RTK or Left-To-Wonder?
Recommended Reading

EPA Office of Pollution Prevention and Toxics Issues Paper #2

Expansion of the Toxics Release Inventory (TRI)
to gather chemical use information: TRI-Phase 3


This issue paper provides an update on the EPA project known as "TRI-Phase 3". The project, conducted by the EPA Office of Pollution Prevention and Toxics (OPPT), is exploring a broad set of issues related to the expansion of the Community Right to Know Program to include facility-level chemical "use" information. While the project is still in its early stages, the agency has made some preliminary findings on the issues, and these are described in this report.

On August 8, 1995, in a memorandum to the EPA administrator, the President directed EPA to develop and implement "an expedited, open, and transparent process for consideration of reporting under EPCRA [Emergency Planning and Community Right-to-Know Act] on information on the use of toxic chemicals at facilities, including information on mass balance, materials accounting, or other chemical use data pursuant to section 313(b)(1)(A) of EPCRA, 42 U.S.C. 11023(b)(1)(A)." The memorandum, issued in conjunction with Executive Order #12969, went on to direct EPA to submit a report on the progress of this effort by October 1, 1995.

EPA previously described the origin of this issue and the value of use data in the first TRI-P3 issue paper, dated 9/2/94. In addition, a public meeting was held on 9/28/94. EPA has reviewed comments from the meeting and has evaluated the issues raised at the meeting. The major finding is that full materials accounting information provides important insights for looking at a variety of environmental issues, and that the agency supports the goal of collecting this data. The purpose of this paper is to report back to stakeholders, to lay out EPA's position on materials accounting, and to focus attention on important unresolved questions. It will also serve to describe an expedited, open, and transparent process for further consideration of TRI-P3 issues.

EPA finds that there are significant differences among stakeholders on the value of chemical "use" information, and this paper includes a variety of questions and a set of data element options designed to facilitate further discussion. These and other issues considered important to stakeholders can then be discussed at the next public meeting to be held on October 18 and 19, 1995. The goal of the meeting will be to provide the public an opportunity to express their views on the types of additional data that should be considered by EPA as project development continues.


TRI, established under the Emergency Planning and Community Right-to-Know Act of 1986 (EPCRA), is considered one of the most successful policy instruments ever created for improving environmental performance. It uses the power of public information to motivate and empower initiatives at all levels, from facility teams, to community groups, to trade associations and state and local governments.

Given the success of TRI, there has been tremendous interest in expanding it to include additional facilities and a more comprehensive data set. The Pollution Prevention Act of 1990 (PPA) created a source reduction and recycling report with the existing TRI Form R to collect information on quantities entering waste streams including amounts treated and recycled, both on and off-site. It also added data elements to describe types of source reduction efforts. In 1993, an executive order expanded reporting to include federal facilities. In 1994, a chemical expansion rule was promulgated, expanding TRI by 286 new chemicals and categories. EPA plans to follow this phase 1 expansion with a phase 2 expansion that adds to the types of industries covered by TRI. In addition to expansion, EPA in 1994 fine-tuned TRI by streamlining reporting for those facilities which generate small quantities of waste.

TRI-Phase 3 began by looking at whether chemical use information could fill in those remaining Right-to-Know gaps in TRI by providing information on chemicals moving through communities and workplaces and in products. In the context of this project, chemical use has referred to "input-output", or "throughput" information on the amounts of toxic chemicals used at a given facility. At the outset of the project, environmental stakeholders identified "materials accounting" (MA) information as having the greatest potential for filling these gaps. Materials accounting involves determining the quantity of a chemical at key junctures in its progression through a facility, using readily derived information such as invoice and product composition data. It is distinct from engineering mass balance, which is a more technically rigorous excercise that includes measurement of process streams. In 1990, the National Research Council concluded that a materials accounting approach would best satisfy the public's Right-to-Know with a minimum of burden.


At the September 1994 public meeting, the central message provided by labor, environmental justice, and public interest groups was that while TRI was an extremely valuable tool, it fell short of providing the complete right-to-know picture needed to fully understand toxic chemical issues. In almost every case, these stakeholders described materials accounting data as the best remedy for filling these gaps. On the other hand, industry and trade association groups were strongly opposed to the need to collect MA information. They disagreed with its value, claiming that such information had limited potential benefits and would confuse the public. In addition, they described concerns about costs of data collection, and the potential for loss of important confidential business information (CBI) to foreign competitors.

Given the different positions on MA data, EPA decided to begin its analysis by moving the issues back one step in an attempt to reframe the underlying issues more clearly. What Right-to-Know issues can be addressed with materials accounting data? What data would be helpful to expand Right-to-Know and how would it be used? How can we balance Right-to-Know with protection of legitimate confidential business information concerns? By using the underlying questions and gaps as a point of departure, EPA and TRI stakeholders might gain a better understanding of the information needed to answer the questions. EPA took the comments from environmental stakeholders and industry and examined them from this perspective. These insights were then used to help define issues for developing a common sense approach.


EPA developed a 3 step framework for evaluating TRI-P3 issues:

Step #1 Evaluate the public meeting comments to identify the underlying Right-to-Know questions. Turn these questions into TRI gap categories. In addition, identify and categorize industry concerns with TRI-P3.

Step #2 Take the TRI gap categories from Step 1 as the starting point for additional analysis.

Step #3 If there is a federal role in collecting this information, and EPA can identify the data elements to provide the necessary information, what are the options and relevant statutes for going about collecting it?

EPA supports the goal of collecting materials accounting data, and has begun the step of evaluating how to best achieve this goal. EPA will incorporate comments from the public meeting in the process of completing step 2, and will continue evaluation of Step 3 during 1996.



After evaluating the public comments from these groups, 7 Right-to-Know information gaps were identified. These so-called gaps are described briefly below. Please note that this review is not intended to provide a comprehensive review of stakeholder comments, or to describe EPA positions in response. EPA is interested in learning more from stakeholders on how these gaps rank in importance to them.

  1. Need for information on the flow and use of toxics at a facility. Several groups indicated that the current TRI data, with its focus on releases and other wastes, does not allow the community to understand the significance of chemical use and its relation to emergency preparedness issues. Paula Forbis of the Environmental Health Coalition of San Diego stated that: "...the critical part of materials accounting data is that it is basically the where-do-we-go-from-here information. It allows the community to look at possibilities for pollution prevention, possibilities for avoidance of accident risk through changes in storage practices, and it also allows them to have better information when they go to some of these industries." Wilma Subra of the Louisiana Environmental Action Network (LEAN) stated that materials accounting data would "provide information to the community on the quantities of chemicals and products shipped through their community". She went on to say that the "community can then utilize the data to determine if special precautions need to be developed to to address various issues such as transportation of the chemicals and the potential vulnerability of the community". Use information was also seen as providing pollution prevention insights as well. Jacqueline Savitz of the Chesapeake Bay Foundation commented that not all companies have looked closely at overall chemical use. She provided an example of a naval aviation depot that made the change from five shops all ordering their own chemicals from different suppliers to a centralized operation. By looking at use and centralizing ordering and dispensing, they reduced hazardous waste generation, reduced the need to stock chemicals by 75%, and increased recycling 200%.

