Data Center Site Report

Data center site selection analysis for any location worldwide — enter a city or address and receive a scored, structured assessment of power availability, natural hazard exposure, cooling efficiency, and overall site viability. Purpose-built for data center developers, colocation providers, and enterprise IT teams evaluating candidate locations before committing to land acquisition, permitting, or infrastructure investment.

The actor queries 8 public data sources in parallel — USGS seismic data, NOAA weather alerts, OpenAQ air quality, EIA energy generation data, FEMA disaster history, weather forecasts, flood warnings, and geocoding — then applies four independent scoring models to produce a composite verdict from PRIME_SITE to NOT_RECOMMENDED. Each score includes the specific signals that drove it, so you understand exactly why a location ranked where it did.

What data can you extract?

Why use Data Center Site Report?

Manual site selection research for a single candidate location can take a data center engineer 2-4 days: pulling EIA reports, querying USGS earthquake catalogs, reviewing FEMA disaster declaration history, estimating PUE from climate data, and synthesizing findings into a coherent recommendation. Multiply that by 5-10 candidate sites and you're looking at weeks of work before a single executive review.

This actor automates the entire process. Eight data sources are queried simultaneously, four scoring models run in seconds, and you receive a structured report ready for a slide deck or a spreadsheet within 60 seconds of clicking Start.

• Scheduling — run quarterly site reviews on a fixed schedule to monitor changing grid conditions or hazard designations
• API access — trigger site assessments from Python, JavaScript, or any HTTP client inside your existing due diligence workflows
• Proxy rotation — public data sources are queried reliably using Apify's built-in infrastructure
• Monitoring — get Slack or email alerts when runs fail or data source availability changes
• Integrations — push results to Zapier, Make, Google Sheets, HubSpot, or any webhook endpoint for downstream processing

Features

• 8 parallel data source queries — USGS earthquake catalog, NOAA severe weather alerts, OpenAQ air quality monitoring, Nominatim geocoding, Open-Meteo weather forecast, EIA electricity generation data, Environment Agency flood warnings, and FEMA disaster declarations, all fetched simultaneously to keep run time under 60 seconds
• Power availability scoring (0-100) — evaluates energy source diversity (up to 5 source types: gas, nuclear, solar, wind, coal) with a 20-point diversity bonus, generation capacity from EIA series records, FEMA-derived grid reliability penalty for power-disrupting disaster types, and weather stability for power infrastructure; levels are INSUFFICIENT, CONSTRAINED, ADEQUATE, STRONG, EXCELLENT
• Natural hazard scoring (inverted, 0-100) — higher scores mean more hazardous; seismic risk uses USGS magnitude and frequency data with bonus for M5.0+ events, flood risk from active warnings and FEMA flood declarations, severe weather risk from NOAA alert severity classifications, and a 20-point FEMA disaster frequency component; an EXTREME rating automatically overrides the composite verdict to NOT_RECOMMENDED regardless of other scores
• Cooling efficiency scoring (0-100) — models free-cooling potential using PM2.5 averages from OpenAQ (40 points; clean air below 12 µg/m³ scores maximum), temperature profile analysis with cool-hour ratio for PUE modeling (40 points), and environmental data stability bonus (20 points); levels are POOR, MARGINAL, ADEQUATE, GOOD, EXCELLENT
• Site viability scoring (0-100) — combines geocoding accessibility (20 points), EIA infrastructure presence (up to 35 points), and inverse disaster resilience from FEMA history (up to 35 points); levels are NOT_VIABLE, MARGINAL, VIABLE, STRONG, PRIME
• Weighted composite score — power availability 30%, hazard score inverted 25%, cooling efficiency 25%, site viability 20%; produces a single 0-100 composite score and one of five verdicts: PRIME_SITE, STRONG_CANDIDATE, ACCEPTABLE, MARGINAL, NOT_RECOMMENDED
• Override rules — locations with EXTREME natural hazard level or INSUFFICIENT power level receive NOT_RECOMMENDED regardless of other scores, preventing misleadingly optimistic composites
• Cooling strategy recommendations — if air cooling is requested but PM2.5 exceeds 35 µg/m³, the actor flags a conflict and recommends hybrid or mechanical cooling
• MW-aware power recommendations — if a power requirement over 50 MW is specified and power level is below EXCELLENT, the actor adds a utility capacity verification recommendation
• Named signal strings — every score produces plain-English signal strings (e.g., "4 energy source types — diversified power grid") explaining exactly what drove the score
• Per-source record counts — dataSources object shows exactly how many records were returned from each of the 8 sources, letting you assess data coverage quality per location

Use cases for data center site selection

Data center development and site comparison

Development teams evaluating 5-10 candidate markets before shortlisting for physical surveys. Enter each candidate city, collect composite scores, and rank locations before investing in site visits and engineering assessments. The PRIME_SITE to NOT_RECOMMENDED verdict scale maps directly onto a stage-gate decision framework.

