// Water — sourced rebuild (2026-08-02). // // The prior version led with an unsourced "About 5% of the water a typical // campus would use" headline over a three-bar chart with no defensible // denominator, and closed with a fabricated "Watershed consultant" quote. // Both are gone. This version leads with a real, dated, attributable // trajectory (one operator's own disclosed water-use-effectiveness numbers), // a sourced air-cooled-vs-evaporative contrast, and a real per-site range — // then keeps this campus's own illustrative figure, clearly badged as a // demonstration number rather than a filed estimate. // See docs/research/2026-08-02-elephants-evidence-briefs.md §1 for sourcing. function Water() { const ref = useRef(null); const [inView, setInView] = useState(false); useEffect(() => { if (!ref.current) return; const io = new IntersectionObserver((entries) => { entries.forEach(e => { if (e.isIntersecting) { setInView(true); io.disconnect(); } }); }, { threshold: .15 }); io.observe(ref.current); return () => io.disconnect(); }, []); // Microsoft's own disclosed fleet-wide water-use effectiveness (WUE), // liters per kWh — a company-specific trajectory, not an industry // average. Source: Microsoft, "Inside Microsoft's two-decade push to cut // water intensity while scaling for growth," Official Microsoft Blog, // June 24, 2026. const REF_MAX = 2.3; const rows = [ { k: 'high', label: 'Early-2000s fleet baseline', sub: 'Microsoft, company-wide disclosure', value: 2.3, unit: 'L/kWh' }, { k: 'low', label: '2025 fleet-wide', sub: 'Microsoft, company-wide disclosure', value: 0.27, unit: 'L/kWh' }, ]; return (
{/* Heading row */}
03 Water

Cooling technology has{' '} changed enormously {' '}since the 2000s.

Early data centers commonly used water-intensive evaporative cooling. The honest question for any specific project is which system it actually uses — not which decade the coverage you've seen was describing.

In {SITE.county}, the aquifer is the well behind the house. Cattle. The family table. It’s not abstract — which is exactly why the cooling system matters more than the decade a data center happened to be built.

{/* Resident question — kept as a calm-trust callout, not a hero shout */}
“Will this campus drain my well?”

A fair question for any specific project — see how to check it below

{/* Chart — one operator's own disclosed trajectory, dated and sourced */}
Water-use effectiveness, liters per kWh Lower is less water per unit of computing. One operator's own disclosed numbers — not an industry average.
{rows.map((r, i) => { const pct = Math.min((r.value / REF_MAX) * 100, 100); return (
{r.label} {r.sub}
{r.value} {r.unit}
); })}
Microsoft, “Inside Microsoft’s two-decade push to cut water intensity while scaling for growth,” Official Microsoft Blog, June 24, 2026. A company-specific trajectory — other operators and colocation providers moved at different paces, and many facilities built in the 2010s still use evaporative cooling.
{/* Supporting sourced stats — the contrast and the real spread */}
Up to 90% less water Air-cooled / closed-loop cooling compared with traditional evaporative cooling towers Microsoft, Official Microsoft Blog, June 2026
500,000 – 3,000,000 gal/day Real range across individual Google data centers in 2024, by climate and cooling design — “data center water use” was never one number Google 2024 Environmental Report
{/* This campus's own projected figure — kept, but clearly badged */}
This campus — demonstration figure
Before this site ~15 million gallons a year. Cattle and irrigation.
This proposal ~18 million gallons a year. Cleaning, makeup water, and employee facilities — not an evaporative cooling loop.

A demonstration number for this template, not a filed engineering estimate. A real project would publish its own water budget with the assumptions behind it.

{/* How you check */}
How you check Quarterly water-use reports filed with the county. Public record.
); } window.Water = Water;