Drilling into intelligence: Using AI to Solve Complex Decisions to Save Time & Money

The Oil & Gas industry is among the frontrunners in its adoption of AI. And that’s been quite a surprise for many. The reason for its rapid embrace? The Industry has recognized the role AI can play in reducing the time to first oil. AI enables the augmentation of many of the judgements and assessments that have to now been the preserve of highly specialized individuals operating in very complex environments.  AI has also allowed organizations to capture dark data and unlock new insights from hidden knowledge and experience.

Oil and gas operations are, of course, hugely complex. No two locations where an operator may choose to drill, for example, are ever the same. A wide range of geological and geophysical factors mean every well requires a unique approach that is painstakingly developed for the characteristics of that location.

Traditionally, to decide where, and optimize how to drill a well, engineers would need to source, access and analyze vast amounts of data from many different sources. These range from seismic data, geological data and well logs, to numerous handwritten, scanned or digital documents and operational reports. AI can help to augment the human intelligence that is tasked today with collating this information, assessing it and defining an accurate execution plan for a prospective drilling location.

With artificial intelligence harnessing technologies such as natural language processing and computer vision, it’s possible to execute much more informed and consistent appraisals of these myriad sources of data. And that consistency can have a major impact on the profitability of the lifecycle of a well. Take well planning, for example. This requires analysis of neighboring wells to understand the timeline of a well, what previously did and did not work, the technical issues that arose, etc. This ‘offset well’ analysis informs the design for the next portfolio of well(s) in a particular location. Depending on the field and its complexity, a well engineer will need to consume a large number of historical documents about neighboring wells. Based on that analysis, they are able to establish the parameters and the characteristics of a new well that they will design.

Accenture’s solution – Well Advisor – uses AI technologies such as Natural Language Processing (NLP), Optical Character Recognition (OCR), and Knowledge Graph (KG) to liberate and connect new and old information. So rather than an individual engineer having to navigate a myriad of disconnected data sources to find relevant information, the solution does this by design and extracts the pertinent entities. Thus, enabling the identification of common patterns and correlations that release new, relevant insights that will bring to light previous wells’ challenges, offer analysis of their root-cause and highlight the techniques that have been applied to solve them. As for the ingestion of new data, this is automatically extracted, classified and hydrated into the underlying Knowledge Graph, connecting new data to old and providing the ability to uncover new insights or lessons.

Where NLP & OCR unlock the data, and Knowledge Graph connects, through Well Advisor, Accenture provides a foundation for the digital transformation of traditional workflows, such as well planning. It also enables the integration of workflows and the creation of recommendation engines to evolve the advisory capability that AI promises.  

Through initial WellAdvisor pilots, we have found that well engineers, by automating laborious and time-consuming document analysis, can save up to 90% in the preparation time and thus reduce well design time, which allows them to work more efficiently and effectively. We have also seen a 15% improvement in well risk assessment associated with better identification of Non-Productive Time (NPT) events and their warning signs associated with hole instabilities. Augmenting human intelligence and experience with AI, liberates well engineers to focus on the key tasks that require their judgement and acumen, to deliver a safe and accurately planned well.  The same approach is also being developed to optimize Exploration and Production workflows; while also adapting the same AI technology to support the execution of capital projects. AI is bringing new possibilities across the entire Upstream Oil and Gas value chain, creating a 360-degree view from exploration through drilling to production and refining.

In an industry where the costs of exploring, finding, producing and refining a barrel of oil run into the 100’s of millions $ and where the drilling of an individual well can run into the millions of dollars having the ability to be more efficient and more accurate across every aspect of the well lifecycle creates a significant financial impact. That’s the scale of rewards that the Oil and Gas industry is achieving from applying AI to highly complex, technical decision-making to augment human expertise and experience.