We lay out where the noise comes from, what other towns have actually measured, and what is and isn't known for the two local sites. Where a figure comes from another city rather than from Prichard or Calvert, we say so.
The hum comes from the cooling, and it carries.
Most of the sound a data center makes comes from cooling. Fans and chillers run around the clock to keep the servers from overheating, and cooling is roughly 40 percent of all the electricity a data center uses. What people nearby tend to notice is not a loud roar but a constant low-frequency hum. That kind of sound travels a long way and, for some people, works its way through walls and windows at night. Reporting on data centers describes neighbors comparing it to a mosquito buzz or a drone that never stops.1,2
It is real, and it is hard to pin down with a meter.
The best-known case is Chandler, Arizona, where a data center gave off a hum that neighbors said rattled windows at night. The company traced it to its cooling fans and added muffling equipment, and the city went on to require new data centers to run a sound study before building, put mitigation in place, and take follow-up readings after opening. In Prince William County, Virginia, residents have measured data-center noise above 60 decibels at their homes. Part of what makes this so slippery is that much of the energy sits in the low-frequency range, which a standard decibel meter can understate, so a facility can look like it is within a limit on paper while still keeping people awake.3,4,5
The cooling that saves water can be the part you hear.
The closed-loop, radiator-style cooling both developers describe is good on water use, and we say so on the Water page. But it moves the heat with large banks of fans instead of by evaporating water, and big fan arrays are exactly the source behind the Chandler hum. That does not make it a dealbreaker. Equipment that is well designed and set back from homes can be kept quiet. It just means noise is a separate question from water, and a low water number is not the same as a low sound number.1,6
Prichard and Calvert.
Prichard (Edged). The Prichard site would sit close to homes in and around Africatown, so the distance to the nearest neighbors is short. Edged's cooling runs on fan-driven dry coolers, and the backup diesel generators are test-run on a schedule (see Diesel & air), which lays a louder, periodic noise on top of the steady hum. No sound study, no fence-line noise limit, and no set hours for generator testing have been made public for the site yet.6
Calvert (Beacon). The Calvert campus would be far larger, which means more cooling and more fans, on rural land where the background is quiet and sound carries farther. Beacon says the site will have no diesel generators, so the periodic generator noise would not apply here, though the round-the-clock fan noise still would. On its Calvert page, the company promises a minimum 1,000-foot setback between critical equipment and the property lines, and says "worst-case calculations predict noise at the property lines well below the allowable limit of 65 decibels… comparable to the noise of a household refrigerator." A household refrigerator is about 40 decibels; 65 decibels is closer to normal conversation, so those two comparisons describe very different levels. And "the allowable limit" is unattributed: unincorporated Mobile County has no noise ordinance we have found. The "worst-case calculations" themselves have not been published. The concrete ask, exactly what Chandler now requires, is to release the sound study and commit to follow-up readings once the site is running.7,8
What to put to the company and the boards.
Will there be a sound study before it is built?
Will the company run a pre-construction sound study and share the results, the way cities like Chandler now require?
What is the noise limit at the nearest home?
What decibel limit applies at the closest house, and does it account for the low-frequency hum instead of just an overall reading?
When are the generators tested, and how loud?
What days and hours will the backup generators run for testing, and how loud are they at the property line?
Every claim, sourced.
- Environmental and Energy Study Institute (EESI), "Communities Are Raising Noise Pollution Concerns About Data Centers."
- Data Center Knowledge, "What Are the 5 Main Causes of Noise in Data Centers?" (cooling as the leading source, roughly 40% of electricity use).
- Tom's Hardware, "AI data centers face increasing complaints about inaudible but 'felt' infrasound," 2025 (low-frequency sound understated by standard meters).
- Prince William Times, "Some cities suffering from data center noise turn to tough limits" (readings above 60 dB at homes).
- Data Center Dynamics, "CyrusOne forced to take action following noise complaints in Chandler, Arizona"; City of Chandler data-center zoning amendment requiring a sound study, effective Jan. 5, 2023.
- MEJAC, "Doing it Differently: Hard Earned Lessons from Engaging with a Proposed AI Data Center on the Africatown side of Prichard," 2026-05-26 (Edged's closed-loop, fan-driven cooling).
- WKRG, "Residents raise concerns over proposed data center in northern Mobile County," 2026-06-30.
- Beacon Data Centers, "Calvert Infrastructure Hub" campus page (published 2026-06-12, updated 2026-06-30): minimum 1,000-ft setback; "worst-case calculations predict noise at the property lines well below the allowable limit of 65 decibels… comparable to the noise of a household refrigerator." Company statement; the underlying study is unpublished. beacondatacenters.com
Keep reading.
Power & your bill
A data center runs around the clock. Who pays to upgrade the grid to feed it, and whether that shows up on nearby power bills.
02Water
How the cooling works, what "closed-loop" actually means on site, and how much water each project would draw.
03Diesel & air
The backup diesel generators, how often they run, and what they add to the air in an area that already lives with a lot of industry.
05Taxes & who pays
The tax breaks projects like these typically seek, the jobs required in return, and what a community trades away in the deal.
06Jobs, the real numbers
How many permanent jobs are actually promised, how many are in writing, and whether they'd go to people who live here.