Data-center infrastructure and externalities
What a large data center draws from the grid, the water system and the aquifer around it can be partially reconstructed from public records — provided the reconstruction is explicit about which quantities are reported and which are modeled.
- Status
- Ongoing public-data modeling
- Period
- 2026 — present
- Themes
- Infrastructure modeling & planning
Public outputs
Research question
How can public data support transparent, reproducible modeling of the infrastructure and resource externalities associated with large data centers when detailed operational telemetry is unavailable?
Why it matters
Large data centers place substantial and concentrated demands on electricity, water and the infrastructure of the communities that host them. Those demands are increasingly the subject of local permitting and utility decisions, which need quantities that can be checked. Operators publish annual site totals; almost everything below that resolution is private.
The challenge
Public records are not a smaller version of operational telemetry. They are assembled on different accounting boundaries, refreshed on different schedules, and report different things: an annual water figure from a sustainability report, a municipal water-system record, weather observations, grid data, building permits and groundwater context each answer a slightly different question. Combining them is only defensible if the reconstruction keeps track of which quantity came from where.
Approach
This project develops a public-data reconstruction framework, beginning with the Meta Prineville campus as a case study. The repository integrates reported annual site quantities with weather, water-system, grid, permit and groundwater-context data while preserving source provenance and the boundary of each data product.
The modeling direction uses physics-constrained reconstruction, chronological validation, and sensitivity and counterfactual analysis. It keeps reported quantities distinct from modeled, proxy and scenario components, and does not treat public data as a substitute for private hourly workload or meter telemetry.
Status
Ongoing public-data modeling. The public repository documents the baseline and its limitations; this page does not report preliminary conclusions.