  2. The need for tracking toxics in products. Another TRI gap pointed out by environmental stakeholders is the lack of relevance for product issues. As stated by Hillel Gray of the National Environmental Law Center: "TRI does not provide information on the amount of toxic chemicals put into products". Product stream data is considered important for two reasons. First, it is a component included in materials accounting, needed for certain P2 measures. Second, product use can also lead to releases and can contribute to waste streams, depending on their properties and functions. Including product reporting in TRI would add an important capability for looking further at these issues. Hillel Gray provided an example of a Massachusetts facility that reported only 8 pounds of arsenic releases to TRI, but where over 300,000 pounds were contained in wood products. This example shows that sometimes it is the product that needs to be followed to find out if there are any important risks or P2 opportunities. As stated by Mike Aucott of the New Jersey Department of Environmental Protection (NJDEP): "For some chemicals, most of what leaves the site is in products, and while the fate can vary, for some chemicals a lot of what gets into the environment is from products. We need to see the whole picture of where chemicals are getting into the environment." Eric Howard of the World Wildlife Federation, commented that product data are ..." useful both to identify prevention options and to assess possible exposure and the potential for accidents during other parts of the commercial cycle such as transport, storage, use, recycling, and disposal."

  3. TRI is an inadequate tool for looking at occupational issues. Labor and environmental stakeholders are increasingly interested in the connections between occupational and environmental issues. Source reduction, with the potential to reduce both worker and environmental exposures, is seen as the logical nexus for this connection. Also, the existing Section 8 TRI data elements, which include items on worker participation in identifying source reduction opportunities, have further stimulated interest in worker-oriented data elements for TRI. Many in these stakeholder groups believe that there are important information gaps on worker exposure, and that TRI, as the primary public database on toxics, is where this information belongs. (Neither the Occupational Safety and Health Administration (OSHA), the National Institute for Occupational Safety and Health (NIOSH), or the Mine Safety and Health Administration (MSHA) has a publicly accessible database for tracking facility performance on toxics).

    Throughput data is seen as an improvement over releases and other wastes as a means to identify facilities and industries where worker exposures may be important. In addition, other data elements to indicate potential or actual exposure have been suggested. Pam Tau Lee of the University of California stated that "Worker exposure is among the highest environmental risks, and there is a need to expand the information that is useful for protecting worker health." Rex Tingle of the AFL-CIO referred to the need to protect both environments inside and outside the workplace and suggested that "occupational exposure information" be added to TRI. Paul Orum of the Working Group on Community Right-to-Know (WGCRTK) supported expanding TRI to look at these issues in order to "raise the visibility of workplace chemical hazards and to improve our ability to promote and track improvements."

  4. The need for a "scorecard" to promote and track pollution prevention/source reduction efforts. This is the primary gap mentioned most often by stakeholders. TRI already functions as a successful scorecard for releases and other wastes, but not as one that allows measurement of facility P2 performance in terms of source reduction. EPA defines pollution prevention as meaning source reduction, and the need for a "P2 scorecard" refers to the need to measure source reduction. Environmental stakeholders claim that MA data is the best way to add this capability to TRI. For example, Elizabeth Collaton of the Northeast-Midwest Institute stated: "Use of TRI data alone is not going to result in significant source reduction of toxic chemicals simply because TRI data focus on waste stream chemicals". Ed Hopkins of Citizens Action stated that his group has tried to use the TRI P2 data to see whether industries are actually making P2 progress, but has been quite unsuccessful. He stated that " cannot make that analysis with the TRI data as they stand", and went on to support MA data, saying that "We think that the ..information that we're discussing here is really essential, because it will provide the information to allow us to determine whether pollution prevention policies at the national and state level are encouraging prevention at the source, and whether individual facilities are succeeding in pollution prevention efforts." Ted Smith, of the Silicon Valley Toxics Coalition, suggested that measuring progress by the unit of product was also an important concept. He reported that Texas Instruments has published data on how for each six-inch silicon wafer produced, 3,200 cubic feet of bulk gases, 20 pounds of chemicals, 22 cubic feet of hazardous gases, and additional amounts of energy and water are used. He indicated that this was the general direction in which P2 measurement should be going.

  5. Lack of a ledger check on TRI estimates. Environmental stakeholders reported concerns about the accuracy of TRI reporting, and indicated that materials accounting reporting provided a means for cross-checking estimates of releases and wastes. For example, Eric Howard of the World Wildlife Fund submitted a report which stated that because MA data "...allow independent calculating of losses to the environment ... they provide - like a financial ledger - a cross-check for both companies and government that estimates are reasonable." John Chelen of the Unison Institute stated that use information: "..will provide a check on the data and make the data more accurate. I think it will implicitly create a tension between reporting releases and reporting substances in product. We have been concerned about people hiding releases by in fact letting these chemicals continue into product."

  6. The need for TRI to serve as a better tool for regulatory integration efforts. The multi-media nature of TRI promotes integration of regulatory efforts at the facility level. Stakeholders indicated that this function would be greatly improved by adding materials accounting data. Because chemical use information allows a more complete understanding of environmental performance, it is seen by some as a necessary prerequisite for consolidation of environmental reporting and for creating the conditions to allow flexibility in facility approaches to environmental compliance. For example, MA data has been found to be valuable for multi-media permitting in both New Jersey and Massachusetts. It gives a better perspective on the facility, and it provides the basis for selection of P2 solutions over end-of-pipe solutions for permitting. As stated by Lisa Doerr of Citizens for a Better Environment of Minnesota: "So if industry wants to look at a new dawning age of regulatory flexibility, if the agencies want to find new ways to do business, we need data that we can tell what's really going on." Mike Leedie of the West County Toxics Coalition of Richmond California described an experience where facility managers were telling coalition members that they were reducing use of the chemical at issue while at the same time requesting a permit to double the current level of incineration of the chemical. Leedie stated that publicly available use information would make such inconsistencies less likely.

  7. Others (research and priority-setting, and improving the accountability of industry and government.) Lastly, several miscellaneous gaps were also mentioned. For example, Connie Tucker of the Southern Organizing Committee for Economic and Social Justice indicated that there is a great need for more answers on the impacts of toxic chemicals. She stated that chemical use data would improve the available framework for beginning to identify and research issues such as multiple chemical exposures. Hillel Gray expressed the view that chemical use data was a more logical choice for setting toxic chemical priorities than the waste data that currently makes up TRI. Lisa Doerr commented that projects such as the Common Sense Initiative and the Great Printers Project were on the right track, with industry, government, labor, and community members working together on environmental issues. She stated that accountability was needed in order to have effective relationships like this, and that use information would be a great help in providing a basis for accountability.


Industry speakers did not focus on TRI gaps and how to fill them. Their positions were related to areas of concern about the collection of chemical use information. The major industry concern categories are listed below:

  1. EPA should not focus on use - it is not an appropriate predictor of exposure, and use reduction is not an appropriate goal. The Chemical Manufacturers Association (CMA) submitted comments expressing the concern that adding materials accounting reporting to TRI was "intended to make "use" the measure of environmental responsibility". CMA contended that activities to reduce use would distract companies from the more relevant work of reducing releases and transfers. Wayne Rausch, speaking for the American Petroleum Institute (API), stated that "...we believe that releases are really the appropriate starting point for investigating exposure". API comments included objections to commercial chemical use being "equated" with pollution and exposure. Sharon Eisel, who represented the CMA at the public meeting, stated that exposure was a key issue, and that "exposure is not related to use, but to a number of factors specific to a process or an application". Patricia Franco of the Electronics Industry Association (EIA) stated that EPA was contemplating "pollution prevention based solely on reduced chemical use". Franco stated that "in particular, information that compares chemical inputs with outputs and accumulations does not necessarily inform the public as to whether there is likely exposure to harmful chemicals from the manufacturing process or from use of the final product". She went on to say "in short, the issue paper does not recognize that it is how the chemicals are used and what avenues of exposure such use causes and not the use itself that is of import." In addition, the CMA comments asked EPA to recognize that unlike chemical releases, there were identifiable benefits from chemical usage that made use reduction an inappropriate goal.

