Hazardous Waste Management System: Identification and Listing of Hazardous Waste: Hazardous Waste Identification Rule (HWIR)
[Federal Register: December 21, 1995 (Volume 60, Number 245)]
[Proposed Rules ]
[Page 66343-66469]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[[Page 66343]]
-----------------------------------------------------------------------
Part II
Environmental Protection Agency
-----------------------------------------------------------------------
40 CFR Parts 260, 261, 266, and 268
-----------------------------------------------------------------------
Hazardous Waste: Identification and Listing; Proposed Rule
[[Page 66344]]
ENVIRONMENTAL PROTECTION AGENCY
40 CFR Parts 260, 261, 266, and 268
[FRL-5337-9]
RIN 2050-AE07
Hazardous Waste Management System: Identification and Listing of
Hazardous Waste: Hazardous Waste Identification Rule (HWIR)
AGENCY: Environmental Protection Agency.
ACTION: Proposed rule, tentative response to Chemical Manufacturers
Association petition and the Hazardous Waste Identification Dialogue
Committee recommendations, and request for comments.
-----------------------------------------------------------------------
SUMMARY: The Environmental Protection Agency (EPA) today is proposing
to amend its regulations under the Resource Conservation and Recovery
Act (RCRA) by establishing constituent-specific exit levels for lowrisk
solid wastes that are designated as hazardous because they are
listed, or have been mixed with, derived from, or contain listed
hazardous wastes. Under this proposal, generators of listed hazardous
wastes that meet the self-implementing exit levels would no longer be
subject to the hazardous waste management system under Subtitle C of
RCRA as listed hazardous wastes. Today's Notice, commonly referred to
as the Hazardous Waste Identification Rule (HWIR), establishes a riskbased
''floor'' to hazardous waste listings that will encourage
pollution prevention, waste minimization, and the development of
innovative waste treatment technologies.
Many of the exit levels are established using an innovative risk
assessment which evaluates potential exposure pathways, both direct and
indirect, from a variety of sources, such as waste piles and surface
impoundments. This assessment focuses on both human and environmental
receptors and is presented for comment in today's Notice. The remaining
exit levels are based on an alternative risk analysis.
The Agency is also proposing to modify some of the land disposal
restriction (LDR) numerical treatment standards listed in subpart D of
40 CFR part 268. This notice proposes to cap technology-based treatment
standards with the risk-based exit levels which minimize threats to
human health and the environment. This notice also takes comment on
several general approaches and one specific approach for conditional
exemptions from subtitle C management. Today's notice also contains the
Agency's tentative response to a petition for rulemaking submitted by
the Chemical Manufacturers Association and the Agency's tentative
response to the recommendations made by the Dialogue Committee on
Hazardous Waste Identification. This committee was formally chartered
in July 1993 in accordance with the Federal Advisory Committee Act
(FACA).
DATES: EPA will accept public comments on this proposed rule until
February 20, 1996. Comments postmarked after this date may not be
considered. However, the Agency recognizes that, because of the
complexity of this proposed rulemaking, some commenters may want to
request additional time for comment submittal. In anticipation of these
requests, EPA will be communicating with the litigants and the court
regarding the implications on our rulemaking schedule of a possible
extension of the comment period for this proposal. If the comment
period is extended, the Agency will provide notice of such in the
Federal Register.
Any person may request a public hearing on this amendment by filing
a request with Mr. David Bussard, whose address appears below, by
January 5, 1996.
ADDRESSES: The public must send an original, two copies, and whenever
possible, a 3.5 inch computer disk containing the comments in a common
word processing format such as WordPerfect version 5.1 1. to: EPA
RCRA Docket (5305W), 401 M Street, SW., Washington, DC 20460.
\1\ This will greatly facilitate EPA's preparation of the
comment responses and will significantly reduce the cost associated
with responding to the comments.
-----------------------------------------------------------------------
Place ''Docket number F-95-WHWP-FFFFF'' on your comments. The RCRA
docket is located at: EPA's Crystal Gateway Office, 1235 Jefferson
Davis Highway, Arlington, Virginia, and is open from 9 a.m. to 4 p.m.,
Monday through Friday, excluding Federal holidays. The public must make
an appointment to review docket materials by calling (703) 603-9230.
The public may copy material from any regulatory docket at a cost of
$0.15 per page. Copies of the background documents, Integrated Risk
Information System (IRIS) chemical files, and other references (which
are not readily available) are available for viewing and copying only
in the RCRA docket.
Requests for a public hearing should be addressed to Mr. David
Bussard, Director, Characterization and Assessment Division, Office of
Solid Waste (OS-330), U.S. Environmental Protection Agency, 401 M
Street, SW., Washington, DC 20460.
FOR FURTHER INFORMATION CONTACT: The RCRA/Superfund Hotline at (800)
424-9346 or at (703) 412-9810. For technical information contact Mr.
William A. Collins, Jr., Mr. Greg Helms, or Ms. Pamela McMains, Office
of Solid Waste (5304), U.S. Environmental Protection Agency, 401 M
Street, S.W., Washington, DC 20460, (202) 260-4770.
Preamble Outline
I. Authority
II. Background
A. Overview of Hazardous Waste Identification Program
B. The Mixture and Derived-From Rules and the Contained-In
Policy
C. Overview of Expected Impacts of the Exit Rule
III. Scope of Revisions to the Mixture and Derived-From Rules
A. Rationale for Retention of the Mixture and Derived-From Rules
B. Revision to Derived-from Rule for Wastes Listed Because They
Exhibit the Characteristics of Ignitability, Corrosivity, or
Reactivity
IV. Development of Exit Levels and Minimize Threat Levels
A. Need for an Exit
B. Overview of the Exit
C. Selection of Constituents of Concern
1. Development of the Master List
2. Development of the Exit Constituent List
3. Constituents of Ecological Concern
D. Risk-Based Information
1. Human Health Benchmarks
a. Non-carcinogens
b. Carcinogens
c. Consideration of MCLs
2. Ecological Benchmarks
3. Sources of Data
a. Human
b. Ecological
E. Risk Assessment
1. The Risk Analysis
a. Introduction
b. How the Analysis was Structured
c. How Uncertainty is Addressed
d. Linkage of the Risk Analysis to the Groundwater Fate and
Transport
e. Risk Targets Used
2. Detailed Overview of the Non-Groundwater Risk Analysis
a. Waste Management Units
1. Use of Subtitle D Survey
2. Fate and Transport
3. Ash Monofill
i. Particle Size Distribution for Air Dispersion Modeling
ii. Monofill Characterization
iii. Vehicle Traffic
iv. Emission Equations for Ash Blown from Trucks and Spreading
and Compacting
4. Land Application Unit
i. Particle Size Distribution for Air Dispersion Modeling
ii. Area of Land Application Unit Relative to Agricultural Field
iii. Application Rate
iv. Waste Characteristics
[[Page 66345]]
v. Depth of Contamination
vi. Partitioning
5. Waste Pile
i. Waste Pile Height
ii. Particle Size Distribution for Air Dispersion Modeling
iii. Waste Characteristics
iv. Vehicle Traffic
v. Emission Equation for Ash Blown from Trucks
6. Surface Impoundments
i. Two-Phase Sludge Formation Model
ii. Dilution of Waste During a Spill
7. Tank
i. Unit Characterization
ii. Volatilization
8. Combustors
b. Fate and Transport
1. Pathways
2. Equations
3. Specific Issues on Pathways and Equations
i. Chemical Transformation
ii. Biodegradation
iii. Meteorological Data
iv. Soil Data
v. Soil Pathways
vi. Surface Water Pathways
vii. Food-Chain Pathways
c. Receptors
1. Human Receptors
2. Ecological Receptors
3. Groundwater Fate and Transport Modeling
a. Fate and Transport Processes
1. Effects of groundwater mounding
2. Transformation products
3. Fate and transport of metals
b. Enhanced solution algorithms
1. Linkage between unsaturated zone and saturated zone modules
2. Numerical transport solution
3. Solution for metals transport
4. Elimination of biases in determination of receptor well
concentrations
c. Revision of Monte Carlo methodology for nationwide
assessments
1. Data sources
2. Finite-source methodology
3. Site-based regional analysis
d. Implementation of EPACMTP
e. Waste management scenarios
1. Landfills
2. Surface impoundments
3. Waste piles
4. Land application units
f. Determination of regulatory limits
g. Chemical specific fate and transport processes
1. Organic constituents
2. Metals
4. Other Risk Assessment Issues
