Date of last update (on RTK NET): 03/2004
First year of data available: 1987
Last year: 2002
Update cycle (how often a new set of data is made):
yearly
See "Changes for 2001 Data" below for important change
information.
The Toxic Release Inventory (TRI) is a database of
information about releases and transfers of toxic chemicals from manufacturing
facilities. Facilities must report their releases of a toxic chemical to TRI if
they fulfill four criteria:
- They must be a manufacturing facility (primary SIC code in 20 -39) or in one
of a number of non-manufacutring industries added for the 1998 reporting year;
- They must have the equivalent of 10 full-time workers;
- They must either manufacture or process more than 25,000 lbs of the chemical
or use more than 10,000 lbs during the year (unless the chemical is a "PBT", see
below);
- The chemical must be on the TRI list of over 600 specific toxic chemicals or
chemical categories.
Therefore, not all, or even most, pollution is
reported in TRI. However, TRI does have certain advantages:
- It is multi-media. Facilities must report the amounts they release to air,
land, water, and underground separately, and must report how much they send
off-site;
- All quantities are reported in pounds. This is an advantage compared to
databases like PCS, which often report releases as concentrations, or other
databases which report releases by volume of waste. These measures are often
impossible to convert into pounds;
- It is congressionally mandated to be publicly available, by electronic and
other means, to everyone. This means that it's relatively easy to obtain TRI
data and that the data is well-known, becoming a national "yardstick" for
measuring progress in pollution and waste generation.
The TRI data is
reported by individual facilities, who send their reports to the federal
Environmental Protection Agency (EPA) every
year. These reports are filled out on a form called "Form R". EPA takes these
forms and converts them into an electronic database. To better understand TRI
data, it is recommended that you order a copy of one of these forms from the TRI
Hotline (1-800-424-9346). You can also order (for free) a national "data
release", or summary on paper, of TRI data every year from the
Hotline.
Changes Beginning with TRI Data for 1995
For the
1995 data year, a Certification Statement (Form C) was added to TRI. This, in
response to lobbying by industry, was added to permit facilities to avoid
reporting if their total releases and production-related waste were less than
500 pounds. A facility filing a Certification Statement must report its name and
address and the name of the chemical but does not have to reveal release,
transfer, or waste information.
Changes Beginning with TRI Data for
1998
There are important changes in the
1998 TRI data. RTK NET users should probably look at our changes document.
Changes for 1999
Data
The "frozen" and "current" versions of the data for 1999 are
identical, except that the "frozen" data do not include delisted chemicals. (See
the explanation of changes for 1998 for more
information on "frozen" versus "current" versions.)
Changes for 2001
Data
The 2001 TRI reporting year adds data for PBTS, persistent
bioaccumulative chemicals. These chemicals of special concern can be dangerous
at low levels, so EPA decreased the TRI reporting thresholds for them. In
addition, they can now be reported using fractional pounds. The most toxic PBT,
dioxin, is reported as total grams. TRI facilities can optionally report the
percentage breakdown of total dioxin into 17 specific chemicals; these
percentages are available through the High detail report. For more information
on PBTs, see the EPA Web site on TRI.
The "frozen" and "current" versions of
the data for 2001 are identical, except that the "frozen" data do not include
delisted chemicals. (See the explanation of changes
for 1998 for more information on "frozen" versus "current"
versions.)
TRI Releases and Release
Ranges
-------------------------------
When facilities report TRI releases using the Form R, they can do so in two
different ways. They can either specify a certain number of pounds released or
specify a range in pounds (such as 0- 500 lbs or 500-100 lbs). It is standard
practice when interpreting TRI data to treat a range of 0-500 lbs as 250 lbs
(the midpoint of the range) and 500-1000 as 750 lbs. Difficulties used to arise
over this issue in RTK NET when users would obtain release totals that did not
count these range indicators. To solve this problem, we created new fields that
hold release numbers in pounds, whether they were reported as ranges or not. The
release ranges and their midpoints are listed below:
| Range |
Midpoint |
| 1 - 499 lbs |
250 lbs (used 1987 - 1989 only) |
| 500 -1000 lbs |
750 lbs (used 1987+) |
| 1 - 10 lbs |
5 lbs (used 1990+) |
| 11 - 499 lbs |
250 lbs (used 1990+) |
RTK NET uses the midpoints
of ranges within all of its release and transfer numbers.