Colocation provider expansion screening

Colo operators evaluating secondary and tertiary markets for new facility builds. The natural hazard scoring surfaces seismic, flood, and severe weather risks that affect insurance costs and uptime SLAs. A MODERATE or HIGH hazard rating early in the process prevents costly land deposits in unsuitable locations.

Enterprise IT disaster resilience assessment

Enterprise infrastructure teams evaluating on-premises data center locations against business continuity requirements. The FEMA disaster history component directly addresses the question of how frequently a location has experienced infrastructure-disrupting events, providing a documented basis for BCP planning.

ESG sustainability reporting for data centers

Sustainability analysts evaluating proposed data center sites for renewable energy availability and PUE potential. The actor surfaces renewable energy indicators from EIA data and cool-hour ratios for economizer feasibility, both of which feed directly into ESG reporting frameworks and PPA strategy documents.

Real estate and infrastructure investment due diligence

Investment analysts screening land parcels or existing facilities for data center conversion. The composite score and named signal strings provide a structured, defensible summary suitable for investment memos and committee presentations.

Multi-market feasibility studies

Consultants building comparative analyses across multiple regions for clients. Run the actor once per location, export all results to CSV, and assemble a side-by-side matrix of power, hazard, cooling, and viability scores without any manual data aggregation.

How to run a data center site analysis

1. Enter the location — type a city name, street address, or lat/lon coordinates in the Location field (e.g., "Ashburn, Virginia", "Phoenix, AZ", "Council Bluffs, Iowa")
2. Set optional parameters — specify your target power requirement in MW if you have one (e.g., 50 for a mid-size hyperscale facility) and select a cooling type (air, liquid, or hybrid) to get cooling-strategy-aware recommendations
3. Click Start — the actor fetches all 8 data sources in parallel and runs the scoring models; most runs complete in under 60 seconds
4. Download your report — open the Dataset tab and export to JSON, CSV, or Excel; the composite score, verdict, all four sub-scores, signal strings, and recommendations are all included in a single record

Input parameters

Input examples

Standard site assessment — most common use:

{
  "location": "Phoenix, Arizona"
}

Full specification with power and cooling context:

{
  "location": "Council Bluffs, Iowa",
  "powerRequirementMW": 100,
  "coolingType": "air"
}

Hyperscale site evaluation with liquid cooling:

{
  "location": "Reno, Nevada",
  "powerRequirementMW": 250,
  "coolingType": "liquid"
}

Input tips

• Use specific city and state — "Phoenix, AZ" returns better geocoding and EIA data coverage than "Phoenix" alone
• Set powerRequirementMW for large facilities — requirements over 50 MW trigger an additional utility capacity advisory in the recommendations array
• US locations produce richest results — EIA, FEMA, and NOAA data are US-centric; international locations still receive seismic, air quality, and weather scoring
• Compare locations in one session — run back-to-back with different locations and use the compositeScore field to rank candidates
• Air cooling on marginal air quality — if you specify coolingType "air" for a location with elevated PM2.5, the actor will explicitly flag the conflict so you can adjust your mechanical cooling plan