a. Difference between groundwater and nongroundwater pathways
1. Infiltration
2. Density of waste applied to land application unit
3. Unsaturated zone characteristics
4. Hydrolysis rates
b. Other groundwater pathway analysis issues
1. Use of 1000 years versus 10,000 years exposure time horizon
2. Implementation of parameter bounds in Monte Carlo procedure
3. Hydraulic conductivity of surface impoundment bottom layer
4. Waste pile infiltration rates
5. Land application unit infiltration rates
6. Aggregate effects of alternative groundwater modeling
procedures
F. Additional Eco-Receptor Consideration
G. Background Concentrations in Soils and other Issues Relating
to Results
H. Constituents with Extrapolated Risk-based Levels
I. Analytical Considerations
1. Development of Exemption Quantitation Criteria (EQC)
2. EQCs and LDR Requirements as Exemption Criteria
a. EQCs as exit levels
b. LDR Requirements in combination with EQC Exit Levels
3. Exemption for Constituents Without EQCs
V. Presentation of Exit Levels
A. Constituents with Modeled or Extrapolated Risk-based Exit
Levels
B. Constituents with Quantitation-based Exit Levels
C. How to Read the Exit Level Tables
VI. Minimize Threat Levels
A. Background
1. Summary of the Hazardous and Solid Waste Amendments of 1984
2. EPA's Interpretation of Standard for Treatment Requirements
B. Risk Assessment and Minimize Threat Levels
1. Rationale
2. Public Policy Considerations
C. Minimize threat levels
1. List of Constituents and Minimize Threat Concentrations
2. Constituents for which Exit Levels are not Minimize Threat
Levels
D. Meeting LDR requirements
1. Wastes Below Exit Levels as Generated
2. Wastes Above Exit Levels as Generated
VII. Dilution
VIII. Implementation of Exit
A. Implementation Requirements
1. Testing Requirements
a. Data Evaluation
i. Compliance with the Exit Levels
ii. Wastewater and Nonwastewater Categories
iii. Totals and TCLP Analyses
iv. Oily Wastes
b. Initial Test
2. Notification Requirements
B. Implementation Conditions
1. Records Maintained on Site
2. Testing Conditions
3. Testing Frequency and Process Change
C. Public Participation
IX. Request for Comment on Options for Conditional Exemptions
A. Legal Basis for Conditional Exemptions
B. Improvements in Management of Non-Hazardous Waste and in Risk
Assessment Methodology
C. Overview of Options for Conditional Exemptions
1. National Approach
a. Eliminate Disposal in Land Application Units
b. Unit-Specific Exit Levels for Each Disposal
c. Consideration of Additional Management Unit Design or
Management Practices
2. State Program Approach
3. Establish Exit Levels that Consider Regional or Site-Specific
Factors that might Affect Constituent Fate and Transport
4. Relief from Land Disposal Restrictions
D. Land Disposal Restrictions for Contingent Management Options
E. Contingent Management of Mixed Waste
X. Implementation of Conditional Exemption Option 1
A. Introduction and Overview
B. When Contingent Management Exemptions Become Effective
1. Placement of the waste in a qualifying unit
2. Point of generation
C. Requirements for Obtaining an Exemption
1. Sampling and Testing Requirements for Contingent Management
Exemptions
2. Requirements for Public Participation in contingent
Management Exemptions
3. Notification Requirements for Contingent Management
D. Implementation Conditions
1. Tracking conditions
2. Qualifying Unit
3. Claimant's Duty to Ensure Compliance with all Conditions
E. Retesting and Recordkeeping Conditions for Contingent
Management Exemptions
F. Compliance Monitoring and Enforcement for Contingent
Management Exemptions
1. Compliance Monitoring
2. Enforcement
G. Exports of Wastes Eligible for Contingent Management
Exemptions
H. Land Disposal Restrictions
XI. Relationship to Other RCRA Regulatory Programs
A. Hazardous Waste Determination
B. Characteristic Hazardous Waste
C. Toxicity Characteristic Level for Lead
D. Hazardous Waste Listings
E. Delisting
F. Requirements for Treatment, Storage, and Disposal Facilities
and Interim Status Facilities
G. Closure
H. HWIR-Media Rule/Subtitle C Corrective Action
I. Land Disposal Restriction Program
J. RCRA Air Emission Standards
K. Hazardous Debris
L. Hazardous Wastes Used in a Manner Constituting Disposal
XII. CERCLA Impact
XIII. State Authority
A. Applicability of Rules in Authorized States
B. Effect of State Authorizations
C. Streamlining Issues
XIV. Regulatory Requirements
XV. References
Appendix A: Background Tables for Risk Analysis
Receptors and Pathways
Appendix B: Table Comparing Groundwater Modeling
Effects of 1000 vs. 10,000 years
Appendix C: Tables Comparing the Modeled or Extrapolated
Risk Levels vs. the EQCs for Each Constituent
Appendix D: Tables Comparing the Exit Levels and the
UTS LevelsRegulatory Language
[[Page 66346]]
I. Authority
These regulations are proposed under the authority of sections
2002(a), 3001, 3002, 3004 and 3006 of the Solid Waste Disposal Act of
1970, as amended by the Resource Conservation and Recovery Act of 1976
(RCRA), as amended by the Hazardous and Solid Waste Amendments of 1984
(HSWA), 42 U.S.C. 6912(a), 6921, 6922, 6924 and 6926.
II. Background
A. Overview of the Hazardous Waste Identification Program
Section 1004(5) of the Resource Conservation and Recovery Act
(RCRA) as amended by the Hazardous and Solid Waste Amendments (HSWA) of
1984, defines ''hazardous waste'' as ''a solid waste, or combination of
solid waste, which because of its quantity, concentration, or physical,
chemical, or infectious characteristics may (A) cause, or significantly
contribute to an increase in the mortality or an increase in serious
irreversible, or incapacitating reversible, illness; or (B) pose a
substantial present or potential hazard to human health or the
environment when improperly treated, stored, transported, or disposed
of, or otherwise managed.''
Section 3001 of RCRA requires EPA to identify those wastes that
should be classified as ''hazardous.'' The Agency's hazardous waste
identification rules designate wastes as hazardous in one of two ways.
First, the Agency has established four hazardous waste characteristics
which identify properties or attributes of wastes which would pose a
potential hazard if the waste is improperly managed. See 40 CFR 261.21-
261.24. Any generator of a solid waste is responsible for determining
whether a solid waste exhibits any of these characteristics. See 40 CFR
262.11. Any solid waste that exhibits any of the characteristics
remains hazardous until it no longer exhibits the characteristics. See
40 CFR 261.4(d)(1).
The other mechanism EPA uses to designate wastes as hazardous is
''listing.'' The Agency has reviewed data on specific waste streams
generated from a number of industrial processes and has determined that
these wastes would pose hazards if mismanaged for one or more reasons,
including the presence of significant levels of hazardous constituents
listed in appendix VIII to 40 CFR part 261, the manifestation of one or
more of the hazardous waste characteristics, or the potential to impose
detrimental effects on the environment. (See generally 40 CFR 261.11).
As discussed in detail in the preambles and in associated dockets
accompanying the listings, EPA has generally determined that these
wastes contain toxic constituents at concentrations which pose risks
which are unacceptable for human or environmental exposure and that
these constituents are mobile and persistent to the degree that they
can reach environmental or human receptors.
On May 19, 1980, as part of the final and interim final regulations
implementing section 3001 of RCRA, EPA published two lists of hazardous
wastes: One composed of wastes generated from non-specific sources
(e.g., spent solvents) and one composed of wastes generated from
specific sources (e.g., distillation bottoms from the production of
benzyl chloride). The Agency also published two lists of discarded
commercial chemical products, off-specification species, container
residues, and spill residues thereof which are hazardous wastes under
specific circumstances. These four lists have been amended several
times, and are currently published in 40 CFR 261.31, 261.32, 261.33(e)
and (f), respectively.
B. The Mixture and Derived-From Rules and the Contained-In Policy
1. Mixture and Derived-From Rules
a. Scope and Purpose of the Rules
In 1980 EPA promulgated its first comprehensive regulatory program
for the management of hazardous waste under RCRA. 45 FR 33066 (May 19,
1980). As part of that rulemaking EPA promulgated several rules to
identify hazardous wastes. Two of these rules clarify the scope of the
hazardous waste listings. Under the mixture rule, a solid waste is a
hazardous waste if it is mixed with one or more listed hazardous
wastes. 40 CFR 261.3(a)(2)(iv). Under the derived-from rule a solid
waste generated from the treatment, storage or disposal of a listed
hazardous waste is also a hazardous waste. 40 CFR 261.3(c)(2)(i).
