The buzz around AI is not a fiction now. It’s a reality that we can feel now knocking at our door, just like the advent of driverless vehicles to robotic surgery in helping the medical professionals. AI, Machine Learning and Big data are now disrupting every industry, and forcing them to keep them in the high corporate priority list to act upon. Seeing the trend, the Oil & Gas companies have too become cautious now as they are heavily investing in the areas like production asset optimization, predictive maintenance, biometric monitoring and fleet management.
Till date the digital revolution that is being seen by the upstream oil & gas industry, is mostly about supporting the E&P (Exploration & Production) division, with specific computational interpretation platform, which heavily relies upon a good number of domain experts like geologist, geophysicist, reservoir engineer’s interpretation skill and knowledge. The low oil price scenario and the shrinking margins on returns, have forced many of the oil & gas operator companies to explore and develop the reserves more cost effectively. This results into deploying small team, with an average domain experience level, but equipped with strong computational knowledge.
The main task of the exploration team is to acquire, process and interpret the subsurface data, resulting into a compact subsurface model, which will be used further by the development and production team to exploit the hydrocarbons. While on initial reconnaissance geophysical surveys wireless sensor and drones are already been used, but yet till date less have been explored of using AI in interpreting the seismic data, or delineating well log facies (specific rock types based on certain measured geological features) and doing correlation taking into consideration of geological uncertainties of the basins.
Subsurface data interpretation by means of AI
Seismic data, which is one of the most key and costly data acquired during the exploration stage, is widely used for getting a first-hand information about the hydrocarbon prospect of the basin at the very early stage. Interpretation of seismic data is mostly done through visual tracking of the seismic horizon section taking consideration of the geological history of the basin. Although most of the traditional petro-technical platform is capable of auto-tracking the seismic horizons, it fails to understand when they come to identification of geological features or events that may have resulted to form those subsurface structures. Highly skilled human interpreters are quick to correlate that geological basin history while tracking seismic horizon or fault planes, and adjust it accordingly. Implementation of AI will be immensely helpful, in not only automating these complex human jobs accurately within less time, but also of enriching the human knowledge by compiling and leveraging the rich database of subsurface models.
Facies identification is based on wireline geophysical logging operations and correlation of litho- facies among the various wells to identify the extent of pay-zone, which basically is an important factor to prepare the field development plan. Till now, these are long and tedious jobs done by petrophysicist (human domain expert having expertise in well logs’ quantitative and qualitative interpretation). Recent progress in AI has been catapulted for finding new techniques of facies classification. Future course of study will not be only to come up with a good algorithm capable of automating the process, but also having deep learning strategies for feature learning and classification, so that geological constraints can be ruled out.
Potential of AI
According to a recent survey done by PwC among Oil & Gas CEOs, it is observed that in the coming one year or so, the C level executives are putting 18% focus on digital innovation (the highest in the category), as their top corporate priorities to drive growth or profitability in the organization. It is of no brainer that in this low oil price scenario, the industry is concentrating more on digital innovation for reducing operational cost and enhancing production. To summarize it all, the oil & gas upstream sector as an immense scope of enhancing its possibility in the field of usage of AI armored with IIoT, big data analytics, and deep machine learning.