From the Danube to the cloud: MOL Group shares data with Axens via PI Cloud Connect
- ChallengeMOL needed a faster way to share catalyst data with its vendor Axens and optimize asset yield performance.
- SolutionA digital bridge using PI Cloud Connect enabled a real-time connection between MOL and Axens’ PI System installations.
- BenefitsPI Cloud Connect helped users on both sides share data securely and create real-time dashboards to make rapid, strategic decisions to maximize asset performance
Nestled just south of MOL Group’s headquarters in Budapest, the company’s Danube oil refinery sits in the town of Százhalombatta. The Danube Oil Refinery is capable of processing 140,000 barrels of crude oil per day, a process that clocks in at 10.6 on the Nelson complexity index. Needless to say, the Danube refinery’s large and complex operations hinge on optimal asset performance with maximum yields.
The Axens catalyst management system lies at the heart of MOL Group’s complex refinery operations. However, because Axens lacked rapid visibility into MOL’s catalyst performance data, Axens struggled to assist MOL in the evaluation of catalyst performance and unit yields. Both MOL and Axens, wanted better, faster access to operational data, so they turned to PI Cloud Connect to share real-time operational data between client and vendor securely and with strong governance.
Optimizing yields and unit performance
Purchasing a catalyst is an expensive endeavor, and a highly complex catalyst can cost anywhere from 10-20 million dollars per year to maintain optimal unit yields. For this reason, the process of running a catalyst is highly controlled. Companies must review catalyst performance relative to operational costs, and develop plans regarding types, prices, and replacement dates. MOL keeps close contact with Axens to support performance control issues, and the companies often hold on-site meetings to review catalyst performance data.
When unplanned events—such as an unexpected drop in yield—arise, MOL must send Axens operational data in order to determine the root cause. This process is often done using spreadsheets and analysis, and this process takes days, if not weeks. During this catalyst review, engineers must reduce catalyst activity to protect the asset during the investigation, costing the refinery thousands of dollars. To improve turnaround time, Axens needed access to MOL’s real-time catalyst data—and that data resides in the PI System.
The Connect’in connection
For over 20 years, MOL has used the PI System as its refinery’s data infrastructure. “The information system of the refinery is based mostly on PI,” said Csaba Galambos, process automation engineer at MOL. “Most of the dashboards are in PI. Everything is connected to PI.” MOL relies heavily on Asset Framework (AF)—the contextualization layer of the PI System—using hundreds of asset templates as no-code operationally focused digital twins and over 25,000 streaming analytics including descriptive, diagnostic, and simple predictive analytics run on real-time data coming from 70,000 sensors. “We have PI Vision which is the backbone of the refinery management now,” Galambos remarked.
Axens uses a tool called Connect’In to view catalyst performance, optimize the lifespan, prevent unplanned events, and assist with predictive maintenance. However, small deviations are difficult and time-consuming to solve, especially if a catalyst is already running at near optimal levels. “In the case to give support, it is really important that we have the actual data, and that’s [what] we developed Connect’In for,” said Arnaud Fouillac, project manager at Axens, during the 2020 PI World online conference.
Axens needed access to MOL’s PI System data to analyze catalyst performance in Connect’In, which required a secure connection between customer and vendor. Since both companies are PI System users, they were able to establish a secure PI System-to-PI System connection in the cloud. When faced with the problem of needlessly-difficult catalyst performance review, Galambos said, “The answer is that we use PI Cloud Connect.”
PI System to “PI in the sky”
Using PI Cloud Connect, the companies embarked on a proof-of-concept (POC) to connect MOL and Axens’ on-premise PI Servers. Data is transferred between both PI Servers, and no process data is stored in the cloud. To protect intellectual property, calculations remain on MOL’s side and Axens can only see the results of the AF analyses. Metadata is managed to determine what data to share, how to share it, who had access, and for how long. User-based permissions and authentication ensure data can’t be shared outside of the ecosystem, and should the partnership end, data and structures will be deleted immediately.
Now, sharing and analyzing data is simple. With the Connect’In web interface, users can continually monitor equipment performance, share catalyst performance reviews, and show trends to help Axens improve services. They can also view displays on PI Vision in a familiar, flexible environment. “If you would ask me two years ago what we will do or with whom we will share the data with I would have laughed in your face because it was just impossible,” Galambos said.
Thanks to PI Cloud Connect, MOL and Axens have access to the same data and dashboards to perform analysis. As a result, both companies learned that there is a lot of room for improvement—even in mature industries. “We were kinda surprised how easy it was to improve with new technology,” Galambos said.
For more information about MOL Plc and the PI System, watch the full presentation here.