Output example

{
  "location": "Ashburn, Virginia",
  "compositeScore": 78,
  "verdict": "PRIME_SITE",
  "power": {
    "score": 82,
    "energySources": 4,
    "renewableIndicators": 1,
    "gridReliability": 26,
    "powerLevel": "EXCELLENT",
    "signals": [
      "4 energy source types — diversified power grid",
      "38% renewable energy — ESG-favorable for DC siting"
    ]
  },
  "hazard": {
    "score": 18,
    "seismicRisk": 4,
    "floodRisk": 5,
    "severeWeatherRisk": 6,
    "hazardLevel": "LOW",
    "signals": []
  },
  "cooling": {
    "score": 65,
    "airQualityIndex": 14,
    "temperatureScore": 30,
    "coolingLevel": "GOOD",
    "signals": [
      "Excellent air quality (PM2.5 avg 14.0) — ideal for free cooling"
    ]
  },
  "siteViability": {
    "score": 75,
    "viabilityLevel": "STRONG",
    "signals": [
      "Location successfully geocoded — accessible site",
      "Strong energy infrastructure data — power grid connectivity likely"
    ]
  },
  "allSignals": [
    "4 energy source types — diversified power grid",
    "38% renewable energy — ESG-favorable for DC siting",
    "Excellent air quality (PM2.5 avg 14.0) — ideal for free cooling",
    "Location successfully geocoded — accessible site",
    "Strong energy infrastructure data — power grid connectivity likely"
  ],
  "recommendations": [
    "Good cooling conditions — air-side economizers can reduce PUE significantly",
    "Renewable energy available — leverage for PPA and ESG reporting"
  ],
  "inputParameters": {
    "location": "Ashburn, Virginia",
    "powerRequirementMW": 50,
    "coolingType": "air"
  },
  "dataSources": {
    "earthquakes": 3,
    "weatherAlerts": 2,
    "airQuality": 12,
    "geocode": 1,
    "weatherForecast": 24,
    "energyData": 8,
    "floodWarnings": 1,
    "femaDisasters": 4
  },
  "generatedAt": "2026-03-20T14:32:00.000Z"
}

Output fields

How much does it cost to run a data center site analysis?

Data Center Site Report uses pay-per-run pricing — you pay approximately $0.15 per location assessed. Platform compute costs are included. The actor calls 8 sub-actors in parallel; your credit consumption covers the orchestration run and all sub-actor calls.

You can set a maximum spending limit per run to control costs. The actor stops when your budget is reached.

Manual site research from a specialist consultant typically costs $1,500-5,000 per location for a comparable data synthesis exercise. Apify's free tier includes $5 of monthly credits — enough for roughly 30 site assessments per month at no cost.

Data center site analysis using the API

Python

from apify_client import ApifyClient

client = ApifyClient("YOUR_API_TOKEN")

run = client.actor("ryanclinton/data-center-site-report").call(run_input={
    "location": "Council Bluffs, Iowa",
    "powerRequirementMW": 100,
    "coolingType": "air"
})

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"Location: {item['location']}")
    print(f"Verdict: {item['verdict']} (composite score: {item['compositeScore']}/100)")
    print(f"Power: {item['power']['powerLevel']} | Hazard: {item['hazard']['hazardLevel']} | Cooling: {item['cooling']['coolingLevel']}")
    for rec in item.get("recommendations", []):
        print(f"  Recommendation: {rec}")

JavaScript

import { ApifyClient } from "apify-client";

const client = new ApifyClient({ token: "YOUR_API_TOKEN" });

const run = await client.actor("ryanclinton/data-center-site-report").call({
    location: "Council Bluffs, Iowa",
    powerRequirementMW: 100,
    coolingType: "air"
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const item of items) {
    console.log(`${item.location}: ${item.verdict} (${item.compositeScore}/100)`);
    console.log(`  Power: ${item.power.powerLevel} | Hazard: ${item.hazard.hazardLevel} | Cooling: ${item.cooling.coolingLevel}`);
    item.recommendations.forEach(rec => console.log(`  > ${rec}`));
}

cURL

# Start the actor run
curl -X POST "https://api.apify.com/v2/acts/ryanclinton~data-center-site-report/runs?token=YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"location": "Council Bluffs, Iowa", "powerRequirementMW": 100, "coolingType": "air"}'

# Fetch results (replace DATASET_ID from the run response)
curl "https://api.apify.com/v2/datasets/DATASET_ID/items?token=YOUR_API_TOKEN&format=json"

How Data Center Site Report works

Phase 1: Parallel data source collection

The actor dispatches 8 sub-actor calls simultaneously using Promise.all, querying USGS for earthquake events near the location, NOAA for active weather alerts, OpenAQ for PM2.5 readings from nearby monitoring stations, Nominatim for geocoding and location validation, Open-Meteo for multi-day weather forecasts (temperature and wind), EIA for electricity generation series data including source type identification, Environment Agency for flood warning records, and FEMA for historical disaster declaration data. Each sub-actor is called with a 512 MB memory allocation and a 120-second timeout. Failures are caught and return empty arrays rather than failing the run, so partial data still produces a scored report.

