On September 30, one of the few national baselines for secure, resilient, and accountable government data centers effectively sunsets, and no clear successor framework is moving to replace it. The law covers federal facilities, not the broader mix of private and commercial data center projects now reshaping power demand, land use, water planning, and local politics across North Carolina. Its lapse still matters, because it shows where data center policy is heading. There is no single national rulebook coming for this buildout.
The rules are being written in state legislatures and utility commissions, one bill and one rate case at a time.
That is the larger shift now underway. The data center rulebook is moving out of Washington and into the places where projects are actually reviewed, challenged, financed, and connected to the grid. Federal standards may be fading, but the policy fights are not. They are showing up in state legislatures, county boardrooms, utility proceedings, water debates, incentive negotiations, and the local land-use decisions that determine whether a project can actually move forward.
That is the backdrop for Senate Bill 730, the Ratepayer Protection Act, which passed the House in early June and now sits in the Senate. The pattern is familiar: the provisions that draw the headlines are rarely the ones that decide who pays.
The demand is real, and lopsided
Start with the scale, because it explains the anxiety driving the bill. Data centers are driving enormous electricity demand, and the grid was not built for this kind of load growth overnight. Duke Energy has signed agreements for 7.6 gigawatts of data center demand, with nearly two-thirds already under construction, and the company has been in advanced discussions for far more.
That pressure is not just a Duke issue. It is becoming one of the central utility-policy questions in the country: how to serve extraordinary new power demand without making ordinary customers carry the risk. In North Carolina, the question is especially immediate because a narrow set of very large customers is driving a wide share of future system investment.
Someone pays for that. The only question worth arguing about is who.
The bill gets the core right
On that question, Senate Bill 730 is strong where it counts. Section 4 requires every data center to sign an electric service contract with a minimum term, minimum billing, and terms that recover the utility’s incremental cost and prevent one class of customers from subsidizing another. Section 8 commissions a study of a large-load tariff. The North Carolina Utilities Commission and its Public Staff already govern how those costs are allocated.
These are the tools that determine who pays, and they are the most durable ratepayer protections in the bill.
They also work. In late June, Duke Energy proposed a large-load tariff before the North Carolina Utilities Commission. Under that approach, data centers and other major power users would face minimum billing obligations for at least a decade, whether they use the full amount of power they requested or not. The proposal would move large-load policy away from confidential, one-off agreements and toward a more transparent framework.
Duke has also lowered its residential rate request from 18 percent to 11.6 percent over two years and removed large-load-related upgrade costs from the pending rate case. That is cost-causation operating in real time, through contracts, tariffs, and regulators, not through a technology mandate.
Where it goes wrong
The bill runs into trouble when it moves from who pays to how to build. A statewide ban on evaporative cooling, written into new G.S. 143-355.5A, can push projects toward more energy-intensive cooling in places where water was never the binding constraint. A blanket prohibition on local economic development incentives strips counties of a tool their competitors still use, even when those incentives are disclosed, clawed back, and tied to measurable public benefits.
Ambiguous language on phased campuses leaves projects that have already broken ground guessing at the rules. None of these provisions adds meaningful ratepayer protection. All three add cost, uncertainty, or both.
The concern from local governments is not imagined. Power costs are only one piece of the issue. Counties are also dealing with water supply, noise, emergency services, land-use compatibility, and public backlash from residents who feel these projects arrive fully formed before the community has any real leverage.
A recent News & Observer op-ed on North Carolina’s data center backlash made the point plainly: a pause is not always a no. Often, it is a community asking for the tools, transparency, and leverage it should have had before the project reached a vote.
A bill that protects ratepayers but ignores those concerns will not solve the political problem. It will simply move the fight back to county commissions.
That is why the better answer is not to strip communities of leverage or pretend every project deserves approval. It is to build a framework that distinguishes responsible projects from weak ones. When data center projects are done right, they can strengthen the tax base, support infrastructure investment, and help large customers carry their proper share of system costs.
But that only happens when the project is sequenced correctly, local concerns are addressed early, and the public can see why the deal makes sense.
This is not simply a partisan divide. The free-market John Locke Foundation has urged the Senate to narrow the bill to its ratepayer core and drop what it calls the command-and-control provisions. Democrats have warned that the energy mandates bundled into the same bill could raise costs rather than lower them. When a measure draws the same structural critique from the right and the left, that convergence is worth reading. The ratepayer protections have broad support. The prescriptive add-ons do not.
Why pushing projects away does not solve the problem
There is a tempting objection. If a data center is expensive to serve, let it build in another state, and the cost leaves with it.
That is half right. The upgrades built for a specific project are never built if the project never comes. But the grid these customers join already faces major investment for reliability, generation, and transmission. A large customer that pays its proper cost of service helps carry those shared costs.
Push it to another state and North Carolina avoids the project-specific upgrades but forfeits the contribution that customer would have made to the costs the system faces regardless, along with the tax base and construction that come with it.
North Carolina ranks thirteenth in the country in data centers, behind Virginia, Texas, Georgia, and Florida. It is competing for this investment, even as it manages the costs and concerns that come with it. Driving projects to those same competitors does not protect ratepayers. It just moves the benefit across the state line.
The stakes are higher because Washington left the field
This is where the expiring federal law returns to the story. When a federal baseline exists, a state can afford to be a follower. With that standard lapsing and no clear replacement in sight, the states are the arena, and the early movers set the template other states copy. North Carolina is on track to write some of the most restrictive data center rules in the country.
It can write rules that protect ratepayers, respect local concerns, and still compete for investment. Or it can write rules that protect ratepayers on paper while exporting the tax base. The difference lives in the details.
That uncertainty is already producing consequences. The pause heard across the Carolinas was not an isolated reaction. It was a warning. When communities do not trust the process, they reach for moratoriums, delays, and broad restrictions. A state framework should reduce that pressure, not add another layer of confusion.
The better approach is narrow, pro-growth, and still protective of ratepayers. Keep the contract requirements, the tariff study, and the Commission’s oversight. Preserve water protection, but tie cooling requirements to local water conditions rather than impose a one-size-fits-all technology mandate. Replace the blanket incentive ban with disclosure, clawbacks, and clearer public accounting so communities can see the costs and benefits before they commit.
Clarify the phased-campus language so projects already in the pipeline are not left guessing at the rules.
The title of the bill is right. The task now is to make its mechanics match it, before the text is final and the moment to shape it is gone.