EPA promulgated the mixture and derived-from rules to close
potentially major loopholes in the subtitle C management system.
Without a ''mixture'' rule, generators of hazardous wastes could
potentially evade regulatory requirements by mixing listed hazardous
wastes with other hazardous wastes or non-hazardous solid wastes to
create a ''new'' waste that arguably no longer met the listing
description, but continued to pose a serious hazard. Such a waste also
might not exhibit any of the hazardous waste characteristics.
Similarly, without a ''derived-from'' rule, hazardous waste generators
and owners and operators of hazardous waste treatment, storage, and
disposal facilities (TSDFs) could potentially evade regulation by
minimally processing or managing a hazardous waste and claiming that
resulting residue was no longer the listed waste, despite the continued
hazards that could be posed by the residue even though it does not
exhibit a characteristic. (See 57 FR 7628).
It is for these reasons that the Agency continues to believe that
the mixture and derived-from rules are extremely important in
regulating hazardous wastes and reducing risk to human health and the
environment. However, EPA acknowledges that the mixture and derivedfrom
rules apply regardless of the concentrations and mobilities of
hazardous constituents in the waste. The purpose of this rulemaking is
to reduce any overregulation of low-risk wastes captured by the mixture
and derived-from rule.
b. Subsequent History
Numerous industries that generate hazardous wastes challenged the
1980 mixture and derived-from rules in Shell Oil v. EPA, 950 F. 2d 741
(D.C. Cir. 1991). In December 1991 the D.C. circuit vacated the rules
because they had been promulgated without adequate notice and
opportunity to comment. The court, however, suggested that EPA might
want to consider reinstating the rules pending full notice and comment
in order to ensure continued protection of human health and the
environment.
In response to this decision, EPA promulgated an emergency rule
reinstating the mixture and derived-from rules as interim final rules
without providing notice and opportunity to comment. 57 FR 7628 (Mar.3,
1992). EPA also promulgated a ''sunset provision'' which provided that
the mixture and derived-from rules would remain in effect only until
April 28, 1993. Shortly after, EPA published the proposal containing
several options for revising the mixture and derived-from rules. See 57
FR 21450 (May 20, 1992). This proposal also included options for
exempting media contaminated with listed hazardous wastes that are
regulated under the ''contained in'' policy.
The May 1992 proposal and the time pressure created by the ''sunset
provision'' generated significant controversy. In response, Congress
included in EPA's 1992 appropriations bill several provisions
addressing the mixture and derived-from rules. Pub. L. No. 102-389, 106
Stat. 1571. First, Congress nullified the sunset provision by providing
that EPA could not promulgate any revisions to the rules
[[Page 66347]]
before October 1, 1993 and by providing that the reinstated regulations
could not be ''terminated or withdrawn'' until revisions took effect.
However, to ensure that EPA could not postpone the issue of revisions
indefinitely, Congress also established a deadline of October 1, 1994
for the promulgation of revisions to the mixture and derived-from
rules. Congress made this deadline enforceable under RCRA's citizen
suit provision.
On October 30, 1992 EPA published two notices, one removing the
sunset provision, and the other withdrawing the May 1992 proposal. See
57 FR 49278, 49280. EPA had received many comments criticizing the May
1992 proposal. The criticisms were due, in a large part, to the very
short schedule imposed on the regulation development process itself.
Commenters also feared that the proposal would result in a
''patchwork'' of differing State programs because some states might not
adopt the revisions. This fear was based on the belief that States
would react in a negative manner to the proposal and refuse to
incorporate it into their programs. Finally, many commenters also
argued that the risk assessment used to support the proposed exemption
levels failed to provide adequate protection of human health and the
environment because it evaluated only the risks of human consumption of
contaminated groundwater ignoring other pathways that could pose
greater risks. Based on these concerns, and based on the Agency's
desire to work through the individual elements of the proposal more
carefully, the proposal was withdrawn.
Meanwhile, a group of waste generating industries challenged the
March 1992 action that reinstated the mixture and derived-from rules
without change. Mobil Oil Corp. v. EPA, 35 F.3d 579 (D.C. Cir. 1994).
EPA argued that the 1992 appropriations act made the challenge moot
because it prevented both EPA and the courts from terminating or
withdrawing the interim rules before EPA revised them, even if EPA
failed to meet the statutory deadline for the revisions. In September,
1994 the D.C. Circuit issued an opinion that dismissed the challenges
as moot under the rationale that the Agency had offered.
In early October 1994 several groups of waste generating and waste
managing industries filed citizen suits to enforce the October 1
deadline for revising the mixture and derived-from rules. The U.S.
District Court for the District of Columbia Circuit entered a consent
decree resolving the consolidated cases on May 3, 1993. Environmental
Technology Council v. Browner, C.A. No. 94-2119 (TFH) (D.D.C. 1994)
Under this decree the Administrator must sign a proposal to amend the
mixture and derived-from rules by November 13, 1995 and a notice of
final rulemaking by December 15, 1996. The decree also specifies that
the deadlines in the 1992 appropriations act do not apply to any rule
revising the separate regulations that establish jurisdiction over
media contaminated with hazardous wastes.
c. Federal Advisory Committees Act (FACA) and Outreach
After the withdrawal of the HWIR proposal, the Agency initiated a
series of public meetings with invited representatives from industry,
environmental groups, hazardous waste treaters, and States. These
meetings focused on three major issues: --RCRA regulation of low hazard
wastes with a particular interest in addressing issues raised regarding
the mixture and derived-from rules; concerns that full RCRA
requirements for contaminated media may unnecessarily impede clean-ups;
and need to regulate additional high-risk wastes outside the scope of
the current listings and characteristics.
A strong and successful effort was made to encourage all the
interested parties to participate in the public meetings. EPA forged a
solid partnership with the States (both ASTSWMO and Environmental
Commissioners under the National Governors Association) and the state
representatives worked closely with EPA as co-regulators in our
analyses of options.
In July of 1993, EPA chartered this group as an advisory committee
under the Federal Advisory Committee Act (Pub. L. 92-463)(58 FR 36200).
The committee rather quickly formed two sub-committees to allow
separate discussion of the low risk waste problem associated with the
mixture and derived-from rules and the rules for managing contaminated
media and other wastes during remediation.
By September of 1994 the low risk waste group had made significant
progress in identifying options for creating exemptions for low risk
wastes. Despite significant investment of time and effort, however, the
group was unable to reach consensus on many key issues.
With the statutory deadline for revisions to the mixture and
derived-from rules approaching, EPA requested that group to present a
final report in late September of 1994. EPA and representatives from
several state environmental agencies then took up the task of selecting
options for creating an exit rule, crafting regulatory language, and
developing necessary supporting materials. The FACA subcommittee's
final report was taken into consideration during the development of
today's proposal.
2. Contained-In Policy
The Agency also has interpreted its regulatory definition of
hazardous waste to extend to mixtures of hazardous wastes and
environmental media (such as contaminated soil and groundwater).2
See 40 CFR 261.3(c)(1) and (d)(2). Media that are contaminated with
listed or characteristically hazardous waste must be managed as
hazardous wastes until they no longer contain such wastes. To date, the
Agency has not issued any general rules as to when, or at what levels,
environmental media contaminated with hazardous wastes are no longer
considered to ''contain'' those hazardous wastes. Media that contain
hazardous wastes with constituent concentrations below the levels
proposed today will be eligible for exemption under the procedures
proposed today. In addition, in a separate rulemaking, the Agency plans
to propose additional rules reducing regulation of contaminated media
during remediation activities.
\2\ EPA's ''contained in'' policy was upheld as a reasonable
interpretation of 40 CFR 261.3(c)(1) and (d)(2) by the D.C. Circuit
in Chemical Waste Management, Inc v. U.S. EPA, No. 869 F.2d 1526
(D.C. Cir. 1989).