Phase 2: Power availability scoring

The power scoring model parses EIA series IDs to classify generation by source type — keywords "solar", "wind", "hydro", and "geotherm" contribute to a renewable bucket; "nuclear" and "natural gas" are tracked separately. Source diversity earns up to 20 points (5 points per unique type) and generation record count earns up to 20 points. FEMA disaster types are scanned for "severe storm", "hurricane", "ice", and "winter" keywords; each power-disrupting type reduces the 30-point grid reliability score by 4 points. Weather forecast extremes (temperature above 40°C or below -20°C, wind above 50 km/h) reduce a 30-point weather stability score by 5 points per extreme reading.

Phase 3: Natural hazard, cooling efficiency, and site viability scoring

The hazard model sums a seismic score (2 points per USGS record, 5 extra per M4.0+ event, 10 extra for any M5.0+, capped at 30), a flood score (5 points per flood warning, capped at 25), a severe weather score (3 points per NOAA alert, 5 extra per "extreme" or "severe" severity, capped at 25), and a FEMA frequency score (3 points per declaration, capped at 20). The cooling model computes average PM2.5 across all OpenAQ readings with a tiered score (PM2.5 below 12: 40 points; 12-25: 30 points; 25-35: 20 points; 35-55: 10 points; above 55: 0) and calculates the fraction of forecast hours with temperature below 18°C as a cool-hour ratio for the 40-point temperature sub-score. The site viability model awards 20 points for a successful geocode, up to 35 points for EIA data density, and up to 35 points for absence of FEMA disaster declarations.

Phase 4: Composite scoring and override rules

The four scores are combined as: compositeScore = power * 0.30 + (100 - hazard) * 0.25 + cooling * 0.25 + viability * 0.20. Two hard override rules prevent misleadingly positive composites: if hazard level is EXTREME, the verdict is forced to NOT_RECOMMENDED; if power level is INSUFFICIENT, the verdict is forced to NOT_RECOMMENDED. The recommendations array is assembled from threshold tests: power score below 40 triggers a utility capacity advisory; seismic sub-score at or above 15 triggers a structural design recommendation; flood sub-score at or above 10 triggers an elevation advisory; cooling score at or above 60 triggers an air-side economizer recommendation; PM2.5 above 35 triggers an air quality advisory; and renewable indicator presence triggers a PPA recommendation.

Tips for best results

1. Specify state or region alongside city name. "Reno" is ambiguous; "Reno, Nevada" returns precise geocoding and ensures EIA queries match the correct regional grid.
2. Run candidate locations back-to-back and export to CSV. The compositeScore, power.score, hazard.score, cooling.score, and siteViability.score fields form a natural comparison matrix for executive presentations.
3. Use powerRequirementMW for large facilities. Requirements over 50 MW with anything below EXCELLENT power level trigger an explicit utility growth advisory, which is often the most decision-relevant output for hyperscale projects.
4. Check dataSources record counts to assess data quality. If energyData returns 0 or 1 records, the EIA has limited coverage for that location and the power score should be treated with lower confidence. US locations typically return 5-15 energy records.
5. Treat EXTREME hazard verdicts as disqualifying. The override rule exists because no amount of favorable cooling or power conditions compensates for extreme seismic, flood, or storm exposure at a critical infrastructure site.
6. Pair with Company Deep Research for acquisition targets. If you are evaluating a site occupied by an existing operator, use Company Deep Research to assess the operator's financial stability alongside the site's physical conditions.
7. Schedule quarterly re-runs for shortlisted sites. EIA grid expansion data, FEMA disaster declarations, and NOAA alert patterns change over time. Scheduled runs on Apify keep your site assessments current without manual effort.
8. International locations still produce partial scores. FEMA and EIA data are US-only, but USGS seismic, OpenAQ air quality, weather forecast, and geocoding data are global. International assessments receive meaningful hazard and cooling scores even without the US-specific components.

