OSIsoft Internet of Things/Big Data, Houston

NOV Max/RigSentry for BOP. Revenos Limitless Well. Element Analytics and CRISP data mining.

A better-late-than-never summary from OSIsoft Internet of Things, Analytics and Big Data forum held late last year in Houston. CTO Helge Kverneland explained how NOV is embedding more automation into its systems. Drill bits and drill pipe may now be connected with (relatively) high data rates. In production systems, flexible risers have been equipped with sensors for decades to detect when things go wrong. But these also generate a lot of data, much of which went unused until now. Kverneland sees this increasing use of data in the context of the ‘fourth industrial revolution.’ Here, oil and gas has been a bit behind in terms of M2M, cyber, robots and ‘we want to be part of this too!’ The internet of things is where the action is. NOV may be a mechanical company but it has been developing software for a long time and has many programmers working on its control systems. To date, the focus has been on asset performance management (in collaboration with the event host OSIsoft).

The BOP stack is a good illustration of the intersection of machinery and control systems. If a 500 tonnes deepwater BOP assembly fails, take up and repair might mean a couple of weeks of downtime. This has been addresses on the mechanical side with a pod of control systems that can be retrieved with an ROV. NOV is now looking to use big data and analytics to predict failures from real time and historical data. Equipment failures tend to follow a ‘bath tub’ curve where failures are highest early and late in the asset’s life. So there are good reasons not to ‘fix what ain’t bust!’ BOP avail-ability is monitored with NOV’s MAX industrial internet of things/big data platform, with a PI System under the hood. ‘Rigsentry for BOP’ is NOV’s first big data breakthrough. When the ‘quad’ safety system detects a failure on a pod, it switches over to another. NOV is now working on a similar ‘RigSentry for top drive.’ Ultimately, data scientists will warn customers that ‘the top drive bearing is about to fail’ and match this fact up with an upcoming window of opportunity for maintenance. Kverneland wound up saying ‘connected products are the future of our industry and data will get us there.’ NOV straddles manufacturing and the internet of things. But ‘we can do this better because we know our equipment.’

Michael Maguire & Wesley Dyk introduced Revenos’ ‘Limitless Well,’ a ‘multi-state well completion and production analytics discovery tool.’ The proof of concept targets a value-case for standardizing disparate data big data and traditional database structures. PI AF and the PI integrator for business analytics are the building bricks. The game plan is to ‘democratize data and deliver actionable insights to decision makers.’

In a similar vein, Sameer Kalwani introduced Element Analytics’ approach to predictions, machine learning, data lakes and data readiness. EA’s web-based platform transforms raw, real-time operational data into a form that data scientists can use to perform analytics quickly and repeatably. EA’s toolset includes an OSIsoft PI System hosted in the Microsoft Azure cloud. The CRISP-DM standardized data mining process also ran. Referring to the data-driven vs. physics-based modeling debate, Kalwani opined that we need both, both models are appropriate and complement each other. In any events, the PI System is foundational to such initiatives.

Read these and presentations from Devon, MOL and OSIsoft on the conference website.

Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.