Risk-Based Waste Classification in California (1999)

In this Chapter, the NRC committee identifies broader programmatic issues that, if addressed, would enhance the risk-based approach of the California Environmental Agency Department of Toxic Substances Control (DTSC.). As described in Chapter 1, DTSC undertook revision of their current waste-classification program to respond to the aims of the California regulatory structure update (RSU). Of the four guiding principles of the RSU, two bear on the scientific or technical aspects of the programs (protecting public health and the environment and fostering compliance through regulatory flexibility and simplicity), and the other two are more policy-based and were not considered by the committee (remove regulations that are ineffective and coordinate regulations with other governmental agencies). The first two principles and DTSC's approach to them by using a risk-based waste-classification scheme are discussed below, along with suggestions for evaluating the success of the new waste-classification regulations.

This chapter then examines whether DTSC's proposed waste-classification scheme would likely achieve the scientific and technological objectives propounded by the DTSC, which are to (1) consider exposure when classifying wastes; (2) develop a mechanism for incorporating new toxicological or technical information; (3) provide guidance on handling specific waste streams rather than specific chemicals; and (4) institute a system for regulating chemicals other than the current 38 total threshold limit concentrations (TTLC) chemicals or the 36 current soluble or extractable thresholds (SERT) chemicals.

Finally, there is a discussion of some other concerns DTSC may want to consider for improving their waste classification system. These include integration of SERTs and TTLCs into a single risk-based value, development of a true multimedia risk-assessment approach for ecological end points, and expansion of the universe of risks that are considered when classifying wastes.

The committee strongly endorses DTSC's proposed use of a risk-based, rather than a simply toxicity-based, waste-classification system. An integral aspect of the proposed DTSC system is a change from a single toxicity-based threshold to distinguish hazardous and nonhazardous wastes to a risk-based system with two thresholds, one to separate nonhazardous from special wastes and a second more conservative threshold to separate special wastes from hazardous wastes. Special wastes will be subject to less stringent disposal and reporting requirements.

Although the current waste-classification system is far from perfect, there is no indication from DTSC or public commenters that it is not protective of the public health and the environment, at least for the 38 chemicals that are included in the system. The proposed classification, with two risk-based thresholds, will increase the number of wastes that may be classified as special. According to DTSC's proposal, wastes currently classified as hazardous may be reclassified as special and be disposed to landfills that are less restrictive. Wastes that are currently classified as nonhazardous may also be reclassified as special and subject to more stringent disposal requirements. Although DTSC indicated at the public meetings that the new waste-classification system will be more protective of human health and the environment, the documentation fails to support this contention. Specifically, it is not demonstrated in the DTSC documentation that the broader special-wastes category will protect human health at the specified risk level. Indeed, for silver, the proposed system is theoretically less protective by removing silver from the TTLC list completely. The inclusion of several examples showing how different wastes would be classified (and subsequently disposed)

under the current and the proposed systems might go some way to demonstrating DTSC's contention that the proposed approach is at least as protective as the current system, if not more so.

DTSC has not fully documented the limitations and safeguards for the disposal of special wastes in class II landfills. The committee is unable to determine whether such disposal is appropriate for the necessary level of health and environmental protection. Documentation for the classes of landfills or other waste management options, such as land conversion, is an essential part of an integrated risk assessment.

There are two issues in the proposed system concerning ecological risks that need to be highlighted. In the first case, the proposed approach states that lower TTLC values, based on the scenarios and parameter values selected to be appropriate to protect humans, will necessarily be protective of wildlife. There are a number of examples in which this is not the case, such as DDT. It is incumbent upon DTSC to provide an analysis that gives the probability that the exit-level TTLCs derived for the protection of human health would be protective of ecological, nonhuman receptors.

