AEP SOFC Data Center Model, $2.65 B Bloom Deal, 1, 000 MW Capacity, and AWS Project (2021-2026)

15 GW in New Requests, AEP’s Shift to On-Site SOFC Generation

Confronted with 15 GW of new data center power requests that overwhelmed its grid capacity, American Electric Power (AEP) executed a strategic pivot from a traditional grid supplier to a provider of utility-owned, on-site generation, using solid oxide fuel cells (SOFCs) to bypass multi-year interconnection queues.

• Between 2021 and 2024, AEP‘s primary challenge was managing a rapidly growing pipeline of data center interconnection requests, which surged to 15 GW by 2024. The utility’s response was conventional, focusing on planned grid upgrades with long lead times and proposing new rate structures to allocate the immense infrastructure costs.
• The strategy shifted decisively in 2025 with the execution of the on-site model. In June 2025, AEP secured regulatory approval from the Public Utilities Commission of Ohio (PUCO) to build, own, and operate on-site fuel cell installations for customers Amazon Web Services (AWS) and Cologix.
• This new direction was cemented in January 2026 with a landmark $2.65 billion agreement for up to 1, 000 MW of SOFCs from Bloom Energy. This move solidified the transition from planning to large-scale procurement and deployment for data center clients, directly addressing their critical speed-to-power requirements.

AEP $2.65 B Bloom Energy Deal, A Pivot to Customer-Sited Assets (2025-2026)

AEP‘s $2.65 billion commitment to Bloom Energy‘s SOFC technology represents a major capital allocation shift, moving investment from centralized grid infrastructure to decentralized, customer-sited generation assets that can be included in the utility’s rate base.

• The investment is structured around a 20-year offtake agreement, providing AEP with a long-term, predictable revenue stream tied directly to the operational needs of high-value data center customers in locations like Wyoming.
• This deal is part of a broader trend, with data center fuel cell agreements totaling $7.65 billion between October 2025 and January 2026. This aggregate investment indicates strong market confidence in this power solution to meet AI-driven demand.
• This large-scale, primary power model contrasts with smaller, more traditional grid-support projects, such as Fuel Cell Energy‘s $160 million, 7.4 MW plant in Connecticut. The difference in scale highlights the massive industrial shift driven by data center power needs.

Table: AEP Fuel Cell Investment in Context

On-Site Power Partnerships, AEP Aligns with AWS, Cologix, and Bloom Energy

AEP‘s strategy is built on a symbiotic ecosystem of partnerships, aligning its infrastructure and regulatory capabilities with a technology provider (Bloom Energy) and hyperscale end-users (AWS, Cologix) to solve the speed-to-power problem.

• The cornerstone of the strategy is the technology and procurement partnership with Bloom Energy. Through this alliance, AEP‘s unregulated subsidiary secured up to 1, 000 MW of SOFCs to create a standardized, rapidly deployable power solution.
• The customer partnership model was validated by PUCO’s June 2025 approval for AEP to build, own, and operate dedicated fuel cell facilities for AWS and Cologix in Ohio. This regulatory decision effectively established a new utility service category for large-load customers.
• This direct utility-customer-supplier model is gaining traction across the industry. For example, Bloom Energy has expanded its master services agreement with Oracle to procure up to 2.8 GW of fuel cell capacity, showing that technology providers are also building direct relationships with hyperscalers.

Table: AEP’s Key Data Center Power Alliances

Ohio as a Proving Ground, AEP’s Data Center Power Model

Central Ohio has become the primary testbed for AEP‘s utility-owned, on-site generation model, driven by its status as a major data center hub where grid constraints have become most acute.

• Between 2021 and 2024, activity in Ohio was characterized by escalating grid congestion as data center power demand surged. AEP‘s focus was on managing its interconnection queue and filing for regulatory changes to address the unprecedented load growth.
• Starting in 2025, Ohio transformed into a deployment theater. The PUCO approval in June 2025 for the AWS and Cologix projects, specifically the planned 72.9 MW facility in Hilliard, provided the legal and financial framework for the on-site model.
• While Ohio is the epicenter, AEP‘s $2.65 billion deal includes a project in Wyoming, indicating the strategy is not limited to one state and is designed to be a replicable template for other regions facing similar data center power deficits.
• The model’s expansion faces local friction, as seen in the lawsuit filed by the City of Hilliard in February 2026. This legal challenge over local zoning authority highlights a potential hurdle for future projects in other municipalities.

