Risk-Based Waste Classification in California (1999)

In the State of California, the management of wastes considered potentially hazardous is first regulated by the state in compliance with the federal Resource Conservation and Recovery Act (RCRA). For wastes that are not regulated under RCRA as hazardous, the state uses an additional classification system to determine whether such wastes pose a threat to human health and the environment. This report focuses on a new approach proposed for use in California to classify wastes that are not considered to be hazardous wastes under RCRA.

As with RCRA, California classifies waste as hazardous or nonhazardous based on four characteristics: reactivity, ignitability, corrosivity, or toxicity. As part of its regulatory structure update process, the California Environmental Protection Agency's Department of Toxic Substances Control (DTSC) has reviewed those characteristics and has undertaken to propose a new waste-classification system for hazardous wastes.¹ The classification of a waste determines how the waste will be managed (e.g., storage, transport, disposal, reporting requirements).

DTSC has spent more than 2 years developing a new approach for the toxicity characteristic. The proposed system has two basic goals: (1)

to move from a classification system that is toxicity-based to a risk-based system that considers both toxicity and potential exposure to the waste² and (2) to replace the one-threshold system that classifies waste as either hazardous or nonhazardous with a two-threshold system that classifies waste incrementally as "hazardous" (high toxicity), "special" (moderate toxicity), or "nonhazardous." The committee notes that the terms hazardous waste and nonhazardous waste, while appropriate for a classification system based on toxicity, might not be technically appropriate for one based on risk, when the definition of the term "hazardous'' refers to the toxicity of an agent and is independent of exposure.

Before the proposed classification system and subsequent regulations can go into effect, DTSC is required by California statute to "conduct an external scientific peer review of the scientific basis of any new rule" (California Health and Safety Code § 57004). Consequently, DTSC requested that the National Research Council (NRC) of the National Academies of Sciences and Engineering conduct such a review of its proposed risk-based approach for the toxicity characteristic used to classify waste and suggest areas where the approach might be improved. This report is the independent review of DTSC's proposed waste-classification system by the National Research Council's Committee on Risk-based Criteria for Non-RCRA Hazardous Waste.

The technical basis for the proposed waste-classification system is documented in DTSC's "Risk-Based Criteria for Non-RCRA Hazardous Waste: A Report to the National Research Council Introducing Proposed Changes to the Definition of Hazardous Waste in the California Code of Regulations." An overview of this report is presented in Appendix D.

This approximately 1,500-page document, which was submitted to the NRC before the first committee meeting, includes a synopsis of the proposed classification system and many of the background materials used to develop the classification criteria.

The committee also examined supplementary DTSC documents, including those found on DTSC's home page on the Internet. The committee requested clarification of numerous aspects of the DTSC documentation both through written responses from DTSC and through verbal responses to questions posed by the NRC committee members at two public meetings. At the public meetings, the committee also gathered information from waste managers, waste generators, environmental groups, and other interested individuals and organizations. Comments on DTSC's approach were given in oral presentations as well as in written materials. Thus, the scientific and technical bases of the proposed classification system that were reviewed by the committee included information received before and during the committee's deliberations. The proposed classification system reviewed in this report is unchanged from the descriptions in the original documentation received by the NRC, albeit with some clarification on specific issues from DTSC and other individuals and organizations.

As presented in its report, DTSC currently uses a one-threshold system to classify wastes either as hazardous or nonhazardous based on the toxicity of the whole waste or the waste components.³ Under the current system, DTSC has established eight toxicity criteria by which a waste might be considered hazardous:

• Federal definition of a hazardous waste according to its toxicity characteristic.
• Exceedence of total threshold limit concentrations or soluble threshold limit concentrations for specific chemicals.
• Exceedence of acute oral toxicity threshold.

• Exceedence of acute dermal toxicity threshold.
• Exceedence of acute inhalation toxicity threshold.
• Exceedence of acute aquatic toxicity threshold (fish).
• The presence of one or more of 16 named carcinogens above a threshold.
• Otherwise shown by experience or testing to be hazardous to public health or the environment (new threats).

For wastes that are not considered hazardous under RCRA, DTSC currently determines whether the waste contains any specific chemicals that exceed thresholds established for soluble threshold limit concentrations, or whether the total amounts of any specific chemicals exceed established total threshold limit concentrations. If the concentration of any of those chemicals exceeds the relevant threshold, the waste is classified as hazardous; otherwise, the waste is classified as nonhazardous. Soluble and total threshold limit concentrations are established to protect humans and other organisms from adverse effects following exposure to those chemicals through contact with groundwater or by any other route (e.g., soil, air, surface water, and food).

