One-time snapshot

EPA Toxic Release Inventory 2022–2023

158,687 EPA TRI form records with exact decimal quantities and 44,385 facility attribute versions for 23,081 distinct facilities.

Snapshot coverage

Records: 158,687 reports; 23,081 distinct facilities

Coverage: 2022–2023 reporting years only

Format: CSV + Parquet

Package uploaded: 2026-09-19

Listing reviewed: 2026-09-19

The upload and review dates do not extend the source coverage. This is a dated snapshot; future updates are not included.

Get this snapshot

$99one time

CSV and Parquet files for the coverage described here. Review the limitations and sample before purchasing.

Buy dated snapshot

Commercial license terms

View 1,000-row sample

New annual update subscriptions are unavailable. Existing customers can use their purchase download links or contact support.

Known limitations

Historical 2022–2023 snapshot only. TRI covers specified chemicals, industries and reporting thresholds; it is not a complete inventory of pollution or environmental risk. Future updates are not included.
Report keys are year plus doc_ctrl_num. Multiple reports may share a facility, chemical and year. Join the two tables by facility_record_id, not TRIFID alone; 277 facility/year groups have multiple attribute versions.
Quantities retain EPA units: Pounds or Grams. Dioxin uses grams; do not sum across units. Form A does not report quantities, and its supplied zeroes are flagged. Other source zeroes can represent blanks or not-applicable values.
Reported totals are preserved with literal EPA-guide formula comparison flags. Primary release totals differ from component sums by at most 0.002 source units. Treatment totals include POTW treatment omitted from the guide formula. Energy and transfer comparisons have unresolved differences up to 142,991 pounds; flags are diagnostics, not proof of invalid reports.
All delivered coordinate pairs pass numeric bounds, but geographic accuracy is not independently verified and horizontal_datum is empty throughout. Raw coordinate text is retained; no datum is inferred or transformed.
This repair changes column names and numeric types and preserves facility attribute versions previously discarded. Text casing, leading-zero identifiers and literal NA remain as supplied. Consult the field mapping and migrate joins to facility_record_id.

Files in this snapshot

Row counts are per table and should not be added as independent events. Expand a file to see its delivered Parquet field names and types.

