There’s a lot of talk these days about data centers. All over the country, people are paying attention, having conversations, and organizing movements. Locally, the discussion has reached both the …
This item is available in full to subscribers.
To continue reading, you will need to either log in, using the login form, below, or purchase a new subscription.
If you are a current print subscriber, you can set up a free website account and connect your subscription to it by clicking here.
Otherwise, click here to view your options for subscribing.
Please log in to continue |
There’s a lot of talk these days about data centers. All over the country, people are paying attention, having conversations, and organizing movements. Locally, the discussion has reached both the Pierce and St. Croix County Boards, and citizens have brought their concerns to the City of River Falls.
Let’s face it, data centers are big.
How big are they?
So big they can’t be tackled in one column, which is why I’ve avoided talking about them until now. It’s also why I’ll dedicate future columns to digging into the minutia, but today I’m keeping it big, as in the big picture, because how can we have a conversation if we don’t understand what we’re talking about?
Contrary to how it sounds, if you’re using “The Cloud” your photos and documents are not being stored up in the sky. “The Cloud” is actually a very physical collection of black boxes located in a massive, mostly deserted, concrete building with perimeter control and lots of air conditioning.
At its most simple, a data center is a large collection of equipment made up of servers, storage systems, and networking equipment. They store and transmit information and perform calculations for everything from banking, online shopping and streaming movies, to running business software, artificial intelligence applications, and bitcoin mining.
Data centers don’t produce a material product. Computing and storage are the product, and electricity and water make it possible.
According to the International Energy Agency, data centers consumed about 1.5% of global electricity in 2024, and they expect this number to double by 2030. While the global number might seem low, a data center’s local impact can be significant because of the immense electrical and cooling needs concentrated in one location.
Every calculation, connection, and stored bit of data requires a small amount of energy. One calculation might use only a little energy. Billions of calculations are another matter.
Additionally, all that computing generates heat, and for the computers to keep computing, that heat must be removed.
Though some data centers are air cooled, most rely on water-based cooling systems. Technology is always changing, and it’s currently shifting from open-loop evaporative methods to more efficient closed-loop systems. These new systems may greatly reduce the on-site water use and waste discharge, which is a good thing, but they require more energy amplifying data centers’ demand for electricity and merely transferring their substantial water use to the power grid.
Speaking of the power grid, all thermoelectric power, whether coal, natural gas, or nuclear, uses water in the process of generating electricity and is the primary source of power in Pierce and St. Croix counties. According to Clean Wisconsin, power generation already accounts for 73% of the state’s water use, more than municipal water systems, agriculture, and industry combined.
The EPA estimates the average American household uses roughly 10,500 kilowatt-hours of electricity and about 110,000 gallons of water in a year. A 25-megawatt data center consumes roughly 219 million kilowatt-hours a year, the equivalent of almost 21,000 households; that’s about four times the size of River Falls! Likewise, that same data center will require roughly 516 million gallons of water per year.
Though not perfect, these comparisons provide a useful sense of scale: when a community decides to allow even a 25-megawatt data center, they are not merely adding another large commercial building to their electric grid; they are adding one customer whose annual electric and water consumption is comparable to that of an entire city.
Additionally, a data center doesn't simply exist on an empty piece of land. Its insatiable appetite for energy might require substations, transmission infrastructure, and potentially new power generation. In addition to requiring access to vast amounts of water, it makes use of local roads, telecommunications, and emergency services. It takes up space and creates pollution: noise, light, and wastewater discharges containing concentrated minerals, dissolved solids, and treatment chemicals.
Data centers are not new; they have been around since the mid-1940’s, so why are they suddenly such a big deal?
To answer that question, we need to look at the technology driving the data center boom. Artificial intelligence (AI) sounds rather abstract, but it is not without weight, nor is it simply digital; its abilities depend on physical systems that require buildings, equipment, electricity, and water.
AI is not one particular application but rather a collection of technologies that allow computers to recognize patterns, make predictions, generate content, and perform tasks that previously required human effort and judgment.
Performing tasks that replace human effort and judgment also require more computations, and more computations means more energy demands.
The amount of electricity and water required depends on the hardware, the size of the AI model, the complexity of the task, how much data is being processed, and how often computations occur.
For example, let’s consider the camera: a traditional camera simply takes photos or records video. Add AI and it can analyze what it sees: identifying vehicles, people, objects or events. A system that continuously analyzes video requires exponentially more computations than one that simply stores footage for a person to review later.
AI and the data centers that make it possible may seem inevitable, abstract, and without physical substance because they exist beyond our screens where we can’t see them. They are none of these. They require a substantial physical infrastructure with very real and significant demands on electricity, water, land, roads, and public services, and more and more communities are starting to experience this reality.
Some say size doesn’t matter, but I beg to differ. Anything that demands the amount of resources and infrastructure that data centers consume should be given a great deal of attention.
Responding to data centers and AI responsibly will require human effort and judgment. Skills and abilities we risk losing if we rely too much on the conveniences that these technologies provide.
CALL OUT BOX
If you’re interested in learning more about data centers and what communities can do to protect themselves mark your calendar for the following events in the Sustainability Speaker Series sponsored by River Falls Public Library, Hope for Earth, and UWRF:
Monday, Oct. 19, 6:30 p.m. VIRTUAL ONLY through the River Falls Public Library Facebook and YouTube pages.
Environmental and Health Harms of AI Data Centers in Wisconsin, A look at the impacts of current AI data center projects in Wisconsin on our air, water, land and energy systems.
Thursday, Dec. 3, 6:30 p.m. In-Person at the River Falls Public Library or virtual through the River Falls Public Library Facebook and YouTube pages.
Intentional Development - from Data Centers to Local Foods: How communities can prepare for, engage with, and interact with development to direct it towards their values, A presentation on how rural communities are being impacted by corporate development such as data centers and how they can try to prepare for and leverage proposed development to meet their needs with an emphasis on resilient local food systems. Presented by Tara Greiman, Wisconsin Farmers Union Director of Conservation and Stewardship