Chemical
Identifiers in TRI
---------------------------
TRI contains information about different types of substances:
- chemicals, which have a name and CAS number;
- chemical categories, which have a name but no CAS number, but do have a
chemical id (see below);
- mixtures, which have a free-form name and no CAS;
- trade secret chemicals, which have a generic name and no CAS.
Since many chemicals have more than one name, it is often safer to search by
CAS (Chemical Abstract Services) number. The CAS numbers in TRI are stored in a
format that is left filled with zeroes to 9 digits, and contains no dashes, so
that the CAS number "50-00-0" (for Formaldehyde) would be stored as "000050000".
There is also another field called "chemical_id". This is an integer which
contains the same value as the CAS number for chemicals. It also contains a
unique number (assigned by RTK NET) for chemical categories. Its major use is
for searches that specify particular chemical categories.
Most of the primary metals listed in TRI also are listed as a class of
"compounds". For instance, Lead can be reported as a chemical (with CAS number
7439-92-1) or as the chemical category "Lead Compounds". If you are doing a TRI
search in which your goal is to find all releases of lead, you should probably
search with chemical_name matching "LEAD*" (the * is a wildcard character). This
search will find both "Lead" and "Lead Compounds".
Data Quality Problems
in TRI
----------------------------
The EPA Office that runs TRI, Office of Pollution Prevention and Toxics (OPPT), has spent considerable effort on
improving TRI data quality. In particular, OPPT has checked that all submissions
filed by the same facility have the same TRI facility id number ( facility-id ).
They also have checked the cities, states, counties, and zip codes of facilities
to see if they match up with each other. However, they did not check the
addresses or ID numbers of off-site locations. These ID numbers, names, and
addresses of off-site facilities and Publicly Owned Treatment Works (POTWs) have
very poor data quality.
OPPT queries some facilities that send in submissions with very high release
numbers to see if the numbers are correct. There is also a program under which
TRI facilities can be inspected (not many are). Other than this, the numbers
that a facility submits for release and transfer quantities are not checked.
Bear in mind in your analysis that TRI release quantities are supposed to be
estimated to only two significant digits. That means that if you add up a number
of TRI releases and get a number like 11,264,586 pounds, only the first two
digits have any meaning. You would be better off writing this number as 11
million pounds.
Explanation of the 1991 Pollution Prevention Act TRI Data
Elements
------------------------------------------------------------------
The 1991+ TRI data release contains a number of new data elements dealing
with waste generation, source reduction, and recycling. These data elements were
mandated by the Pollution Prevention
Act, which added to TRI. They can be confusing, especially when compared
with the old TRI, so this message is going to attempt to explain the new data
and its uses. This message was written by RTK NET staff and should not be
considered in any way to be an official EPA interpretation of the data or to
reflect EPA policy. I also recommend the TRI Public Data Release to all RTK NET
users; it can be ordered for free from the TRI Hotline at 1-800-535-0202.
An outline of this message is given below:
- Transfers for Recycling and Energy Recovery: Closing the "Recycling
Loophole".
- What is Source Reduction? ; Methods of dealing with waste.
- What do "Waste Generation" numbers mean?
- Waste Generated = Amount recycled + Amount treated + Amount released +
Amount burned for energy recovery
- Production vs. Non-Production waste
- Production Ratio
- What is the relationship between "Releases and Transfers" and "Waste
Generation"?
I. Transfers for Recycling and Energy Recovery: Closing the "Recycling
Loophole".
In previous years, some environmentalists complained that TRI does not
require reporting transfers to "energy recovery" facilities such as cement kilns
or industrial boilers where hazardous waste is burned. In previous years, these
transfers were considered to be for the purpose of recycling and did not have to
be reported. Starting with the 1991 TRI data all transfers off-site had to be
reported whether for the purposes of recycling, energy recovery, treatment, or
disposal. Therefore the total quantity of transfers will go up dramatically
because of the new reporting requirements.
