Why AI Data Centers Lower Energy Costs and Save Water

Key Takeaways

• Crusoe CEO Chase Lochmiller disputes the popular claim that modern AI data centers drain local reservoirs, pointing to closed-loop liquid cooling designs.
• A 140-megawatt Crusoe building in Abilene, Texas consumes the same annual water volume as roughly 10 single-family homes.
• Large compute projects amortize grid transmission upgrades across more megawatts, which tends to reduce retail electricity rates for nearby residents.
• Crusoe raised a $3.9 billion Series F to scale AI data center sites while vertically integrating electrical equipment manufacturing.

Closed Loops Replace Municipal Water Drains

Public headlines often paint AI clusters as industrial sponges draining municipal water reserves dry. Lochmiller hears this myth constantly. “The number of people that have told me that data centers use all the water in the world is crazy and you know it's also just false,”▶ 25:30 he says.

The reality comes down to plumbing architecture. Early data centers relied heavily on evaporative cooling towers that continuously pulled fresh water and evaporated it into the air to shed heat. Modern AI facilities do not operate like 1990s server rooms. They run on closed-loop liquid cooling.

“There is water in the building. We are using water to cool the GPUs. But the the critical aspect here is we we've actually designed a closed loop architecture,”▶ 26:13 Lochmiller explains. Once cooling fluid fills the pipes, it cycles through radiators and cold plates repeatedly without constant replenishment.

Consider Crusoe's buildout in Abilene, Texas. Lochmiller notes: “One of our giant buildings in Abalene, Texas that consume, you know, it's a call it 140 megawatts of power is like what's budgeted for each of those buildings on the on the first eight buildings. They use about the same water annually as about 10 single family homes.”▶ 25:46 Ten suburban houses consume roughly the same municipal water as a facility pulling 140 megawatts of compute. The water panic simply does not match modern engineering data.

Why Mega Data Centers Lower Local Energy Bills

The second common fear is that massive GPU clusters drive up utility bills for everyday residents. Critics assume that adding hundreds of megawatts of demand to a regional grid creates immediate scarcity, pushing power costs higher for families living nearby.

Historical data shows the opposite effect. “When you look at markets where data centers have made investments and have built big data centers, typically energy prices for communities have come down,”▶ 26:52 Lochmiller notes. He adds: “The data has shown that energy prices actually come down. It's actually the opposite of the narrative that's being told.”▶ 27:48

This price drop happens because data center operators do more than draw power from existing lines. They pay for major grid updates, subsidize transmission capacity, and create predictable base demand that attracts new power generation to the region. The fixed costs of regional substations, high-voltage lines, and grid maintenance get spread across vastly more total megawatt-hours. Because the data center operator pays for the heavy electrical equipment, the per-unit cost of transmission distribution for residential customers drops.

What to Do With This

If you are evaluating AI infrastructure vendors or negotiating data center colocation agreements, ask providers for their exact water efficiency metrics and local grid interconnection agreements. Verify whether their facilities run on closed-loop liquid designs or evaporative towers, and check if their regional utility contracts include local transmission cost offsets.