The second issue concerning ecological risks is identical to that of using a two-tiered system for human health. Does a two-tiered system have any advantage for environmental protection? The proposed system also uses an upper and lower acute toxicity threshold based on an acute fish toxicity test for the upper threshold (nonhazardous versus special wastes). This toxicity value is then divided by a liner protection factor to distinguish between special and hazardous wastes. The advantage of using this system over one with a single threshold is not specified. Similarly, the advantages of a two-tiered system for protecting human health and the environment are not specified in, nor are they obvious from, the DTSC documentation. This suggests that the more intrinsic goal of the two-tiered system is to increase regulatory flexibility and simplify the regulatory requirements.

A second principle of the RSU is to increase regulatory flexibility and simplify regulations. DTSC intends to accomplish this by increasing the number of wastes that will be categorized as special based on the two risk thresholds. The advantages of having a waste designated as special

is not obvious from DTSC's documentation; the waste classification system given in the report (DTSC 1998a, p. 36) suggests that the new system is more complex than with the one-tiered system.

Given the proposed exposure scenarios, and in particular their lack of definition of protection goals and appropriate model parameters, it is possible that a significant number of compounds will move from the nonhazardous to the special-waste categories. At the public meetings, DTSC indicated that the volume of hazardous and nonhazardous wastes will not change significantly under the new proposal (DTSC, personal commun., September 10, 1998 and November 20, 1998); however, they have not presented a convincing argument supporting this statement. DTSC should reevaluate the flexibility created by the addition of the special-waste category after it has better defined the populations to be protected, protection goals, exposure scenarios, and model parameters consistent with those goals.

The proposed system appears to lack flexibility with regard to the addition or deletion of chemicals. As discussed below, there is no clearly defined mechanism for selecting new chemicals for SERT or TTLC development or for removing chemicals from the SERT or TTLC lists should there be new evidence that they pose little or no risk. This might also be a consideration for chemicals already on the list if speciation is an important factor in the risk posed by the chemical. For example, metals are regulated as elemental forms with no distinction between inorganic or organic forms (with the exception of lead). These forms might have significant differences in toxicity and subsequently risk. Chromium is a good example for which toxicity varies considerably with the speciation of the metal.

DTSC indicated that it had several goals when revising the current waste-classification system. California has determined that the federal Resource Conservation and Recovery Act (RCRA) classification system was not sufficiently protective of human health or the environment and that it was necessary for the state to identify and manage those wastes that might not be classified as hazardous under RCRA but that might pose risks to California's. The current waste-classification system, as summarized in Chapter 1, is limited. Specifically, DTSC wished to improve the system by doing the following:

• Considering exposure in classification of waste.
• Incorporating new toxicological or technical data.
• Handling specific waste streams rather than specific chemicals.
• Developing a mechanism for regulating chemicals other than the 36 SERT and 38 TTLC chemicals.

DTSC has prepared a good prototype system for incorporating risk into the classification of waste. Using models, such as CalTOX, it has factored exposure in with chemical hazards to develop risk-based thresholds for selected waste components. This has been done for human health risks from exposure via groundwater, surface water, air, soil, and food exposure pathways. Incorporating exposure data including information on the environmental fate and transport of chemicals, as well as the identification of potential receptors (human and ecological), can only increase public confidence that the proposed regulations will protect their health and the environment. Risk-based thresholds for human health have been developed for 36 chemicals for groundwater exposure (SERTs) and 38 chemicals for other exposure pathways (TTLCs).

Exposure pathways, however, have not been included in the development of ecological thresholds. The aquatic toxicity values are toxicity-based with little or no consideration of the various pathways by which numerous receptors might be exposed to chemicals. Exposure models are available for ecological risk assessments and their use would significantly improve DTSC's waste-classification system.

However, DTSC has not been clear as to what human or ecological populations are being protected and what level of protection is required or acceptable. Without such specification, it is difficult to develop the appropriate exposure scenarios and to choose the appropriate parameter values for the exposure models.

The incorporation of new chemical or model-specific information into the regulatory process is an important goal for any regulation designed to protect human health and the environment, and is one of the primary aims of DTSC's proposed system. The documents provided to the com-

mittee do not describe the mechanism by which DTSC will collect, review, and update such information in the future. No system was presented for periodic or episodic review of SERTs or TTLCs should new toxicity or exposure information become available. For example, if the U.S. Environmental Protection Agency (EPA) were to change the maximum concentration limits on which a TTLC or SERT was based, how would DTSC account for this change in value? Would the new information be considered on an ''as available basis,'' or would there be a systematic review similar to that provided for reviewing the minimum risk levels developed by the Agency for Toxic Substances and Disease Registry for chemicals found at Superfund sites?

