Co-op Webinar: How to analyze grid impacts of EVs and Solar in just 10 weeks. Register now
The U.S. power grid is under stress. The explosive growth of AI and cloud computing is accelerating electricity demand—just as utilities are racing to deal with interconnection backlogs, aging infrastructure, and limited visibility into local constraints.
For data center developers, this creates a painful bottleneck. Today, it can take 5 to 8 years to get a new data center connected to the grid. That’s assuming you're even in a position to proceed—many projects are delayed indefinitely due to upstream constraints on generation, transmission, or local distribution networks.
In response, some developers are turning to behind-the-meter fossil generation as a workaround. In Memphis, a proposed xAI data center recently made headlines after plans emerged to power the site with on-site gas generators in lieu of waiting for grid power. But this approach isn’t fast either: GE Vernova is reportedly sold out of industrial gas generators through 2028. It’s also expensive, carbon-intensive, and increasingly out of step with public and regulatory expectations.
Whether you wait for grid upgrades or go off-grid, the result is the same: slower deployments, higher costs, and more carbon emissions.
At Camus, we’re working with utilities and developers to implement flexible interconnection as a faster, smarter way to bring new data centers online. This blog outlines how Camus' software platform supports flexible interconnection—what it looks like in practice, what it requires from both parties, and why it’s emerging as a critical enabler of speed, reliability, and decarbonization.
It’s important to understand that flexible interconnection is not demand response. Demand response is a cost-savings mechanism—typically used after a site is connected—to reduce demand during peak hours. Flexible interconnection, by contrast, is a planning and approval pathway. It enables utilities to say “yes” to projects that infrastructure constraints would otherwise stall, by incorporating flexibility into the interconnection study itself. It’s the difference between waiting 5–8 years… and turning on in 12-18 months.
Flexible interconnection can sound abstract at first, but it’s built on a few simple principles. In the sections below, we walk through how it works in practice—from the structure of the connection itself, to the signals sent to the site, to the data and coordination required to make it successful.
Flexible interconnection enables developers to unlock significantly more capacity by agreeing to be flexible during rare periods of grid stress. In the most dynamic model, grid operators forecast available capacity 48 hours in advance and update operating limits hourly. These limits are sent to the site’s EMS (Energy Management System), which orchestrates local assets like batteries, generators, or load shifting.
Simpler versions are also possible, such as seasonal or time-of-day schedules. In every case, Camus’ orchestration platform provides the data, forecasts, and visibility needed to maximize the use of existing grid infrastructure safely.
Some developers may opt for a phased approach. For example, a 75 MW site might start with a 25 MW firm grid connection, using behind-the-meter generation to serve the rest. Over time, as upgrades are made or constraints ease, the utility increases the grid service level. This model gives developers a path to scale—even in constrained locations.
This flexible approach also enables cleaner models of service. With solar difficult to site directly behind-the-meter, some developers are exploring “energy park” models, where nearby generation or storage is deployed on the distribution network—not behind the same meter, but still in close proximity. These assets serve the broader grid and help mitigate constraints that would otherwise limit interconnection.
Operating limits are typically sent in the form of MW caps per hour, updated 24–48 hours ahead and refined hourly. Where ramp rates are a concern, limits can also be sent in 5-minute intervals to prevent sudden swings in load that could disrupt grid stability.
The EMS receives these limits and orchestrates site resources accordingly. Most sites will rely on local flexibility—dispatching batteries, solar, or generators—to keep within their defined cap during constrained periods.
To enforce limits and build utility confidence:
Failsafe mechanisms are also evolving. Tripping an entire 100+ MW site offline isn’t practical—it could destabilize the grid. Instead, utilities and developers are working together to define granular “big red button” controls, such as cutting power to a row of servers or a single building. These are true last-resort measures, only used after multiple alerts and recovery steps.
Most data centers already have significant behind-the-meter assets—natural gas generators, battery storage, and increasingly, solar generation. Under flexible interconnection, these systems can be used to deliver flexibility, supplying power during constrained hours when grid access is temporarily limited.
This isn’t a full off-grid model—it’s about using the grid when it’s available and dispatching local resources only when needed. It allows developers to speed up timelines, reduce infrastructure costs, and maintain alignment with long-term decarbonization goals.
The specific amount of capacity that can be unlocked through flexible interconnection varies by site and constraint type. But examples and studies point to massive potential:
The grid can handle far more than it seems—if we manage it dynamically.
Flexible interconnection is made possible by a solid foundation of data and coordination—but it doesn’t require a massive overhaul of existing systems. In fact, most of what’s needed is already available or already underway. Camus' orchestration platform is designed to connect the dots—bringing together data, analytics, and control to make flexible interconnection fast, safe, and scalable.
Camus integrates data from both utilities and data center operators to build a shared view of grid conditions and site performance.
From utilities, we pull in:
From data centers, we incorporate:
Together, this data foundation powers time-series simulations, dynamic operating limits, and real-time monitoring—all within the Camus platform.
Flexible interconnection doesn’t just speed up connections—it improves how utilities and data center operators manage the grid overall.
For utilities, Camus provides the ability to simulate flexible interconnections in planning studies, forecast grid constraints, generate operating envelopes, and monitor site compliance—all using tools that build on existing operational workflows. Our platform integrates with utility SCADA and EMS systems and supports closed-loop controls when needed, giving operators full confidence without requiring manual oversight.
For data center operators, flexible interconnection works best with an energy management system (EMS) that can receive operating limits and coordinate behind-the-meter resources like batteries, solar, and generators. The Camus platform shares real-time limit updates and compliance signals with the site controller and supports visibility into both firm and flexible capacity delivery over time. These capabilities not only enable flexible interconnection but also strengthen internal resource management and reliability.
In short: flexible interconnection isn’t about building something new from scratch. It’s about using the right data and tools to unlock the full potential of the grid—and connecting faster, more affordably, and more cleanly as a result.
Flexible interconnection should be considered as soon as it’s clear that a standard interconnection isn’t viable—whether due to long queue times, study results showing expensive upgrades, or site-level constraints.
Right now, flexibility is a strategic advantage. Most utilities don’t yet offer a formal “flexible queue,” but many are eager to work with developers who bring viable flexibility plans. Acting early gives developers a chance to secure high-value sites—and connect years ahead of standard timelines.
Over time, flexible interconnection will likely become the default model in constrained areas. But for now, it’s an edge that can unlock otherwise inaccessible locations.
The energy transition demands urgency, and the grid can’t be rebuilt overnight. But with flexible interconnection, we can make smarter use of the capacity we already have.
Instead of overbuilding diesel systems or waiting 5–8 years for upgrades, developers can connect now, scaling clean energy, reducing emissions, and accelerating digital infrastructure.
At Camus, we’re helping data center developers and utilities make this future real.
👉 If you're planning a new data center and want to connect faster, get in touch.
