Exploring the Toxics Release Inventory (TRI) for New Jersey

TRI.NET is a defunct database program, previously accessible to search the EPA Toxics Release Database.  It is not longer updated by the EPA.  An archived version, allowing queries from 1987-2013 is available here.  This was last updated in 2015.

In 2000, 757 New Jersey facilities were identified in the TRI database (EPA TRI Folder > TRI_NET_Queries > TRI2000.shp).  545 of those facilities report toxic weighted emissions (toxicity x emissions) that were greater than zero (TRI2000TW.shp).

The top five facilities with the highest on-site hazard score (or toxic weighted pounds equivalent emissions) for 2000 in New Jersey were:

1. 3M Co., Belle Meade, NJ, 313 billion TWPE (Toxic Weights Pounds Equivalents)
2. DuPont Chambers Works, Deepwater, NJ, 122 billion TWPE
3. Gac Kearny Inc., Kearney, NJ, 38 billion TWPE
4. BL England Generating Station, Beesleys Point, NJ, 33 billion TWPE
5. Custom Ally Corp., High Bridge, NJ, 16.7 billion TWPE

The minimum TWPE was 1.5 TWPE, the mean was 1.15 billion TWPE, with a standard deviation of 14 billion TWPE.  The sum of all TWPE was 638,868,266,587.80029.  Most facilities score low on reported TWPE, as the histogram shows.  In fact, the top five TWPE emissions account for 82 percent of all emissions; the top 10 facility emissions account for 87 percent of total state emissions.

[The source for the hazard score is from the US EPA Risk Screening Environmental Indicators (RSEI).]



This makes me wonder about the usefulness of proximity based analysis used in environmental justice research that considers the TRI.  Clearly all TRI facilities are not equivalent.  The relative risks to health, if they are associated with the TWPE measure, imply the very largest emitters are of greatest concern, and so the communities that are proximate to these sites would be at much greater environmental risk than the full facility data set.   The EPA DMR tools adds the caveat about TWPE: "this value is not a measure of risk or potential for human health impacts."  The reason being that TWPE does not account for pollutant fate and transport, meaning how it travels and where it ends up is not evaluated by TWPE.  I think this is likely an overly cautious comment about interpreting the hazard score.  Clearly hazard score has some association with risk - otherwise it wouldn't be called a hazard score and wouldn't be a part of the "risk screening environmental indicators" process.


Note about parcels for New Jersey:

I think I might like to link the EPA TRI point data to New Jersey parcels to assess the relative size of the facilities in the TRI.  This began with acquiring the parcel data.

Parcels for the state of New Jersey were downloaded by county from NJGIN.  These were unziped with 7 Zip to be merged into a single state file for parcels.  7 Zip permits unzipping multiple zipped files in batch, and saving all files into a single folder (instead of multiple folders).  This speeds up processing for merging multiple files in ArcGIS.  The union, merge, and append tools were all used to attempt to combine all county-level parcel data into a state-wide parcel data set, but failed.  There may have to do with different data structures of the different county parcel files.  I will investigate this further.