Data Centers Can Help Lower Electric Bills—If We Make Them Pay Their Fair Share

By T.L. Headley, MBA, President
The Hedley Company

America's data-center boom has created a fair and urgent question for families and businesses: if giant new facilities require enormous amounts of electricity, will everyone else end up paying more for power?

The honest answer is that they could. If a utility builds new power plants, transmission lines, substations and transformers to serve data centers—and then spreads those costs across the bills of households, small businesses and existing industries—rates will rise. Dress it up however you like: that's a subsidy paid through the monthly electric bill.

There's another way this plays out, though.

When a data center is added to the grid on the right terms, it can help reduce the upward pressure on electric bills and, in some circumstances, actually lower the costs borne by families. It can bring a large, reliable new revenue stream into an electric system that already needs enormous investment. It can help pay for modernizing an aging grid. It can support dependable generation. And it can help spread fixed costs over more customers rather than leaving a shrinking or stagnant customer base to shoulder the entire burden.

The key is simple: data centers must pay for the infrastructure and reliable power they require, while also contributing fairly to the broader system they will use.

That's an argument for making data centers large, long-term, creditworthy customers who pay their own way, with nothing hidden in everyone else's bills.

The Grid Already Needs Modernization

The discussion often begins with the premise that data centers are creating the need for grid investment. In many places, that is incomplete.

America's electric grid is old by modern technological standards. Much of the transmission network, distribution system, substations, transformers and control equipment in service today was built decades ago. Some of it has performed remarkably well. But age, weather exposure, population shifts, cybersecurity risks and changing patterns of electricity use are all forcing utilities to replace or upgrade major portions of the system.

That work is necessary whether or not a single new data center is constructed.

Utilities must replace aging poles and wires. They must modernize substations. They must install more advanced control and monitoring systems. They must harden facilities against storms and outages. They must address shortages of large transformers and other critical equipment. In many regions, they must also expand transmission capacity so electricity can move reliably from power plants to homes, businesses and industries.

All of that costs money.

If the customer base is flat or declining, the same unavoidable costs are divided among fewer people. The result is predictable: higher rates for those who remain. A household may be using no more electricity than it used five or ten years ago, yet still see its bill rise because the cost of maintaining and rebuilding the system is being spread over a limited base of customers.

A large new customer can change that equation.

A properly structured data-center project can provide real new revenue to help finance upgrades that were already coming—upgrades families would otherwise have to fund largely on their own.

The size of a data center is a given. What matters is whether that size becomes a burden on existing customers or a source of additional revenue for the electric system.

Electricity Must Exist Before It Can Be Shared

There is one non-negotiable condition: the region must have enough dependable electricity to meet the new demand.

A data center cannot lower anyone's bill if its arrival creates an electricity shortage. If utilities do not have sufficient reliable generating capacity, or cannot deliver that power over the transmission system, then a major new load can drive up wholesale prices, force emergency purchases, increase the use of expensive short-term generation and weaken reliability.

That is the wrong way to grow.

The first responsibility of a utility and its regulators is to ensure there is enough dispatchable, dependable power to meet the needs of homes, hospitals, manufacturers, small businesses, public safety agencies and existing employers. Only then should the system commit to major new loads—or it should require the new customer to finance the generation and infrastructure needed to serve itself.

This is why reliable baseload and near-baseload generation remains central to the discussion. Coal, natural gas, nuclear and hydroelectric resources provide the dependable capacity needed to keep the lights on regardless of weather or time of day. A grid that lacks sufficient reliable generation cannot solve its problem simply by adding more wires, more computer servers or more policy promises.

Data centers can be part of the solution only if they are connected to a grid capable of serving them, or if they help pay for the additional reliable capacity required.

The Central Rule: The Customer That Causes the Cost Pays the Cost

Electric utilities recover their costs through rates. Those costs include fuel, power plants, transmission lines, substations, local distribution systems, maintenance crews, debt payments and a reasonable return on investment. Regulators often refer to the total amount a utility must collect as its “revenue requirement.” In plain English, it is the annual amount needed to operate the system, pay its bills and make the investments required to provide service.

The central fairness principle should be cost causation: the customer that causes a new cost should pay that cost.

If a data-center campus requires a dedicated substation, the data-center developer should pay for that substation. If it requires a larger transformer, a new transmission connection, an expanded switching station or a new generating resource, the developer should bear the cost through upfront contributions, special rates or binding long-term contracts.

The utility should not be permitted to build a costly new facility for a private customer and quietly shift the expense to everybody else.

At the same time, data centers should pay their fair share of the broader grid they use. They benefit from the existing network of transmission, grid-control systems, reliable generation, emergency reserves and the skilled workforce that keeps electricity flowing. Like any other large customer, they should contribute to maintaining and modernizing that shared system.