epa_tri_releases.parquet — 158,687 rows · 147 fields

Field	Type
year	int64
trifid	string
frs_id	string
facility_name	string
street_address	string
city	string
county	string
state	string
zip	string
bia	string
tribe	string
latitude	double
longitude	double
horizontal_datum	string
parent_company_name	string
parent_company_dun	string
standard_parent_company_name	string
foreign_parent_co_name	string
foreign_parent_co_db_num	string
standard_foreign_parent_co_name	string
federal_facility	string
industry_sector_code	string
industry_sector	string
sic_1	string
sic_2	string
sic_3	string
sic_4	string
sic_5	string
sic_6	string
naics_1	string
naics_2	string
naics_3	string
naics_4	string
naics_5	string
naics_6	string
doc_ctrl_num	string
chemical_name	string
elemental_metal_included	string
tri_chemical_compound_id	string
cas_number	string
srs_id	string
clean_air_act_chemical	string
classification	string
metal	string
metal_category	string
carcinogen	string
pbt	string
pfas	string
form_type	string
unit_of_measure	string
release_fugitive_air	decimal128(30, 7)
release_stack_air	decimal128(30, 7)
release_water	decimal128(30, 7)
release_underground	decimal128(30, 7)
release_class_i_wells	decimal128(30, 7)
release_class_ii_v_wells	decimal128(30, 7)
release_landfills_legacy	decimal128(30, 7)
release_rcra_c_landfills	decimal128(30, 7)
release_other_landfills	decimal128(30, 7)
release_land_treatment	decimal128(30, 7)
release_surface_impoundment_legacy	decimal128(30, 7)
release_rcra_surface_impoundment	decimal128(30, 7)
release_other_surface_impoundment	decimal128(30, 7)
release_other_land_disposal	decimal128(30, 7)
onsite_release_total	decimal128(30, 7)
transfer_potw_release	decimal128(30, 7)
transfer_potw_treatment	decimal128(30, 7)
transfer_potw_total	decimal128(30, 7)
transfer_m10	decimal128(30, 7)
transfer_m41	decimal128(30, 7)
transfer_m62	decimal128(30, 7)
transfer_m40_metal	decimal128(30, 7)
transfer_m61_metal	decimal128(30, 7)
transfer_m71	decimal128(30, 7)
transfer_m81	decimal128(30, 7)
transfer_m82	decimal128(30, 7)
transfer_m72	decimal128(30, 7)
transfer_m63	decimal128(30, 7)
transfer_m66	decimal128(30, 7)
transfer_m67	decimal128(30, 7)
transfer_m64	decimal128(30, 7)
transfer_m65	decimal128(30, 7)
transfer_m73	decimal128(30, 7)
transfer_m79	decimal128(30, 7)
transfer_m90	decimal128(30, 7)
transfer_m94	decimal128(30, 7)
transfer_m99	decimal128(30, 7)
offsite_release_total	decimal128(30, 7)
transfer_m20	decimal128(30, 7)
transfer_m24	decimal128(30, 7)
transfer_m26	decimal128(30, 7)
transfer_m28	decimal128(30, 7)
transfer_m93	decimal128(30, 7)
offsite_recycled_total	decimal128(30, 7)
transfer_m56	decimal128(30, 7)
transfer_m92	decimal128(30, 7)
offsite_energy_recovery_total	decimal128(30, 7)
transfer_m40_non_metal	decimal128(30, 7)
transfer_m50	decimal128(30, 7)
transfer_m54	decimal128(30, 7)
transfer_m61_non_metal	decimal128(30, 7)
transfer_m69	decimal128(30, 7)
transfer_m95	decimal128(30, 7)
offsite_treated_total	decimal128(30, 7)
transfer_unclassified	decimal128(30, 7)
transfer_offsite_total	decimal128(30, 7)
total_releases	decimal128(30, 7)
production_releases_legacy	decimal128(30, 7)
production_onsite_contained	decimal128(30, 7)
production_onsite_other	decimal128(30, 7)
production_offsite_contained	decimal128(30, 7)
production_offsite_other	decimal128(30, 7)
production_energy_recovery_onsite	decimal128(30, 7)
production_energy_recovery_offsite	decimal128(30, 7)
production_recycling_onsite	decimal128(30, 7)
production_recycling_offsite	decimal128(30, 7)
production_treatment_onsite	decimal128(30, 7)
production_treatment_offsite	decimal128(30, 7)
production_waste_total	decimal128(30, 7)
one_time_release	decimal128(30, 7)
production_ratio_or_activity	string
production_ratio	decimal128(30, 7)
source_file	string
source_row	int64
latitude_raw	string
longitude_raw	string
coordinate_status	string
quantity_reporting_status	string
facility_record_id	string
onsite_release_total_difference	decimal128(30, 7)
onsite_release_total_matches_components	bool
offsite_release_total_difference	decimal128(30, 7)
offsite_release_total_matches_components	bool
total_releases_difference	decimal128(30, 7)
total_releases_matches_components	bool
transfer_potw_total_difference	decimal128(30, 7)
transfer_potw_total_matches_components	bool
offsite_recycled_total_difference	decimal128(30, 7)
offsite_recycled_total_matches_components	bool
offsite_energy_recovery_total_difference	decimal128(30, 7)
offsite_energy_recovery_total_matches_components	bool
offsite_treated_total_difference	decimal128(30, 7)
offsite_treated_total_matches_components	bool
transfer_offsite_total_difference	decimal128(30, 7)
transfer_offsite_total_matches_components	bool
production_waste_total_difference	decimal128(30, 7)
production_waste_total_matches_components	bool

epa_tri_facilities.parquet — 44,385 rows · 41 fields

Field	Type
facility_record_id	string
year	int64
trifid	string
frs_id	string
facility_name	string
street_address	string
city	string
county	string
state	string
zip	string
bia	string
tribe	string
latitude	double
longitude	double
horizontal_datum	string
parent_company_name	string
parent_company_dun	string
standard_parent_company_name	string
foreign_parent_co_name	string
foreign_parent_co_db_num	string
standard_foreign_parent_co_name	string
federal_facility	string
industry_sector_code	string
industry_sector	string
sic_1	string
sic_2	string
sic_3	string
sic_4	string
sic_5	string
sic_6	string
naics_1	string
naics_2	string
naics_3	string
naics_4	string
naics_5	string
naics_6	string
latitude_raw	string
longitude_raw	string
coordinate_status	string
source_record_count	int64
facility_year_variant_count	int64

Inspect the public CSV

import pandas as pd

df = pd.read_csv(
    "https://huggingface.co/datasets/claritystorm/epa-tri-toxic-release-inventory/resolve/main/sample_1000.csv"
)
print(df.shape)
print(df.columns.tolist())
print(df.head())

This CSV is a sample, not the full package. It does not establish complete historical coverage or represent every field’s missingness. License details are on the Hugging Face card.

Source: government source portal · Dataset changelog