European Industrial Emissions Portal (E-PRTR) Data

The register of what Europe's heaviest industry actually emits, moves and burns - not a survey, not a sustainability-report digest, the reporting obligation itself turned into tables. The European Environment Agency runs the portal jointly with the European Commission, and since replacing the former E-PRTR website in 2021 it has been the single entry point for data on the impacts of industrial activities in Europe. Two flows feed it. The EU Registry on Industrial Sites supplies the who-and-where: production sites, facilities, addresses, coordinates and administrative detail reported under Implementing Decision 2018/1135. Integrated E-PRTR and Large Combustion Plant reporting under Regulation (EC) No 166/2006 supplies the how-much: pollutant releases to air, water and land, off-site pollutant transfers, off-site waste transfers, and combustion plant fuel input and stack emissions.

The scope numbers deserve a moment. More than 60,000 industrial sites across 65 economic activities, the chemical industry sector included. Ninety-one pollutants in seven groups - greenhouse gases, other gases, heavy metals, pesticides, chlorinated organics, other organics, inorganics - each subject to an Annex II release threshold, so the register records crossings, not trivia. Facilities report annually to national competent authorities, which compile and quality-check submissions before the European record is assembled. And every quantity declares how it came to be: measured, calculated or estimated, stamped right on the row.

Within Datadory's commodity-chemicals shelf this is the European emissions layer. EPA's Toxics Release Inventory is its American counterpart, Chemical Data Reporting maps US production volumes, and the TSCA Inventory enumerates the substances themselves. When a question concerns who released what, where, in Europe - this is the record that answers.

Three consecutive rows captured whole during the August 2026 research pass - Europe's lignite belt, as it happens:

# three consecutive facility rows -- Annex I activity 1(c), combustion installations >= 50 MW
facility_id    : PL.MŚ/000000349.FACILITY            # INSPIRE facility key, stable year over year
facility       : PGE Górnictwo i Energetyka Konwencjonalna S.A. - Oddział Elektrownia Bełchatów
city           : Rogowiec                country : PL (Poland)
main_activity  : 1(c) - Thermal power stations and other combustion installations >=50MW

facility_id    : de.bb.inspire.pf.eureg/45025564     # INSPIRE facility key, German federal registry branch
facility       : LEAG, Kraftwerk Jänschwalde
city           : Teichland               country : DE (Germany)
main_activity  : 1(c) - Thermal power stations and other combustion installations >=50MW

facility_id    : de.sn.sax4inspire.pf/80011277      # INSPIRE facility key, Saxony registry branch
facility       : LEAG Lausitz Energie Kraftwerke AG  Kraftwerk Lippendorf
city           : Neukieritzsch           country : DE (Germany)
main_activity  : 1(c) - Thermal power stations and other combustion installations >=50MW

Read them as one argument. Three named thermal power stations - a Polish giant and two German plants under the same parent - each resolving to a persistent INSPIRE facility key, which means a 2007-through-2023 trajectory for a single plant is a filter, not a fuzzy-name reconciliation project. Each carries its Annex I activity code (1(c), combustion installations of 50 MW or more), its city, its country. Beneath the identity layer sit the quantity rows: releases by pollutant and medium with their method codes, off-site waste transfers with hazard class and treatment destination, combustion plant fuel inputs in terajoules. Stack a facility's rows and you have its environmental file; repeat across 60,000 sites and you have the register.

Twenty-two columns do the heavy lifting, each definition verified against the released database during the research pass - nothing inferred from headers:

• Identity and location - Facility_INSPIRE_ID as the join key that survives renaming and rereporting, parentCompanyName, city, countryCode, and pointGeometryLat/pointGeometryLon for mapping work.
• Activity - mainActivityCode and mainActivityName on the E-PRTR Annex I list, plus NACEMainEconomicActivityCode, which is the bridge to industrial statistics; facilityType flags whether the installation falls under E-PRTR scope.
• Time - reportingYear, one reporting year per row, the axis every trend rides on.
• Releases - pollutantName/pollutantCode from the 91-substance list, medium (air, water, land), totalPollutantQuantityKg, accidentalPollutantQuantityKG kept separate from routine operation, and methodCode declaring measured, calculated or estimated.
• Waste transfers - wasteClassificationCode (hazardous or not), wasteTreatmentCode (recovery or disposal), totalWasteQuantityTNE, and transboundaryIndicator, the export trail in one boolean.
• Combustion plant - fuelInputName and energyInputTJ for declared energy input by fuel, totalPollutantQuantityTNE for stack emissions in tonnes.

