Digitalisation and use of data analytics will come under the spotlight at this June’s Underwater Technology Conference (UTC) in Bergen.
Our day two panel debate, focusing on data and digitalisation, will feature speakers including Google Cloud Nordics Regional Manager Otso Juntunen, Jeff Jensen, CTO at Arundo Analytics, and Cognite’s Director of Customer Success Magnus Muri Boberg. In advance of the conference, we asked Muri Boberg for some of his views.
How advanced is the subsea oil and gas industry in terms of its adoption of data analytics?
The oil and gas industry is, and has clearly been, quite advanced in data visualization and data analytics in certain areas, such as, e.g., for subsurface data.
However, these activities have not really scaled across the industry, since the data is stuck in separate, disconnected data silos and the corresponding workflows are old fashioned and highly manual. For example, users have typically stored their own personal data repositories for individual reporting and analytics.
Also, subsea systems are complex, expensive and difficult to repair. Although subsea systems can generate large amounts of data, these data have typically not been liberated from the individual data silos, nor been put into a usable context for wide-/large-scale visualization and analytics.
With the recent developments in modern data processing and contextualization technologies, our experience shows that there is a big potential for data visualization and analytics to be applied more widely to the subsea business.
We would also like to emphasize that – even before moving into advanced data analytics – there is a lot to be gained in easy and user-friendly data reporting and visualization. In some cases, as much as half of the return can be gained by empowering people with the right, contextualized data to make better decisions.
Are there areas where the industry is good at data analytics?
The oil and gas industry has used advanced visualization analytics for decades, but typically in silos and in specific areas. For example, data-driven analytics and statistics like Monte Carlo simulations, multivariate regression or principal component analysis have been widely used in exploration and reservoir modelling.
Now, fuelled by the combination of cheaper, faster storage and near limitless computing power
on the cloud, we see a focus on more “data intensive” machine learning methods, like neural networks or even deep learning. Many oil and gas companies are investing and testing these new opportunities, but often within silos or as stand-alone proof of concepts or pilots.
Hence, there are still very few wide-scale operationalized uses of advanced analytics with a clear return on investment. However, we are optimistic that we will see a number of operationalized use cases happening over the next 6-12 months, as oil and gas companies are getting their hands on data that is liberated and contextualized from the core, usually siloed, data systems.
Where do you see the greatest potential for adoption?
When you liberate data from a subsea system, you bring together all the relevant information for operations, including sensor values, environmental values, maintenance history, 3D models, ocean models, etc. Typically, this data is locked into separate systems, possibly at multiple locations, with no simple way to access and link it back together. Once liberated, the data can be contextualized, i.e., linked back together.
Once liberated and contextualized, this data can be deployed with great benefits for a wide variety of applications, from simple data visualization and reporting, via standard analytics (gaining insights on real-time system behavior combining production data with sensor data and maintenance data), to more advanced machine learning applications predicting failures, or even proposing operating conditions to ‘fly the system home’ to the next scheduled intervention.
With high-quality data finally readily available to be shared, a new relation between original equipment manufacturers (OEMs) and clients will emerge. There is both monetary and non-monetary values to unlock if the proper insight is developed by correct visualization and data analytics. This can result in better operations or radically new performance-based business models. This will aid both brownfield life-time strategies and green-field development.
The great potential here is improved environmental outcomes, increased uptime and to unlock better performance. Furthermore, OEMs can also use this to further improve product fit and quality.
What is the most common question you get asked by the industry?
We usually find that the suppliers of subsea systems are eager to get holistic real-time operational data on the equipment they supply. Today, OEMs deliver equipment and typically get a call/mail when it is performing sub-standard. They are quite eager to use their expertise to deliver a better experience for the end-user, and also to learn to improve its products. It is extremely rare that subsea supplier has online systematic access to their equipment operational conditions and performance.
In parallel, every operator wants improved uptime and lowered cost over time: We argue that these are possible when you liberate and contextualise data at scale and enable advanced analytics for fault diagnostics, predictive maintenance and increased production.
What have the unbelievers got to lose?
We would argue that this applies to both suppliers and operators. For suppliers it’s clear, data analytics provides a market differentiation in terms of cost, quality and uptime. For operators, the large-scale use of data analytics will lower the breakeven point for producing from subsea wells, which will allow the believers to produce from fields that are not viable for anyone today, and not viable for non-adopters in the future.
About the Underwater Technology Conference
UTC 2018 will be the 24th Underwater Technology Conference, with 700 professionals and 40 exhibitors expected to attend at Bergen’s Grieghallen from 12-14 June 2018. Register here
Interview by Elaine Maslin