Energy Recovery is now defined as a situation in which waste with a
significant BTU (heat generated) value is burned in a facility which recovers
the heat generated for some useful purpose. In terms of what actually happens to
the waste, energy recovery facilities are similar to incinerators.
On RTK NET, different types of releases and transfers are distinguished by
their "environmental medium", as listed below:
Releases on-site-
- Fugitive Air
- Stack Air
- Water
- Underground Injection
- Land
Transfers off-site-
- To Publicly Owned Treatment Works (POTWs)
- Other Off-site Transfers
All of the newly reported transfers to recycling and energy recovery
facilities, as well as the old-style transfers to treatment and disposal
facilities, are reported as medium 7 above. The way to distinguish between a
transfer to disposal and one for recycling is to look at the transfer's
treatment type code. Each specific transfer must have an associated destination
for the waste and a code indicating what happens to the waste once it gets
there. These codes (for 1991+) are listed below:
Recycling Transfers:
- M20 SOLVENTS/ORGANICS RECOVERY
- M24 METALS RECOVERY
- M26 OTHER REUSE OR RECOVERY
- M28 ACID REGENERATION
- M93 TRANSFER TO WASTE BROKER-RECYCLING
Energy Recovery Transfers:
- M56 ENERGY RECOVERY
- M92 TRANSFER TO WASTE BROKER-ENERGY RECOVERY
Treatment Transfers:
- M40 SOLIDIFICATION/STABILIZATION
- M50 INCINERATION/THERMAL TREATMENT
- M54 INCINERATION/INSIGNIFICANT FUEL VALUE
- M61 WASTEWATER TREATMENT (EXCLUDING POTW)
- M69 OTHER WASTE TREATMENT
- M95 TRANSFER TO WASTE BROKER-WASTE TREATMENT
Disposal Transfers:
- M10 STORAGE ONLY
- M71 UNDERGROUND INJECTION
- M72 LANDFILL/DISPOSAL SURFACE IMPOUNDMENT
- M73 LAND TREATMENT
- M79 OTHER LAND DISPOSAL
- M90 OTHER OFF-SITE MANAGEMENT
- M94 TRANSFER TO WASTE BROKER-DISPOSAL
- M99 UNKNOWN
Note that metals or metal compounds should never be reported as being sent to
"energy recovery", since they have no BTU value. However, significant quantities
of metals were reported as being sent to "energy recovery" in the 1991 data.
This indicates that many facilities did not understand the instructions for
filling out the "Form R"; the form on which they report their TRI numbers to
EPA. I recommend that you examine these "treatment type codes" when you analyze
transfers, since the data quality is apparently bad. In particular, many
companies will argue that transfers for recycling and energy recovery should not
be counted into their release and transfer total. If you do decide to treat
recycling and energy recovery transfers separately, you should make sure that
these transfers have been appropriately labelled and are not actually transfers
for treatment or disposal. Whenever RTK NET shows a data field like "Total
Transfers" or "Total Releases and Transfers", this datum will include the sum of
all transfers, no matter what their treatment type codes, as well as all
transfers to POTWs.
II. What is Source Reduction? ; Methods of dealing with waste.
The Pollution Prevention Act sets out a hierarchy of methods which should be
used to handle waste: source reduction, recycling, treatment, and disposal.
Source reduction refers to the process of preventing waste from being generated.
This is the best method of "handling" toxic waste since the waste is never
generated in the first place. According to the Pollution Prevention Act, waste
should first be prevented; waste which can not be prevented should be recycled;
that which can not be recycled should be treated and destroyed; that which can
not be treated should be disposed safely.
TRI has always required that facilities report which methods they were using
for on-site and off-site treatment and disposal. To encourage the "Pollution
Prevention Hierarchy" above, facilities are now required to report the methods
they use for source reduction and recycling as well. Due to pressure from
industry, "energy recovery" was also added by EPA as a category that comes in
between recycling and treatment on the hierarchy.
