Methodology
InfraTrack’s principle is that a score without a reason is not intelligence. Every index on the map exposes its components, weights, data vintage, and scoring version. Nothing is a black box; if a number moves, you can see which input moved it.
Speed-to-Power Index (S2P)
S2P ranks how quickly and realistically a location can support new AI/data-center or electrified-industry load. Higher is faster/safer to build. It is computed per cluster — a named region with a defined radius — from the datasets on this map.
| Component | Weight | Measures | Data |
|---|---|---|---|
| Water availability | 30% | Mean baseline water stress across the cluster’s data centers (abundance-scored) | WRI Aqueduct 4.0 spatial join |
| Permitting & community | 25% | Documented opposition cases (moratoriums, lawsuits, zoning fights) within the cluster radius | datacenterfights.org + own international compilation |
| Grid quality | 20% | National grid composite: access, T&D losses, consumption per capita | World Bank indicators × Natural Earth |
| Nearby generation | 15% | Operating capacity (all technologies) within the radius, log-scaled to a 20 GW reference | Global Energy Monitor trackers |
| Interconnection delay | 10% | Ordinal assessment of connection-queue reality (e.g. PJM backlog, EirGrid moratorium) | Curated, source-noted per cluster |
Weights are a deliberate analytical position: physical water limits and political buildability are weighted hardest, because they are the constraints that most often surprise siting decisions late. Behind-the-meter power strategies join as a component in a future version. Tiers: ≥70 Fast · 45–69 Moderate · 25–44 Slow · <25 Gridlocked. Current scoring version: s2p-0.1.0.
Composite Risk Index
Computed per data center (4,253 facilities), shown as halo layers and in facility popups:
Measured factors (water, documented opposition) are weighted above modeled ones (concentration, grid proxy). Opposition and concentration use percentile normalization to handle outliers like Northern Virginia. Full factor definitions are shown in each facility popup; tiers: Low <30 · Moderate 30–49 · High 50–69 · Critical ≥70.
Data sources & licensing
Every dataset ships through a source registry that records license, commercial-use rights, attribution, and review dates. A build fails if a dataset is unregistered or its license is incompatible. Current registry:
| Source | Covers | License | Vintage |
|---|---|---|---|
| Global Energy Monitor | Pipelines, LNG, nuclear, solar, wind, hydro, geothermal | CC BY 4.0 | 2025–2026 trackers |
| OpenStreetMap (Overpass) | Data centers (4,253) | ODbL 1.0 | 2026-03 snapshot |
| WRI Aqueduct 4.0 | Baseline water stress | CC BY 4.0 | 2023 release |
| datacenterfights.org | US opposition cases (660) | CC BY 4.0 | 2026-03 |
| Own research compilation | International opposition (96 cases, 32 countries), subsea incidents, DC attacks, East-Med corridor | Compiled from public reporting; per-item source links | re-verified 2026-08 |
| World Bank Open Data | Grid quality indicators | CC BY 4.0 | 2018–2023 |
| Natural Earth | Country boundaries | Public domain | 110m |
| World Port Index (US NGA) | Ports (3,804) | Public domain | 2026-08 snapshot |
| Wikidata | Desalination plants | CC0 | 2026-04 |
| Protomaps / © OpenStreetMap | Basemap (self-hosted planet build, z0–10) | ODbL (attributed) | 2026-09-02 build |
Update discipline
versioned Every score run carries a scoring version and an as-of date, displayed in its popup. change-logged The app version (top-left panel) identifies the exact build. curated with care Curated event datasets (opposition, sabotage, corridor status) are re-verified against live reporting before any client deliverable is issued.
What this is not
InfraTrack provides exposure, dependency, and buildability analysis at cluster/regional level for research and commercial decision support. It deliberately excludes operational vulnerability detail about specific facilities. Scores are comparative screening tools built on public data — not investment advice, engineering due diligence, or a substitute for site-specific studies. Data may not reflect real-time conditions.
InfraTrack · methodology last updated 2026-08-19 · infratrack map