Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124

Project Ruby has formally requested 600 megawatts of electricity — more than the entire current consumption of the city of Columbus. At that power load, a facility using only 330,000 gallons of water per day would be achieving efficiency roughly ten times better than Google’s best-reported comparable facility. Nobody has explained where that number comes from.
Project Ruby’s water demand figure has no disclosed source — and the math doesn’t support it
When Columbus residents ask how much water Project Ruby will use, they get a number: roughly 330,000 gallons per day. Columbus Water Works CEO Jeremy Cummings has said the project will not impact customer bills. Choose Columbus CEO Missy Kendrick has called it manageable. The overlay committee has cited the developer’s consent to pay for water infrastructure as evidence that ratepayers — the people who pay monthly water bills — are protected.
There is one problem. Nobody has explained where that number comes from.
That is not a minor detail. It is the central question. And as this series will show, the closer you look at that number — where it originated, what the Chattahoochee River basin looks like upstream, and what Georgia communities have discovered after approving projects backed by similar assurances — the less confidence it inspires.
The first thing to notice is that the official figure is not even consistent.
In February 2026, Choose Columbus told local media that Project Ruby would use nearly 90,000 gallons daily when construction is complete in 2028, rising to nearly 260,000 gallons per day by 2035. By April 2026, Missy Kendrick was telling Inside Climate News the figure was 330,000 gallons per day at “full buildout.”
The official number shifted from 260,000 to 330,000 between February and April — a 27 percent increase — with no public explanation of why. That inconsistency alone should prompt questions about the rigor of the underlying analysis.
Neither figure came with a source. No engineering study was cited. No comparable facility was named. No methodology was described. The numbers appear to have originated with the developer — Habitat Real Estate Partners — and been accepted and repeated by Choose Columbus without independent verification.
Here is where the math gets uncomfortable.
Project Ruby has formally requested 600 megawatts of electricity from Flint Energies, delivered in three phases: 200 MW beginning in 2029, growing to 600 MW by 2034. That figure is not in dispute — it comes from Flint Energies’ own vice president of power supply, Jake Hopkins, who confirmed the request and said the grid study shows the transmission facilities can handle it.
To put 600 megawatts in perspective: it is more than the entire current electricity consumption of the city of Columbus. It is an enormous amount of power.
Water and electricity consumption in a data center are not independent variables. They are directly linked. Every megawatt of power a data center draws generates heat — think of the warmth that comes off a laptop, multiplied hundreds of thousands of times. That heat has to be removed continuously or the computers fail. The primary way heat is removed in most large facilities is water-based cooling. The more power a facility draws, the more cooling it needs, and the more water that cooling typically consumes.
The industry measures this relationship using a metric called Water Usage Effectiveness, or WUE — essentially, how many gallons of water are used per unit of computing power. Google publicly reports the water consumption of its individual data centers. Its largest facility, in Council Bluffs, Iowa, draws roughly the same power load as Project Ruby’s planned capacity and withdraws an average of 3.9 million gallons of water per day, consuming 2.8 million gallons after evaporation.
At 600 MW of power demand, a facility consuming only 330,000 gallons per day would be achieving water efficiency roughly ten times better than Google’s best-reported comparable facility. That is not impossible — but it would be extraordinary, and it would require a specific, verified, and legally binding commitment to a particular cooling technology to achieve.
No such commitment has been publicly documented for Project Ruby.
The phrase you will hear from Choose Columbus and Columbus Water Works is “closed-loop cooling system.” The claim is that a closed-loop system recirculates water rather than discharging it into the air through evaporation, which dramatically reduces water demand.
That claim is technically defensible — a genuinely closed-loop system can achieve very low water consumption. But there are several things that phrase does not tell you.
First, “closed loop” is not a single technology. It describes a design principle that can be implemented in many different ways, at many different scales of water efficiency. A facility can truthfully be called closed-loop and still require hundreds of thousands of gallons of makeup water per day to replace what is lost through normal system operation.
Second, there is no binding commitment in any publicly available document specifying what cooling technology Project Ruby will actually use, what chemicals it will circulate, or what water consumption rate it is legally required to achieve. “We plan to use closed-loop cooling” is a statement of intent, not a legal obligation.
Third, and most critically: the engineering study that would validate any of these claims has not been done, or at least has not been made public. Columbus Water Works has acknowledged that the city has the treatment capacity to draw from the Chattahoochee River, but does not have the pipes to deliver water to the Project Ruby site — and that infrastructure would have to be built. You do not plan and fund a major water infrastructure extension without knowing what demand you are building for. If the pipes are sized for 330,000 gallons per day, and actual demand turns out to be 1.5 to 3 million gallons per day, the infrastructure is undersized from day one — and someone will pay to fix it.
