How Chevron Is Using Generative AI to Revolutionize the Energy Industry

Chevron, a multinational oil and gas company, is leveraging generative AI to unlock valuable insights from its massive datasets and enhance operations in the energy sector.

Harnessing AI for complex data processing: With vast amounts of data generated from oil and gas operations, Chevron has been utilizing GPUs since 2008 to handle computationally intensive workloads, and is now integrating generative AI tools to further analyze and extract value from its datasets:

• A single seismic survey in New Mexico can generate a petabyte-sized file, requiring 100 exaflops of computing power to process and visualize for decision-making purposes.
• Chevron is one of the largest landholders in the Permian Basin, which accounts for 40% of U.S. oil production and 15% of natural gas production, with an estimated 20 billion barrels remaining.
• The publicly available data from the Permian Basin allows companies to learn from each other, with generative AI enabling rapid analysis and insights.

Enhancing collaboration and safety with AI: Generative AI is being applied to fill in data gaps, enable proactive collaboration, and prioritize worker safety:

• Large language models (LLMs) are used to generate engineering standards, specifications, and safety alerts, with constant fine-tuning by AI scientists to ensure accuracy and prevent creative deviations.
• AI is being evaluated for its potential to predict the location of new oil basins by informing models about geology and equipment.
• Robotic models are being explored to perform dangerous tasks, allowing human workers to stay safe while overseeing the operations remotely.

Fostering cross-functional collaboration: Chevron is working to bridge the divide between field teams and office teams by embedding them together and encouraging cross-functional skill development:

• The company has sent engineers back to school to obtain advanced degrees in data science and system engineering, updating their skills to align with the evolving industry.
• Data scientists, or “digital scholars,” are embedded within work teams to catalyze new ways of working and promote collaboration between machine learning engineers and mechanical engineers.

Addressing environmental concerns with synthetic data and digital twins: Chevron is utilizing synthetic data and digital twin simulations to support carbon sequestration efforts and reduce the environmental impact of its operations:

• The company operates some of the world’s largest carbon sequestration facilities and is using digital twin simulations to predict how carbon reservoirs will perform over time, ensuring the captured carbon remains securely stored.
• Synthetic data is being generated to support these predictions and improve the effectiveness of carbon sequestration processes.
• Chevron is also focusing on managing the energy consumption of its data centers and AI operations as cleanly as possible to minimize environmental impact.

Broader implications: Chevron’s adoption of generative AI and advanced technologies showcases the transformative potential of AI in the energy sector. By leveraging these tools to optimize operations, enhance safety, and address environmental concerns, the company is setting an example for how AI can drive innovation and sustainability in traditionally data-rich industries. However, the long-term impact of these initiatives on the environment and the workforce remains to be seen, and continued monitoring and responsible deployment of AI will be crucial to ensure a balanced and sustainable future for the energy industry.