-----------------------------------------------------------------------
C. Overview of Expected Impacts of the Exit Rule
1. Listed Wastes
The purpose of this rule is to exempt from hazardous waste
regulation those solid wastes currently designated as hazardous waste
even though they contain constituent concentrations at levels that pose
very low risk to human health and the environment. While facilities
generating such wastes can petition for delisting by rulemaking under
the provisions of 40 CFR Sec. 260.20 and 260.22, EPA believes that the
detailed waste-stream specific review required under delisting is not
necessary for the low risk wastes that are identified by today's
proposal. The alternative, generic exit rule proposed today will be
faster and less resource-intensive for both the Agency and the
regulated community. By providing an opportunity for a more selfimplementing
exemption, the Agency intends to create incentives for
effective and innovative waste minimization and waste treatment and to
reduce unnecessary demand for Subtitle C disposal capacity, without
[[Page 66348]]
compromising needed environmental protection.3
\3\ As will be discussed further in this notice, the Agency
believes that the delisting process will continue to be valuable for
certain types of wastes which are not eligible for an exemption
under this proposal. Thus the Agency is not proposing to eliminate
or modify the delisting program as a result of this proposal.
-----------------------------------------------------------------------
By proposing a risk-based ''floor'' to listed wastes, today's
proposal should give a very strong incentive to generators of listed
hazardous waste to apply pollution prevention to their processes to
avoid Subtitle C control. This action should also give incentive for
the development of innovative treatment technologies to render wastes
less risky.
Today's proposed rule specifies sampling and analysis requirements,
public participation, reporting and record keeping requirements. Most
of these provisions are alternatives to the safeguard of waste-specific
review provided under the delisting program. The exit levels are riskbased
concentrations at which a human or wildlife species could be
directly or indirectly exposed to the exempted waste, and would be
unlikely to suffer adverse health effects. The exposure scenario used
to develop these levels assume that the exempted waste will no longer
be subject to Subtitle C control, but will be managed as a solid waste
in one of a variety of non-hazardous waste management units regulated
under Subtitle D.
2. Characteristic Wastes
Listed hazardous wastes exempted under today's proposed rule which
exhibit any of the characteristics will continue to be regulated as
hazardous wastes until the characteristic is removed. In a number of
cases, wastes were listed on the basis of containing both toxic
hazardous constituents and exhibiting one or more of the hazardous
waste characteristics that do not relate to chemical toxicity (e.g.,
ignitability, corrosivity, or reactivity). If such a waste still
exhibits any characteristic after complying with the exemption criteria
proposed in today's proposed rule, it must continue to be managed as a
characteristically hazardous waste.
III. Scope of Revisions to the Mixture and Derived-From Rules
The mixture and derived-from rules promulgated in 1980 and
reinstated in 1992 require Subtitle C regulation of all mixtures of
listed hazardous wastes and solid wastes and all residuals from
treatment of hazardous wastes. The rules proposed today, however, allow
rapid exemptions for mixtures and derived-from wastes that present no
significant threats to human health and the environment. Those wastes
that would remain subject to the mixture and derived-from rules
typically will pose risks that warrant regulation under Subtitle C. To
the extent that this is not true for a particular mixture or treatment
residual, the delisting process remains available (at least at the
state level) to exempt wastes with constituents at more site- and
waste-specific levels. Consequently, EPA has tentatively determined
that further revisions of the mixture and derived-from rules, with the
exception of the one minor change to the derived-from rule discussed
later in this section, are not warranted in this rulemaking. However,
EPA requests comment on this conclusion.
A. Rationale for Retention of the Mixture and Derived-From Rules
EPA continues to believe that it had ample statutory and regulatory
authority to promulgate the original rules and that it also has ample
authority to maintain the rules without further revisions. The mixture
and derived-from rules, particularly with the revisions proposed today,
ensure that hazardous wastes that are mixed with other wastes or
treated in some fashion do not escape regulation so long as they are
reasonably likely to continue to pose threats to human health and the
environment. They thus retain jurisdiction over listed hazardous wastes
and clarify that such wastes are not automatically eligible for exit
when they are mixed or treated. Although RCRA sets out criteria for the
identification of hazardous wastes to enter the subtitle C system, it
is silent on the question of how to determine that a waste is eligible
to exit the system. EPA's interpretation of the statute is thus
entitled to deference so long as it is reasonable and consistent with
RCRA's purposes.
EPA believes that its decision to retain jurisdiction over major
portions of the universe of waste mixtures and treatment residues is
consistent with its authorities under sections 3002-3004 of RCRA to
impose requirements on waste handlers until wastes have ''cease[d] to
pose a hazard to the public''. Shell Oil Corp. v. EPA, 959 F.2d 741,
754 (D.C. Cir. 1991). See also Chemical Manufacturers Assoc. v. EPA,
919 F.2d 158, 162-65 (EPA may regulate the disposal of nonhazardous
wastes in a hazardous waste impoundment under section 3004) and
Chemical Waste Management, Inc. v. EPA, 976 F.2d 2, 8, 13-14 (D.C. Cir.
1992) (EPA may require further treatment of wastes under section 3004
even though they cease to exhibit a hazardous characteristic).
The mixture and derived-from rules are also valid exercises of
EPA's authority to list hazardous wastes under section 3001. That
provision gives EPA broad authority to promulgate listing criteria.
EPA's 1980 criteria authorize the listing of classes of hazardous
wastes when it has reason to believe that wastes in the class are
typically or frequently hazardous. See 40 CFR 261.11(b). Such class
listings are permissible even if some members of the class do not
actually pose hazards. Nothing in the section 1004(5) definition of
hazardous waste, in section 3001, or in EPA's listing criteria require
EPA to prove that every member of a class poses a hazard. In fact, many
waste listings describe ''classes'' of hazardous wastes because they
cover a range of materials that are not identical in composition. The
mixture and derived-from rules thus are fully authorized as class
''listings'' under section 3001.
EPA has also made a reasonable factual determination that these
classes of waste warrant regulation under sections 3002-3004 and
section 3001. In 1980 EPA determined that the hazardous constituents
contained in these wastes are not generally eliminated or rendered
nontoxic simply because a waste is mixed with other wastes or managed
in some fashion. In 1992, when EPA repromulgated the mixture and
derived-from rules, it documented numerous instances of mixed and
derived-from wastes that continued to pose hazards. See 57 FR 7629
(March 3, 1992). Today, EPA is proposing that members of this class of
wastes that pose low risks will be eligible for an expedited, selfimplementing
exemption from Subtitle C regulation. Accordingly, EPA has
an even better basis for believing that wastes which remain within the
scope of the mixture and derived-from rules pose threats warranting
regulation.
Additionally, EPA continues to believe, as it did in 1980, that it
would be virtually impossible to try to identify all possible waste
mixtures and treated wastes and assess their hazards individually.
EPA's rule reasonably retains jurisdiction over both broad classes and
places the burden of proof on the regulated community to show that a
particular waste has ceased to present a hazard. Today's selfimplementing
exit proposal will reduce that burden significantly,
ensuring that the mixture and derived-from rules represent a reasonable
approach to regulating these classes of wastes.
[[Page 66349]]
B. Revision to Derived-from Rule for Wastes Listed Because They Exhibit
the Characteristics of Ignitability, Corrosivity, or Reactivity
In 1981 EPA responded to a number of comments on the scope of the
original 1980 mixture rule by promulgating a number of exemptions for
mixtures of solid wastes and listed hazardous wastes which, according
to information submitted by commenters, posed no significant risk to
human health and the environment. See 46 FR 56582 (Nov. 17, 1981). The
1981 rule included an exemption for mixtures of solid wastes and
hazardous wastes listed solely because they exhibited one or more of
the hazardous waste characteristics, if the resultant mixtures no
longer exhibited a characteristic. The exemption was based on a finding
that such mixtures did not pose threats to human health and the
environment warranting Subtitle C regulation. See 46 FR 56568 and the
current text of the exemption at Sec. 261.3(a)(2)(iii). EPA notes that
it has never promulgated any listings for wastes solely on the basis
that they exhibit either the 1980 EP toxicity characteristic or the
1990 toxicity characteristic; consequently, only mixtures containing
wastes listed because they exhibit the characteristics of ignitability,
corrosivity, or reactivity have been eligible to exit Subtitle C when
they no longer exhibit the characteristic.