As we will document in the third article in this series, the same “closed-loop” assurance was made by a data center developer in Fayette County, Georgia — whose facility then drew nearly 30 million gallons from the county water supply without paying for it, at volumes that exceeded the limits the developer had agreed to during the planning process.
For a sense of what a fully built-out hyperscale campus actually demands, consider two Georgia projects drawing comparable power loads to Project Ruby.
The data center backed by Lt. Governor Burt Jones in Butts County is projected to use more than 4.5 million gallons of water per day at full build — more than tripling that county’s entire current water usage. Project Sail in Coweta County has requested up to 6 million gallons per day alongside 900 megawatts of power.
Neither comparison is proof of what Project Ruby will consume. Cooling technology, local climate, and site-specific factors all play a role. But they establish a reference point. Facilities drawing comparable power in comparable Georgia climate conditions are projected to need between thirteen and eighteen times what Project Ruby’s promoters are claiming.
That gap demands an explanation.
The People’s Overlay ordinance — the citizen-drafted alternative to the Chamber of Commerce ordinance currently before the council — contains a specific provision addressing exactly this problem. Before any building permit can be issued, the developer must submit a water demand study prepared by a licensed civil engineer. That study must project daily water demand for each phase of development and for the full buildout of the entire 865-acre parcel, reference comparable facilities that are already operating at hyperscale — not startup-phase projections — and be reviewed by Columbus Water Works before any permits are granted.
The ordinance’s own drafting notes explain why this standard is needed: “The Chamber presented 88,000–330,000 gallons per day as the site’s water demand. That figure reflects a single building at startup. Google’s own filings show comparable hyperscale sites using 2–3.9 million gallons per day. Requiring a study referenced to comparable operating facilities prevents the city from being presented with startup-phase figures as permanent projections.”
That is precisely what appears to have happened here.
Jeremy Cummings has said publicly that the project will not impact customer bills and that the city has adequate water capacity. Those statements may both be true — at 330,000 gallons per day, the demand is easily absorbed by a system that currently handles 31 million gallons daily.
But they are only true if the 330,000 gallon figure is accurate.
The question that has not been publicly asked — or answered — is this: what independent engineering study, prepared by a licensed civil engineer and referenced to comparable facilities operating at 600 megawatts of power load, actually supports that figure?
Not the developer’s representation. Not Choose Columbus’s summary of the developer’s representation. An independent study, with a disclosed methodology, citing comparable operating facilities, covering the full four-building buildout across the 865-acre parcel.
If that study exists, it should be published. If it does not exist, then Columbus Water Works is making public assurances about ratepayer protection based on a number it has never independently verified.
That is not accountability to the public. That is accountability to the developer.
At every public meeting, at every council session, in every conversation with elected officials and utility executives, residents should be asking four specific questions:
One. Where does the 330,000 gallon per day figure come from, and who calculated it?
Two. Has any engineer independent of the developer validated that figure against the facility’s 600 MW power request using industry-standard water efficiency benchmarks?
Three. What specific cooling technology is the developer legally required to use, and what is the maximum water consumption that technology is contractually permitted to reach?
Four. What happens — in the contract, financially, and operationally — if actual water demand exceeds the projected figure by two, five, or ten times?
If Columbus Water Works and the overlay committee can answer those four questions with documentation, the 330,000 gallon figure deserves to stand. If they cannot, it does not.
The next article in this series examines a risk Columbus Water Works has not publicly addressed: what happens to the city’s water supply when the dozens of data centers already approved upstream on the Chattahoochee begin drawing at full capacity — especially during drought.
Sources
Project Ruby water demand figures and Choose Columbus statements: WTVM, February 16, 2026; Inside Climate News, April 13, 2026.
Project Ruby power request and Flint Energies confirmation: Ledger-Enquirer / Yahoo News, February 25, 2026.
Project Ruby site dimensions and building plan: Georgia Public Broadcasting, May 4, 2026; Data Center Map / Habitat Real Estate Partners.
Columbus Water Works system capacity and infrastructure acknowledgment: WTVM, February 16 and March 18, 2026.
Google Council Bluffs water withdrawal data: MOST Policy Initiative, “Data Center Water Use,” April 2026, citing Google 2025 sustainability disclosure.
Butts County data center water projection: Georgia Recorder, December 2025.
Project Sail water demand: Georgia Recorder / CBS Atlanta, December 2025 and May 2026.
People’s Overlay ordinance water demand provision and drafting note: Community-Protective Technology Overlay District — Proposed Alternative Draft, Section (i), June 2026.
This Series: Water & Project Ruby
Preface: Can Columbus Trust the Information It Has Been Given?
Article 1: Where Does the 330,000 Gallon Number Come From?