Because these names hold across all four row families, a facility's air history, waste trail and combustion detail join on the same keys without a mapping project.

• Geography: the EU27 plus Iceland, Liechtenstein, Norway, Switzerland, Serbia and Northern Ireland. The reporting obligation draws the boundary, and it pays to know the edge: some countries submit late or incompletely for the most recent years, so a latest-year ranking deserves a completeness check before publication.
• Temporal: three clocks run side by side - E-PRTR facility reporting from 2007 through 2023, Industrial Emissions Directive installations from 2017, and Large Combustion Plant detail from 2016 - each reaching 2023 in current versions, with later versions extending the record. Analyses that splice the flows should respect their different starting points rather than pretending one series begins in 2007.
• Granularity: individual facility or installation, one reporting year per row, with aggregate views at national, E-PRTR sector and Annex I activity level for the questions facility resolution would only clutter.

Scale check: tens of thousands of sites times seventeen reporting years times up to ninety-one pollutants times three media - far past spreadsheet range, comfortably inside a warehouse. That gap between unmanageable and trivially queryable is precisely the space a delivered extract occupies.

API, files, or your warehouse. Daily, weekly, or hourly.

You pick the channel and the cadence; the field dictionary above travels unchanged across all three. Files suit the analyst loading a full European benchmark once a quarter, payloads suit products that surface one facility's record inside a dashboard or a sourcing tool, and warehouse delivery suits teams running exposure models in SQL beside their own supply-chain and finance tables. Changing cadence is a settings conversation, not a re-integration project.

Six uses recur, and they map cleanly onto the register's structure:

• Polluter ranking and exposure screening - rank facilities by release volume within any activity, country or pollutant group; the defensible answer to which plants dominate European tonnage.
• Facility-level trend tracking - 2007-onward trajectories per plant on stable keys, watching whether releases bend after permit changes or merely get restated.
• Waste export trails - follow off-site transfers flagged transboundary to see whose hazardous waste leaves the country and whose stays for recovery or disposal.
• ESG and regulatory-risk factors - screen portfolio companies' installations for rising releases, growing hazardous-waste volumes or heavy combustion-plant emissions.
• Competitive footprint intelligence - track competitor-owned installations' reported releases and transfers year over year, ahead of anyone's curated sustainability summary.
• Citation-grade research and journalism - cite the official European register itself when a story or paper needs facility-level pollution figures nobody can quibble with.

The first and last of these anchor the extremes: rankings need the quantity columns taken seriously, citation needs the provenance taken seriously, and the register happens to supply both. The journalists academics and data scientists pages develop the workflows.

Seven Datadory personas carry this record, ranked by assigned relevance:

1. Journalists, Academics & Students (relevance 3) - investigate Europe's largest polluters facility by facility with the official register behind every figure.
2. Data Scientists & ML Engineers (relevance 2) - model facility-level emission trajectories from a tidy tabular spine that joins cleanly to industrial-site registries.
3. Developers & Data-Product Builders (relevance 2) - wire environmental dashboards onto a documented column set that holds across all four row families.
4. Competitive Intelligence & Product Teams (relevance 2) - monitor competitor-owned installations as reported figures accumulate, year over year.
5. Investors & Quant Researchers (relevance 2) - screen portfolio plants for release trends and combustion exposure as an ESG risk factor.
6. Market Researchers & Consultants (relevance 2) - quantify emissions footprints by country and sector for chemicals-market sustainability chapters.
7. Sales & Growth Teams (relevance 1) - build prospect lists of large European industrial operators reporting under E-PRTR and the Directive, for environmental-services outreach.

Persona pages collect the angles: investors quants, competitive intel product teams, developers builders, market researchers.

Because the register presents itself as a portal - one facility, one pollutant, one year at a time, interactive until the question becomes longitudinal - and most serious questions want the other direction: every plant in a sector, every year at once, every row family flattened into the dictionary above, every facility resolved on its INSPIRE key so joins never depend on spelling. Version drift is handled on our side; the completeness quirks of recent reporting years get confirmed against the exact delivered cut rather than promised blind. Get a sample of this dataset scoped to the activities, countries and row families you actually work with, then schedule deliveries that keep the file current as new reporting years land.

Five companions round out the picture. The EEA source profile explains the institution and the reporting disciplines the register inherits. The EPA Toxics Release Inventory provides the American releases counterpart, and the comparison against EPA Chemical Data Reporting sets Europe's emissions layer against America's production census. The E-PRTR glossary entry defines the regime before you build on it, and the best commodity chemicals datasets list shows where this record ranks for the vertical.