Developers want airport land, grid capacity and water. Airports should know what they need before they sign anything.

The New Neighbour Wants Your Power

Airports have always attracted neighbours. Flat land, good access, large parcels under one owner. Logistics parks, hotels, rental car facilities.

The new one is different. The data center developer wants land, yes. But what it really wants is electrical capacity. That is something airports have never had to ration.

This is not hypothetical. In Ames, Iowa, LightEdge has proposed leasing about 11 acres of city-owned airport land along Aviation Way. Full buildout could draw up to 25 MW. The city was clear that any deal needs FAA approval, including a land release or land-use change and confirmation that the proceeds benefit the airport.

The U.S. Air Force has gone further. It is soliciting AI data centers on underutilized land at airfields including Davis-Monthan, Edwards and Robins. Minimum load: 100 MW. Minimum investment: $500 million.

Most airport operators have no method of evaluating these proposals. Economic development agencies and local governments push hard for the tax base. The airport is often the last to be consulted and the first to live with the consequences.

ACRP has recognised the gap. New research will start with a First Look on the issues airports should weigh, followed by work on planning, utility coordination, operational impacts, mitigation and legal questions. Good. It is overdue.

Two Demand Curves, One Grid

Start with the numbers. Lawrence Berkeley National Laboratory puts U.S. data center consumption at about 176 TWh in 2023. That is 4.4% of national electricity.

By 2028, the range is 325 to 580 TWh. That is 6.7% to 12% of the total. Demand more than doubled between 2017 and 2023 alone. AI is the driver.

Now take one project. A 100 MW campus at full load draws about 876 GWh a year. That is more than 80,000 average American homes. Utilities cannot serve that without new substations, new lines and sometimes new generation. Those queues run for years.

Airports are on their own curve. GSE electrification. PCA and 400Hz at every gate. Electric buses and fleet vehicles. EV charging for passengers and rental cars. AI-enabled operations. Terminal expansion. Further out, electric and hybrid-electric aircraft and AAM vehicles needing high-power charging on the apron.

Very few master plans have quantified this properly. Fewer still have secured the utility capacity to serve it.

Put a data center on the same substation or feeder and you have two customers competing for the same scarce capacity. If the data center takes the headroom today, the airport waits years and pays more when it needs to electrify tomorrow.

That is the single biggest risk. And it is easy to miss when the proposal arrives dressed up as a land lease.

Water, Plumes and Land

Water comes next. Berkeley Lab estimates U.S. data centers used about 17 billion gallons directly for cooling in 2023. That could double or quadruple by 2028.

For airports in water-stressed regions, especially the Southwest, this matters. It competes with terminal growth, ARFF reserves and community goodwill. In Tucson, local groups are already campaigning against regional data center proposals on water grounds.

Plumes are specific to us. Evaporative cooling towers and large heat-rejection systems push warm air upward, sometimes with visible vapour. The FAA treats thermal exhaust plumes from power plants and cooling towers as incompatible with airport operations during critical phases of flight. It rates overall risk as low but directs airport sponsors to assess them with the MITRE Exhaust Plume Analyzer.

A large campus under an approach or departure path, or close to a GA or helicopter traffic pattern, needs that assessment. Along with the usual Part 77 review of building and stack heights.

Then land. A data center is a long-life, high-value, immovable asset. Put it on land the ALP will need for cargo, MRO, fuel or apron expansion and you lose that option for decades.

Add the smaller items. Diesel backup generators and air permits. Detention ponds that attract wildlife. Rooftop equipment and lighting that need checking for glare and interference with NAVAIDs.

Know Your Federal Obligations

For a federally obligated airport, this is not just a local planning call.

Grant assurances apply. Airport revenue stays with the airport. Non-aeronautical land is leased at fair market value. The airport stays as self-sustaining as possible. Land needed for aviation is protected. Land acquired with federal funds, or shown as aeronautical on the ALP, will typically need FAA concurrence and possibly a formal release. The ALP must be updated.

These obligations are not only constraints. They are leverage. They give the airport solid ground to insist on market rent, escalation, protective lease terms and full cost recovery on utility upgrades.

Ames makes the point. City officials noted that leasing airport property gives the city tools that zoning alone does not, including making the developer, not existing customers, carry the infrastructure costs.

Airports that understand this negotiates from strength. Airports that do not get handed a deal already shaped by the developer and the city.

There Is an Upside

We are not saying every proposal should be turned away.

Structured well, a data center can be a very good non-aeronautical tenant. Long-term ground rent with escalation diversifies revenue away from passenger volumes.

A developer-funded substation, sized for the airport's future load as well, delivers capacity the airport could never fund on its own. Shared generation, battery storage or a microgrid can improve resilience for critical airport systems. Waste heat recovery, still rare in the U.S., could serve terminal buildings or de-icing facilities.

Good outcome or bad outcome, it is decided by the terms agreed before construction starts. To write those terms, the airport has to know its own future needs for power, land and water.

Six Tests Before You Say Yes

Until ACRP's First Look is out, we recommend airports screen every proposal on six tests.

Power. Does the project use utility capacity the airport will need over the next 20 to 30 years? Will the developer fund upgrades sized for both parties?

Plume. Has a thermal exhaust plume assessment been done against current and future approach, departure and pattern airspace? Has the Part 77 review been completed?

Place. Is the parcel truly surplus to the ALP's long-term aeronautical needs? What will the FAA require to release or reclassify it?

Pipes. What is the water source, peak demand and discharge? Does it compete with airport or community needs?

Permits. Are grant assurance obligations met, including fair market value, revenue use and ALP approval? How are generators, noise and wildlife handled?

Partners. Are the utility, FAA, local government, airlines and community engaged early? Is the allocation of cost and risk clear?

Fail any one of these and the proposal should be restructured or declined. However attractive the headline rent.

Avinia's View

Do your homework before the developer walks in.

Build a long-range electrical load forecast. Cover terminal growth, GSE and gate electrification, AI and digital systems, and an electric aviation scenario. Agree a capacity reservation strategy with your utility.

Identify which parcels are truly surplus under the ALP and which must be protected for aviation. Put a plume and airspace screening process in place.

With that done, a data center enquiry becomes a structured negotiation. Without it, it is a reactive scramble.

Conclusion

Data centers want to be near airports for the same reasons airports are valuable: land, connectivity and access to power.

That is exactly why airports need to be careful. The power, land and water a data center uses are the same resources the airport needs to grow and decarbonize.

Treat it as a real estate transaction and you may quietly mortgage the airport's future. Treat it as an infrastructure partnership, with the airport's long-term needs protected first, and it can make the airport stronger.

Source: Avinia analysis of ACRP's newly scoped research on data center development and airports (First Look and follow-on research); Lawrence Berkeley National Laboratory, 2024 United States Data Center Energy Usage Report; FAA thermal exhaust plume guidance