If a waste does not contain chemicals that exceed any soluble threshold limit concentrations or total threshold limit concentrations, it can still be classified as hazardous based on its acute toxicity (oral, dermal, inhalation, or aquatic) or the presence of one of 16 specific carcinogens. There are thresholds for both acute toxicity and the 16 specific carcinogens. The acute toxicity of a waste is determined either by testing the whole waste or by summing the acute toxicity values of the individual waste components. A waste also might be considered hazardous if data are available to suggest that a new chemical or route of exposure might pose a threat to humans or other organisms.

The proposed system will continue to classify wastes as hazardous or nonhazardous; however, it will also include a category of special wastes that are considered to be less than hazardous but still might cause adverse effects if disposed of as nonhazardous waste. Special wastes might be subject to less restrictive reporting and disposal requirements than hazardous wastes.

In the proposed system, DTSC will modify five of the eight toxicity criteria listed in this summary by establishing two thresholds for each of the following: total threshold limit concentrations or soluble threshold limit concentrations; acute oral toxicity; acute dermal toxicity; acute inhalation toxicity; and acute aquatic toxicity. One threshold will distinguish hazardous wastes from special wastes (upper threshold). The other will distinguish special wastes from nonhazardous wastes (lower threshold), as shown in Figure ES-1.

DTSC also proposes to replace the soluble threshold limit concentrations with a new category referred to as soluble or extractable regulatory thresholds, and to develop upper and lower thresholds for that category as well. The classification criteria for wastes containing carcinogens and for wastes posing new threats would remain the same as in the current system.

Unlike the present system, which classifies wastes based on their toxicity, the proposed system classifies wastes based on their potential risks to the health of humans and other organisms. The risk posed by the waste is estimated from (1) its long-term risk (chronic toxicity and long-term exposure potential), (2) its short-term risk (acute toxicity and short-term exposure potential), or (3) both its long-term and short-term risks. A variety of models are used to determine the exposure potentials for the chemicals in the waste.

To calculate the two new total threshold limit concentrations, DTSC used four scenarios by which humans or other organisms might be exposed to chemicals in the wastes: (1) residents living near a hazardous waste landfill; (2) waste workers at a landfill; (3) residents living on land to which waste had been previously applied; and (4) ecological (nonhuman) receptors. Those scenarios were designed to protect workers, residents, and ecological receptors from chronic adverse effects resulting from long-term exposures. Multipathway exposure models were used to estimate the risks associated with each of the human health scenarios. Risks to ecological receptors were based primarily on risk assessments developed in the technical support document for the U.S. Environmental Protection Agency's proposed Hazardous Waste Identification Rule.

The proposed upper and lower soluble or extractable regulatory thresholds were derived by a different method from that used to calculate the total threshold limit concentrations. For soluble or extractable regulatory thresholds, the only exposure scenario for humans was as-

* Under the current system, a very small number of hazardous wastes receive variances for disposal as special wastes.

Figure ES-1

DTSC's waste-classification systems.

sumed to be ingestion of contaminated drinking water. Ecological receptors were assumed to be in direct contact with the groundwater.

For acute toxicity thresholds, acute exposure scenarios were used. The upper acute oral and dermal toxicity thresholds were based on exposure of adults, and the lower toxicity thresholds on exposure of children. For acute inhalation toxicity, upper and lower thresholds were developed separately for volatile chemicals and for particles. Acute toxicity values are applicable to whole wastes, either by direct testing of the whole waste or by appropriate summation of acute toxicity values for all the waste constituents.

In its report, DTSC also proposes to use the federal RCRA extraction test, the toxicity characteristic leaching procedure, rather than the waste extraction test, which it currently uses for extracting wastes to determine if the wastes have the potential for contaminating groundwater.

Improved flexibility in the management and disposal of wastes while protecting human health and the environment is a primary goal of the regulatory structure update process under which DTSC's proposed approach was developed. Flexibility is to be improved through the use of a two-threshold rather than a one-threshold classification system, a new risk-based procedure for developing soluble or extractable regulatory thresholds and total threshold limit concentrations for chemicals (those currently regulated as well as additional ones), the ability to incorporate new toxicity and exposure data for chemicals, and the potential to review the risks posed by entire wastes. The lower the risk posed by a waste, the greater the number of disposal options.