  2. Materials accounting would create an excessive burden - it is costly to collect and it puts confidential business information (CBI) at risk. According to John Defazio of the Chemical Specialties Manufacturers Association (CSMA): "Creation of a CUI would impose a significant burden on both the agency and industry without any commensurate benefit". Defazio stated that materials accounting reporting for a Massachusetts CSMA member required 150 hours of staff time for one chemical. The CMA provided a cost estimate for a large chemical manufacturer as $1.5 million the first year, and $800,000 per year thereafter. Patricia Franco of EIA suggested that tracking inputs and outputs would be very costly, and would increase the paperwork burden for TRI reporters. She went on to describe concerns that use reporting would result in disclosure of a great deal of confidential business information that is of keen interest to competitors. Sharon Eisel of CMA also stated that "confidential business information is certainly a concern. There is a whole multi-million dollar industry of competitive intelligence and reverse engineering." The CMA comments admitted that few CBI claims were filed under the New Jersey law, but stated that this was due to the burden involved with filing claims under the statute, and the lack of appreciation at the plant level that such information could compromise products and company business plans.

  3. Materials accounting is of limited utility, and presents counting problems not seen with releases. The CMA expressed concern that the collection of material accounting and occupational demographics data would not result in the benefits that EPA has anticipated. For example, CMA stated that collecting and reporting volumes of use would not necessarily measure source reduction efforts. In addition to causing "information overload", the CMA claimed that collected materials accounting data would not be complete enough to produce an accurate picture of what was really happening inside a facility's many processes. CMA also claimed that MA data did not provide a useful check on the accuracy of TRI reports, or for direct comparisons of facility efficiencies. The CMA was concerned about the possibility of "multiple counting" because while releases are only counted once, chemical volumes can be counted several times, for example starting out as benzene, then being turned into phenol, and then some other product before leaving the site. According to CMA, this might result in a substantial overstatement of perceived hazards posed by chemical use at a facility.

  4. EPA lacks authority to collect MA or worker exposure information. Most of the industry speakers mentioned this issue. According to John DiFazio of CSMA, "the TRI program is release specific and does not provide for a broad base collection of use information". Wayne Rausch of API stated that EPA needed to specifically state the legal basis for proceeding with the project. He also stated that he thought that OSHA and EPA already collected much of the relevant information, and suggested that a better approach might be to find some way to make the existing data more accessible. The CMA also took issue with the authority for collection of materials accounting and occupational demographics data, and added that they did not believe that such data would prove effective at furthering the aims of EPCRA or the PPA.


EPA is evaluating the issues raised by stakeholders as part of Step 2. In addition, EPA seeks additional comments on a variety of issues, and is interested in evaluations that stakeholders may have performed since the last public meeting. A partial review of some of the important issues and preliminary findings is provided below:

* No EPA data sources were found that fully address the identified TRI gaps.

EPA examined 15 different data systems (See Appendix 1 for listing). None provide a direct match for the TRI-P3 data needs. In a few cases, other data came close, but the scope of chemicals covered, or the public availability of the data did not match. For example, the EPCRA Section 311/312 Tier I and Tier II reports do include range codes for the average daily amount and the number of days on site. Taken together, this could provide a range for the yearly "use" volume of a chemical. Setting aside the issue of range data, the accessibility to this data is limited because it is provided to Local Emergency Planning Committees (LEPCs) for planning purposes, not to the general public. Similarly, Material Safety Data Sheets (MSDSs) for products are submitted under EPCRA Section 311. While MSDSs can provide data on toxics contained in products manufactured by facilities, the available information does not include data on the overall volume of toxics in products coming from a facility. Performing a mass balance is considered a basic planning step in preparing an operating permit under the Clean Air Act, but the information is not included in the minimum information set submitted for review. EPA will consider the potential for modification of existing databases in its Step 3 review of options for pursuing materials accounting information.

One important overlap was found with the Biennial Report System (BRS) used for reporting the generation, management, and disposition of hazardous wastes under the Resource Conservation and Recovery Act (RCRA) . Form GM includes a data element on the quantity of individual waste streams reduced by source reduction, which is relevant for the TRI gap concerning the need for a P2 scorecard. However, the focus of the BRS is on total amounts of materials in waste streams, so that quantities reported include toxics plus other materials such as soil, water, and other substances considered hazardous due to properties other than toxicity (e.g. flammability). A system of waste codes is used that both classifies and describes constituents and characteristics of the wastes. The BRS data does not allow identification of the proportion of the wastes that are toxic chemicals as defined by TRI. This limits the ability to link and compare the two databases. In addition, the BRS data are not accessible via a public database. The BRS data are important, but they address a different set of issues and goals than TRI.

EPA requests that stakeholders share ideas about other data systems that might overlap with TRI-P3. Data integration issues are important, and EPA has begun a long-term agency-wide project to address this issue in a comprehensive fashion. EPA intends to forward findings of potential overlaps to this project, and to continue to address this issue as the TRI-P3 project moves forward.

* EPA agrees that there is a need to be able to quantify pollution prevention (P2) performance.

This gap was the most common one mentioned, and the agency does consider it an important area for consideration. The PPA directs EPA to "establish standard methods of measurement of source reduction" as part of a strategy to promote source reduction. If P2 is to have meaning as a national policy, it is necessary to have goals and measures to gauge national progress. In turn, goals require a means to measure P2. TRI may be the most appropriate tool for creating a national P2 scorecard. While it already reflects levels of P2 activity, it does not allow EPA to understand the amount of any reductions that are due to source reduction.

EPA believes that at a very fundamental level, any method to measure P2 should address definitions and relevant techniques defined in the PPA. Section (5)(A) of the PPA defines source reduction as any practice which:

The Act goes on to say that the term includes the following techniques:

The techniques can be further grouped into three fundamental activities:

  1. changing toxic chemical inputs, (e.g. substitution of raw materials)
  2. modifying production process factors, (e.g. equipment or technology modifications)
  3. changing product outputs (e.g. reformulation/redesign of products)

TRI does not currently include data elements for raw material inputs and product outputs. As for process data elements, pollution prevention occurs "at the source" where incoming raw materials are transformed into products. A generalization can be made that there are only three outcomes for a toxic chemical once it enters a production process, and a case can be made that all need to be tracked if P2 is to be measured:

  1. the toxic can continue unchanged as an ingredient in a product,
  2. the toxic can be transformed into another chemical product, (i.e. consumed).
  3. the toxic can wind up in the waste stream as a release or transfer on or off-site.

TRI does already measure the components of the waste stream, but it does not currently address toxics that are consumed or continued into products. The value of current TRI data elements for measuring source reduction is discussed in greater detail in a following section. In sum, the PPA itself provides an important framework for looking at P2 measurement needs. To quantify source reduction, some form of information is needed on toxic chemical inputs, waste stream data (already in hand from TRI), and output data addressing toxics consumed during product creation along with those that continue into products. It is important to note that these data needs generally coincide with the information provided by materials accounting. EPA is interested in hearing from stakeholders on how they view this series of conclusions, or other possible interpretations on P2 measurement needs derived from the PPA.

* Product issues are important and are very relevant to tracking source reduction.

