Does AI have the power to refine oil and gas efficiency? Does AI have the power to refine oil and gas efficiency?

By EY Global

Ernst & Young Global Ltd.

6 minute read 4 Jun 2019

Artificial Intelligence and associated technologies could be more than just a way for oil and gas companies to improve operations.

The oil and gas industry isn’t shy of technology. Today, it manages three of the world’s most powerful supercomputers, each one of them a room-sized machine, able to process trillions of equations per second.

A good portion of this processing power helps build predictive models that shape how the industry explores, produces and eventually refines oil and gas. Yet there’s potential to do much more – especially when it comes to Artificial Intelligence (AI).

As margins in the industry tighten, oil and gas companies are already looking to AI to optimize costs. So it’s not surprising that close to half of the respondents to a recent EY survey say the technology is a top investment priority.

An AI-led efficiency boom

Despite its computing power relative to other industries, the oil and gas sector has so far been slower to adopt and integrate AI technologies. But advances are being made.

Across its upstream and downstream operations, the industry is ramping up AI investments. More than half the oil and gas companies we surveyed are currently implementing AI applications. As a whole, these companies invest about $65 billion a year in R&D. From 3D seismic imaging to horizontal drilling, over its recent history the industry’s deep research capabilities have led to game-changing innovation. This expertise is now also driving the development of in-house AI applications.

One Spanish energy supplier has applied AI and data tools across its refineries. These measure pressure, temperature and processing rates to ensure the entire plant is operating at optimal efficiency – while also reducing downtime by preventing breakages and accidents. According to the firm, these analytics could increase profits by as much as 30 cents per barrel.

AI is also improving the distribution of refined products. One of the world’s largest oil companies uses analytics to better anticipate fluctuations in demand and supply, maximizing the value generated by each barrel of oil produced. A US oil major is working to capture all operating data generated by its refining and chemical operations. The amount of data – up to 1 billion bits per minute – is unprecedented, and can be used to unlock unique operational efficiencies.

While we know AI is starting to transform downstream operations, it’s also moving into other parts of the value chain. AI and associated technologies – including deep learning – are transforming upstream operations. Recently a Paris-based International Oil Company (IOC) struck a partnership with Google Cloud to develop AI applications, including Computer Vision technology to improve the interpretation of sub-surface images. Some of the innovations will also automate the analysis of technical documents.

The next frontier: AI-enabled exploration

Exploration and production (E&P) is a capital-intensive element of the oil and gas business, where potentially massive discoveries justify large investments – though those finds can be few and far between.

AI offers a pathway to optimize E&P spend. Companies are investing in robots to streamline exploration, and one company has even partnered with the Massachusetts Institute of Technology (MIT) to develop AI-powered robotic technology to detect natural oil seeps flowing out of deep sea rock formations. Devices are also in the works to track emergencies and abnormal equipment activity at exploration sites. Such efficiencies not only cut E&P costs, but they also help oil and gas companies reduce their environmental footprint.

AI is also transforming drilling. State-of-the-art drilling software informed by a variety of data – vibration, thermal gradients, pressure differentials among others – are making real-time decisions on the speed and direction of drilling. Such optimization prevents failure and keeps progress as safe and smooth as possible. Drilling is expensive, so any way of improving the process is going to generate substantial savings.

Keeping portfolios performing

There are plenty of further applications of AI and machine learning when it comes to managing upstream portfolios.

For example, oil and gas companies oversee thousands of wells. Some are operating but many others lie idle until market conditions make it worthwhile to extract oil and gas out of them.

Maintaining these assets is time-consuming and costly. Yet wiring wells with sensors can help build predictive models to optimize field deployments, so only those wells needing maintenance get a visit. The same is true for pipelines and other midstream transportation and storage systems – sensors can reduce time-consuming and expensive maintenance checks, and prevent spot leaks and spills before they become a significant problem.

Developing the future

The strategic application of AI and smart, connected tech can already enable more efficient refineries and optimized E&P campaigns, strengthening the bottom lines of oil and gas companies across the industry’s value chain.

The challenge is in identifying the right strategy, as one size does not fit all. In fact, there are a number of advantages to developing applications in-house. The very nature of the technology means in-house development – or partnering with others in the value chain – can help force the data silos that often prevail in the oil and gas industry to be broken down. That’s because AI, deep learning and other applications are only as good as the data they are fed. This reality motivates a silo-free approach that accumulates large data pools, which can generate value for the whole company and not just a single business unit.

Breaking down silos and promoting greater data sharing to enable the development of in-house capabilities can create unique monetization opportunities for oil and gas companies with the capability and scale to oversee such initiatives. Access to this data – consolidated from partners from upstream to downstream – could unlock insights and efficiencies that could revolutionize operations.

There is even the possibility that oil and gas companies could turn their data and technology into a standalone revenue stream, by licensing access and usage of in-house assets to third parties. After all, they do say data is the new oil!

Summary

Oil and gas companies have applied AI and analytics to their downstream technologies for some time now, driving significant efficiencies. Today, these solutions are also beginning to move upstream.

About this article

By EY Global

Ernst & Young Global Ltd.