AI Data Centers Are Turning Energy Companies into New Infrastructure Partners
As demand for artificial intelligence computing continues to grow rapidly, the data center industry is facing an increasingly clear challenge: power availability is becoming a key constraint on the expansion of AI infrastructure.
Recently, Project Kilby, promoted by Microsoft and Chevron, has attracted industry attention. According to public reports, the project plans to build approximately 2.67 GW of co-located natural gas power generation facilities in West Texas to provide long-term dedicated power for Microsoft’s AI data center campus. The project is expected to include natural gas power generation facilities, together with supporting step-up, distribution, control, and future grid interconnection infrastructure.
The significance of this type of project is that it is no longer a traditional “backup power” configuration. Instead, it represents a GW-scale behind-the-meter power system designed to meet the long-term operating needs of AI data centers. Public reports indicate that the project is expected to begin supplying power in 2028 and gradually scale up to 2.67 GW.
From Backup Power to Dedicated Power Infrastructure
In the past, data center power systems were mainly built around utility grid access, UPS systems, battery systems, and diesel backup generators. Backup power was primarily used to ensure business continuity during unexpected power outages.
However, in the AI data center environment, power demand is undergoing a structural shift. High-power GPU servers, large-scale cooling systems, and continuously operating AI training workloads are significantly increasing data centers’ dependence on stable, continuous, and large-capacity power supply.
As a result, more projects are moving away from a “backup power” mindset and toward a “dedicated power infrastructure” mindset.
Co-located generation projects such as Project Kilby reflect this trend: the relationship between energy companies, data center operators, and power equipment suppliers is becoming increasingly close.
New Opportunities for the Power Equipment Industry
For the power equipment industry, the opportunities created by AI data centers are no longer limited to the procurement of individual pieces of equipment.
Future demand may increasingly focus on complete power system solutions, including:
- Natural gas generator sets or gas turbine systems
- Step-up transformers
- Medium-voltage switchgear
- Protection and control systems
- Power distribution systems
- Energy storage systems
- Microgrid control systems
- Future grid interconnection infrastructure
Together, these systems form the critical power infrastructure behind AI data centers.
Under this trend, the ability to supply individual products remains important. However, greater opportunities may come from integrated, modular, and fast-deployable power solutions.
Impact on the Transformer and Power Generation Equipment Supply Chain
The rapid development of AI data centers will further increase demand for power transformers, distribution transformers, dry-type transformers, switchgear, and power generation equipment.
In large data center campuses, transformers are not only responsible for voltage conversion. They also directly affect system efficiency, power reliability, scalability, and the overall project delivery schedule.
At the same time, the development of on-site generation and behind-the-meter power systems will create more demand on both the generation side and the distribution side. This includes equipment such as step-up transformers, medium-voltage switchgear, paralleling control systems, and microgrid management systems.
This means that future data center power system construction will place greater emphasis on:
- Faster delivery
- Higher reliability
- Lower losses
- Stronger scalability
- Better system integration capabilities
Industry Outlook
The expansion of AI infrastructure is reshaping the cooperation model between the energy industry and the power equipment sector.
Energy companies are no longer only fuel or electricity suppliers. They may become long-term infrastructure partners in data center development. For power equipment companies, data centers are no longer just a standard end-use market. They are becoming a fast-growing application field with higher technical requirements and stronger demand for system integration.
In the future, companies that can provide stable, scalable, low-loss, and large-scale deployable power infrastructure more quickly may gain greater opportunities in the AI data center power supply market.