As described above, product data are necessary for measuring facility-level P2 efforts. In addition, products can have environmental impacts once they leave the facility. In some cases, it may be waste streams at the consumer level that pose the greatest challenges for reducing the entry of a toxic material into the environment. The important role of products is reflected in the increasing use of life cycle assessment/management programs. Several methodologies exist, but all include a broader view of pollution prevention that looks beyond the factory site to analyze all stages of product life, from raw materials acquisition to manufacturing to use and service, to product retirement and disposal. Life-cycle evaluations explicitly recognize that the product must be followed beyond the plant gates in order to have a complete understanding of the potential for pollution. In a similar vein, the goal of product stewardship programs is to continually improve the design and environmental performance of products, and to extend producer responsibility beyond the manufacturing plant to influence product use, recycling, and disposal by customers.

EPA believes that P2 performance should not be limited to the factory site. The PPA defines source reduction as reducing the "amount of any hazardous substance, pollutant, or contaminant entering any waste stream or otherwise released into the environment.." (emphasis added) and includes reformulation or redesign of products among source reduction techniques. P2 improvements in products can reduce releases and affect waste disposal practices at the facility of origin, at downstream commercial customers, all the way to consumers as well. TRI captures the releases and wastes that occur when products are used at downstream reporting industrial facilities. However, it does not address the potential for "deferred" releases and wastes at the consumer level.

One illustration of the importance of products is the case of mercury used in batteries. While batteries do not involve exposure to consumers, the overall contribution made by discarded batteries to the waste stream is significant, far outweighing the contribution of the industrial facilities where the batteries are manufactured. A major P2 effort by battery manufacturers has been to redesign batteries to reduce the need for mercury. According to data from the National Electrical Manufacturers Association, use of mercury in batteries declined 92%, from 778 tons in 1984 to 62 tons in 1990 (EPA, 1991). While the data were provided in this case, they are not typically available, and the genesis of such data are not readily "reviewable" by the public.

Another example involving ozone depleting chemicals can be taken from the INFORM Toxics Watch 95 review of New Jersey MA data. Ozone depleters are being phased out by various deadlines mandated by the Clean Air Act and the Montreal Protocol and the example provides insights into how national level product data could be useful. INFORM found 131 forms reporting on ozone depleters for 1992. Of the 124,174,299 pounds of input statewide, only 3% ended up as waste (3,835,893 lbs). This compares favorably with the state TRI chemical average of 33% of input as waste for all TRI chemicals. However, the destination of 58% (71,571,546 pounds) of the input amount was to products containing the ozone depleters. This amount is much larger than the waste volume, and it helps to show that the product life cycle needs to be followed for a comprehensive picture of overall releases and wastes.

The product stream data alone does not allow one to determine the amount of deferred releases and waste streams or the types, characteristics, and names of the products involved. But together with other information such as the volatility of the substances and the identity of the facilities reporting products, one can pursue the issue to eventually estimate total releases, or to track facility performance with the phase-out requirements. If the data indicate that a particular toxic is not found in the product stream, one can confidently rule out the need to evaluate product issues further. If toxics are reported in the product stream, one has the screening level information and context for prioritizing the role of products, improved ability to focus additional questions, and a roadmap for finding the answers.

Advocates of materials accounting claim that adding this type of product reporting to TRI would promote product benchmarking and similar efforts that would have the overall effect of stimulating the rate of development of products that are environmentally superior. By providing a standardized source of data for use in life cycle and Design for the Environment (DFE) assessments, it would also be likely to spur the development and use of improved analytical tools for looking at products as well.

* The existing TRI cannot reliably quantify P2.

Some industry stakeholders maintain that existing Form R Section 8 data can already be used to track pollution prevention. For example, Jamro (1994) describes a P2 index created by comparing "current year wastes" to "prior year wastes" and then using the TRI production ratio to adjust the quantities for variation in production levels. If the resulting index is positive, then it could be used to demonstrate pollution prevention when combined with the source reduction method data from Section 8.10. This type of method assumes that once you account for variations in production, that any reduction in the waste quantity generated from one year to the next must be due to source reduction. EPA is aware of this type of assessment, and an equivalent method is described and discussed on pages 148-152 of the 1993 EPA TRI Public Data Release Annual Report. EPA believes that these types of methods do yield an "indicator" of P2 performance, but that there are a number of important limitations involved that reduce the value of this type of calculation for actual quantification of source reduction. These limitations are described in brief below:

These methods are silent on some important P2 aspects. In defining source reduction, the PPA includes techniques such as substitution of raw materials, and reformulation or redesign of products. Accordingly, these techniques are currently reflected in the source reduction activity codes for Section 8.10 reporting on types of P2 measures. Similarly, any system to quantify source reduction must also account for use of these techniques. The waste stream focus of the current TRI does not allow substitution of raw materials or reformulation of products to be tracked and measured. It is unclear how they can be reliably quantified without some type of input and output information.

Other variables can impact the waste numbers. Waste numbers may drop from year to year for a variety of reasons. For example, changes in estimation methods or clarifications in reporting instructions can have a significant impact on waste estimates.

The production index may not be accurate in all situations. TRI currently includes a production ratio or activity index for each chemical so that the impact of production changes on TRI data can be seen. The purpose of adjusting, or "normalizing" pollution prevention data is to separate out the effect of year to year production fluctuations from other changes such as source reduction. Calculating an index may be relatively straightforward when the toxic has only one use at a facility. However, when the toxic has multiple uses involving different products, and the resulting production rates vary, the situation is more complicated. Additionally, a source reduction measure may affect one, but not all particular uses. A loss of accuracy can occur when data from the various uses is aggregated to develop a facility-wide number. Depending on how the averaging is done, a loss of precision can distort source reduction findings when the production index is used.

The production ratio may not accurately reflect the relationship between changes in the production level and the generation of wastes. Methods using the current Section 8 TRI data are based on the assumption of a linear relationship between production activity and creation of wastes. While a linear relationship may hold in some cases, it may not be true for others. Variations in the efficiency of the production process may be unrelated to waste generation. Changes in inputs may affect the amounts going to waste streams differently than product streams. Because fugitive releases from storage tanks and pipes are relatively constant, they will affect linear assumptions as well. Even if the process generates less wastes after a source reduction intervention, the relationship may not be linear since storage-related losses will remain independent of the production level.

INFORM (1995) examined the reliability of using existing Section 8 TRI data to measure P2 in their Toxics Watch 1995 report. INFORM took a small sample of 19 chemicals from 8 New Jersey facilities and compared the amount of source reduction reported by the facilities on their state forms with theoretical source reduction amounts derived using TRI Form R waste data and facility production indices. The closest match was a 12% difference (8000 pounds reported to NJ vs 7040 pounds calculated with the TRI indicator method), and the greatest discrepancy was -1639% (13,800 pounds reported to NJ vs 239,979 pounds calculated with the TRI indicator method). In half of the cases, the theoretical amounts derived using TRI data exceeded the amounts reported to the state by the facilities themselves. This example helps to illustrate why EPA does not view existing TRI-based methods as reliable enough for quantifying source reduction. EPA is interested in learning of any other methods or examples that shed more light on both the potential and limitations of P2 measurement methods that use current Section 8 TRI data.

* Lessons from the New Jersey and Massachusetts experience.

Because both states collect materials accounting data, they can be viewed as "pilot" studies on how such reporting can work. Both states combine materials accounting reporting with requirements that facilities use the information for P2 planning. Each state has set statewide goals for reducing wastes, and ask facilities to publicly report their goals and progress on pollution prevention. Each state uses slightly different methods and data elements but both ask facilities to examine whether inputs and outputs balance. The two programs are described briefly below:

Massachusetts - The Massachusetts Toxics Use Reduction Act (TURA) of 1989 requires that large-quantity users of toxic materials: 1) submit an annual Toxics Use Report to the Massachusetts Department of Environmental Protection (MADEP); and 2) engage in a planning process designed to reduce their wastes and usage of toxic chemicals. The state uses the term "byproduct" to describe the total wastes equivalent to the sum of Part II, Section 8.1-8.8 of EPA Form R. TURA includes a state goal that byproduct generation be reduced 50% over ten years, using 1987 as the base year. Initial reporting began in 1991 for 1990 data. Facilities engage in a systematic pollution prevention planning process to review source reduction opportunities, to evaluate feasibility and cost issues, and to set goals for future reductions. The plans themselves are kept at the facility.

