PlainEnviro

Data Methodology

Download the environmental statistics extract cited on this page: plainenviro-environmental-statistics.csv (EPA public records · CC BY 4.0).

Data Sources

All data comes from EPA public programs and federal datasets:

Data Vintage

TRI facility and county pages show the 2023 reporting year as the current snapshot; facility trend pages show the full 2019-2023 history. SDWIS data reflects the most recent EPA release. NPL data reflects the current EPA National Priorities List as of the database build date. Each data type is refreshed when EPA publishes updated files.

Processing Pipeline

  1. TRI bulk data files are downloaded from the EPA TRI program data portal. Each record represents one facility-chemical-year combination.
  2. SDWIS data is downloaded from the EPA ECHO system and processed to extract active violations, health-based concerns, and public water system metadata.
  3. Superfund NPL data is downloaded from EPA CERCLIS and filtered to active NPL sites.
  4. All three datasets are linked to geographic identifiers (state, county) and loaded into a structured SQLite database.
  5. State and facility pages aggregate TRI data by chemical, release medium (air, water, land), and carcinogen/PBT classification.
  6. Violation severity is categorized following EPA's health-based vs. monitoring/reporting distinction.

TRI Reporting Thresholds

Not all chemical releases appear in TRI data. Facilities must report only when they meet all three criteria:

  • Have 10 or more full-time employees
  • Fall within one of the covered industry sectors, determined by 6-digit NAICS code (manufacturing, mining, electric utilities, and others listed at 40 CFR 372.23)
  • Exceed an activity threshold for a listed chemical: 25,000 pounds manufactured or processed in the year, or 10,000 pounds otherwise used (40 CFR 372.25)

Persistent, bioaccumulative and toxic (PBT) chemicals carry far lower thresholds under 40 CFR 372.28, so the two figures above do not apply to them. Most PBTs report at 10 or 100 pounds, and dioxin and dioxin-like compounds report at 0.1 grams. This is why the dioxin figures on this site are small numbers rather than the thousands of pounds seen for common industrial chemicals, and why a dioxin release can appear here when a 10,000-pound chemical from the same facility does not.

Many smaller facilities and non-covered sectors are not required to report. Absence from TRI does not mean a facility has zero chemical releases.

Data Vintage and Update Frequency

TRI facility and county pages use the 2023 reporting year, the most recent complete year available; facility trend pages show the full 2019-2023 multi-year history. Facilities submit TRI reports annually by July 1 for the prior calendar year, and the EPA publishes compiled data in the fall. SDWIS data reflects the most recent quarterly release from the EPA ECHO system. NPL data reflects the current EPA National Priorities List as of the database build date. Each dataset is refreshed when the EPA publishes updated files.

Accuracy Commitment

PlainEnviro reproduces EPA data exactly as published without editorial modification or advocacy framing. Release quantities, violation counts, and site statuses are presented as reported by the EPA and its partner agencies. Geographic aggregations (state and county totals) are computed directly from the underlying facility-level records. When data is unavailable for a particular geography or time period, the site displays this clearly rather than interpolating or estimating values.

Derived Scores (transparent indexes, not official EPA ratings)

PlainEnviro publishes two named composites. Both are transparent derived indexes from EPA TRI / SDWIS / Superfund extracts: not official EPA ratings, not health forecasts, and not interchangeable with each other.

County profile burden score

Computed by a version-controlled scoring routine, so every rebuild reproduces the same result. Fixed national caps (not percentiles): facilities ≤25 pts (50+ = max), log10(releases+1) ≤25 pts (108 lbs = max), Superfund ≤25 pts (5/site), health-violation water systems ≤15 pts (3/system), high-chemical facilities ≤10 pts. Bands: High ≥60, Moderate ≥30, else Low. Rebuilds from the same page inputs on every SSR pass.

EJ Screening percentile score

Computed by a separate version-controlled screening routine. Each of up to 500 counties is percentile-ranked on four rates: toxic lbs per capita (35%), health-based SDWIS rows per water system (25%), Superfund sites per 100k residents (20%), carcinogen-facility share (20%), then weighted into a 0–100 burden score with letter grades. Counties without population take a neutral 50th percentile on the per-capita indicators. Expect different numbers than the county-profile score for the same county.

Limitations

  • TRI data is self-reported by facilities. The EPA conducts periodic audits but does not independently verify all reported quantities. Actual releases may differ from reported values.
  • SDWIS violation data reflects violations identified through state and local monitoring programs. Compliance history depends on the frequency and rigor of inspection programs, which vary by jurisdiction.
  • Superfund site status changes as remediation progresses. Current NPL status reflects the most recent EPA update, not real-time environmental conditions at the site.
  • Environmental conditions are complex and cannot be fully captured by any single dataset. TRI, SDWIS, and NPL each cover specific aspects but do not represent a complete picture of environmental quality.
  • Facility counts, release totals, and water-system counts are heavily influenced by how many EPA-regulated entities happen to be located in an area and how consistently they report, not solely by underlying environmental risk. A county with more TRI facilities is not necessarily more hazardous than one with fewer; it may simply host more of the industrial activity TRI covers. Compare figures within the same category (facility-to-facility, county-to-county) rather than treating raw counts as a danger ranking.
  • Both derived scores (see Derived Scores) document regulatory and disclosure history. A past violation or an elevated score does not itself mean current conditions are unsafe. Neither score is population-health epidemiology.
  • PlainEnviro presents data without advocacy framing. Readers are encouraged to consult local environmental agencies and public health departments for community-specific context.
  • PlainEnviro is not affiliated with the EPA or any government agency.