So, I was catching up with a friend from the transmission team the other day, and he told me something that sums up where our industry is right now. He has two different piles on his desk. One is a stack of interconnection requests from renewable developers that he hasn’t touched in weeks because the grid is full. The other is a folder from a massive data center developer who wants to bypass the grid and connect directly to a nuclear power plant.
This is what folks are calling “co-location,” and it’s the hottest debate in the industry this week. If you haven’t been following the FERC dockets, here’s the gist: big tech companies are tired of waiting. They need huge amounts of power for AI, and they need it immediately. They’ve realized that if they build their data center next to a large power plant and connect behind the meter, they won’t have to wait for new transmission lines.
It sounds like a win-win: the tech company gains power, the plant owner secures a steady customer, and the grid doesn’t have to worry about the extra load. But for those of us in leadership and management, it’s a major headache. It forces us to confront a fundamental question about what a utility truly is: are we a platform for everyone, or can the biggest players opt out?
This isn’t a technical issue; it’s a management and fairness challenge. If the largest, wealthiest customers “direct connect” to the top power plants, who bears the cost of the rest of the grid? Regulators are already beginning to push back, and some of your fellow CEOs will go to war over this. As a leader, you must decide where your company stands. Do you pursue quick profits and immediate connections, or do you safeguard the long-term stability of rates for everyone else? This is the kind of decision that will shape your career in the coming years.
The Allure of the Direct Connect
Let’s consider this from a data center manager’s perspective. Your board has instructed you to get five new AI training clusters operational by 2027. You contact the local utility, and they inform you that the transmission upgrades required for your site will take ten years and cost half a billion dollars. By then, your project is essentially dead.
Then you find out there’s a nuclear plant or a large gas station 30 miles away with a few hundred megawatts of capacity. If you can buy that land next door and run a private line, you skip the ten-year wait. You’re online in two years. You’re beating your competitors to the punch.
For the power plant owner, it’s equally tempting. We’ve spent the last decade worrying that these plants will be “pushed out” of the market by cheap renewables or low prices. Now, suddenly, they have a customer willing to sign a twenty-year contract at a premium price. It’s the ultimate financial safety net.
This is why we’re seeing large deals—like the one with Amazon and the Susquehanna plant—making headlines. It’s a pursuit of certainty in an uncertain world. As a manager, you understand the logic. You see the revenue. You see the jobs. It’s tough to turn down that kind of economic momentum. But the logic begins to fall apart when you consider the grid as a whole. The grid isn’t just a bunch of wires; it’s a shared system for recovering costs. When you remove a large portion of generation and load from that system, you create a gap that everyone else must fill.
The Shifting Burden of Grid Costs
This brings us to the core of the “fairness” argument. The electric grid operates on the principle of socialized costs. We all share the expense of transmission lines and substations because we all benefit from a dependable system. When a data center co-locates at a power plant, they often claim they shouldn’t have to pay for those shared transmission costs because they aren’t “using” the grid.
But is that really true? If the data center’s direct connection fails, where do they get backup power? They rely on the grid. If the power plant they’re connected to goes down for maintenance, how do they stay operational? They draw power from the grid. Essentially, they use the entire power system as a giant, free insurance policy while refusing to pay the premium.
Other utilities are raising alarms. They’re warning regulators that if this trend continues, residential customers and small businesses will see their bills rise. If data centers aren’t covering their share of the “transmission and distribution” charges, the cost has to be subsidized somehow.
As a leader, you must navigate this political minefield. Supporting co-location could lead to accusations of letting “Big Tech” avoid paying the bill. Opposing it might label you as the bottleneck preventing the next economic boom. Some management teams are starting to suggest “hybrid” models—where data centers pay a “grid access fee” even if they are directly connected. It’s an attempt to find common ground, but it’s a tough sell to customers who believe they already found a way to save millions. This tension is becoming the main conflict in state regulatory hearings from Ohio to Virginia.
FERC and the New Rules of the Road
If you were hoping the federal government would stay out of this, you’re out of luck. FERC just issued new orders this month directing PJM and other grid operators to establish clear rules for co-location. They’ve realized the “wild west” era of these deals is over.
The regulators are concerned about two main issues: reliability and funding. Regarding reliability, they want to understand what happens if that power plant trips offline. If a 1,000-megawatt plant and a 500-megawatt data center are both connected at the same point, that’s a huge amount of power moving in an instant. The grid must be able to handle that sudden surge.