(NOTE: Location of Word Perfect Table on Massachusetts Materials Accounting Details - See Appendix 2 for non-table equivalent)

A key feature of the Massachusetts law is the focus on process level planning and reporting. Facilities identify these production units and the unit of product to be used for normalization purposes. Facility-level materials accounting data is reported yearly, along with production unit information. Range codes are used to afford CBI protection for the amounts of toxic chemicals used in the production units. Two additional performance indicators, the Byproduct Reduction Index (BRI) and Emissions Reduction Index (ERI), are also reported each year for the production units. Every other year, facilities report on facility level changes compared to planning estimates, and project future estimates of changes as well. The TURA law established the Office of Technical Assistance (OTA) and the Toxics Use Reduction Institute (TURI) to provide technical assistance to industries in the state. These programs have used the data to set priorities and to target assistance to facilities and industry sectors. A variety of target measures are used, from input and output volumes, to derived "efficiency" ratios of byproduct to use, to more general findings that focus in on common chemical uses among facilities in the state. State officials believe that toxic chemical use data can be effectively collected, and that confidential business information can be protected by combining conventional trade secret protections with the use of data collection and reporting strategies designed to allow tracking without the need for public release of sensitive business data. In addition, use data is seen as providing significant benefits to state agencies, industrial firms, and the general public. These benefits include improved public and governmental understanding of the volumes and patterns of toxic chemical use, improved capacity to target state resources, increased ability to measure progress in pollution prevention, improved toxic chemical management in firms, and a better basis for promoting technology transfer among firms (McCarthy, 1995).

New Jersey The New Jersey materials accounting program is founded in the Worker and Community Right to Know Act of 1984. Since 1987, the state Department of Environmental Protection (NJDEP) has collected throughput data from industrial facilities. The state uses the term "nonproduct output" to describe the quantity of a toxic chemical which is generated as a result of a production process prior to storage, handling, treatment or disposal that is not intended for use as a product. In addition to materials accounting, the original reporting requirements also included data elements on the amounts reduced by source reduction.

(NOTE: Location of Word Perfect Table on New Jersey Materials Accounting Details - See Appendix 3 for non-table equivalent)

In 1991, the New Jersey Pollution Prevention Act (NJPPA) was passed, requiring facilities to perform P2 planning. The law created a statewide public policy goal of 50% reduction in nonproduct output by 1997, using 1987 as a baseline. The planning process is designed to help facilities to identify source reduction opportunities at the production process level. Detailed plans are kept at the facility. Reporting for facilities in SIC codes 26, 28, 30, 33, and 34 began with 1993 data reported in 1994. The other SIC codes will begin reporting on 1995 data in 1996. The new reporting also includes process level data collection. Facilities can select any combination of "targeted" processes, as long as they represent at least 90% of use, generation of nonproduct output, or releases at the facility. The NJPPA also addressed performance measures, and changed from a reporting of the amounts reduced by source reduction in pounds to indicators based on annual changes (in percent) compared to the previous year. One performance index is based on usage, and the other is based on changes in nonproduct output. State officials maintain that the program experience has been positive, and that collection of materials accounting data is essential to the success of P2 planning, equipping facilities to identify areas of cost savings and environmental improvements. A 1995 pilot survey performed by the NJDEP (described in the following section) has found that the overall process can be considered beneficial to facilities. The state has also found the information to be useful, especially for regulatory program integration. For example, the state is working with a number of facilities in a pilot project to replace single media permitting with a streamlined multi-media system that includes materials accounting data. In one case, 60 separate permits were replaced with a single permit. The data is also seen as useful for identifying errors in Form R reports, allowing citizens access to information about the efficiency of facilities, and for looking at emergency planning and response issues.

* Materials accounting provides essential information.

EPA finds that materials accounting data are valuable for facilities to collect, whether required or not. Materials accounting provides a system for tracking the transformation of toxic materials into products and wastes, which is at the core of the relationship between an industrial facility and the environment. The collection of this information provides a common sense foundation for all other environmental data. When inputs equal outputs in a materials accounting, it is likely that important flows and destinations have been identified and that no major mistakes have been made in estimating their magnitude (INFORM, 1995). Because MA data are rooted in production, they insure that environmental and managerial data systems are linked. The picture provided by MA data has been found to help facilities and agencies to identify source reduction opportunities and other innovative ways of meeting environmental goals. End-of-pipe and monitoring data are valuable, but they are of limited utility for providing insights into avoiding the creation of wastes in the first place. Also, not all toxics are monitored on a regular basis, and when they are found in the environment it can be difficult to trace them back to their sources. In contrast, facility data about inputs, wastes, and product streams point a business and its community neighbors to sources of pollution in choices of materials, processes, or practices (Erwin, 1995). High performing companies are already collecting MA data and using it as an asset for implementing cost-effective environmental improvements. Materials accounting data is good business, and is beneficial for looking at a variety of environmental issues, including those described by the TRI gaps. EPA has not uncovered any reports describing true alternative systems to materials accounting information. In contrast, a variety of reports reflect the utility of MA data. Examples of such reports are described below:

* Materials Accounting should be viewed as within the Right-to-Know.

Should materials accounting data be considered "Right-to-Know" information to be made available to the public? EPA supports the general concept of providing the public better information on the use of chemicals, and believes that materials accounting data is certainly within the bounds and the spirit of "Community Right to Know". Chemical use information is integral to understanding the presence and movement of toxics through the community. It provides an important foundation for comprehending the magnitude of releases, managed wastes, and emergency planning needs.

For example, toxics input data can provide community members with insights about the flow of toxics through their community, so that the possibility of transportation-related mishaps can be identified, prioritized, and addressed. This information directly supports the goal of emergency planning and prevention of spills and accidents. Knowledge of facility throughput volumes helps to provide initial insights into the need for precautions and the level of program sophistication required to safely manage chemical operations. Throughput data can provide similar insights into the need for safeguards to protect those community members who work at the facility. With this information, communities can then follow-up with facilities to obtain additional detail. Product stream volumes leaving the plant can also point to transportation-related emergency planning needs.

In addition to emergency planning uses, MA data is also key to understanding the facility context for pollution prevention approaches. By describing the dimensions of chemical use that give rise to pollution, they allow the community a better understanding of the setting for releases and wastes, and the opportunities for avoiding them in the first place. Providing a basic understanding of this relationship is a prerequisite for bringing pollution prevention into the mainstream of environmental dialog at the community level. EPA believes that community-based approaches to environmental protection hold great promise, but that this potential is greatly contstricted if limited to end-of-pipe information and strategies. Information directly affects the way that problems are identified, options are developed, resources are allocated, and accountability is created. The lack of chemical use information amounts to an institutional barrier blocking the further adoption of source reduction practices. By law, source reduction is identified as the preferred national approach for improving environmental performance. Given this national policy and the importance it confers on source reduction, it defies common sense for the agency to view the information needed to understand and track source reduction progress as falling outside the public's Right-to-Know. Adding to this potential contradiction is the fact that TRI already provides the public with data on the less preferred paths for managing pollutants - recyling, treatment, and disposal or release.

