ProcIndustries Enterprise-Wide Rollout Results

The annual Proclndustries' leadership meeting with Bill Roberts (vice president of operations) and Peter Argus (continuous-improvement manager) was scheduled for the upcoming week. They were to review progress made since Proclndustries decided to proceed with the enterprise-wide EIDI deployment almost two years before. Since then, all of the refineries had deployed the EIDI and were using it as part of their daily work routine. By leveraging the South Texas refinery's methodology, configuration templates, and improved work processes, the remaining refineries were able to quickly use the EIDI data in an effective manner to reduce costs, optimize asset health, improve safety conditions, and resolve operational issues as they occurred. This was in part because of the hard work and vision of the original digital transformation team. But the personnel at the other refineries made it happen by their willingness to adapt and utilize what had been accomplished at the South Texas refinery to achieve the strategic goals described in Chapter 11.

In preparation for the meeting, Bill and Peter set aside a full day to strat- egize and prepare. Bill noted that all but one of the Proclndustries refineries had met the target of returning the initial EIDI investment within two years. The one refinery that did not meet the goal had unexpected personnel turnover and were on track to reach their target in the next six months.

Cloud Strategy

Bill and Peter convened to prepare for the management meeting. Bill suggested that they start with the most pressing issue: defining cloud policy, implementation, data flow, and security for the next few years. Peter related that he had attended the latest OSIsoft conference and heard several companies discuss their specific approach. Peter also attended an oil and gas industry dinner during the conference, where he was able to informally discuss cloud strategy with his peers to and try to pick their brains on what actually constitutes best practices, if such things yet exist.

Peter found that most companies were grappling with several key issues: cloud design to facilitate more analytics offerings in the market, Internet of things (IoT) data integration with existing systems, and moving as many analytics as possible from the corporate environment to the edge, close to the refinery equipment to provide more timely value.

When Peter asked a fellow refiner, a strategic planning IT manager, about cloud design, she said that their company had embarked on a strategic initiative to architect their cloud to take advantage of the plethora of analytics coming onto the market that utilize production, operations, and equipment data. The first issue is always about security. Most operating companies, such as a downstream refining company, initially deployed a private cloud, where their data would be secure. However, this approach limited their ability to use newly available third-party packages that contain IoT sensors and analytics as single packages where data is sent to the public cloud.

The IT manager related that these new solutions are quickly installed and are very cost-effective, compared with the expense of connecting these sensors to existing control systems, both from the vendor cost, manual labor involved, and length of time to plan and implement such an initiative. She told Peter they had embarked on such a project to improve their vibration analysis on some of their rotating equipment. The vendor quickly installed the IoT sensors in one refinery and sent the results to their cloud analytics solution. The insights from this short-term project led to better operation of the units, and the maintenance team learned how to prevent abnormal situations.

She added that it was beneficial for them; however, after embarking on several of these successful projects, they had a lot of meaningful IoT data in individual vendor cloud environments. This resulted in siloed vendor clouds without a method of integrating the IoT data, analytics results, and vendor-added insights to their EIDI, which prevented many people being able to visualize and analyze the data in context to the rest of the process, production, and equipment data.

Because of this and many other business reasons, her company implemented a hybrid cloud, where much of the data is private but some of the data is public, mostly for analytics and public-facing information. Once that was decided, her company wrestled with how best to integrate the siloed analytics with their existing EIDI systems. She told Peter that after attending a cloud presentation at a conference, the answer might be a multi-tenant cloud environment, where different analytic and IoT sensor vendors might coexist in an environment where IoT data plus contextualized EIDI data can be easily and securely sent to this new cloud for analysis. At the same time, data and results from any of the analytic offerings can be integrated to the EIDI environment (and perhaps the legacy control systems) for further analysis, reporting, and collaborative notifications.

Peter took note of her key points and planned to talk to Pat Verlaine, Proclndustries' IT director about how this could tie in to their existing cloud plans. Peter thought that if Proclndustries could seamlessly extend the EIDI system to the cloud without needing to shift data to new systems, the time to value would decrease and the utility could easily increase.

Bill Roberts was intrigued by Peter's cloud conversation and asked Peter if he had any other takeaways from the conference. Peter shared that he learned a few interesting trends: The first was edge analytics and the hybrid cloud vision he saw and the second was the changing role of the process engineer at the modern plant or refinery. The two, he said, were intertwined.

 
Source
< Prev   CONTENTS   Source   Next >