29/08/2026
THE INDUSTRY YOU THINK USES THE MOST WATER ACTUALLY DOESN'T. WHO DOES MIGHT SURPRISE YOU.
For many people in Stockton, the words "data center" now bring up a set of familiar ideas: huge water use, high electricity demand, not many permanent jobs, and a large industrial presence that can seem disconnected from the everyday economy.
The concern about electricity is real. The latest AI-focused data centers can need a massive amount of power, and any city considering one should ask tough questions about substations, transmission capacity, backup power, protecting ratepayers, and who will cover the costs for the infrastructure needed to support it.
But when it comes to water, public understanding is falling behind the technology.
What might surprise some is that industries Californians have lived near for decades, especially those in food and agricultural processing, can move much more water through their operations than most people realize. For instance, a Pacific Coast Producers tomato-processing plant in Woodland reported about 1.72 million gallons of wastewater per day at the height of its processing season, with allowed flows reaching up to 4 million gallons each day.
Wastewater flow is not the same as net fresh-water use, and it should not be described that way. Still, it shows just how much water is involved in traditional industrial processing. These industries are well-known in the Central Valley, important to the economy, and deeply woven into the region, yet their resource needs often get much less attention than a proposed data center.
That is where the comparison gets interesting.
Older data centers could also use a lot of water. Google once reported average water use of about 450,000 gallons a day at its data centers, mostly because evaporative cooling was for a long time one of the most efficient ways to cool servers.
But that is not the only approach anymore.
The newest data-center designs are now using direct-to-chip liquid cooling, closed-loop systems, and dry cooling technologies that greatly reduce or, in some cases, almost eliminate ongoing water use for cooling after the initial system fill. Microsoft has said its newest design uses basically no water for cooling once the closed system is filled, while other recent projects have estimated ongoing water use for cooling in the hundreds of gallons per day, not hundreds of thousands.
That does not mean every new data center is good for the environment, nor does it mean every developer should be trusted just because they mention "closed loop" in a presentation. It means the discussion needs to be more thoughtful.
The question is no longer whether data centers use water.
The question is how much water a specific facility will use, what kind of cooling technology it will have, if potable water is needed, if recycled water can be used, what happens during droughts, and whether actual water use will be measured and made public once the facility is running.
For Stockton, that difference matters because the city is not choosing between having development or not. It is deciding among different kinds of development, each with its own mix of jobs, taxes, water use, traffic, energy needs, and long-term value for the economy.
A traditional processor might hire hundreds of people and support a key agricultural supply chain, but it may also need a lot of process water, wastewater treatment, and truck movement. A large warehouse might create more permanent jobs than a data center, but it can also bring heavy daily truck traffic and wear on roads. A maritime manufacturer might offer high-value jobs, but it needs waterfront access that cannot be easily found elsewhere.
A modern data center has a different set of features. It usually employs fewer people once it is up and running, but it can bring billions in private investment, need very little daily truck traffic, and, if built with modern cooling systems, use much less water than many older types of industry.
So, its economic impact is less about how many people it employs and more about how much capital it brings.
A large, modern data center can hold extremely expensive electrical equipment, cooling systems, networking gear, and servers that are regularly updated as technology moves forward. That creates a tax base that is very different from a standard warehouse or industrial building.
For Stockton, the demand for electricity itself could also bring in significant city revenue, since Stockton has a 6% Utility Users Tax on taxable electricity charges. When that is added to property taxes and possibly large sales and use-tax revenue from server and equipment purchases, the financial impact can be large even if the number of permanent employees is not.
Based on a sample 20-year fiscal model, a well-structured 60-megawatt data center could generate about $11 million each year in steady, service-supporting fiscal capacity for Stockton. That is about $114 per Stockton household annually, or roughly $9.50 per month.
That does not mean residents would get a monthly check. It means that millions of dollars needed for police, fire, streets, parks, libraries, and other city services could come from a major commercial taxpayer instead of falling entirely on current homeowners, renters, and businesses.
At 100 megawatts, the projected yearly fiscal capacity nears $19 million. At 250 megawatts, it gets close to $47 million.
These are planning estimates, not guaranteed income, and they depend on the site, tax rules, electricity rates, equipment values, and development agreements. Still, they show why counting only permanent jobs is not enough to judge this kind of project.
The bigger lesson is not that Stockton should pick data centers over food processing, manufacturing, or logistics. It should not.
San Joaquin County's agricultural and food-processing sector is still one of the region's most important economic drivers. Stockton should keep working to attract manufacturers, protect land for port-dependent industry, develop aerospace and autonomous-systems opportunities around the airport, and support logistics where it fits.
The aim should be a mixed economy where different industries are matched with the locations and infrastructure that suit them best.
For data centers, that means choosing sites with very high electrical capacity and fiber connections that do not take up land better used for port, aviation, or labor-heavy manufacturing.
It also means setting higher standards before approving a project.
Stockton should require applicants to reveal their expected water use, separate potable water from recycled water, specify the cooling system, document electricity needs, pay for infrastructure costs tied to the project, protect current ratepayers, keep backup-generation emissions low, and offer clear local training and apprenticeship options.
Then the city should make the results public.
If a developer claims a facility will use only a few hundred gallons of water per day, the city should install a meter.
If the project is expected to bring in millions in tax revenue, the city should report the actual amount.
If local hiring is promised, count the jobs.
That is a better way to judge modern industry than relying on assumptions left over from older technology.
The most surprising part of the data-center discussion may end up being this: some of the industries we see as normal and familiar can actually use much more water than the next generation of computing infrastructure.
Modern data centers still have environmental impacts, especially in their huge need for electricity, and those impacts must be managed carefully. But the latest technology also proves that very large economic investments do not always require equally large water use, heavy truck traffic, or the same environmental footprint as traditional industrial development.
Stockton should judge the next generation of industry by what it actually uses, what it actually pays, and what it truly leaves behind for the community.
Not by what people think it does.