In sum, EPA believes that a strong Right-to-Know case can be made for MA data. The agency is interested in stakeholder views on this subject, and in finding a way to design a system that puts chemical use data into the public domain, while meeting industry concerns about the loss of intellectual property.

* Performance measures derived from materials accounting.

MA data directly address a variety of the listed TRI gaps, such as providing a ledger check, or providing basic data on product streams. However, MA data also serve as the starting point for calculation of additional performance measures, especially for tracking source reduction. As described in a previous section, New Jersey and Massachusetts each collect MA data, but both also require facilities to calculate and report P2 performance measures. This leads to several questions:

EPA would like stakeholders to articulate which gaps require MA data, which are best addressed by measures derived from use data, and how these measures fit within the overall TRI-Phase 3 discussion.

* Facility vs Process reporting and the role of normalization

Several stakeholders recommended that EPA consider process level reporting for TRI-Phase 3, because it enables a more accurate measurement of P2 progress. In addition, both New Jersey and Massachusetts collect process level data. While process reporting should not be ruled out, EPA would prefer to first explore those options that are consistent with TRI's current facility-wide format. The additional burden potentially associated with process level reporting is an important consideration for this preference.

However, EPA does recognize that a number of issues related to facility-level reporting need additional discussion. One important issue involves the lack of precision provided by facility-level production indexing. As described earlier, TRI currently includes a production ratio or activity index so that the impact of production changes on TRI data can be seen. EPA provides TRI guidance on this topic along with several examples and does state that the ratio or index must be based on the variable that most directly affects the quantities of the toxic chemicals generated in waste. The resulting number provides important context for the other TRI data, but it may not be accurate enough for use in calculating source reduction. The selection of a production measure, and the manner in which multiple uses of a toxic across process and product lines are taken into account is important to minimizing error in the resulting index. Some methods can introduce other variables that might then distort resulting P2 performance calculations. As discussed earlier, EPA believes that the current index has several drawbacks.

Materials accounting information can be used to improve normalization, and there are methods that compare throughputs as a way to normalize production rate variations. In addition, both New Jersey and Massachusetts have developed specific guidance in order to minimize normalization problems. In Massachusetts, materials accounting reporting is facility-wide, but P2 progress reporting is done only at the production unit level. New Jersey does require that P2 progress be reported for both the facility and targeted processes. While EPA's TRI guidance currently includes an example how production ratios can be determined based on a weighted average of the contribution to the waste stream, the New Jersey instructions go into additional detail. Specific instructions are provided on selecting an appropriate unit of product, and both the units of production and the quantity of products are reported on the reporting form. In addition, when a toxic chemical is used for producing more than one product, the instructions call for weighting the involvement of each product based on the use of the toxic chemical per unit of product. Taken together, these measures serve to reduce the error factor involved in creating a facility level index. Both states have moved towards measuring P2 progress "per unit of product" as the best way to insure accuracy in normalization. The New Jersey program has found that the process-level focus of P2 planning has led to refinements in the reporting of facility-level production indices compared with those historically reported.

Normalization is an important issue that has ramifications for facility reporting and calculation of P2 performance measures. EPA seeks comments from stakeholders on how they view this issue, and whether it should be addressed as part of TRI-Phase 3.

* Occupational issues

EPA agrees with environmental and labor stakeholders that TRI does not directly address occupational issues, and that this can be viewed as an important information gap. EPA coordinates closely on occupational issues with OSHA, NIOSH, and MSHA, and generally defers to these agencies on occupational research, risk reduction, and regulatory matters. EPA does have certain unique information collection and dissemination authorities that cut across toxics issues, and the agency does take the lead on interagency efforts to fill acknowledged data gaps where appropriate. EPA considers the collection of an occupational exposure indicator as among the core group of data elements that should be included when full materials accounting data is collected.

The issue of worker exposure to toxics is important, and source reduction shows promise as a framework for correcting both occupational and environmental problems at the same time. One of the lessons learned from the TRI experience is that toxics cut across all media and there are gains to be made in taking a multi-media perspective on environmental issues. In a very simple way, worker (and consumer) exposures can be viewed as if they were additional media that need to be accounted for to insure against the possibility of risk-shifting. EPA supports the need to better understand the linkage and fit between occupational and environmental performance, and views the TRI-Phase 3 discussion of this issue as one that will inform all stakeholders as well. There are at least two issues for consideration:

The utility of chemical use information for looking at occupational issues Because there are no other publicly accessible national databases for facility reporting of occupational exposure or chemical use, there is increasing interest in the using TRI as a basic tool to identify a list of candidate facilities for follow-up on occupational issues. As stated by Geiser (1994), use data will never be viewed as an accurate record of exposure (because it does not account for exposure control features in place at the facility), but it does provide a better indicator than release data because it reflects the chemicals actually employed on the shop floor. EPA believes that chemical use data will improve the usefullness of TRI as a tool for occupational purposes.

The case for occupational exposure indicators EPA described several occupational indicator data elements in the first issues paper. These types of data elements were intended to provide basic but up-to-date estimates on the number of potentially-exposed workers, and the extent to which worker exposures have been assessed for a given toxic. Such information is valuable for priority-setting on many levels, and for tracking trends and performance over time. For example, it would supplement the most current available estimates on the number of potentially exposed workers, which date from 1983. Environmental and labor stakeholders submitted additional data elements, including some that directly address exposure levels. Industry stakeholders pointed out that where employee exposure monitoring has been performed, the data is already available to affected workers upon request. From this perspective, there is no gap to fill. However, there is no mechanism for employees or others, including state and federal government agencies, to have access to such information via a public database. From this vantage there is a gap, since dispersed data sources can be considered as functionally inaccessible.

The TRI experience shows that public access to relevant data creates a wave of benchmarking activity, and helps individuals and groups at many levels to understand, prioritize, and follow-up on toxic chemical issues. If EPA's experience to date is any indication, it is clear that including additional occupational indicators in a publicly accessible database would focus additional attention and activity on occupational safety and health performance, and how it overlaps with environmental performance. EPA is interested in what community, labor, and industry groups have learned about integrating the two areas, and on views about occupational exposure indicator data elements.

Step 3 of the TRI-Phase 3 project will involve looking at the appropriate role of the federal government in collecting TRI-P3 information, and the options and relevant statutes for going about collecting it. This will be a critical step for supplemental occupational data elements. EPA is also considering occupational data needs as part of the Chemical Use Inventory/Inventory Update Rule (CUI/IUR), and will explore this vehicle in more detail as part of Step 3. In addition, EPA will consult further with its occupational agency partners in order to explore other possible options for collecting this information.

* Confidential Business Information issues.

EPA understands the basis for the concern that reporting of materials accounting data could lead to loss of CBI, with resulting economic losses for the company, state, or the U.S. economy in general. EPA considers CBI a critical issue, one that must be addressed and resolved in any system for reporting on chemical use. In addition, EPA considers the current low CBI claims rate for TRI to be an essential feature contributing to its overall value as a public database.