The 1981 notice focused exclusively on issues concerning the
mixture rule. Consequently, EPA did not propose any parallel exemption
for such wastes for the separate derived-from rule (codified at
Sec. 261.3(c)(2)(i)), even though the derived-from wastes would appear
to present similarly low risks if they no longer exhibited a
characteristic and were treated to meet LDR standards before land
disposal. Recent inquiries from the public have highlighted the
discrepancy in the scope of the mixture rule and the derived-from rule
for wastes listed solely because they exhibit characteristics. EPA
believes it has no reason to treat derivatives of wastes listed solely
because they exhibit the characteristic of ignitability, corrosivity,
or reactivity any differently from the way it treats mixtures of such
wastes because both present similar low risks to human health and the
environment.
Consequently, EPA is today proposing a revision to the derived-from
rule that will closely resemble the 1981 revision to the mixture rule.
Since no listings to date have been based on the toxicity
characteristic, EPA is proposing to limit the new revision to the
derived-from rule to wastes listed because they exhibit only the
characteristics of ignitability, corrosivity, or reactivity. EPA is
also not proposing to exempt wastes that might in the future be listed
only because of the toxicity characteristic because (as this rule
proposal indicates) there can be risk concerns with the TC constituents
below TC levels. EPA requests comment on this proposal to create a new
exemption to the derived-from rule for this limited category of listed
wastes.
The proposed exemption will also remind the regulated community of
the separate duty to comply with requirements imposed by the part 268
regulations implementing the LDR program. In CWM v. EPA, 976 F.2d 2
(D.C. Cir. 1992), the U.S Court of Appeals for the D.C. Circuit
interpreted RCRA section 3004(m) as requiring treatment of decharacterized
hazardous wastes to meet LDR treatment standards even
after the wastes cease exhibiting a characteristic. EPA believes that
de-characterized derived-from residues from wastes listed because they
exhibit characteristics also must meet LDR requirements, unless they
are either delisted or are exempt at the point of generation pursuant
to other provisions proposed in this rule (e.g., meeting HWIR levels at
the point of generation).
In 1992 EPA amended the 1981 exemption to mixture rule to provide a
similar cross-reference and clarification for mixtures containing decharacterized
listed wastes. See 57 FR 37194, 37210-11 (Aug. 18, 1992).
That 1992 clarification, however, only covers nonwastewater mixtures.
As explained in that mixture rule preamble, EPA then regulated decharacterized
wastewaters much less stringently under the LDR program.
Consequently, EPA did not believe it was necessary to remind the
regulated community to comply with LDR requirements for wastewater
mixtures.
Later in 1992 the CWM v. EPA decision invalidated most of the
distinctions between the LDR rules for wastewaters and nonwastewaters.
EPA is now revising the LDR program to comply with that decision in the
LDR Phase III and Phase IV rulemakings. To reflect the changes in LDR
regulation of wastewaters, the derived-from rule exemption proposed
today reminds the regulated community of the need to comply with part
268 LDR requirements for all types of derived-from residues. EPA
requests comment on this clarifying language. EPA also requests comment
on whether it should revise the LDR clarification for the mixture rule
as well.
IV. Development of Exit Levels and ''Minimize Threat'' Levels
A. Need for the Exit
The primary purpose of this rule is to address listed hazardous
wastes, mixtures of listed hazardous wastes and solid wastes, and
residues derived-from managing listed hazardous waste that, under
current rules, continue to be designated as ''hazardous waste''
although they are either generated with constituent concentrations that
pose low risks or treated in a manner that reduces constituent
concentrations to low levels of risk.
EPA notes that there are currently exemptions, both codified and
contained in policy directives, from the hazardous waste identification
system, particularly the mixture and derived-from rules, for certain
types of wastes or wastes with certain constituent concentrations. See
e.g. 40 CFR 261.3(a)(2)(iv)(A) through (E) and policy memorandums such
as the ''Skinner Memorandum'' dated August 23, 1995. EPA is not
proposing to modify or replace any of these exemptions and policy
statements.
B. Overview of the Exit
For 191 of the 376 constituents of concern, EPA conducted a
detailed human health risk analysis to develop risk-based levels for
either the wastewater or nonwastewater form of a constituent (or both).
To conduct this analysis, EPA identified five types of units actually
and rather frequently used to manage nonhazardous wastes that covered
the full range of environmental releases needing analysis. The May 1992
proposal of exit levels for listed wastes, like many previous RCRA
rules, assessed only risks from releases to groundwater. In response to
complaints that such an assessment would not protect human health and
the environment from other types of releases, EPA also assessed
potential releases to air, surface water and soil in this proposal.
For each category of releases, EPA evaluated both relatively simple
pathways (such direct human ingestion of contaminated groundwater) and
more complex pathways (such as the deposition of windblown waste
particles on agricultural land, followed by crop uptake, consumption of
the crop by cattle, and consumption of contaminated beef or milk by
humans). EPA assessed approximately 8 to 27 release pathways depending
on the type of waste management unit.
Additionally, EPA screened the same group of 191 constituents to
identify the highest priorities for assessment of
[[Page 66350]]
ecological receptors. In addition, EPA considered for its assessment
the toxicological effects of silver on ecological receptors. EPA
conducted a specific assessment of ecological risks for 47 constituents
using the same five units and the same pathways (modified to reflect
ecological exposures) for each unit. This risk assessment is described
in more detail in sections V.B. and C.
Data limitations and resource constraints prevented EPA from
conducting a risk analysis for the remaining constituents of concern.
For each of these constituents, EPA extrapolated exit levels from
levels derived-from the risk assessment for similar chemicals. EPA's
extrapolation methodology is described in section IV.F.
The current capabilities of analytical chemistry constrain EPA's
ability to use some of concentrations as exit levels. For approximately
one-fourth of the constituents, EPA found that available methods could
not routinely measure the constituent at the modeled or extrapolated
risk-based exit level.
C. Selection of Constituents of Concern
1. Development of the Master List
EPA developed an initial ''Master List'' of 506 constituents to be
evaluated for purposes of establishing exit criteria. This master list
was developed by combining the constituents specifically listed in the
following appendices of 40 CFR part 261: Appendix VII, Basis for
Listing Hazardous Waste; Appendix VIII, Hazardous Constituents; and
appendix IX of part 264, the Ground-Water Monitoring List. The master
list includes the full list of constituents referenced in appendix VII,
including the F039 constituents.
Appendix VII to part 261, which was originally promulgated on May
19, 1980 (45 FR 33084) sets out the chemical constituents found to pose
threats to human health and the environment that served as the actual
basis for each of EPA's original hazardous waste listings. Appendix
VIII to Part 261, also promulgated in 1980, is a more general listing
of chemicals found to pose potential threats to human health and the
environment. (45 FR 33084). EPA considers wastes containing appendix
VIII constituents to be candidates for listing determinations. EPA
amends appendix VII from time to time as EPA identifies additional
potentially toxic constituents.
EPA later promulgated appendix IX to part 264 to identify those
appendix VIII constituents which it could routinely expect owners and
operators of permitted hazardous waste treatment, storage and disposal
facilities to monitor in groundwater. EPA also included in this
appendix 17 additional constituents found to pose significant risks
that the Superfund program routinely monitored in groundwater. (52 FR
25942, July 9, 1987).
EPA established in these rulemakings that each of these
constituents had significant potential to threaten human health, and,
by implication, potential to threaten the environment. (Most of the
data EPA utilized predicted toxic effects on humans.) EPA finds it
reasonable to include each of these constituents on the list of
chemicals of concern.
Further, EPA believes that, with the exception of the six chemicals
identified below, the three appendices identify the chemicals of
current concern to EPA that are likely to be found in listed wastes.
The Agency requests comment on whether the master list should also
include six constituents that are not listed in any of the above
sources. These six constituents, which are listed in Table 1, are found
in six ''U'' listed wastes (commercial chemical products that become
hazardous wastes when discarded). See 40 CFR 261.33(f). EPA originally
listed these wastes because they routinely exhibited the characteristic
of ignitability. Since the original listings, however, sufficient
toxicity data have become available for these constituents. (The risk
number for dimethylamine was recently withdrawn; however, EPA
understands that it will shortly be replaced). Because of the toxicity
data associated with these constituents, the Agency is taking comment
on whether exit levels should be established for these six constituents
in today's rulemaking. The Agency also requests comment on whether
these six constituents should be added to Appendix VIII.
Table 1.--Constituents Not on Appendices VII, VIII, or IX
-----------------------------------------------------------------------
Nonwastewater
CAS # Constituent Wastewater -------------------------
Totals Leach
-----------------------------------------------------------------------
75-07-0............................. Acetaldehyde (ethanal)............ ........... ........... ...........