99.999% Uptime, SOFC Technology Matures to Primary Data Center Power

Solid oxide fuel cells (SOFCs) have transitioned from a niche technology for backup power to a mature, utility-grade solution capable of providing primary, baseload power for megawatt-scale data centers, a shift validated by AEP‘s multi-billion dollar procurement.

• Through 2024, fuel cells were often viewed as a supplementary or backup power source, competing with diesel generators but not yet widely adopted by utilities as a primary, grid-independent solution for large industrial customers.
• In 2025, AEP‘s adoption of [Bloom Energy]’s SOFCs for baseload power signifies a critical validation of the technology’s reliability. Vendors now market systems with up to “five nines” (99.999%) uptime, a key requirement for mission-critical facilities.
• The technology’s key advantages for this application include 15% to 20% higher efficiency than gas turbines, modular scalability, and lower criteria pollutant emissions, which directly address operator requirements for reliability, speed, and sustainability.
• The industry is coalescing around standardized offerings, such as Fuel Cell Energy‘s launch of a packaged 12.5 MW power block in March 2026, signaling that the technology is moving towards off-the-shelf, rapid-deployment products.

SWOT Analysis, AEP’s Utility-Owned On-Site Generation Model

AEP‘s on-site generation model is a powerful strategic response to data center demand, but its execution and scalability face regulatory, financial, and local jurisdictional risks.

• Strengths: The model directly solves the “speed-to-power” bottleneck for high-value customers, creating a new, long-term revenue stream for the utility while mitigating strain on existing transmission infrastructure.
• Weaknesses: The model’s initial success is heavily dependent on the availability and price of natural gas, exposing it to fuel price volatility and scrutiny over carbon emissions until a viable green hydrogen supply chain materializes.
• Opportunities: This approach is highly replicable and can be deployed by AEP in other states or by other utilities like [Duke Energy] and Southern Company facing similar grid constraints, positioning the utility as an indispensable partner to the tech industry.
• Threats: Local municipal opposition, as seen in the Hilliard lawsuit, poses a significant threat. This could create a patchwork of conflicting regulations that slows or halts project deployments and undermines the model’s key advantage of speed.

Table: SWOT Analysis for AEP’s On-Site Power Model

AEP Execution Risk, Hilliard Project and 1, 000 MW Bloom Option in Focus

The primary signal to watch in the coming 12 to 18 months is AEP‘s ability to execute on the 72.9 MW AWS project in Hilliard, Ohio, as its success or failure will determine the near-term scalability of the utility-owned on-site power model.

• If this happens: The Hilliard lawsuit is resolved in AEP‘s favor and construction begins on schedule in January 2026. Watch this: AEP will likely exercise a significant additional portion of its 1, 000 MW option with Bloom Energy for projects in other locations. These could be happening: Other utilities in data center-heavy regions will begin filing similar tariff structures with their state regulators.
• If this happens: The Hilliard lawsuit is upheld, granting the municipality veto power over the project. Watch this: AEP‘s deployment timeline for on-site generation in Ohio will be severely delayed, and the company may shift focus to its Wyoming project or other jurisdictions with more favorable state laws. These could be happening: Data center operators will revert to lobbying for faster grid upgrades or pursuing self-build power solutions.

The questions your competitors are already asking

This report covers one angle of the emerging utility-owned model for data center power. The questions that matter most depend on your work.

• What is actually happening with AEP’s AWS and Cologix fuel cell installations since the Ohio approval?
• What is the outlook for utility-owned SOFC deployment for data centers by 2030?
• What are the opportunities for other utilities to replicate AEP’s on-site generation model for data centers?

This report does not answer these. Enki Brief Pro does.

Your question, your angle, your framework. SWOT, PESTL, scenario modelling. The same niche depth, built around the decision your work actually depends on.

Run your first brief in Enki Brief Pro