Given this importance, EPA seeks to fully understand the practical experience and lessons learned by stakeholders who have had first-hand experience with the state reporting laws in Massachusetts and New Jersey. This experience is likely to provide the most relevant insights on how choices of data elements, reporting measures, and CBI provisions can address CBI concerns. EPA has heard conflicting views in this area and seeks clarification on the CBI experience in these states. For example, industry commenters have expressed concern that materials accounting data can reveal proprietary information to competitors. However, the experience to date in New Jersey indicates that few businesses have requested materials accounting data about other facilities. There have not been any requests from foreign business firms. Another issue relates to the meaning of CBI claim rates. CBI claim rates in both New Jersey and Massachusetts are low - between 1 and 2%, even with some reporting of process-level information. It is true that the largest chemical user in Massachusetts - a Monsanto facility, has asserted CBI claims every year since reporting began (Jamro, 1994). The low rates seems to suggest that state reporting and CBI provisions are successfully addressing CBI concerns. However, some industry commenters have stated that the low rates are deceiving, because the burden of making these claims is great under state procedures, thus discouraging the filing of claims in the first place. Available data for New Jersey indicates that the state has granted all CBI claims submitted to date, which does not seem to square with concerns that the state procedures are overly burdensome. Here again, EPA would like to receive additional information and analysis.

There is some evidence that the state laws do include provisions that adequately protect CBI. According to Glenn Roberts, Government Relations Director for both FEMA (Flavor & Extract Manufacturers' Association of the United States) and FMA (Fragrance Materials Association of the United States) the low rate of CBI claims in New Jersey is not surprising. Industry efforts during the drafting of the state Pollution Prevention law were intended to protect trade secrets from needing to be reported to the state. According to Roberts, the fragrance and flavor industry holds the bulk of trade secrets related to mixtures and chemical products in New Jersey, and their experience with previous state Right-to-Know reporting provisions was that pre-filing of trade secret claims was problematic. Thus industry input into the more recent state laws was intended to avoid the need to make such claims in the first place. For example, the law includes a provision allowing facilities to group similar products made with similar ingredients for reporting as one process, so as to shield trade secret information from disclosure. As stated by Roberts:

"The New Jersey experience shows that trade secrets must be protected, and also that they can be protected without impairing environmental controls or a chemical use reporting program. The New Jersey experience is based on the assumption by industry that confidential business information claims are not a workable device (as we have seen with the prefiling of trade secrets under the [state] Right to Know Program) and that the best way to truly protect trade secrets is to exclude them from the need to be reported in the first place. Any new federal program should adopt the same approach." (Roberts, 1994, emphasis added)

EPA seeks additional comments from stakeholders on experiences with the existing state laws, examples of cases where CBI was lost, or suggestions for provisions to protect CBI disclosure. It would be very useful to obtain actual case studies from industry stakeholders in order to better understand the CBI issues. From environmental stakeholders, it would be valuable to hear their views on how features to protect CBI, such as the use of ranges or reporting measures derived from MA data, impact the Right-to-Know value of the resulting information. EPA believes that on the federal level, as in New Jersey, that it should be possible to develop a common sense approach to MA reporting at a level of detail that expands the Right-to-Know, while respecting legitimate confidentiality concerns.

* TRI-Phase 3 Cost issues

While a full discussion of cost issues is beyond the scope of the paper, they are important and a preliminary review is provided here. EPA seeks additional cost estimates and comments 1from stakeholders on these and other cost-related issues that EPA should consider for TRI-P3.

Existing cost estimates. CMA comments included a cost estimate for a large chemical manufacturer that indicated costs approaching $1.5 million for first year reporting, and $800,000 for following years. The figure included breakdowns for each activity for each "unit" - for example it was estimated that 34 hours would be needed to calculate beginning inventory, and 24 hours to calculate quantities consumed on-site, manufactured on-site, and shipped off-site. The total time estimate was 120 hours per unit. The number of TRI chemical records was not indicated, but the facility size is reflected reflected by the fact that it involved over a hundred units (CMA, 1994).

A study performed by the NJDEPE of 14 facilities (4 small, 6 medium, and 4 large) from the specialty inorganic (SIC code 2819), pharmaceutical (SIC 2842) and cleaning products industries (SIC 2841-43) also looked at costs. The study reported that 7.5 hours on average was needed per chemical form (NJDEQ-114) to complete the materials accounting information. This figure reflected the time needed above and beyond the amount needed for filling out the existing TRI forms, and responses ranged from 2 hours to 2 to 3 days (NJDEPE, 1993).

What is the extent to which MA data is already being collected? Data collection costs are likely to be greater than data reporting costs, and there are indications that many facilities already have this data on hand at sufficient level of detail, whether for completing existing TRI forms, or as part of good operating practices. Ray Jones, a technical advisor for Citizens for a Better Environment and a former corporate manager, stated at the 9/28 public meeting that based on his experience, all companies collect this type of data in one form or another. John Defazio of the CSMA, in response to a question at the public meeting, also stated that to a large extent, member facilities would already know input and output amounts. Dorfman et al (1992) looked in depth at source reduction practices at 26 chemical manufacturing and processing facilities in a 1992 study, and found that 22 of the 26 were collecting materials accounting information in some form.

Once collected, does MA data have other benefits for a facility? If so, how should they be evaluated? Environmental benefits and cost savings have been found when MA data is used to guide P2 interventions. Examples of economic benefits were previously described in the section on the utility of materials accounting information. The emerging area of environmental accounting is another example of how such information might provide the basis for overall savings on environmental expenditures. In addition, the data may be collected for other beneficial but non-environmental uses as well. Dorfman et al. (1992) found that two of the 26 facilities collected MA data as part of their "Statistical Process Control" (SPC) program. SPC data is collected to track and support improvements in product quality at facilities. It is considered a basic component of Total Quality Management (TQM) programs.

How might costs vary among manufacturers, processors, and users? It is likely that use reporting costs will vary depending on how the chemical is used. For materials accounting reporting, companies that "otherwise use" a given toxic would need to report input quantities, but would generally not need to report on quantities in products. This is because by definition, no toxic chemical is intended to go into the product for this type of use. For 1993, 38% (30,344) of the total number of 79,976 Form R's for TRI fell into the category of "otherwise used". On the other hand, reporting would be more complicated for the 2% (1457) of forms where use included manufacture, processing, and otherwise used.

Significant digit issues TRI currently requires that data be collected to 2 significant digits. In essence, this means that a figure of 21,940 pounds can be reported as 22,000 pounds. The current level of precision may not support the use of the data for balancing inputs and outputs. How important is such precision to stakeholders? What are the cost implications for collecting more precise data? The EPA currently believes that a greater level of precision may not be justifiable, given that a materials accounting approach cannot supply a precise "ledger check" for releases and outputs.


EPA believes that a "full" materials accounting approach should be the appropriate goal, where full refers to the ability to provide a balance of inputs and outputs. However, EPA is still reviewing data element options. There are a variety of options available for creating data elements (especially for toxic chemical inputs), and EPA is interested in selecting elements that balance goals of providing the public with meaningful data, protecting legitimate trade secrets, and minimizing resource requirements. EPA is listing a variety of options below in order to stimulate discussion. All options are based on facility-wide reporting. EPA encourages stakeholders to comment on the pros and cons of these options, to test them against their own priorities, and to report back to EPA with additional data element suggestions or refinements.

Input options

Set A

Set B

Set C

Output options

Set A

Occupational exposure indicator options

Set A

Set B

Set C

Other supplemental options beyond full materials accounting

Waste-related source reduction performance measures

Normalization refinements


Legislative Authority issues - A broad array of statutory authorities are available to EPA to accomplish some or all of the goals described in this report. Some of the materials accounting data that would be valuable input data into a total materials accounting system is already being collected on TRI. These data allow EPA, facilities and the public to track releases and transfers of toxic materials and wastes. These particular data elements however fall short of allowing EPA or the public to assess quantitative progress in source reduction.

