Surviving the current downturn will require oil and gas companies to adapt. While the pandemic will relent, sustainability pressures will not. Oil and gas companies know they must invest; most majors are targeting carbon-neutrality by 2050 or earlier. AI can help oil and gas companies become more sustainable.

Sustainability concerns are an existential threat to the O&G industry. Public tolerance for environmental damage is very low. Collaborating companies are withdrawing from the sector; in 2020, Google pledged to stop building custom tools for crude oil extraction. Endangering personnel is seen as unacceptable.

GlobalData’s new report, Thematic Research: AI in Oil & Gas, details several ways AI can help companies be more sustainable.

More efficient production

If production is more efficient, less of it needs to be done. By analyzing seismic and subsurface data, AI can improve site discovery. Saudi Aramco invested in Earth Science Analytics, whose software predicts rock and fluid properties in the subsurface. Various drilling-assistance tools improve drilling efficiency. Shell, for example, automates its drilling systems by training them on historical data with reinforcement learning. AI tools can also make drilling faster, reducing economic and environmental costs by saving time. Rosneft’s automated drilling management system improved the mechanical penetration speed by 15%, reducing mechanical well drilling time by a day. Considering all these opportunities together, it’s clear that AI can significantly boost production efficiency. Better efficiency on each well means fewer wells are needed.

Reduced leakage

Leakages are catastrophic for oil and gas companies, from both a sustainability and a PR perspective. AI can pre-empt and contain leakages. BP was an early adopter. In 2017 it saw that methane leaks contributed significantly to total greenhouse gas emissions. It partnered with Kelvin, a US-based software provider specializing in automating physical systems using AI. A vast number of sensors were installed at the Wamsutter gas wells. The sensors relayed real-time field data to Kelvin’s AI system, which combined it with historical site data to inform optimization simulations. The simulations were able to predict leakages and alert BP engineers to the need for maintenance before the leak occurred. Six months after implementation, methane leaks from the wells had been reduced by 74%. Production volumes were up 20%, and operating costs were down by 22%. After the success, BP sought to install similar sensors at all its wells.

Additionally, some predictive maintenance products, such as SparkCognition’s SparkPredict, allow users to set control variables in alignment with their targets. Oil and gas users could set the model to maximize production while minimizing emissions.

Safer personnel

Rig work is dangerous. AI can automate dangerous manual tasks to remove the need for personnel to put themselves at risk. Smart robots can perform some maintenance tasks. BP trialled the Spot robot at a refinery, which gathered data, detected abnormalities and removed workers from hazardous spaces. Additionally, AI can analyse accident records and illuminate the historic causes, thereby helping avoid future accidents.

With sustainability pressures growing, the AI capabilities mentioned represent a serendipitous and unmissable opportunity for the oil and gas industry. All companies should invest.