China Pushes AI in Energy Sector, Data Center Electricity Becomes New Challenge

High-powered broadband equipment is used at a site for the China National Petroleum Corp.'s 3D intelligent full-wave inversion data acquisition (Photo: China Daily)

China is accelerating the use of artificial intelligence or AI in the energy sector. The goal: more efficient industrial operations, safer energy supply, and the ever-growing data center electricity needs can be managed.

According to a report by China Daily, quoted on Saturday, May 30, the latest action plan of the China National Energy Administration or NEA targets a large increase in the clean energy supply capacity for AI computing infrastructure by 2030. The use of AI in the energy sector will also be expanded.

AI is a technology that makes computer systems able to read data, recognize patterns, and help make decisions. In the energy sector, AI can be used to manage the electricity grid, monitor drilling, read risks, and help plant operations.

NEA noted that by 2025 China built 42 large-scale intelligent computing clusters. The national computing center's electricity consumption reached 170 billion kilowatt-hours.

This figure is important. AI data and computing centers require huge amounts of electricity. Globally, data center electricity consumption is expected to almost double by 2030 compared to 2025.

Lin Boqiang, head of the China Institute for Studies in Energy Policy at Xiamen University, said the surge in electricity demand was a major challenge for the network. However, China's energy infrastructure and renewable energy growth provide an opportunity to turn that pressure into an advantage in the AI competition.

"The interaction between AI and the energy sector has moved from one-way support to deep integration," Lin said, quoted by China Daily.

According to the NEA, China's energy sector has launched dozens of large industry-specific AI models. The model is used in the electricity grid, renewable energy, thermal and nuclear power plants, coal, oil, and gas.

One example is China National Petroleum Corp's large Kunlun AI model. The latest version of Kunlun is said to move from a passive question-and-answer system to "active intelligence". This means that AI not only answers questions, but can also plan, choose tools, analyze data, and perform tasks on the production line.

China Daily reported that Kunlun is now used in 152 scenarios in the energy and chemical industries.

The impact is starting to be seen. Kunlun has reduced the processing of three-dimensional acoustic wave inversion from 20 days to three days and reduced costs by more than 30 percent. This technology is used to read underground conditions, including in energy exploration activities.

The model also has a drilling risk warning system with an accuracy of more than 85 percent. In the last six months, the system has issued more than 300 early warnings to help prevent accidents.

Lin assessed that this development shows a change in the way the Chinese energy sector works. From manual supervision, the industry is starting to move to AI-based active management.

For China, the challenge is not only expanding the use of AI. A sufficient, clean, and stable electricity supply is also an important part so that the technology can support the energy industry sustainably.