98-82-8............................. Cumene............................ .67 18,000 2.5
124-40-3............................. Dimethylamine..................... ........... ........... ...........
110-00-9............................. Furan............................. .16 1300 .06
79-10-7............................. Acrylic acid...................... (\1\) (\1\) (\1\)
98-01-1............................. 2-Furancarbox- aldehyde (furfural) (\1\) (\1\) (\1\)
-----------------------------------------------------------------------
\1\ No exit levels because no EQC is available for this constituent. The criteria for exit would be to meet LDR
treatment standards in Sec. 268.
Full documentation concerning the selection of constituents of
concern is available in the docket under The Background Document to
Support Development of the Final Constituent List under the Waste Exit
Rule.
2. Development of the Exit Constituent List
The Agency narrowed the list of 506 constituents to consist of 376
constituents that are included in the exemption list. 130 constituents
were deleted from the master list. Criteria for constituent deletions
from the master list include: Reactivity in air, analysis as a
different constituent, reactivity in water, hydrolysis in soil or
water, or is part of a chemical class with a specific constituent
represented on the list. Because different methods and quantitation
limits are necessary for solid and liquid matrices, two separate
analyses were conducted. The Background Document to Support Development
of the Final Constituent List under the Waste Exit Rule in the docket
further justifies deletions of constituents from the master list and
lists the deleted constituents.
Molybdenum is not on the Appendices VII, VIII, or IX, which
provided the scope of today's master list of constituents. In
anticipation of the Petroleum listing, due to a Drinking Water Sewage
Sludge regulatory level, and due to available toxicity information, the
Agency has included molybdenum on the exemption list. Due to modeling
time constraints, Molybdenum was not modeled for groundwater risk. The
groundwater
[[Page 66351]]
leach level was estimated by assuming a DAF of 10 and using the RfD.
The Agency requests comment on whether molybdenum should be on the
list.
Two modeled constituents do not have estimated quantitation
criteria (EQCs--see section IV.G.) and therefore do not have associated
exit levels. These constituents are ethylene thiourea and phenyl
mercuric acetate. The Agency requests comment on how to deal with these
two constituents. The following table represents modeled results for
these constituents for comment.
Table 2.--Modeled Constituents Without EQCs
-----------------------------------------------------------------------
CASNUM Constituent NWW totals NWW leach WW totals
-----------------------------------------------------------------------
96-45-7............................... Ethylene thiourea............. 0.51 0.00017 .00053
62-38-4............................... Phenyl mercuric acetate....... 0.0093 0.0045 0.012
-----------------------------------------------------------------------
EPA modeled chromium VI in the risk assessment. However, totals
chromium appears on the exit tables based on the exit levels calculated
from modeling chromium VI. This approach is consistent with the
Toxicity Characteristic approach to chromium. The Agency asks for
comment on this approach.
The cyanide exit level was extrapolated. It is meant to be totals
cyanide. The Agency requests comment on whether testing for totals
cyanide is appropriate.
The values in the exit tables for silver do not represent results
of human toxicity data for silver, rather they represent ecological
results from the risk assessment. The Agency has determined that the
effect of silver on humans is not a human health problem, rather it is
an aesthetic problem. The groundwater model did not model ecological
exposure, therefore, there is no groundwater risk level for silver.
3. Constituents of Ecological Concern
As explained above, EPA established in previous RCRA rulemakings
that the constituents on the exit list (376) present significant
threats to human health. Numerous comments submitted on EPA's May 1992
proposal to establish exit levels urged EPA to conduct a more specific
and detailed analysis of threats to non-human species. Consequently, in
this rulemaking EPA determined the constituents it believed to also be
reasonably likely to pose risks to ecological receptors.
EPA has not set benchmarks for ecological impacts for a large
number of constituents under any of its programs. Establishing such
benchmarks for this proposal would be a resource-intensive and timeconsuming
task. Accordingly, EPA narrowed the list of exit constituents
for which ecological receptors would be evaluated. First, EPA decided
to consider only the 191 constituents which it had already targeted for
analysis to protect human health. Second, the Agency developed a
methodology for screening the 191 constituents to identify those most
likely to pose significant risks to ecological receptors.
Based on an extensive review of available literature, EPA developed
five criteria to indicate the potential for ecological risks:
(1) Constituents that bioaccumulate (and possibly biomagnify) in
the food chain that can present elevated exposures to certain
predators;
(2) Persistent constituents that are likely to increase long-term
multi-generational exposures in wildlife;
(3) Constituents that cause reproductive and developmental effects
that can elicit adverse effects at sensitive life stages;
(4) Constituents that may cause ecological effects that have no
human analog (e.g., eggshell thinning); and
(5) Constituents that may cause effects to ecological receptors
continuously exposed.
EPA also developed operational definitions for each criterion. The
definitions were quantitative where possible. Further details can be
found in appendix B of the Technical Support Document for the Risk
Assessment for Human and Ecological Receptors.
EPA decided to designate as constituents of ecological concern the
47 constituents that exhibited at least two of the five criteria. The
Agency believes these constituents present the highest priorities in
terms of environmental risk. An additional 36 constituents exhibited
only one criterion. EPA, however, chose not to designate them as
constituents of concern because time and resource constraints would
prevent the Agency from completing an analysis with these constituents.
EPA, nevertheless, believes it has identified and analyzed sufficient
constituents of concern to ensure that the exit levels proposed today
provide for reasonable protection of the environment. Only 83 of 191
screened constituents showed any significant potential to pose threats
to the environment at levels protective of human health. Further, as
discussed in more detail below, of the 47 constituents that EPA
actually assessed for ecological impacts, only 6 wastewater
constituents and 18 nonwastewater constituents required exit levels to
protect environmental receptors lower than those necessary to protect
human health under the baseline proposal. Consequently, EPA believes it
is unlikely that all of the remaining constituents will present
significant threats to ecological receptors at levels that would
adequately protect human health.
D. Risk-Based Information
The Agency's proposed option for establishing exit values is based
on risk modeling to a hazard quotient of 1 and a 1 x 10-6 cancer
risk. The Agency chose a hazard quotient of 1 as its toxicity benchmark
value for non-carcinogens because evaluation of these compounds
presumes there is a threshold exposure above which individuals would be
at significant risk of suffering the adverse effects attributable to
the compound. The HQ is the Agency's best attempt to estimate that
level. Therefore, the Agency believes all exposures should remain below
HQ 1. Some Agency programs rely on HQ values less than 1 in standard
setting (the drinking water program uses an HQ of 0.20 to provide a
safety factor which allows for exposure to the constituent from sources
other than drinking water).
The Agency chose a toxicity benchmark of 1 x 10-6 cancer
risk for carcinogens for several reasons. A cancer risk level of 1 x
10-5 risk was used as a clearly hazardous level in establishing
the toxicity characteristic. Second, in the listings program, a 1 x
10-4 cancer risk is used as the presumptive listing risk, and a 1
x 10-6 as the presumptive no-list level. A cancer risk of 1 x
10-5 represents a level of initial concern about risk. Therefore,
in allowing listed hazardous waste to exit the requirements of Subtitle
C, the Agency was targeting waste that is clearly not hazardous. Thus,
the Agency believes the risk level should be at the
[[Page 66352]]
low end of the risk range used to bring waste into the hazardous waste
system.
Similarly, the Agency sought to be protective of public health in
developing its fate and exposure modeling. For the groundwater
evaluation, the Agency used a DAF 10 (which represents an approximate
90th percentile protection level) for infinite source type
constituents. (Constituent-specific DAFs were developed using the same
input assumptions, and different DAFs result from modeling of
degradation or retardation factors in the environment). This is the
generic DAF used in the delisting program for large volume wastes.
Since this is a national program which will largely benefit the largest
volume generators, the DAF 10 assumption is consistent with delisting
practice. Also, the toxicity characteristic used a DAF of 100
(representing an approximate 85th percentile protection level) for
identifying clearly hazardous waste (for infinite source type
constituents; regulation of hydrolysers was deferred). Again the policy
goal of exits was to strive to be well below clearly hazardous levels.
The Agency also modeled exposure at the nearest downgradient well. The
TC rule restricted well placement to within the plume. Today's proposal
attempts to balance the protectiveness level and well placement by
requiring a more protective level than the TC rule, but is less
restrictive in well location, e.g., wells outside of the plume, at
significantly lower risk, are averaged in.