Combine with other Apify actors

Limitations

• FEMA and EIA data are US-centric. The grid reliability and infrastructure presence components of the scoring models are most accurate for US locations. International locations receive 0 on FEMA-dependent components, which lowers their power and viability scores relative to equivalent US locations — not because they are worse, but because data is unavailable.
• Earthquake data reflects recent catalog entries, not long-term seismic maps. The USGS query returns recent events; a location in an active fault zone may show low seismic scores if there have been no recent events. For high-stakes seismic assessment, validate against USGS national seismic hazard maps.
• Air quality readings depend on monitoring station proximity. OpenAQ requires a monitoring station near the target location. Remote or rural locations may return 0 air quality records, which defaults the PM2.5 score to 50 µg/m³ (a conservative middle-ground). Check dataSources.airQuality to see how many readings were returned.
• No physical site survey data. The actor uses publicly available datasets and cannot assess soil bearing capacity, fiber connectivity, water availability, zoning status, or proximity to substations — all of which are critical for final site selection.
• Flood warnings reflect current or recent active warnings, not long-term floodplain designation. A location with no active warnings may still be in FEMA Flood Zone A. Cross-reference with FEMA's Flood Map Service Center for regulatory flood exposure.
• Weather forecast data covers 7-10 days, not annual climate averages. The cool-hour ratio used for PUE modeling is an approximation from recent forecast data, not a 30-year climate normal. For precision PUE modeling, use ASHRAE climate data for the target location.
• The composite score is a screening tool, not a definitive siting decision. The actor is designed to narrow a list of candidate locations from 20 to 5, not to replace a full engineering feasibility study. Use the output as input to a more detailed investigation of shortlisted sites.
• Sub-actor timeouts. Each of the 8 sub-actors has a 120-second timeout. If a data source experiences high latency, that component returns an empty array. Check dataSources counts to identify which sources returned data for a given run.

Integrations

• Zapier — trigger a site assessment automatically when a new candidate location is added to a spreadsheet or CRM record
• Make — build multi-step workflows that run site assessments, filter by verdict, and post PRIME_SITE results to a Slack channel or project management tool
• Google Sheets — export all site scores to a shared spreadsheet for team review and multi-site comparison
• Apify API — integrate site assessment into automated real estate or infrastructure due diligence pipelines
• Webhooks — receive a webhook notification when a site report completes, enabling downstream processing without polling
• LangChain / LlamaIndex — feed site report JSON into an LLM-powered analysis pipeline that synthesizes site scores with market demand data or financial models

Troubleshooting

Composite score seems low for a well-known data center market. Check dataSources record counts. If energyData or femaDisasters return 0 records, the EIA or FEMA query did not match the location string. Try a more specific format: "Ashburn, Virginia" rather than "Ashburn" or "VA". The location string is passed directly to each sub-actor's query parameter.

International location scored NOT_RECOMMENDED despite low apparent risk. International locations receive 0 on FEMA-dependent components (grid reliability and disaster resilience), which pulls down the power and viability scores. This is a data availability limitation, not a judgment about the location. Check hazard.seismicRisk, cooling.coolingLevel, and dataSources counts to see which components had coverage.

Run completed but allSignals array is empty. Signals are only generated when specific thresholds are crossed. A location with moderate scores across all dimensions may not trigger any named signals. The scores themselves are still valid — examine the individual sub-scores and levels rather than relying on signals alone.

dataSources.airQuality is 0. No OpenAQ monitoring station returned data for the location. The cooling score defaults PM2.5 to 50 µg/m³ in this case, resulting in a lower-than-accurate cooling score. For locations with no OpenAQ coverage, the cooling score should be treated as a conservative lower bound.

Run timed out or returned partial data. The actor has a 120-second timeout per sub-actor. On rare occasions when upstream data sources are slow, some sub-actors may return empty arrays. Re-run the actor — results are typically consistent on retry. If failures persist for a specific data source, check the dataSources counts to identify which one is affected.

Responsible use

• This actor only accesses publicly available government and environmental data from USGS, NOAA, FEMA, EIA, OpenAQ, and similar open data sources.
• Data is used for site analysis and reporting purposes only — do not use output to misrepresent site conditions in regulatory filings or investment disclosures.
• FEMA disaster declarations and NOAA alerts are used as proxies for infrastructure risk; they do not constitute professional engineering assessments.
• For guidance on web scraping legality, see Apify's guide.

FAQ

How accurate is the data center site score for real site selection decisions? The composite score is calibrated as a screening tool to narrow a list of 10-20 candidate locations to a shortlist of 3-5 for detailed engineering review. It draws on real government datasets (USGS, FEMA, EIA, NOAA, OpenAQ) but cannot replace a geotechnical survey, utility capacity study, or regulatory review. Treat a PRIME_SITE verdict as "proceed to detailed investigation" and a NOT_RECOMMENDED verdict as "strong reason to deprioritize."

What locations can I analyze with the data center site report? Any location worldwide — enter a city name, street address, or lat/lon coordinates. US locations produce the richest results because EIA, FEMA, and NOAA data are US-centric. International locations still receive seismic risk, air quality, weather, and geocoding scoring from USGS, OpenAQ, and Open-Meteo, which are global datasets.