As presented in the DTSC report, the derivation of risk-based values would appear to be a one-time event. The development of the upper and lower TTLCs and SERTs are based on an analysis of recently available toxicological information, including reference doses and maximum contaminant levels. Reference doses and maximum contaminant level values are developed by the EPA and are subject to periodic revisions as new toxicological data become available. Supporting documents, which present the toxicological database upon which the health-effects values are based, are also developed for each chemical.

In tandem with the need to update TTLCs and SERTs with new toxicological information is the need to incorporate new exposure assessments. The exposure assessments are based on exposure models such as CalTOX and the preliminary endangerment assessment model. Exposure models are subject to multiple iterations, being updated and refined as new fate, transport, and other exposure information becomes available. The exposure scenarios currently being used could subsequently be found to be inappropriate and would then need to be modified to reflect real-world situations. DTSC has not proposed any method to identify and collect new exposure data based on further environmental monitoring or testing, or additional modeling.

DTSC indicated that its new approach will be useful for regulating new chemicals and wastes in addition to those for which TTLCs and SERTs have already been developed. Essentially any chemical might be consid-

ered for TTLC or SERT development; however, DTSC does not indicate whether there will be an attempt to select chemicals with potentially greater risk because of greater toxicity, greater exposure, or other criteria. This is an important consideration and has been used with ranking and screening criteria development in systems such as EPA's toxics release inventory and the Occupational Safety and Health Administration's permissible exposure limits. Selecting chemicals for TTLC or SERT development might require the concurrent development of screening mechanisms and ranking processes to identify those chemicals found in wastes that pose the greatest risks. Indeed, this is one criticism of the proposed list of TTLC and SERT chemicals. Many of the TTLC and SERT chemicals are no longer in commerce and will occur in wastes primarily from dredging or recycling of contaminated media, not from industrial wastes. Those chemicals that are highly toxic or persistent in the environment should be distinguished from those known to be in commerce or commonly found in waste.

DTSC should develop a process for selecting chemicals to be considered for SERT and TTLC development. This process should be based on population and environmental protection goals outlined at the beginning of the waste-classification process.

Other areas where DTSC might expand the utility of its proposed waste-classification system or where the system might be improved are briefly discussed below.

Integration of TTLCs and SERTs to give just one value protective of human health and the environment via groundwater or other possible exposure pathways.

The distinction between TTLCs and SERTs is not entirely justified. In a true multimedia, multipathway approach, only one risk level should be developed for each chemical, any new chemicals, and preferably, for entire wastes. In real-life situations, people and other organisms might be exposed to chemicals in the groundwater through ingestion of groundwater as drinking water. However, they might also be exposed to chemicals from the same source by inhalation of vapors evaporating

from such water, and by these and other routes through entirely different pathways from the same source. Therefore, the committee recommends that the exposure modeling include all media and all exposure pathways, and not separate out groundwater exposure. For example, the CalTOX multimedia model is fully capable of including exposure assessment for groundwater along with soil and air. A single regulatory value for all exposure pathways will simplify the waste-classification system.

Multimedia risk assessment for ecological impacts.

Ecological risk assessments are subject to the same multimedia and multipathway considerations as are human risk assessments. However, DTSC's proposed approach is seriously lacking ecological exposure scenarios and subsequent risk assessment values. Use of a toxicity value such as an LC₅₀ for fish or EPA's ambient water quality criteria does not constitute an ecological risk-assessment with some exposure assessment and, therefore, should not be called such in the DTSC documentation. Fish are not always the most sensitive species nor is surface water always the primary exposure pathway. As has been demonstrated in the past with numerous pesticides, birds are frequently at a greater risk from exposure to soil pathways, and thus might be the most affected species. Given the exposure scenarios presented in the DTSC report, particularly the land conversion scenario, the presence of contaminants in soil and subsequent ingestion by birds or other wildlife that consume soil (inadvertently or otherwise) might be of significant concern. Therefore, DTSC is encouraged to be more explicit and comprehensive in developing its ecological risk assessments. As presented, ecological concerns are not paramount, and it is impossible to ascertain whether wildlife will be adequately protected under this waste-classification approach.