For modeling of the non-groundwater pathways, the Agency used four
high-end parameter values for which the modeling outcome is most
sensitive as inputs to the analysis to be protective of public health
and the environment. These include: Two high-end parameters in the
waste management unit characterization and fate portions and two highend
parameters in the exposure portions of the model. The remaining
input parameters were evaluated at typical values or central tendency
values. The Agency sought to be protective of a high percentile exposed
population (at least 90th percentile).
1. Human Health Benchmarks
For each constituent on the master list, the Agency evaluated the
existing toxicity information to determine whether there were
sufficient toxicity data to establish a benchmark. For those
constituents with adequate data, the data were evaluated either by the
Agency's CRAVE (Carcinogen Risk Assessment Verification Endeavor)
Workgroup, Reference Dose/Reference Concentration (RfD/RfC) Workgroup,
or the Office of Research and Development. This approach is consistent
with the approach used in the Agency's other risk-based RCRA programs
such as the Toxicity Characteristic, delisting petition evaluations,
listings, as well as the CERCLA program. See Section 4, ''Benchmarks,''
of the Technical Support Document for the Hazardous Waste
Identification Rule: Risk Assessment for Human and Ecological Receptors
for more details.
a. Non-carcinogens
The Agency proposes to use oral reference doses (RfDs) and
inhalation reference concentrations (RfCs) as the basis for developing
the exit criteria for non-carcinogenic constituents. An RfD or RfC is
an estimate (with uncertainty spanning perhaps an order of magnitude)
of a daily exposure to a constituent for the human population
(including sensitive subgroups) that is likely to be without an
appreciable risk of deleterious effects during a lifetime.
The approach used to derive an RfD or RfC is to identify the
highest test dose of a constituent associated with no effects or
effects that are not considered adverse in an appropriate animal
bioassay test. These experimental no-observed-adverse-effect-levels
(NOAELs) or no-observed-effect-levels (NOELs) are considered to be an
estimate of the animal population's physiological threshold for adverse
effects. The RfD or RfC is derived by dividing the NOAEL or other
toxicity benchmark by suitable uncertainty and modifying factors. In
the event that an appropriate NOAEL or NOEL is not available, the
lowest-observed-adverse-effect level (LOAEL) may be used with
additional uncertainty factors.
It is important to note that the contributions of the constituent
from various sources in the environment (e.g., air, food, water) are
not considered in the development of an RfD or RfC. Rather, the RfD or
RfC reflects the estimated total permissible daily human exposure from
all sources of exposure. RfDs and RfCs have been calculated for many,
but not all, of the non-carcinogenic constituents for which the Agency
is establishing exit criteria.
The Agency prefers to use only RfDs and RfCs that have been
evaluated and verified by the RfD/RfC Workgroup as the basis for
setting regulatory levels. However, for some constituents, the Agency
has not yet completed its verification process; thus, RfDs and RfCs
under development are being used for purposes of this proposal for
those constituents. If the final verified RfDs and RfCs differ from the
RfDs and RfCs under development proposed in today's notice, the Agency
will adopt the new (i.e., verified) values for the final rule after
noticing the data in the Federal Register.
b. Carcinogens
The Agency proposes to use the oral cancer slope factor and
inhalation cancer unit risk as the basis for developing exit levels for
carcinogenic constituents unless the non-carcinogenic effects occur at
lower levels. EPA's CRAVE Workgroup and Office of Research and
Development have estimated the carcinogenic slope factor (CSF) (i.e.,
the slope of the ''dose-response'' curve) and inhalation unit risks for
humans exposed to low-dose levels of carcinogens in the environment.
The slope factors indicate the upper-bound confidence limit estimate of
excess cancer risk for individuals experiencing a given exposure over a
70-year lifetime. In practice, a given dose multiplied by the slope
factor gives an upper estimate of the lifetime risk to an individual of
developing cancer. By specifying a level of lifetime risk (no matter
how small), one can also estimate the corresponding dose using the
slope factor.
EPA proposes to quantify on a weight-of-evidence basis, as
described below. EPA promulgated ''Guidelines for Carcinogen Risk
Assessment'' on September 24, 1986 (51 FR 33992), which defined a
scheme to characterize substances based on experimental data and the
kinds of responses induced by a suspect carcinogen. These guidelines
specify the following five classifications:
Group A--Human carcinogen (sufficient evidence from epidemiologic
studies)
Group B--Probable human carcinogen
Group B1--Limited evidence of carcinogenicity in humans
Group B2--A combination of sufficient evidence in animals and
inadequate or no evidence in humans
Group C--Possible human carcinogen (limited evidence of carcinogenicity
in the absence of human data)
Group D--Not classifiable as to human carcinogenicity (inadequate human
and animal evidence of carcinogenicity or no data available)
Group E--Evidence of non-carcinogenicity for humans (no evidence of
carcinogenicity in at least two adequate animal tests in different
species or in both adequate epidemiologic and animal studies).
The weight-of-evidence basis was used to eliminate Group D and E
constituents from further consideration as carcinogens.
[[Page 66353]]
Under each of the regulatory options presented in today's proposal,
the Agency is using the same risk level for Groups A, B, and C
carcinogens. This approach is consistent with the way carcinogens were
treated in the 1990 Toxicity Characteristic rule, hazardous waste
listing determinations, and the delisting program. The rationale for
this approach is that while the classifications indicate the type
(human or animal) and strength of the studies available which reflects
upon the uncertainty about the carcinogenic potential, the severity of
the effect, cancer, warrants equal treatment. It is important to note
that a few Group C carcinogens do not have slope factors or unit risks.
In these cases the Agency used the benchmark developed for the noncancer
endpoint.
c. Consideration of MCLs
The Agency is proposing two approaches for setting human healthbased
levels for carcinogens and non-carcinogens in routes of exposure
involving water ingestion. For the first approach, the Agency is
proposing to use Maximum Contaminant Levels (MCLs) promulgated under
the Safe Drinking Water Act (SDWA) of 1974, as amended in 1986, as the
human health-based levels for the constituents for which they have been
established. In general, MCLs for non-carcinogens are derived from the
Reference Doses (RfDs), while MCLs for most carcinogens are set as
close to zero as technically and economically feasible; this normally
corresponds to risk levels that range from 10-4 to 10-6.
(Note that, although the derivation of MCLs considers feasibility of
treatment, analytic chemistry, and cost factors in addition to health
effects, it also considers other routes of exposure. The Agency's
policy has been to use MCLs, when available, in other similar
concentration-based programs.) For those constituents which do not yet
have MCLs, the Agency is proposing to use oral reference doses (RfDs)
for non-carcinogens and oral slope factors for carcinogens as described
above. However, if new MCLs are finalized under the SDWA prior to the
promulgation of today's rule, the Agency proposes to substitute the new
MCLs for the RfDs and slope factor-derived human health-based levels
for water ingestion presented in today's notice.
For the second approach, the Agency intends to propose to use only
RfDs and slope factors in deriving human health-based levels for water
ingestion. The Agency requests comment on these two approaches.
2. Ecological Benchmarks
Ecological benchmarks were developed for a variety of ecological
receptors based on the availability of data. Benchmarks were needed for
mammals, birds, plants, soil fauna, fish, aquatic invertebrates,
aquatic plants, and benthos (sediment-dwelling organisms). A much
smaller number of constituents have been evaluated by the Agency for
ecological effects than have been for human health effects, as
discussed under V.A. In general, measurement endpoints were selected:
(1) For consistency with the Agency's Framework for Ecological Risk
Assessment (U.S. EPA 1992x), the Great Lakes Initiative, and other
ecological efforts within the Agency, and (2) relevance to the
ecological receptor. As discussed in ''Section D--Risk Assessment'' the
ecological assessment focussed on inferring the sustainability of
populations and communities within ecosystems. Therefore, benchmarks
were derived from measurement endpoints (i.e., reproductive,
developmental, growth, survival, and mortality) from which such
inferences could be made. Reproductive studies (e.g., number of viable
young per female) were preferred over other endpoints. For some
constituents, acute or mortality studies were used, however, this
occurred only for developing benchmarks for fish, aquatic
invertebrates, and benthos where protocol exists (AWQC development) for
using such data. The Agency seeks comment on the measurement endpoints
selected for each ecological receptor.