This is where a properly designed arrangement can benefit households.

If a utility's system already needs modernization, a large new customer paying a fair share of the shared costs brings in new revenue that was not previously available. That revenue can help cover necessary investments, reducing the amount that must be collected from existing customers.

The principle is no different from a well-run local water system. If a town must replace old pipes regardless, adding a large industrial customer that pays its full share can help spread the cost. But if the town must install an oversized new line exclusively for that customer and gives it a discount, homeowners are left with the bill. The outcome depends entirely on the terms.

Why a Large, Steady Customer Can Help

Data centers use a great deal of electricity, but their demand is often steady and predictable. They operate around the clock, every day of the year. Their servers do not work only from nine to five, and they generally do not disappear from the system during evenings and weekends.

That steady demand can be valuable to an electric utility.

The electric system has many fixed costs. Power plants, transmission corridors, substations, control rooms, skilled employees and maintenance programs must be maintained whether a particular customer uses a little electricity or a great deal. When a new customer pays for service over a long period, it helps cover a larger share of those fixed costs.

That is particularly important in places where manufacturing has declined, population has stagnated or major industrial customers have closed. The system may still contain infrastructure built to serve a larger economic base. Without new customers, the costs of maintaining that system must be recovered from the customers who remain.

A large data center can also give a utility more certainty. A binding 15- or 20-year service agreement can support financing for a new substation, transmission upgrade or modern grid-control system. The utility knows the customer will be paying for access to the system over time. That is far better than requiring households to finance infrastructure based on speculation that future growth may someday arrive.

But this works only if the commitment is real. A data center should not be allowed to reserve huge amounts of capacity, induce expensive construction and then walk away without financial responsibility. Contracts should include minimum-payment obligations, exit charges and other protections so ordinary customers are not left holding the bag if a project is delayed, reduced or cancelled.

A Plain-English Illustration: The “No Data Center” Case

Consider a simplified utility service territory with one million residential customers. The numbers below are hypothetical, meant only to illustrate how the economics can work.

Assume the utility must undertake $600 million per year in grid-modernization spending. This could include replacing aging transmission facilities, rebuilding substations, improving storm resilience, upgrading control systems and replacing critical equipment.

Importantly, assume this $600 million is needed even if no data center is built. The system is old. The upgrades are already necessary.

Under the utility's normal cost allocation, residential customers are responsible for 40 percent of those costs. Commercial and industrial customers pay the remaining 60 percent.

Table 1 — Baseline: No Data Center

Read the table from the top down. The utility needs $600 million a year for modernization no matter what gets built next door. Residential customers carry a fixed 40 percent of every dollar the utility collects, which puts their share at $240 million. Spread that across one million households and each one owes $240 a year — about $20 a month — before a single data center shows up on the grid.

Of course, a real electric bill contains many other charges: fuel, generation, transmission, distribution, taxes and various regulatory riders. The point of the example is narrower. It isolates one unavoidable category of cost: modernizing an aging grid.

The “Properly Added Data Center” Case

Now assume a large data center wants to locate in that service territory.

The facility requires new infrastructure specifically because of its size: a dedicated substation, additional transformers, transmission improvements and reliable generating capacity or contracted power. The annualized cost of those data-center-specific facilities is $200 million.

The first rule is that the data center pays the full $200 million. It does so through a combination of direct construction contributions, demand charges, long-term service payments and enforceable minimum bills.

That protects existing customers from the new costs the data center itself created.

But the data center also uses the existing and modernized shared grid. It benefits from the reliability improvements, controls, transmission system and other upgrades already needed for the region. Suppose its rate is designed to contribute an additional $40 million per year toward the broader $600 million modernization program.

The calculation then looks like this:

Table 2 — Properly Structured Data-Center Deal

Walk it line by line and the logic holds together. The $600 million baseline from Table 1 doesn't move. The data center's own $200 million in dedicated facilities — its substation, its transformers, its transmission upgrades — gets billed to the data center in full, so it never touches the shared modernization number at all. Its separate $40 million contribution toward the broader $600 million program is what actually changes the math: subtract that from $600 million and the shared cost still needing to be recovered drops to $560 million. Residential customers still carry their 40 percent of that reduced figure — $224 million, or $224 a year per household, about $18.67 a month.

The last two rows do the comparison for you: line that $18.67 up against the $20.00 households paid in Table 1 with no data center at all, and the gap is $1.33 a month — $16 a year — saved. That $16 isn't a separate number pulled from thin air; it's simply 40 percent of the data center's $40 million shared contribution, the same 40 percent share residential customers carry on every other cost in this example.

That might not sound dramatic in isolation, but it matters for two reasons.

