Contributed by Isaac Smith, Principal, PPM Consultants
In follow-up to a series of articles I wrote on Artificial Intelligence (AI), data centers, electric vehicles (EV), and the rapidly increasing demand for electricity, the issue has continued to evolve quickly. When the earlier articles were written, the main question was whether the U.S. electric grid could keep up with the growth in AI and data center development. That question still remains, but it is no longer the only one.
The more immediate question for many states, utilities, regulators, communities, and industrial customers is becoming much more practical: if data centers require major new generation, transmission, distribution, water, and backup power infrastructure, who pays for it?
That question is now being asked here in Alabama as well as other states across the country. The answer will have significant consequences for residential customers, industrial power users, utilities, data center developers, economic development agencies, environmental regulators, and the consultants who support them.
U.S. Data Center Power Demand Is Still Accelerating
The basic trend has not changed. Data centers are becoming one of the largest drivers of new electricity demand in the U.S. The scale of the projected growth is difficult to overstate. The Department of Energy and Lawrence Berkeley National Laboratory reported that U.S. data center electricity use increased from approximately 58 terawatt-hours (TWh) in 2014 to approximately 176 TWh in 2023, and could increase to a range of approximately 325 to 580 TWh by 2028. A 2025 update from Lawrence Berkeley National Laboratory extended the analysis to 2030 and estimated that data centers could account for approximately 11.8 percent of total U.S. electricity use by the end of the decade.
According to the Electric Power Research Institute (EPRI), an independent, nonprofit research organization focused on the electric power industry, their 2026 updated projection presents an even wider range of potential outcomes, predicting that data centers could consume approximately 9 to 17 percent of U.S. electricity by 2030, depending on how quickly AI demand, server density, efficiency gains, and project execution develop. Those numbers are not just abstract national projections. They are showing up in utility integrated resource plans, generation procurement decisions, transmission planning, local zoning meetings, water-supply discussions, and state public service commission dockets.
For industrial and environmental professionals, this means data center development should not be viewed only as a technology or economic development issue. It is now directly connected to air permitting, grid reliability, utility rate design, water supply, emergency generators, community engagement, and long-term infrastructure planning.
Alabama Is Now a Front-Line Example
Alabama has become a useful case study because the state is now seeing the same collision of economic development, power demand, utility regulation, and public concern that has already emerged in Northern Virginia, Ohio, Georgia, Texas, Pennsylvania, Arizona, Illinois, and other states.
A recent Inside Climate News article reported that more than a dozen data center developments have been proposed across Alabama. According to that analysis, if those proposed campuses were built to full capacity, they could use more electricity than all Alabama homes combined. The article also noted that some proposed projects are not minor load additions. In Lowndes County, representatives for the developer of Project Red Clay reportedly requested 1,500 megawatts (MW) of capacity from Alabama Power. In Bessemer, the proposed Project Marvel campus was described as requiring approximately 90 times more electricity than all Bessemer residences combined.
Those numbers explain why this issue is getting attention. Alabama is not just asking whether data centers bring investment and jobs. The state is now asking whether existing customers could end up paying for generation, transmission, distribution, fuel, taxes, stranded infrastructure, or other costs needed primarily to serve very large new loads.
Why Ratepayer Protection Is Becoming the Central Issue
Large customers generally pay large bills. In many circumstances, that can benefit a utility system because a large, stable customer may spread fixed costs across more sales and improve utilization of existing assets. That is the argument often made in favor of large industrial load growth.
The concern with data centers is different. Many proposed projects are extremely large, may require new infrastructure before they reach full operation, and may have uncertain buildout schedules. If a utility builds or buys resources for a speculative load that is delayed, downsized, or never fully materializes, the costs may not disappear. They can become stranded costs unless the contract, tariff, or regulatory order clearly assigns those costs to the data center customer.
That is why terms like minimum bills, minimum contract length, financial security, upfront payments, load ramp schedules, exit fees, contribution in aid of construction, fuel-cost recovery, and cost-of-service allocation are becoming central to the data center debate. These terms may sound like utility accounting details, but they are the mechanisms that determine whether costs stay with the large load customer or migrate to everyone else.
Alabama Senate Bill (SB) 270: A Step, But Implementation Matters
In Alabama, SB 270 was enacted in 2026 to add a specific review framework for large-load data center contracts. The statute applies to a large load data center customer with an expected total peak demand of 150 MW or greater at one or more contiguous parcels of land and directs the Alabama Public Service Commission (PSC) to consider whether the contract is in the public interest.
Importantly, the law focuses on incremental cost recovery. It defines incremental costs as additional costs associated with the generation, transmission, distribution, sale, or furnishing of electricity to the large load data center customer, including fuel costs and taxes, that the utility would not incur but for the data center contract. The law also identifies pricing, contract terms, minimum term length, upfront payments, security, collateral, and minimum payment as relevant considerations.
That structure is important because it puts the right concept on the table: the data center should pay the costs that exist because the data center exists. However, the real effectiveness of SB 270 will depend on implementation. If the review process relies heavily on redacted contracts, confidential terms, or general assurances that costs will not be shifted, the public may not be able to independently determine whether the protection is meaningful. If, however, the PSC requires transparent evidence of incremental cost recovery, binding minimum-payment obligations, enforceable exit protections, and clear allocation of fuel and infrastructure costs, the statute could become a practical tool for balancing economic development with ratepayer protection.
What Other States Are Doing
Alabama is not alone. Around the country, states and regulators are trying to solve the same basic problem: how to welcome economically attractive projects without asking residents and existing businesses to subsidize the infrastructure needed to serve them.
