data – getting the right data, analyzing that data to gain new insights, and then automating the use of that data to prevent or predict issues in the future. The IIoT will transform the chemical industry.

How is the chemical industry different? Here are three examples.

• Complexity and unpredictability of the manufacturing processes: In working with chemical reactions, very small changes in raw materials, operating conditions, or operator technique can have significant impact on product quality, production throughput, or equipment integrity. Detecting these variations and determining the cause of them can be really hard.

• Very conservative risk tolerance: When you are working with toxic, corrosive, and flammable materials, it is appropriate to take a very conservative approach to making changes to the manufacturing process. It’s uncomfortable to change to predictive maintenance instead of tried-and-true preventive maintenance techniques. Similarly, it feels less risky to rely on employees with years of experience instead of technology using a machine-learning algorithm.

• Expensive and aging equipment:  The high capital costs of building a new chemical plant means that chemical companies are more likely to upgrade existing processes in an incremental fashion. It isn’t practical to completely rewrite the control software, and only a few experts are allowed to modify it, slowing the rate at which improvements can be introduced.

However, advances in IIoT software present excellent opportunities to address these challenges.

• Complexity and unpredictability of the manufacturing processes:  You need to work with many sources of data (EAMs, ERPs, control systems, MESs), and the more sources you can aggregate, the greater the insights you can discover. But you need more than just a BI tool, because once you have analyzed the data, you need to take action on your discoveries – including implementing advanced analytics that can prevent problems, or providing advanced warning of issues so you can respond proactively and avoid fire drills.

• Very conservative risk tolerance: Consider that doing things “the way we’ve always done it” isn’t necessarily less risky. Reducing process variability generally leads to improved quality, production throughput, and equipment integrity. Relying on scheduled maintenance can actually increase risk versus using predictive analytics to detect early signs of failure that give you time to properly plan an equipment overhaul. And analytics that automatically provide a diagnosis and recommended action will reduce human error.

• Expensive and aging equipment: For decades, companies have been putting calculations and advanced analytics into their DCS or SCADA systems because there wasn’t another option. This is time-consuming and requires the resources of the controls engineer. The process engineers aren’t allowed to just do it themselves. Edge connectivity, with capabilities for edge analytics, are now a cost-effective option. Process engineers can augment the existing control system by doing analytics on the edge, including feedback to the control system itself.

As a process engineer, I spent more than half my time looking for relevant data and getting it into a format that I could analyze. Then, I could finally do my real job of making improvements in quality, throughput, and equipment integrity. Getting more data collected electronically was cost prohibitive at that time.

Now that costs have decreased exponentially for wireless devices, edge software, and cloud software to monitor and control the plant, we have the opportunity to fundamentally change how we improve the chemical process.  We can finally get the right data to understand the impact of tiny variations in the process. With this knowledge, we can deploy analytics to detect and correct these variations in the future. When used to augment the existing control systems, these analytics will reduce risk.

According to Gartner, “Asset performance management (APM) is becoming a core competency to improve overall business operations in asset-intensive industries.” To learn more, check out the 2019 Gartner APM Market Guide* here.

The future is finally here for the chemical industry. Learn more about how GE Digital solutions can help you on this journey.