An imbalance has emerged between the amount of energy needed to meet demands in the United States and the ability to quickly bring energy projects online to supply that need. As generative AI continues to advance and become more mainstream, so too do announcements that AI developers are investing in data centers and energy generation projects to help fuel their technological innovations. Some hyperscaler data centers can require more than a gigawatt of power for a single location—enough power to serve approximately three quarters of a million average homes. These circumstances present a tremendous opportunity for the energy industry, but it faces significant headwinds from a generation interconnection process that remains inefficient, slow, and costly.
Generative AI is extremely energy intensive, with a single query estimated to require nearly 10 times the amount of electricity needed for a traditional internet search. As AI becomes more popular, energy demand to fuel these queries will continue to rise and in turn more energy generation projects will need to interconnect to the electric grid.
Recognizing that interconnection queues around the country are already bloated, the Federal Energy Regulatory Commission (Commission), which regulates generation interconnection in most of the continental United States, and regional transmission organizations (RTOs), which study and approve generator interconnection requests through their “queuing” process, are actively searching for solutions to expedite those interconnections.
Several factors contribute to the interconnection process run by RTOs across the country being slow and costly. For example, in the past requests to interconnect to the electric grid were mainly submitted by large-scale energy generators. Today, more and more interconnection requests are made by smaller-scale energy generators, frequently renewable generation, resulting in a significant spike in the total volume of requests in the interconnection queue. This relative decreased capacity-per-project results in a larger total number of projects to be processed.
More projects in the queue also means more studies to be conducted for each project as each interconnection request generally must receive satisfactory results in at least two studies before any project will receive an agreement setting forth the terms of interconnection to the electric grid. The rising complexity of assessing grid reliability has resulted in projects being subject to progressively longer wait times in the queue and overall delays before those projects interconnect to the grid. For example, projects built in 2023 are experiencing on average five years between when its interconnection request is made to commercial operation. This is a significant increase over the typical time in years past.
Other contributing factors to these delays are increasing baseload capacity retirements, which creates resource adequacy concerns, and rising costs due to inadequate and aging transmission infrastructure that often needs to be constructed or rebuilt to accommodate new generation. These changes both increase the cost of network upgrades needed to interconnect new generation and also contribute to the complexity of the engineering studies needed for new interconnections.
The Commission has recognized that longer timelines between interconnection request and commercial operation are creating a bottleneck effect, preventing much-needed new generation from coming online. It sought to address this issue, and interconnection shortfalls more broadly, with Order No. 2023. Through the order the Commission imposed several requirements on both project developers and RTOs with the ultimate goal of making interconnection more certain.
Importantly, the order requires transmission providers to study projects in “clusters” instead of individually and implements penalties on providers who fail to complete studies by a certain deadline. The order also increases the preparedness requirements for new generation, particularly with strict site control obligations, and increases the financial security that must be posted through the interconnection study process.
These are significant costs and, in theory, will preemptively weed out less certain projects, reducing the sheer number of projects that need to be studied. At the same time, this should also increase the likelihood that projects in the process will move forward and enter commercial operation.
While it has been almost two years since Order No. 2023 was issued, the Commission is still in the process of reviewing and approving utilities’ and RTOs’ plans for how to comply with its requirements. As such, how effective Order No. 2023 will be in making the interconnection process more efficient remains to be seen.
Like the Commission, RTOs are also trying to improve their interconnection processes. Many have proposed several solutions meant to alleviate some of the pressures on the electric grid and interconnection:
Another potential solution to remedy long interconnection queues and inefficient interconnection processes, particularly for data centers, is generator co-location wherein data centers are strategically sited alongside power generation facilities in so-called “behind-the-meter” arrangements. This type of configuration allows projects to bypass the need to interconnect to the electric grid, and thus the interconnection process entirely, by getting energy from generation facilities directly.
Even as behind-the-meter configurations become more popular, the need to interconnect to the electric grid is not going away and will remain the path forward for the majority of energy projects coming online in the future. This reality means that efforts to improve the interconnection process will remain critical as energy demand continues to rise.