Methodology
How the Data Center Water Stress Index is calculated
DCWSI is a county-level screening metric for journalists, researchers, and residents tracking where data center growth overlaps with water-system pressure.
Water score
60%
WaterByCounty 0-100 county score
Density rank
40%
Facility-count percentile among counties with mapped data centers
Output
0-100
Higher DCWSI means higher combined scrutiny priority
Formula
DCWSI = (water stress score x 0.6) + (data center density percentile x 0.4)
Each county receives a WaterByCounty water score from 0 to 100. Separately, data center facility counts are ranked as a percentile among counties with mapped data center records. The weighted average produces the county's DCWSI score.
Sources
| Source | Use in DCWSI | Limitation |
|---|---|---|
| EPA ECHO | Active facility records filtered by NAICS 518210 and data-center-like 517110 records. | ECHO identifies regulated facilities and permits; it does not fully measure water consumption. |
| EPA ECHO CWA | NPDES-linked data center records and available actual average flow fields. | Flow fields are treated as discharge estimates, not withdrawal or consumptive use. |
| OpenStreetMap Overpass | Supplemental mapped data center nodes, ways, and relations geocoded to county FIPS. | Coverage depends on public OSM tagging completeness and may miss private campuses. |
| WaterByCounty county dataset | Existing 0-100 county water score used as the water stress component. | The score is a county-level proxy and should be interpreted with local utility context. |
What DCWSI can show
DCWSI is strongest as a first-pass county comparison. It highlights places where mapped data center concentration is unusually high and the local water score makes additional reporting worthwhile.
What DCWSI cannot show
DCWSI is not a facility-level water-use audit. It does not prove consumptive use, water-right exposure, cooling technology, or utility-specific capacity constraints.