EPA added the following new data elements to meet the requirements of the
Pollution Prevention Act: methods of source reduction used, on-site recycling
methods, on-site energy recovery methods, and methods used to identify
opportunities for source reduction. Of these new data elements, only the Methods
used for Source Reduction have been included in data available through the
Standard Reports; the rest are available through the TRI Form Query.
III. What do "Waste Generation" numbers mean?
A. Waste Generated = Amount recycled + Amount treated + Amount released +
Amount burned for energy recovery
In addition to requiring that
facilities report what methods of source reduction they were using, the
Pollution Prevention Act also requires that facilities report numerical measures
of progress in source reduction. The only way to measure progress in source
reduction is to see how much toxic waste is generated every year. "Waste" in
this context is defined as any waste that is generated and must be handled in
some manner, even if it is later recycled. To understand this concept a bit
better, imagine a waste stream at a facility:
Waste Generated --
Release or Transfer |
Recycling --
!
Amount Recycled |
Energy Recovery --
! ! ! |
Treatment
!
Amount Destroyed |
|
|
Amount Destroyed |
|
Any amount of waste that the facility generates is either going to be
recycled and reused, destroyed by being burned in an energy recovery facility,
destroyed in a treatment facility, or released or transferred for disposal
elsewhere. Of course, there can be releases from any stage of the process -- for
instance, energy recovery facilities only destroy part of the waste that they
burn and release the rest. But this diagram shows in an idealized way all of the
possible final destinations of the waste once it is generated. Any waste that is
not recycled or destroyed by energy recovery or treatment will be released (or
disposed of off-site). Previous TRI data elements only cover the rightmost part
of this diagram; the Pollution Prevention Act adds information about the
rest.
The Pollution Prevention Act adds the following data elements to let you keep
track of what is happening to the waste once it is generated:
- Amount Recycled On-site
- Amount Sent Off-site for Recycling
- Amount Destroyed by Energy Recovery On-site
- Amount Sent Off-site for Energy Recovery
- Amount Treated (Destroyed) On-site
- Amount Sent Off-site for Treatment
- Amount Released On-site or Disposed Off-site
These data elements cover all of the branches of the diagram above. You can
figure out the total amount of waste generated by adding up the amounts sent to
all of these branches. In other words, the "Waste Generated" amount at the head
of the diagram was not directly reported (due to interference from the Office of
Management and Budget). Instead, you must find it by adding up the waste sent to
all of the possible destinations (RTK NET has done this already and placed the
result in a separate data field).
B. Production vs. Non-Production waste
There is a further
complication. The waste generated is divided into two different categories
according how it was generated. Waste that was generated because of an accident
(like an explosion or tornado), because of a remedial action (like when a plant
cleans up a Superfund site on-site and sends the collected waste off-site) or
from some other "one-time event" is reported as a separate category from all of
the data elements above. This is because this type of waste is assumed to be
less amenable to source reduction -- assuming that the tornado damage could not
have been prevented, that the waste from the Superfund site had been spilled
years ago, etc. The drafters of the Pollution Prevention Act wanted to keep
track of waste that could be prevented and waste that couldn't separately. So
they required a separate data element for waste from accidents, remedial
actions, or other one-time events. Waste reported as part of this data element
should not be double counted and reported as part of any of the other Pollution
Prevention Act data elements above.
On RTK NET, the waste that was generated because of an accident, remedial
action, or other one-time event is referred to as "Non-Production-Related
Waste", since it is waste that is not related to the design or operation of the
normal production process at the facility. The total waste generated from the
other categories is referred to as "Production-Related Waste".
Production-related waste is assumed to be possibly amenable to source reduction,
since in theory the production process could be re-designed to generate less of
this waste. Finally, RTK NET has a calculated data element called "Total Waste"
which simply adds the production-related and non-production-related waste
together.
Confused yet? Let's go back over the data elements:
Production-related waste-
- Amount Recycled On-site
- Amount Sent Off-site for Recycling
- Amount Destroyed by Energy Recovery On-site
- Amount Sent Off-site for Energy Recovery
- Amount Treated (Destroyed) On-site
- Amount Sent Off-site for Treatment
- Amount Released On-site or Disposed Off-site
- Total Production-Related Waste (sum of 7 numbers above) (This is the amount
of waste generated that is potentially amenable to source reduction).
