Alabama communities are discovering that data centers are not ordinary commercial buildings. They are long-term commitments involving electricity, water, roads, tax policy, utility upgrades and specialized industrial equipment.
The building is the part residents can see. The contract is the part that determines who carries the risk.
A moratorium is not necessarily opposition
Tarrant’s City Council approved a one-year moratorium on data-center business licenses on September 8. City officials said the purpose was to study effects on electrical capacity, water resources, transportation, public safety, utilities, municipal services and surrounding neighborhoods.
Harpersville has taken a similar approach. Its town council adopted a six-month moratorium on August 3, expiring February 3, 2027, and scheduled informational sessions as part of a formal study.
Neither action establishes that data centers cannot operate in those communities. Both acknowledge a more basic problem: local governments may be asked to approve facilities whose most important characteristics have not been translated into ordinary zoning terms.
A typical zoning review asks where a building sits, how tall it is, how much traffic it creates and whether its use is compatible with nearby properties. For a large computing facility, those questions are necessary but incomplete.
Officials and residents also need to know:
- What is the facility’s peak electrical demand?
- Who pays for substations and transmission upgrades?
- What happens if the customer uses less power than promised?
- How much water does its cooling system consume?
- What emissions come from backup generators?
- How long is the utility contract?
- What happens if the facility is delayed, downsized or abandoned?
Those are not primarily architectural questions. They are infrastructure and finance questions.
The grid is where the project becomes real
WBC’s recent reporting on proposed Brookwood developments found that public discussion was focused on square footage, investment and jobs while key technical information—including electrical load, cooling design and water use—remained unclear.
That distinction matters because square footage is a poor measure of a data center’s effect on a community.
A large facility is effectively a concentrated industrial load. Its computers, cooling systems, power-distribution equipment, backup generators and communications links must operate as one system. The building may be quiet from the outside while placing demands on infrastructure that was not designed for such a customer.
Alabama’s Public Service Commission recognized this problem in July when it opened a generic proceeding to establish how contracts between Alabama Power and large-load data-center customers should be reviewed.
The commission’s action moves part of the debate away from local speculation and toward the terms of the electric-service agreement.
A data center does not simply connect to the grid in the same way as a store or small office. A sufficiently large facility may require new generation, transmission lines, substations and distribution equipment. Those investments can benefit the wider system, but they are also made because a particular customer wants service at a particular scale.
The question is not whether the facility consumes electricity. It is whether the facility pays for the additional system it requires—and whether the agreement protects other customers if the project does not develop as planned.
Alabama is trying to make “customer pays” enforceable
Alabama’s Act 2026-610 takes effect October 1. For data-center agreements involving loads of 150 megawatts or more, the pricing and terms must be expected to recover the incremental costs created by serving the facility and promote positive benefits for other retail customers. The law also directs regulators to consider economic growth in the community where the facility is located.
“Incremental costs” are expenses the utility would not incur without the new customer. Alabama Power identifies possible categories including generation, fuel, transmission, distribution and taxes.
The accounting can become complicated. If a utility builds a substation primarily to serve one facility, the arrangement may protect existing customers if the data center pays enough to cover the cost. If the facility later scales back, delays its opening or leaves, the utility may still own infrastructure with limited alternative use.
That is why large-load contracts commonly include minimum bills, minimum terms, upfront payments, security or collateral. Alabama Power says those provisions are intended to prevent project-specific investments from becoming the responsibility of households and smaller businesses if a customer changes plans.
The existence of those provisions is encouraging. It is not proof that every project will produce a good deal for the public. Their value depends on the numbers in the contract: the committed load, the payment schedule, the size of the required upgrades and the consequences of default.
The public may not see all of those numbers. Alabama Power says regulators and the attorney general receive complete, unredacted agreements, while public copies omit confidential commercial and financial information.
That arrangement may be understandable from a negotiating standpoint. It is also a limit on public oversight. Residents may be asked to accept road, water or land-use consequences without seeing the complete financial terms that explain the project’s public value.
Tax incentives make the timing more important
Alabama’s existing Chapter 9B program permits substantial tax abatements for qualifying data-processing centers. Projects granted abatements before January 1, 2027, may receive property-tax abatements lasting 10, 20 or 30 years depending on their investment levels, along with corresponding sales-tax abatements.
That does not mean every proposed project receives the maximum benefit. It does mean local officials may be evaluating long-lived infrastructure commitments while negotiating long-lived tax reductions.
The financial timeline matters. Construction jobs may be numerous but temporary. Permanent staffing may be far smaller than the facility’s capital cost or physical size suggests.
WBC’s Brookwood reporting noted that the public should distinguish construction employment, permanent on-site jobs, indirect employment, wages and the duration of each category.
A project can therefore be economically significant without becoming a major employer. Its value may lie in tax revenue, utility investment, construction activity or broader industrial development. Those benefits should be measured separately rather than combined into one impressive investment figure.
The next rule should be technical, not rhetorical
Tarrant’s moratorium and Harpersville’s study will be useful only if they lead to specific standards.
A meaningful local framework would require developers to disclose, before approval:
- Expected peak and average electrical demand
- The source and quantity of water required for cooling
- Backup-generation capacity, fuel storage and emissions
- Road, traffic and emergency-access plans
- Stormwater and drainage impacts
- Construction and permanent employment estimates
- Tax abatements and expected public revenue
- Utility-upgrade costs and who will pay them
- Minimum operating commitments
- Financial protections if the project is delayed or abandoned
Some information may legitimately remain confidential. The public should not be asked to accept the consequences of a project while being told that its most important obligations are private business information.
Brookwood’s rezoning decision will not settle these questions. A rezoning approval is not the same as a final construction commitment, and a permit is not proof that a facility will be built. But each decision can make future commitments easier to approve and harder to reverse.
The technology is not the difficult part. A data center is a building full of computing equipment supported by power, cooling, communications and security systems.
The difficult part is assigning responsibility for everything around that building.
Alabama’s moratoriums and utility proceedings show that local and state governments are beginning to understand the difference. The question now is whether they will finish the work before communities commit themselves to infrastructure contracts, tax abatements and public costs that last longer than the original sales pitch.
A data center may be advertised as an economic-development opportunity. Technically, it is first a promise about electricity.
The public should be allowed to read that promise before it signs on.