Ohio has taken one of the clearest utility-tariff approaches. American Electric Power (AEP) Ohio’s data center tariff, approved by the Public Utilities Commission in 2025, requires large new data center customers to pay minimum demand obligations tied to their contracted capacity for up to 12 years, including a ramp-up period. AEP Ohio has stated that the tariff is intended to ensure large-load customers pay their fair share and that existing customers are protected from costs associated with grid upgrades or infrastructure built for those large loads. AEP also reported that speculative data center requests dropped substantially after binding tariff obligations were adopted.
Michigan has also moved toward stronger contract protections. The Michigan Public Service Commission approved terms for very large customers that include exit-fee provisions, annual reporting, and commission review of large-load contracts. In a separate DTE Electric data center contract approval, the commission imposed mandatory safeguards and directed the utility to file a large-load tariff for very large customers such as data centers.
Pennsylvania has taken a broader permitting and policy approach. Governor Josh Shapiro’s 2026 executive action and Governor’s Responsible Infrastructure Development (GRID) requirements direct state agencies to require data center proposals to make legally binding commitments before state permit review proceeds. Those commitments include paying the full cost of new electricity generation, transmission, distribution, and related infrastructure needed to serve the project, while also addressing local approval, water use, transparency, labor, and environmental requirements.
Arizona and Illinois have addressed the issue through tax incentives. Arizona’s 2026 budget included a three-year moratorium on new data center tax credits while the state develops a more responsible path forward. Illinois paused processing new data center incentive applications beginning July 1, 2026, while the state evaluates longer-term reforms. Those actions do not directly set electric rates, but they recognize an important point: incentives can attract development, but if the public receives the cost risk while the developer receives the benefit, the policy may not be sustainable.
Georgia has also moved toward more customized treatment of very large loads. New AI data center and other very large customers at or above certain demand thresholds are being directed toward customized contracts and additional regulatory oversight rather than standard large-power tariffs alone. The concept is consistent with what is emerging elsewhere: very large, high-impact loads need a rate and contract structure that reflects their actual cost to serve.
The Air Permitting and Environmental Side of the Power Question
For environmental professionals, the power-cost question is only one side of the issue. New generation and transmission decisions can also affect air permitting, greenhouse gas emissions, water withdrawals, emergency engine permitting, stormwater, wetlands, land use, and environmental justice concerns.
The practical challenge is that the fastest path to power is not always the lowest-emission path. Data centers can often be designed and constructed faster than new generation or major transmission projects can be permitted and built. That mismatch can lead to renewed reliance on existing fossil generation, new natural gas resources, diesel-fired backup generators, or behind-the-meter generation. Each of those options can create environmental permitting and compliance obligations.
For example, a data center campus with large numbers of emergency engines may need careful evaluation under Clean Air Act (CAA) permitting rules, including potential applicability of New Source Performance Standards (NSPS) and National Emission Standards for Hazardous Air Pollutants) requirements, state air permitting thresholds, fuel-use limits, maintenance and readiness testing restrictions, recordkeeping, and whether generator operation could become more than emergency use. If the solution involves new simple-cycle combustion turbines, combined-cycle natural gas generation, or other dedicated power resources, then Prevention of Significant Deterioration (PSD), Nonattainment New Source Review (NNSR), Title V, state construction permitting, and air dispersion modeling may become part of the development timeline.
This is where environmental consultants and facility professionals can provide value. The question is not simply whether a data center is good or bad. The better question is whether the project has a defensible plan for power, water, emissions, reliability, cost recovery, and community impacts before commitments are made.
What This Means for Alabama
For Alabama, the timing is significant. The state already has relatively high average residential electric bills compared to many other states, and the public is understandably sensitive to claims that massive new loads will not affect household costs. Alabama Power has stated that data centers will be required to pay the full cost to serve their energy needs. SB 270 provides a legal framework for the PSC to evaluate whether that is true in specific contracts.
The next step is the critical one. The PSC’s implementation process will determine whether the law becomes a meaningful ratepayer-protection tool or simply a procedural checkpoint. In my opinion, the strongest approach would combine the following elements: a distinct large-load or data-center rate structure, transparent cost-of-service support, minimum payment requirements, enforceable financial security, clear allocation of fuel and infrastructure costs, annual reporting, and enough public disclosure to build confidence that residential and small-business customers are not subsidizing speculative load growth.
Alabama does not have to invent the entire framework from scratch. Ohio’s data center tariff, Michigan’s contract safeguards, Pennsylvania’s GRID requirements, and Arizona and Illinois’ pause on incentives all provide examples of tools that can be adapted. The best approach for Alabama may not be identical to any one of those states, but the principle should be the same: if a project creates the need for new utility infrastructure, the project should pay for the infrastructure.
Conclusion
The rapid growth of data centers presents real opportunities. These projects can bring investment, tax revenue, construction activity, and long-term technology infrastructure. They may also help accelerate investment in new generation, grid modernization, storage, nuclear, geothermal, renewables, and other power solutions.
But opportunity does not eliminate risk. Data centers can also create enormous new electricity demand, require major infrastructure investments, increase pressure on water resources, complicate air permitting, and create cost-allocation questions that are difficult to unwind after contracts are signed and facilities are built.
The practical path forward is not to ignore the demand or assume every project should be rejected. The practical path is to require good data, transparent review, enforceable contracts, appropriate rate design, environmental due diligence, and a clear answer to the question residents are already asking: who pays?
If data center developers, utilities, regulators, and communities can answer that question clearly before projects are approved, then data center growth can be managed in a way that supports economic development without unfairly shifting costs to households and existing businesses. If they cannot, the public backlash we are already seeing in Alabama and other states is likely just beginning.
If you’d like to see more articles like this on a weekly basis, consider subscribing to the PPM Environmental Journal. Click HERE to subscribe.