TRI itself could be expanded to collect additional information. For example, section 313(g)(1)(i) and (ii) could be used to collect information about a facility's manufacturing, processing, or other use of a chemical by requiring more detailed reporting on categories of use, and the amounts (in ranges) of the chemical present at the facility for each of those categories. EPA is also in the process of evaluating whether to refine the scope of TRI reporting on source reduction and waste management activities under the Pollution Prevention Act (PPA).

Environmental statutes other than EPCRA and the PPA may also be used to collect relevant materials accounting data. For example, section 8(a) of the Toxics Substances Control Act (TSCA) authorizes EPA to collect data on the amounts of a chemical substance manufactured or processed by category of use, byproducts, worker exposure, and disposal methods. There are significant differences, however, between the scope of coverage for EPCRA and TSCA, as well as information access issues and facilities coverage issues. EPA is also exploring whether information currently being reported under TSCA or any other environmental statute could help meet the materials accounting goal.

In pursuing the issue of statutory authority further, EPA will aim to assure that the most appropriate statute or combination of statutes is selected, that duplication of information requirements are avoided to the maximum extent possible, and that in considering the Agency's authorities, EPA will identify areas where legislation may need to be enhanced or expanded.

Development of an open and transparent issue resolution process - The EPA will proceed with evaluation of this issue in an expedited manner, while maintaining a commitment to full and open participation. The next scheduled step in this process is the public meeting on October 18 and 19, 1995. The purpose of this meeting will be to lay out EPA's position on materials accounting and to engage the public and private sector in a continuing dialog to move forward. A transcript of the meeting will be prepared and will be available to the public in disc form in the EPA Non-confidential Information Center, along with copies of all written comments submitted at the meeting. EPA welcomes follow-up comments after the public meeting, but requests that such comments be submitted as soon as possible. Following the meeting, EPA plans to consider a variety of actions:

EPA will also be exploring other regulatory and voluntary approaches to collection of materials accounting data, as well as any efforts needed to create principles for new legislation. EPA will also consider the option of independent or third party validation of submitted materials accounting data.


In conclusion, EPA welcomes public dialog on this important issue. Since its inception in 1986, TRI and the Community Right-to-Know Program has had a significant influence on improving the environmental quality of the nation. EPA believes that assuring a full and comprehensive Right-to-Know program is one of its primary missions. EPA believes that there are common sense approaches that can enhance Right-to-Know while preventing the disclosure of trade secrets. EPA is committed to assuring that this Program continues to move forward, ensuring that the public has a compete picture of toxics in the environment, and that they have the information available to them to understand and influence decisions that affect their lives, their families, and their communities.


Dorfman, M.H., Muir, W.R., and C.G. Miller. 1992. Environmental Dividends - Cutting More Chemical Wastes. INFORM.

DuPont, 1993. DuPont Chambers Works Waste Minimization Project. EPA/600/R-93/203.

Environmental Protection Agency. 1991. Pollution Prevention 1991: Progress on Reducing Industrial Pollutants. Washington, D.C. USEPA Office of Pollution Prevention.

Erwin, Frances, et. al. 1995. A Benchmark for Reporting on Chemicals at Industrial Facilities. World Wildlife Fund, Baltimore, Maryland.

Geiser, K. 1994. The Benefits of Chemical Use Data: Suggestion from Early Massachusetts Experience. Presentation at 1994 TRI Data Use Conference . Proceedings: Building TRI and Pollution Prevention Partnerships. Boston, Massachusetts, December 5-8, 1994. page 56.

INFORM, 1995. Toxics Watch 1995. Edited by Carroll Bastian. INFORM, NY, NY

Jamro, E. S. 1994. Chemical Use Data is Not Needed to Track Pollution Prevention. Presentation at 1994 TRI Data Use Conference . Proceedings: Building TRI and Pollution Prevention Partnerships. Boston, Massachusetts, December 5-8, 1994. page 157.

McCarthy, G. 1995. Letter to M.McDonnell, EPA from Gina McCarthy, Executive Director, Administrative Council on Toxics Use Reduction, Executive Office of Environmental Affairs, Commonwealth of Massachusetts. March 10, 1995.

New Jersey DEPE. 1993. Materials accounting as a potential supplement to the Toxics Release Inventory for pollution prevention measurement purposes. Office of Pollution Prevention, State of New Jersey Department of Environmental Protection and Energy. December, 1993.

New Jersey DEP. 1995. Early Findings of the Pollution Prevention Program. Part 1: On-site Reviews of Pollution Prevention Plans. February 1995.

Pojasek, R.B. and L.J. Cali. 1991. Measuring Pollution Prevention Progress. Pollution Prevention Review. Spring 1991.

Porter, M.E. and C. van der Linde. 1995. Green and Competitive: Ending the Stalemate. Harvard Business Review, September-October 1995.

Roberts, G. 1994. The New Jersey Experience with Confidential Business Information - An Industry Perspective. Attachment with cover letter to M. Gillen, November, 16, 1994.

White, A.L., Dierks, A., and D.E. Savage. 1995. Environmental Accounting: Principles for the Sustainable Enterprise. Presented at the 1995 TAPPI International Environmental Conference, May 7-10, Atlanta, Georgia.




Massachusetts Materials Accounting Details

Applicable Law
Massachusetts Toxics Use Reduction Act (TURA, M.G.L. c.211), 1989

Reporting Forms

Materials Accounting Data Elements

Inputs / Outputs

Special performance measures

Byproduct Reduction Index (BRI) = (A - B) x 100

Emissions Reduction Index (ERI) = (C - D) x 100

A= Byproduct quantity in base year divided by the number of units of product produced in base year.
B= Byproduct quantity in reporting year divided by the number of units of product produced in reporting year.
C= Emissions quantity in base year divided by the number of units of product produced in base year.
D= Emissions quantity in reporting year divided by number of units of product in reporting year.

Facility level reporting

Annual - Materials accounting data elements

Optional question on reasons for lack of materials balance

Biennial - Changes (in pounds) in use and byproduct from planning year amounts

Changes (in percentage) in use and byproduct from base year

Process level reporting (For each Production Unit)

Annual - Range code for Quantity of chemical Toxics Use Reduction Techniques Codes Byproduct Reduction Index Emissions Reduction Index

Biennial - Projected BRI change from base year
* 2 year and 5 year estimates


New Jersey Materials Accounting Details

Applicable Laws

Reporting Forms

Materials Accounting Data Elements

Starting Inventory of the substance Quantity consumed on site
Quantity produced on site Quantity shipped off site as (or in) product
Quantity brought on site Ending inventory
Quantity of brought on site Quantity of ending inventory (item above) as recycled substance that is nonproduct output
Quantity recycled out of process Total nonproduct output on site and used on site

Special performance measures

Annual percentage reduction for use = (A x E) - B x 100
                                    A x E

Annual percentage reduction for nonproduct output = (C x E) - D x 100
                                    C x E

A= starting inventory + quantity produced on site + quantity brought on site + quantity recycled out of process and used on site - ending inventory from the previous year
B= starting inventory + quantity produced on site + quantity brought on site + quantity recycled out of process and used on site + waste stored on site - ending inventory from the reporting year
C= total nonproduct output generated from the previous year
D= total nonproduct output generated from the reporting year
E= production ratio (weighted for multiple uses per unit of product)

Facility level reporting

Materials accounting data elements Annual percentage reduction for use Annual percentage reduction for nonproduct output

Process level reporting (for targeted processes)

Annual percentage reduction for use Annual percentage reduction for nonproduct output Codes for pollution prevention techniques used and planned

Source: USEPA

Contact MapCruzin

Back to the Top

Website maintained by

Michael R. Meuser

Copyright  1996-2009
All Rights Reserved

Visit us for Free GIS Maps and Resources