- Non-Production-Related Waste (accidental, remedial, or one-time)
- Total Waste Generated (sum of Total Production-Related & Non-Production
Related Waste)
There's even more data elements than this. To track trends in source
reduction from year to year, the Pollution Prevention Act requires that the
"production-related" data elements be reported for the current data year, for
the previous year, and projected for two years into the future. This means that
all 7 of the production-related data elements above were reported for 1990 and
1991 and projected for 1992 and 1993. (For this first year of data, facilities
did not have to report back one year to 1990 if they didn't still have data
records for that year.)
C. Production Ratio
Finally, facilities also had to report a
"Production Ratio" or "Production Index" for each chemical. This number
represents the ratio of production or some other activity related to waste
generation in the current year to that in the previous reporting year. For
example, if a facility made 1000 cars in 1990, and only 900 in 1991, then their
production ratio would be 0.9 . If they made 1000 cars in 1990, and production
rose to 1100 in 1991, then their production ratio would be 1.1 . This permits a
user of the data to see whether changes in waste generation from year to year
were due to source reduction or simply to production changes. For instance, if
production in a facility went down by 10%, you would also expect the amount of
waste produced by the plant to go down by 10% -- if you assume a direct
relationship between production and waste generation. The production ratio can
give you an idea whether production levels varied or whether the process of
production really changed so that less waste was produced per unit of
production.
These PPA data elements are available through the Medium and High levels of
detail in the TRI Standard Reports on RTK NET. (Only totals of waste generated
are available through Low and Summary detail; and only High detail has the full
breakdown of quantities through four years.) All of these data are available
through the TRI Form Query on RTK NET.
D. What is the relationship between "Releases and Transfers" and "Waste
Generation"?
Some users of TRI data may be confused about the
relationship of releases and transfers with waste generation. I think that it is
most helpful to think of these as two completely different categories of
numbers. In theory, all TRI releases and transfers should constitute part of the
"waste generation" numbers, either as production-related or non-production
related waste. However, some facilities appear to have misinterpreted the Form R
instructions. EPA added up numbers for all facilities and found that many totals
which should have matched each other did not. For instance, the total quantity
of waste reported sent offsite for recycling in the "transfer" section of the
Form R does not equal the total quantity reported recycled offsite in the "waste
generation" section. (These two numbers should be the same). If there is no
non-production-related waste to complicate things then, in theory, the
relationships between these elements should be as follows:
| Releases & Transfers |
|
Waste Generation |
| Releases to air, land, water, underground + Transfers for disposal
off-site |
= |
Amount "Released" |
| Transfers to POTWs or to treatment off-site |
= |
Amount Treated Off-site |
| Transfers to recycling facilities off-site |
= |
Amount Recycled Off-site |
| Transfers to energy recovery off-site |
= |
Energy Recovery Off-site |
Note that if non-production-related waste does exist, then the above
relationships do not necessarily hold since any release or transfer could be
part of the non-production-related waste.
The TRI public data release has a good section on these relationships and how
they do not in general hold in the data due to reporting errors. In general,
remember that what RTK NET reports is whatever the facility had written down on
its Form R, whether that makes sense or not. The only exceptions are the "total"
quantities, where RTK NET has added up various numbers that the facility
reported. These are:
- Total Releases: all releases to media 1 through 5
- Total Transfers: all transfers to media 6 and 7 (and medium 8: see part IV
of this message below)
- Total Releases and Transfers: all releases/transfers to all media
- Total Production-Related Waste: Amount "released" (section 8.1 of Form R)
plus amount recycled on-site (section 8.2) plus amount sent off-site for
recycling (8.3) plus amount destroyed in energy recovery on-site (8.4) plus
amount sent off-site for energy recovery (8.5) plus amount treated on-site (8.6)
plus amount sent off-site for treatment (8.7)
- Total Waste Generated: Total Production-Related Waste as defined above plus
Non-Production-Related waste (section 8.8 of Form R)