The Industrial Internet of Things (IIoT) plays a more prominent and extensive role in oil & gas production applications. At the 2022 ARC Industry Forum, ARC Advisory Group’s Tim Shea hosted a session, Leveraging IIoT and Operational Analytics to Realize Digital Transformation.
Tim opened by highlighting several main applications, including autonomous drilling, well monitoring & optimization well spacing and “frac zone” multi-pad well drilling visibility, production planning & monitoring, asset integrity management, and supply chain management & optimization.
The main challenges in deploying and maintaining IIoT solutions include cybersecurity considerations, technology standardization, platform & protocol interoperability, ESG pressures, and overall system complexity. Having strong technology partners is vital to successful implementations. Many producers lack the in-house expertise to develop integrated, enterprise-wide analytics and asset management system capability.
The benefits of developing an integrated analytics and asset management system working with key technology partners include:
- Integration of different data types-structured/unstructured, missing/incomplete, quality, ongoing maintenance
- Operational visibility improvements for better control & monitoring, reporting, visualization
- Role-based visibility, improved workflows, increased collaboration, better decisions
- Barriers & disconnects remove and closer connection with ecosystem partners
One advantage of IIoT measurements in oil & gas production is reducing carbon intensity by significantly reducing routine operator & maintenance rounds requiring driving onshore well pads and helicopters or boats for offshore production.
I’ll recap one of the sessions I found interesting with their use of reinforcement learning artificial intelligence.
In well completions, better visibility to the surface and downhole data can help oil & gas producers optimize operational and engineering processes. The key to successful fracturing treatments is a successful execution. The value is in reducing rig time, reducing non-productive operations activities, and increasing well production performance. The goal of improving this visibility is better-informed decision-making.
AI learning algorithms address this challenge, including a family of machine learning applications—supervised, unsupervised, and reinforcement learning. Reinforcement learning was used to address this challenge. It’s a step-by-step iterating process where the right move to avoid constraints is reinforced to arrive at the most optimal path. The Microsoft Bonsai platform is used for reinforcement learning.
The solution worked by taking oil & gas field data and comparing the predicted value versus the actual value. That result feeds the Bonsai AI brand and the output producers’ actions such as modulus, toughness, stress, and leakoff coefficient.
The presenter shared some results and lessons learned. The first is that most regions’ brains give a good result with only a few outliers. This work encourages them to progress to real-time decision support and scale to other areas and applications. A key lesson learned is that an SME champion and management sponsorship will make or break these projects. These projects require a champion with data science skills, statistics expertise, and subject matter expertise.
For those who attended the conference, this presentation and many more are available on the ARC Industry Forum’s online event pages (login required).