Summary of Key Points
The accelerated construction of AI data centers in North America has exposed a mismatch between the US power infrastructure (slow approval processes, aging facilities, and insufficient cross-state capacity) and the rapid growth of AI. This has elevated the role of energy storage from a supporting technology for renewable energy to a critical component of computing infrastructure. Chinese energy storage companies are shifting their strategy from simply selling products to focusing on localized manufacturing, system integration, and long-term operation in North American markets. New technologies such as SOFC (Solid Oxide Fuel Cells) and gas turbines complement rather than replace lithium-ion batteries in energy storage solutions. The North American market has become a testing ground for the competitiveness of energy infrastructure in the AI era, where the ultimate winner will be those capable of providing comprehensive solutions.
Why Have AI Data Centers Suddenly Made Energy Storage More Valuable in North America?
AI data centers have an enormous demand for electricity—both in quantity and stability (even a brief outage can result in significant losses). However, the US power grid is struggling to meet these demands:
- Low Grid Efficiency: The US grid is fragmented, with separate systems for power generation, transmission, and distribution, and each state has its own rules. Building new transmission lines requires years of approval processes.
- Fading Infrastructure: Over 70% of the transmission infrastructure is more than 25 years old and prone to failures, similar to aging water pipes.
In response, technology companies are opting to build their own energy storage systems (combined with natural gas generation and solar power), turning energy storage into a buffer that can help manage peak demand and provide backup power in case of outages. As a result, energy storage has evolved from a supporting technology for renewable energy to a essential component of AI data centers.
Why Can't Chinese Companies Simply Sell Products in North America?
While Chinese energy storage companies once dominated the market due to their cost-effective solutions (with 80% of global Tier 1 suppliers being Chinese), the situation has changed in North America:
- Policy Barriers: The Inflation Reduction Act requires local manufacturing and procurement to qualify for tax incentives, and high tariffs have significantly increased the cost of pure exports.
- Changing Customer Needs: AI data centers require not just individual energy storage devices but comprehensive solutions that include design, installation, and maintenance.
Chinese companies are now focusing on more substantial investments in North America, such as establishing local teams and service networks (for example, SunPower), forming joint ventures to build battery factories (with EVE Energy), and expanding production capacity (like Envision Power). These efforts aim to integrate into the local supply chain, including managing labor relations, engaging with communities, and working with utility operators—a true test of their ability to operate locally.
Will AI Data Centers Abandon Lithium-Ion Batteries?
Some suggest that new technologies like SOFCs and gas turbines will replace lithium-ion batteries, but they are more likely to complement them rather than replace them:
- Advantages of Lithium-Ion Batteries: They respond quickly (within milliseconds), making them ideal for backup power and demand management.
- Advantages of New Technologies: SOFCs and gas turbines can provide steady, continuous power, suitable as a primary energy source.
- Technological Challenges: However, new technologies are still in the development stage; for example, SOFCs are expensive, and natural gas infrastructure is not well-established, while small modular reactors (SMRs) are still being tested.
In the short term, lithium-ion batteries will remain the primary choice for AI data centers due to their mature supply chain and proven track record.
What Really Matters in the North American Energy Storage Competition?
The focus has shifted from comparing the cost and performance of individual products to evaluating the ability to provide comprehensive solutions:
- Localization: Can companies produce, deliver, and maintain systems locally?
- System Integration: Can they integrate energy storage, power generation, and grid management to meet the unique needs of AI data centers?
- Long-Term Operation: Can they ensure stable performance for 10 or even 20 years, regardless of unexpected circumstances?
Customers are looking for solutions that guarantee uninterrupted data center operations. Those companies that can deliver these will have a competitive advantage in North America.
Conclusion: The Competition in Energy Storage in the AI Era is About Solutions
The changes in the North American energy storage market reflect how AI is reshaping the infrastructure of energy supply. Energy storage is no longer a secondary component but a core element of computing infrastructure. For Chinese companies, this represents both an opportunity (due to the surge in AI demand) and a challenge (the need to improve their local capabilities). The future winners will be those who can transform products into services and provide comprehensive solutions. What happens in North America today may set the precedent for the global energy storage industry tomorrow.