How does the actor score power grid reliability? Power grid reliability is proxied from FEMA disaster history. The model scans FEMA declarations for incident types including "severe storm", "hurricane", "ice storm", and "winter weather" — all of which historically cause extended power outages. Each power-disrupting disaster type reduces the 30-point grid reliability component by 4 points. Separately, weather forecast extremes (temperatures above 40°C or below -20°C, wind above 50 km/h) reduce a weather stability score by 5 points per extreme reading.

What is PUE and why does the cooling score matter? Power Usage Effectiveness (PUE) is the ratio of total facility power to IT equipment power. A PUE of 1.0 is perfect; hyperscale facilities target below 1.2. Locations with cool temperatures and clean air allow air-side economizers — free cooling using outside air — which can reduce PUE from 1.6 to 1.15 or below. At 10 MW of IT load, the difference between PUE 1.6 and 1.2 is roughly 4 MW of wasted power, or $1.4M annually at $0.04/kWh. The cooling score predicts how many hours per year a location can use free cooling.

Can I compare multiple sites using this actor? Yes. Run the actor once per candidate location and compare the compositeScore values directly. The four sub-scores (power.score, hazard.score, cooling.score, siteViability.score) allow dimensional comparison — one site may score higher on power while another scores higher on cooling. Export all results to CSV via the Dataset tab for side-by-side matrix analysis.

How long does a typical data center site report run take? Most runs complete in 30-60 seconds. The actor dispatches all 8 data source queries in parallel, so total run time is roughly equal to the slowest individual sub-actor response rather than the sum. On rare occasions when upstream data sources are slow, runs may take up to 2 minutes.

Does the data center site report flag renewable energy availability? Yes. The power scoring model identifies renewable generation series in EIA data by scanning for "solar", "wind", "hydro", and "geotherm" keywords in series IDs. If renewable generation is detected, the actor calculates the renewable percentage of total generation and flags it in the power.signals array. A recommendation to leverage the renewable capacity for PPAs and ESG reporting is added when renewables are present.

How is this different from paying a site selection consultant? A traditional site selection consultant providing comparable data synthesis typically charges $1,500-5,000 per location for initial screening and $15,000-50,000 for a full site feasibility study. This actor provides automated data aggregation from the same government datasets at approximately $0.15 per location, in under 60 seconds, with structured output ready for programmatic processing. It is not a substitute for the engineering judgment a consultant provides, but it covers the data gathering phase completely.

What happens if a data source returns no records? Each of the 8 sub-actor calls is wrapped in a try-catch block. If a data source fails or returns no records, its contribution defaults to empty arrays and the relevant scoring components receive their default scores. The dataSources object in the output shows exactly how many records each source returned, so you can assess data coverage quality for any given run.

Is it legal to use this data for site selection decisions? Yes. All 8 data sources used by this actor are publicly available government and environmental datasets published for public access. USGS, NOAA, FEMA, and EIA data are released under open government data licenses. OpenAQ data is released under open licenses from national monitoring agencies. Using these datasets for site analysis and reporting is entirely lawful. For more information, see Apify's guide on web scraping legality.

Can I schedule this actor to monitor site conditions over time? Yes. Use Apify's built-in scheduling to run the actor on a weekly, monthly, or quarterly schedule for any number of locations. This is useful for tracking whether a shortlisted site's hazard profile or energy infrastructure data changes over time — for example, if new FEMA disaster declarations are added or EIA data shows grid capacity changes in a target market.

How does the composite scoring formula weight each dimension? The composite score is calculated as: power availability (30%) + natural hazard inverted (25%) + cooling efficiency (25%) + site viability (20%). The natural hazard score is inverted (100 minus the raw hazard score) so that lower-hazard locations receive higher composite scores. Two hard overrides apply: if hazard level is EXTREME or power level is INSUFFICIENT, the verdict is forced to NOT_RECOMMENDED regardless of the composite score calculation.

Help us improve

If you encounter issues, you can help us debug faster by enabling run sharing in your Apify account:

1. Go to Account Settings > Privacy
2. Enable Share runs with public Actor creators

This lets us see your run details when something goes wrong, so we can fix issues faster. Your data is only visible to the actor developer, not publicly.

Support

Found a bug or have a feature request? Open an issue in the Issues tab on this actor's page. For custom solutions or enterprise integrations, reach out through the Apify platform.