In the following sections, the committee discusses areas where DTSC's proposed approach is successful in meeting these goals, as well as areas where there are significant failings in either the scientific and technical aspects of the system or in the documentation of DTSC's approach. The committee makes recommendations for improving both the scientific underpinnings of the approach and its presentation.

The committee concurs with DTSC's decision to develop a risk-based, multimedia, multipathway approach to classifying wastes. Such an approach considers not only the toxicity of the waste, but also the potential for people or other organisms to be exposed by multiple pathways in a variety of settings. In addition, the approach can be applied to protect humans and other organisms at levels of protection required by California policy. In order to illustrate the differences between the current and the proposed systems, the committee recommends that DTSC select several known wastes and show how they would be classified under each system.

The committee recommends that DTSC adopt a comprehensive approach to its proposed multimedia implementation by integrating risks from all

exposure pathways into a single risk characterization. The exposure models used by DTSC should integrate all pathways because humans and other organisms might be exposed simultaneously through numerous pathways. This might require a slightly different approach by DTSC that would combine the results of the extraction tests (for groundwater protection) and the measurement of total concentrations in the waste.

DTSC's proposed approach lacks clear, explicit definitions of the human and ecological populations or groups to be protected and the desired levels of protection for these populations, including sensitive subpopulations such as children. By linking human and ecological protection goals to specific exposure scenarios, the selection of appropriate parameters for the multimedia transport and exposure models would be improved.

The DTSC documentation should describe how all exposure scenarios were developed for residents, workers, and nonhuman organisms, and screened for acceptability and completeness. The rationale for choosing the long-term exposure scenarios for the total threshold limit concentrations and for the soluble or extractable regulatory thresholds should be clearly explained, as should the short-term exposure scenarios used to develop the acute toxicity thresholds.

DTSC should explain in its documentation how the model parameters were selected to ensure consistency with the appropriate exposure scenarios. A systematic approach is required to identify the appropriate parameters. Furthermore, DTSC should indicate how population or environmental shifts might be incorporated into future exposure scenarios. For example, how will changes in subpopulations, such as children, be factored into the scenarios?

The committee recommends that DTSC's proposed approach explicitly describe the physical processes presumed to occur in each environmental pathway of each exposure scenario, to allow determination of whether the assumptions are realistic.

In its review of the component models used in the exposure models,

the committee found that specific models were not well defined, and that it was difficult or impossible to determine the original sources of many of the parameters used. The DTSC documentation should indicate how the specific models, with their chosen parameter values, match the physical processes occurring in each particular exposure scenario and what the original sources for all the parameter values were.

The committee found many errors and inconsistencies in: (1) the selection of models for the scenarios (application); (2) the documentation of the scenarios, models, and parameters; (3) the implementation of the models (i.e., what was actually done was not what DTSC says it did); and (4) the parameter values, which were often incorrect or mismatched to the scenarios. The committee recommends that a more thorough quality-control review and validation be conducted as these are critical for the development of a scientifically defensible classification system.

The DTSC proposal does not adequately distinguish between variability and uncertainty in its modeling efforts. With a more complete statement of policy goals, including the desired levels of protection and the populations to be protected, the correct treatment of uncertainty and variability should become apparent and should be explicitly discussed in the context of the policy goals.

The committee recommends that DTSC conduct a thorough sensitivity analysis to identify critical parameters and dominant pathways in the exposure models. Both the method for and results of the analysis should be clearly communicated, particularly given DTSC's statement that one exposure pathway dominated each scenario. The committee recommends that DTSC also reexamine the statistical distributions and determine what effect assumptions of lognormality have on the modeling results.

DTSC's proposal to use twice the estimated quantitation level, an analytical end point, when the calculated concentration of the risk-based solu-

ble or extractable regulatory threshold is less than the estimated quantitation level, is not a risk-based approach. There is no connection between the sensitivity of an analytical method and the sensitivity of exposed organisms. The committee recommends that DTSC establish the protectiveness of this default value.

DTSC proposes to use the federal RCRA toxicity characteristic leaching procedure to extract wastes rather than the current California waste extraction test to detect soluble chemicals in landfill leachate. The committee finds that DTSC is generally on the right path, but recommends that DTSC work more closely with stakeholders and the U.S. Environmental Protection Agency to address the shortcomings of the toxicity characteristic leaching procedure and the waste extraction test, and to explore development of a new extraction test that does not have the shortcomings of these tests, before final action is taken.