Expand consideration of potential risks.

DTSC's proposed approach is a risk-based waste-classification system that results in a change in how some wastes are managed after generation, and possibly, in how they are disposed. The committee commends DTSC's use of risk-based analyses, and the integration of these analyses into waste regulations that protect human health and the environment. However, it might be advantageous for DTSC to consider the risks presented to humans and other organisms through the transport of

wastes. Although DTSC has indicated that there will be no net effect on the total volumes of waste that are classified as hazardous, special, or nonhazardous (DTSC, personal commun., September 10, 1998), there might be substantial changes in the pattern of movement of wastes. The consequences of such changes should be considered in DTSC's proposed system. If there is substantial additional transport of wastes to more distant landfills (for example, class III landfills are to have a 10-km buffer zone), then there might be a larger risk resulting from the additional transport. The increase in the number of miles traveled might lead to an increase in the number of vehicular accidents, resulting in possible personal injuries and a potential increase in waste spills.

Changes in the pattern of waste movement also raise the question of the acceptability of trade-offs in short-term versus long-term risk. Reclassifying a non-RCRA nonhazardous solid waste to a non-RCRA hazardous waste might ultimately result in a more secure disposal site (e.g., at a landfill with double synthetic liners and leachate monitoring), thus reducing long-term or chronic risk to the environment. Does that reduction, however, warrant the increased potential for more on-site waste handling, longer transport distances from the point of generation to disposal, or greater waste volumes having to be transported and disposed? These are difficult questions to address, but they are reasonable to consider when incorporating a risk-assessment system into the classification of wastes because some action may reduce risk at one point of the waste-management process and simultaneously increase risk at another. Although these issues are policy related, scientific methods may be used to analyze short-term and long-term risk at different points in the waste management system, and can allow multipathway, multimedia comparisons of risk at one waste-management location versus another. The committee supports the use of the multipathway, multimedia risk analysis, but points out that this is currently only being proposed for the location of disposal.

An essential part of any regulatory program should be a method for determining the success of the program. DTSC's proposed approach fails to incorporate a mechanism to qualitatively or quantitatively monitor the success of the new approach. DTSC provides a process to categorize

waste as hazardous or not based on risk estimates associated with specific waste components. The RSU process, which is driving DTSC's review and changes, has as one of its guiding principles the elimination or modification of regulations that are duplicative, ineffective, or do not provide needed protection or information. That goal will be difficult to meet without a feedback mechanism in the program to determine success.

In Chapter 4, the committee criticized DTSC's lack of a mechanism to validate the various attributes of the models that are used to calculate DTSC's regulatory thresholds. On a larger scale, determining the results of the proposed non-RCRA hazardous-waste program by inclusion of a meaningful evaluation system would lend significant credibility to DTSC's desire to meet the objectives of the RSU. To validate the intended effects on protection of human health and other organisms as well as the environment and the regulated community, DTSC should consider the development of a formal, periodic evaluation process.

At a minimum, DTSC should develop a tracking system that provides the data necessary to allow a meaningful assessment of the success of its approach. The development and implementation of a minimal database designed to track various programmatic effects should include the following:

• The actual increase or decrease of waste volume for each non-RCRA hazardous, special, and nonhazardous waste affected by this program.
• Changes in waste movement frequency and routing.
• The use or nonuse of the variance system that DTSC is increasingly relying on to allow program flexibility.

The scientific and regulated communities, as well as the general public, are aware that local, state, and federal regulations are continually modified as new risks are identified or duplication is being eliminated. As a result, the adoption of a regulatory system that has the flexibility to respond to such changes is critical.