Surge in Power Consumption#
China's power consumption from data centers is projected to increase dramatically from about 300 billion kilowatt-hours to 1 trillion kilowatt-hours by 2030. This surge is expected to account for around 7% of the country's total power consumption, according to a senior director at a research institute under the National Development and Reform Commission (NDRC).
Cost of AI Computing#
The cost of AI computing in China is estimated to be below $3 per million tokens, which is significantly cheaper than the approximately $15 per million tokens in other countries. This cost advantage is likely to contribute to the increased demand for power as AI technologies continue to expand.
Renewable Energy Growth#
From 2026 to 2030, over 80% of the new electricity demand in China is expected to be met by renewable energy sources. Wind and solar power installations are anticipated to exceed 250 gigawatts annually during this period, with specific projections of around 120 gigawatts for wind and 180-220 gigawatts for solar in 2026. Additionally, offshore wind energy is expected to recover, reaching over 10 gigawatts per year, aided by improved project readiness and clearer policies.
Green Certificates and Energy Costs#
China's green certificate market is also growing, with average transaction prices expected to rise by about 90% in the second half of 2025 compared to the first half. By 2030, green certificate prices could reach 1-2 Chinese yuan cents per kilowatt-hour. The levelized cost of energy for new utility-scale solar projects is estimated to be between 0.136-0.242 yuan per kilowatt-hour, while wind power is projected at 0.12-0.25 yuan per kilowatt-hour. After including transmission fees and government support, the total power cost for northern data centers is expected to stabilize around 0.30-0.35 yuan per kilowatt-hour.
Future Projections#
By 2030, renewable energy is forecasted to support about 20% of peak load, a significant increase from the current 10%. This growth will be facilitated by better forecasting and intelligent energy management. The expert suggests that regional guidelines should replace strict nationwide limits on energy curtailment, allowing for more flexibility in energy management across different regions.