In DTSC's proposed approach, the chronic-risk estimate for a waste is based only on those chemicals for which total threshold limit concentrations or soluble or extractable regulatory thresholds have been developed. Chronic risks posed by other components of the waste or combinations of components are not considered, even though chronic risks from a combination of chemicals are likely to be greater than from any individual chemical. The committee, therefore, recommends that DTSC's proposed approach consider the chronic risk of all waste components, including those chemicals with total threshold limit concentrations or soluble or extractable regulatory thresholds, when assessing the total risk posed by the waste.

The committee also recommends that the DTSC approach take into account metal speciation, as the toxicity of a metal might depend greatly on the metal species present in the waste.

For wastes that do not contain chemicals for which a total threshold limit concentration or soluble or extractable regulatory threshold has been established, DTSC proposes that the toxicity characteristic for the waste

would be based on risks from its acute toxicity. The committee recommends that chronic toxicity data, when available, be considered by DTSC when assessing the risks posed by a waste. If only acute toxicity data are available for the risk assessment, the committee recommends that DTSC follow standard practice and use an uncertainty factor to account for the lack of chronic toxicity data. Failure to consider chronic effects could seriously underestimate the risks posed to human health and the environment from long-term exposure to wastes.

The committee further recommends that DTSC consider including acute irritation and sensitization (allergenicity) testing in its proposed approach. Respiratory irritation can exacerbate existing health conditions, such as asthma, and repeated exposures to some compounds can lead to sensitization of some people.

Clear documentation of the proposed DTSC approach is vital. The report reviewed by the committee lacks a roadmap to the documentation (almost 1,500 pages) and a synthesis into a manageable form. The documentation should parallel the process used by DTSC in developing its risk-based approach, beginning with a comprehensive discussion of the California human and ecological populations to be protected and at what levels, and followed by a description of all exposure scenarios and an explanation of the rationale for selecting representative scenarios, model selection and modifications, and parameter values. DTSC should include in its documentation the following: (1) background materials, including unpublished materials; (2) all implementations (in this case, spreadsheets); and (3) adequate references for all data sources. The committee believes that the lack of clarity in DTSC's documentation, including a clear description of the actual process by which a waste is classified, is a serious failing.

The committee recommends that DTSC reexamine its claim of improved flexibility of the proposed approach. DTSC does not specify in its documentation how reclassifying some wastes into new categories will reduce

the regulatory or economic burdens or increase flexibility with regard to waste disposal while still being protective of humans and other organisms. It behooves DTSC to propose a feedback or adaptive management mechanism to ensure that human health and the environment are protected at the desired levels.

The committee further recommends that DTSC's proposed approach include a clear process by which (1) additional chemicals can be evaluated for inclusion on the list of regulated agents, and (2) regulated chemicals can be reevaluated for possible reassignment of thresholds or even deletion from the state's list. Such a process must include a mechanism for response to new information as it becomes available. The inclusion of several examples using chemicals with widely varying characteristics (e.g., organic or inorganic, hydrophobic or hydrophilic) or new data (e.g., new toxicity studies, information on persistence in the environment) would considerably increase public confidence in the utility of the new waste-classification system.

In principle, the committee agrees that the use of a risk-based multimedia approach for classifying wastes is appropriate. In its assessment of the proposed DTSC waste-classification system, the committee concluded that, at the most fundamental level, DTSC can improve its approach by (1) providing a clear statement of its protection goals and their regulatory context; (2) explicitly defining and justifying the exposure scenarios; and (3) applying these scenarios consistently throughout the subsequent modeling. Furthermore, DTSC should conduct a much more extensive and detailed internal review of its proposed classification process to ensure scientific and technical accuracy and clarity.

Although the committee has identified a number of specific technical concerns with regard to DTSC's classification proposal, these concerns should not be interpreted to suggest that the approach has insurmountable flaws. The committee believes that DTSC's basic approach of considering the likely risks associated with waste disposal rather than just the toxicity of the waste, and using multimedia, multipathway exposure models for assessing these risks, represents a significant improvement in developing waste management regulations. The committee appreciates the difficulty faced by DTSC, or any organization, when

developing a risk-based approach to waste classification that has the flexibility to incorporate new scientific knowledge. The committee's review suggests that the proposed approach would benefit from additional development and thorough documentation. When fully developed and documented, DTSC's approach could be a useful tool for other states and the U.S. Environmental Protection Agency in addressing similar waste classification issues. The committee, therefore, urges DTSC to address the concerns presented in this report, and submit the revised report to a detailed and comprehensive scientific (and editorial) evaluation by California Environmental Protection Agency staff and other experts before any further external review.