Data centres · what goes out of the watershed and into the air

Gigawatt Ledger

Fifty-six of the largest data centre sites on earth, measured on two things only: the water they take out of a basin and never give back, and what their turbines and power plants put into the air people breathe.

Scenario
Display
Nearest to you

Finding the sites closest to you…

    Waterconsumed, not returned
    Removed from watersheds bn gal/yr
    From stressed basins % of on-site draw
    Airwhat the stacks emit
    NOx · SO₂ · PM2.5 kt/yr
    Health damages $bn/yr
    CO₂e Mt/yr

    Where it lands

    Circle area is the impact itself — gallons taken, or tonnes emitted — not how big the building is.

    No sites match
    SiteLocation Water sourceBasin stress M gal/yrt NOx/yr t PM2.5/yr$M health

    The twenty worst

    Both charts split each bar by where the harm comes from. Totals follow the filters above.

    Water taken out of the basin

    Cooling towers evaporate water on site. Power stations evaporate more, somewhere else, to make the electricity — that half almost never appears in a company's own reporting.

    Nitrogen oxides, sulphur dioxide and fine particulates

    Distant power plants versus turbines running on the site itself. On-site combustion is a small share of the total and a large share of what the nearest neighbourhood actually breathes.

    How these numbers are made

    No operator publishes per-site water or air emissions. Everything here is modelled from disclosed capacity rather than copied from a report that does not exist. The whole model is five lines:

    facility_MWh = IT_MW × 8760 × load_factor × PUE
    water_basin  = facility_kWh × WUE          (evaporated on site)
    water_power  = facility_kWh × grid_water   (evaporated generating it)
    NOₓ SO₂ PM   = grid_MWh × grid_rate + onsite_MWh × turbine_rate
    health_$     = Σ tonnes × damage_value

    Water consumed is not water withdrawn. A cooling tower takes water in and evaporates most of it; what evaporates leaves the basin. That is the number here — the water the watershed does not get back — and it is far smaller than the withdrawal figures utilities publish.

    Seventeen sites publish a real annual water figure. For those, WUE is back-solved so the model reproduces the published number exactly, and the site is marked fitted to disclosure. Fitting them showed the remaining engineering estimates ran roughly 1.7× high, so they were corrected by that factor.

    Air pollution here means criteria pollutants, not carbon. Nitrogen oxides, sulphur dioxide and fine particulates are what damage lungs, and they are what an air permit actually regulates. CO₂e is reported alongside, but it is a climate number, not an air-quality one. Emission rates come from the generation mix of the grid a site draws on. Where a campus runs its own gas turbines, that share is charged at an uncontrolled simple-cycle rate of 0.45 kg NOₓ/MWh — roughly fifteen times a modern combined-cycle plant with selective catalytic reduction, which is the entire substance of the permit fights in Memphis and Abilene.

    Health damages apply US marginal damage values of $18,000 per tonne of NOₓ, $48,000 per tonne of SO₂ and $420,000 per tonne of PM2.5. These are national averages. Real damages scale with how many people live downwind and vary by more than tenfold between a rural site and one upwind of a city — which is exactly why siting next to a neighbourhood is the thing worth arguing about.

    What this misses Diesel backup generators are not counted. Every site has them, they are tested monthly, and they emit heavily for those hours — so the on-site figures here are floors, not ceilings. Grid rates are annual averages, not the marginal plant that actually answers a new gigawatt of demand, which is usually dirtier. Site-level error is roughly ±40%.
    GridNOxSO₂ PM2.5WaterCO₂e

    Pollutant rates in kg per MWh generated; water in litres per kWh; CO₂e in kg per kWh. Hydropower reservoir evaporation is included in the water column, which is why the Pacific Northwest and Norway look water-expensive despite being nearly carbon-free.

    Gigawatt Ledger · compiled September 2026 from company announcements, utility rate filings, air and water permit applications, environmental court records and press reporting.
    Capacity, tenancy, water source and approval histories are sourced. Water, air emissions and health damages are modelled as described above. Coordinates are approximate campus centroids. The model is deliberately legible so it can be argued with.