On the financial side, they’re examining “cost shifting.” They’re asking the tough questions that utility managers have avoided: how much of the existing grid was originally built to support that power plant? If that plant is now dedicated to a single customer, shouldn’t that customer cover the costs of the infrastructure connecting the plant to the rest of the grid?
For those of us in leadership, this means our “burden of proof” has shifted. It’s no longer enough to demonstrate that a co-location deal benefits your company’s bottom line. You must also prove that it doesn’t harm others on the system. This requires better data, improved modeling, and a transparent relationship with your commissioners. You need to show how much “grid insurance” that data center is actually using and its value. If you can’t quantify it, regulators will set the price for you—and you might not agree with their decision.
Managing the Hyperscaler Relationship
One of the most interesting management shifts I’ve seen this year is how we communicate with tech companies. For a long time, we treated them like any other large industrial customer. But companies like Microsoft, Google, and Amazon are not like a steel mill. They have their own energy teams that are often larger than our planning departments. They are sophisticated, move quickly, and have deep financial resources.
They are also beginning to realize that the “grid bypass” strategy carries its own risks. If they become seen as responsible for rising residential power bills, the political backlash could severely harm their reputation. We’re witnessing a new form of stakeholder management. Microsoft, for instance, recently committed to a framework that ensures its data centers “pay their way” so they don’t increase rates for neighbors.
This is a great opportunity for utility leaders. Instead of viewing these customers as opponents trying to cut costs, we should see them as partners who can help finance the grid of the future. The best managers are having “honest” conversations with these tech giants. They’re saying: “Look, we want to help you build fast, but we can’t let our other customers suffer. Let’s find a way for you to fund the transmission upgrades we both need.”
It’s a shift from a “transactional” relationship to a “strategic” one. It demands a different set of executive skills—part negotiator, part economist, part community leader. The utilities that succeed are those that have established dedicated task forces reporting directly to the CEO. They aren’t just processing orders; they are co-creating a new energy system that balances speed with fairness.
The Engineering of Flexibility
Even if we resolve the issues of money and politics, we still face the challenge of physics. Co-location presents a unique operational difficulty for our engineers. When you have a large load and a substantial generator “behind the meter,” traditional grid management rules don’t always apply.
We need software and hardware to control these “islands” of power. We must access the data center’s energy management system to monitor its activity in real time. If they suddenly reduce 200 megawatts of load, we have to know so we can balance the rest of the grid.
This requires a level of integration that most utilities are still struggling to achieve. Our control rooms are accustomed to managing assets we own. Now, we have to oversee assets that our customers own, which are deeply integrated into our system.
Middle managers serve as the link here. You need to persuade your operations team that allowing a customer this influence on the grid is safe. You also have to persuade the customer that sharing data and control with you benefits them. It’s about cultivating a trust as strong as the physical wires.
We’re also seeing a rush for new technologies like “islanding” breakers and fast-acting flow controllers that can isolate a co-located site if something goes wrong. These are the tools that make co-location possible without risking a regional blackout. The managers who invest in these technologies now will have a significant competitive edge when the next hyperscaler seeks a site. They will be the ones who can say “yes” when others are still trying to figure out the safety protocols.
Conclusion
The co-location debate is more than just a regulatory disagreement. It signals that the old “one-size-fits-all” utility model is breaking down under pressure. We are moving into an era where the grid is no longer a single, unified system but a network of interconnected hubs and “energy parks.”
As a leader, you can’t afford to be a bystander in this conversation. You must be the one defining what “fairness” looks like in 2026. You need to explain to your board that a quick win on a co-location deal today could lead to a decade of regulatory issues if the costs fall on the wrong people.
The way forward is rooted in transparency and collaboration. We must be truthful about the costs of the grid, the advantages of data centers, and the compromises we are making. We need to partner with tech companies to create a system that is rapid enough for their needs and affordable for everyone else.
The “co-location conundrum” isn’t going away. It will be the key management challenge for our industry over the next five years. The decisions you make this month will determine whether your utility is seen as a leader of the new economy or a barrier to it. It’s time to choose a side—and ensure you can defend it to your customers, regulators, and your own conscience. Let’s get to work and build a grid that benefits everyone, not just those with the shortest lines.