First, this is only one large customer and one category of cost. A region that adds several properly structured large-load customers—or that requires a larger contribution from a single exceptionally large facility—can create more substantial savings or avoid larger future rate increases.

Second, the more realistic outcome may be avoiding a bill increase that would otherwise occur. If the grid needs modernization, the cost is coming one way or another. A new, creditworthy customer can help make the increase smaller.

In utility policy, preventing a $10-a-month increase is every bit as meaningful to a family as cutting an existing bill by $10.

What Happens If the Deal Is Poorly Structured?

Now consider the bad version of the same deal.

The data center requires $200 million in new facilities, but it receives a discounted rate or an inadequate contract. It pays only $80 million toward the costs it created. The remaining $120 million is shifted into the general rate base.

Assume residential customers pay 40 percent of that shifted amount. Their share of the subsidy is $48 million annually.

Table 3 — Poorly Structured Data-Center Deal

Same starting point as Table 2 — $200 million in facilities built specifically for the data center — but this time the data center only pays $80 million of it. The unpaid $120 million doesn't disappear; it gets folded into the utility's general rate base like any other cost of doing business, and residential customers absorb their usual 40 percent share of it, or $48 million. Divided across one million households, that's $48 a year, or $4 a month, added to every family's bill — and that's before counting any further costs from inadequate generation, higher wholesale power prices or emergency purchases.

That is why transparency is indispensable. The public should be able to see the proposed project, the necessary infrastructure, the cost of that infrastructure, the amount paid by the company and the amount left for ordinary customers.

The Game Plan: How to Make Data Centers a Benefit Rather Than a Burden

There is no mystery about what public officials, regulators and utilities must do. The state should welcome data centers, but it should welcome them on terms that protect ratepayers and improve the system.

First, require a credible power-supply plan before the project is approved. A data center should not receive firm service unless the utility can demonstrate that enough dependable generation and transmission capacity exist to serve the facility without jeopardizing reliability for existing customers.

Second, require the data center to pay for every direct cost it causes. That includes dedicated substations, new transformers, interconnection facilities, transmission upgrades and any new generating capacity needed specifically for its service.

Third, require long-term binding contracts. A 15- or 20-year contract with minimum-payment requirements gives the utility and its lenders confidence to finance infrastructure. It also protects households if the developer cancels, scales back or delays the project.

Fourth, establish a special large-load rate that reflects the true cost of service. The rate should include charges for energy, demand, transmission, generation capacity and grid modernization. Data centers should not receive below-cost “economic development” rates disguised as incentives.

Fifth, ensure large customers contribute to the shared modernization of the grid. If a utility must replace aging equipment anyway, the data center should pay its fair proportional share of those upgrades. That new contribution is the source of potential savings or avoided increases for families.

Sixth, use enforceable demand-response provisions where technically practical. Some computing workloads can be delayed, shifted or reduced during short periods of extreme grid stress. If a data center can reduce its demand during critical hours, it may help avoid expensive emergency generation or new peak-demand facilities. Any discount for this service should be tied to actual, measurable performance.

Seventh, make the deal public. The public service commission should review the contract, cost allocation and ratepayer protections. Commercially sensitive details may sometimes deserve limited confidentiality, but the central facts should not be hidden: what infrastructure is being built, who is paying and what protections apply if the project does not materialize.

Growth Must Strengthen the System

A state can attract major investment and protect electric customers at the same time.

Disciplined growth is the right approach: bring in new customers that help finance the electric system, but insist they pay for the reliable power and infrastructure they require, without rejecting every project out of fear or handing developers a blank check.

The American grid needs modernization whether data centers are built or not. Families and existing businesses should not have to carry that entire cost alone. Properly added data centers can bring new revenue, support the financing of critical upgrades, improve use of the existing system and help distribute unavoidable costs more fairly.

But the order matters.

First, make certain there is enough reliable electricity. Second, make the new customer pay for the costs it causes. Third, require a meaningful contribution to the shared system. Fourth, protect ratepayers with transparent contracts and enforceable obligations.

If those rules are followed, data centers can become more than large consumers of electricity. They can become part of the financial foundation for a stronger, more modern and more reliable grid—one in which families pay less than they otherwise would have paid to keep the lights on.

About the Author

Terry L. Headley, MBA, is President of The Hedley Company, a Charleston, W.Va.-based communications and research firm serving the energy sector, and founder of The Seneca Center for Energy and Critical Minerals Policy. He has more than 30 years of experience in Appalachian energy communications, including service as Communications Director for the West Virginia Coal Association and the American Coal Council. He publishes several weekly intelligence briefings covering coal, natural gas, the electric grid and critical minerals, and is the author of seven books.