The toxicological benchmarks were established using the more
conservative no effects level (or concentration) approach for
ecological receptors as compared to a 20% effects level. The 20%
effects level is the lowest level for ecological effects that can be
detected in field population analyses (Suter et al., 1992). Although
the 20% effects level may indeed be the lower limit that could be
reliably confirmed in field studies, this level reflects our current
analytical abilities and not necessarily the ecological significance of
the effects level. The no effects approach was taken because the
ecological analysis infers the sustainability of various populations
under the assumption that if a sufficient number of populations within
an ecosystem is protected, then the likelihood of adverse effects that
are causally related to the chemical stressor will be reduced at the
ecosystem level. The Agency was concerned that if an effects approach
was taken, then the assumption underlying the ecological analysis would
no longer be valid. The Agency seeks comment on the approach taken for
setting toxicological benchmarks.
Given the number and variety of ecological receptors included in
the analysis (predatory birds to soil fauna) as well as the variety of
effects and endpoints considered, the benchmark development process
required an approach that was internally consistent and acknowledged,
at least qualitatively, the uncertainty involved in estimating
ecological benchmarks. The Agency, therefore, developed a benchmark
classification scheme to incorporate both the relationship of the
benchmark to the entire toxicity data set and the adequacy of the
database used to derive the benchmark. Three classifications were
established: Adequate, provisional, and interim. These classifications
were developed on a receptor group-specific basis (i.e., fish and
aquatic invertebrates, benthos, mammals, birds, soil fauna, and
terrestrial plants) and represent a weight-of-evidence designation for
the toxicological benchmark. In many respects, this classification
scheme is similar in meaning to the human carcinogen weight-of-evidence
groups and the difference between ''verified'' values on IRIS and
''unverified'' values in HEAST. The classifications relate to the
certainty assigned to a given ecological benchmark. The benchmarks were
treated the same in the analysis regardless of classification. See
Section 4 in the ''Technical Support Document for the Hazardous Waste
Identification Rule: Risk Assessment for Human and Ecological
Receptors'' for details on each classification and how they were used
for each ecological receptor group. The Agency seeks comment on the
classification developed for the analysis.
Below is a discussion of how benchmarks were developed for each of
the receptor groups. For a detailed discussion of each of their
developments, see Section 4, ''Benchmarks,'' and Appendix B,
''Toxicological Profiles for Ecological Receptors,'' of the ''Technical
Support Document for the Hazardous Waste Identification Rule: Risk
Assessment for Human and Ecological Receptors.'' The Agency seeks
comment on the overall development of each of the ecological benchmarks
generated for this proposed rule.
For populations of birds and mammals, the overall approach used to
establish toxicological benchmarks was similar to the methods used to
establish RfDs for humans as described in IRIS. Each method uses a
hierarchy for the selection of toxicity data (e.g., no effects
[[Page 66354]]
levels are generally preferred to lowest effects levels) and
extrapolates from a toxicity benchmark for the test species to a
toxicity benchmark for the desired species. However, the procedures
used to develop benchmarks (i.e., RfDs) for the protection for human
health establish an acceptable daily dose for all individuals
(including sensitive sub-populations) while the development of
ecological benchmarks for this analysis establish a level that will
sustain the reproductive fitness in a local population. Consequently,
benchmarks for birds and mammals were established using three key
guidelines. First, because the reproducing population was selected as
the assessment endpoint, the benchmarks were developed from measures of
reproductive success or, if unavailable, other effects that could
conceivably impair the maintenance of the population.
Second, the taxon of the test species was matched to the taxon of
the wildlife species to the greatest extent possible. The evolutionary
processes that result in obvious differences in taxa (e.g., morphology)
also result in differences in the physiological processes that govern
chemical response. Moreover, taxonomic similarities are generally
associated with similarities in feeding habits, physiology, and
chemical sensitivity at the family classification and, to a lesser
extent, the order classification. For example, herbivores are generally
more resistant to toxicants than predators because they are exposed to
plant toxins, and the enzymatic system that detoxifies plant toxins
also detoxifies pesticides and other organic chemicals.
Third, a default safety factor of 10 was adopted only for
extrapolating from an lowest-observed-effects level (LOEL) to a noeffects
level (NOEL). A ten-fold safety factor was not applied to subchronic
studies since reproductive and developmental toxicity studies
are frequently short-term. Even among target organ toxicity studies,
there are many instances where sub-chronic studies are actually more
sensitive than chronic studies carried out on the same substance. Also,
for mammals and birds, differences in interspecies uncertainty were
indirectly addressed through the use of the species-scaling equation
described in Section 4 of the ''Technical Support Document for the
Hazardous Waste Identification Rule: Risk Assessment for Human and
Ecological Receptors.'' The Agency requests comment on the use a safety
factor of 10 when extrapolating from a LOEL to a NOEL. The Agency also
requests comment on the use of a scaling approach to address
interspecies uncertainty as described above. Furthermore, the Agency
seeks comment on the inability of the Risk Assessment to evaluate the
inhalation and dermal routes of exposure for birds and mammals.
For the terrestrial plants, the approach used to establish
toxicological benchmarks was adapted from the Effects Range Low (ER-L)
approach developed by the National Oceanographic and Atmospheric
Administration (NOAA). The NOAA ER-L approach estimates a percentile of
the distribution of various toxic effects thresholds. The measurement
endpoints were generally limited to growth and yield parameters because
(1) they are the most common class of response reported in
phytotoxicity studies and, therefore, will allow for benchmark
calculations for a large number of constituents, and (2) they are
ecologically significant responses both in terms of plant populations
and, by extension, the ability of producers to support higher trophic
levels. It should be noted that these benchmarks were limited to soil
concentrations and do not explicitly consider the adverse impacts on
plants from ambient contaminant concentrations in the air. Further
details can be found in section 4.3.3 of the ''Technical Support
Document for the Hazardous Waste Identification Rule: Risk Assessment
for Human and Ecological Receptors.'' The Agency solicits comment on
the overall approach taken to develop benchmarks for the terrestrial
plant community.
For the soil fauna, the toxicological benchmarks were established
based on methods developed by the Dutch National Institute of Public
Health and Environmental Protection (RIVM). The RIVM approach estimates
a confidence interval containing the concentration at which the no
observed effects concentration (NOEC) for p percent (95th percentile
was selected) of the species within the community is not exceeded 50%
of the time. A minimum data set was established in which key structural
and functional components of the soil community (e.g., decomposer and
grazing organisms) encompassing different sizes of organisms (i.e.,
microfauna, mesofauna, macrofauna) were represented. As with the
Ambient Water Quality Criteria, measurement endpoints included
reproductive effects as well as measures of growth, survival,
mortality. The Agency requests comment on the use of the RIVM
methodology, and protecting 95 percent of the community 50 percent of
the time. The Agency also requests comment on its inability to fully
quantify the effect of soil characteristics on toxicity of constituents
to soil organisms.
For populations of fish and aquatic invertebrates (represented by
daphnids), a hierarchical approach was taken for use of data sources in
deriving benchmarks. The first choice was final chronic values (FCVs)
from the Sediment Quality Criteria effort by the EPA Office of Water,
followed by values from the Great Lakes Initiative (GLI) effort, and
finally, the Ambient Water Quality Criteria (AWQC). If these benchmarks
were not available, then a benchmark was developed using AWQC
procedures or, if data were inadequate, the GLI Tier II procedures for
establishing chronic values (termed secondary chronic values--SCVs).
The AWQC ranked third since many years have passed since their
establishment and the SQC and GLI efforts re-evaluated the toxicity
data sets of several of these. The Agency solicits comment on the
hierarchical approach described above for deriving toxicity benchmarks.
For aquatic plants, the approach used to establish toxicological
benchmarks was adapted from the ER-L approach developed by NOAA. The
NOAA ER-L approach estimates a percentile of the distribution of
various toxic effects thresholds. However, due to the general lack of
toxicity data, the default ER-L approach was used wherein the lowest
LOEC for either vascular plants or algae was used. The Agency solicits
comment on the overall approach taken to develop benchmarks for aquatic
plants.
For the sediment organisms, the approach used to establish
toxicological benchmarks for non-ionic, hydrophobic organic chemicals
was based on sediment quality criteria methods for non-ionic
constituents. Two key assumptions form the basis for the proposed
sediment quality criteria. First, benthic species, defined as either
epibenthic or infaunal species, have a similar toxicological
sensitivity as water column species. As a result, FCVs (or SCVs)
developed for the fish and aquatic invertebrates can be used for the
benthic community. Second, pore water and sediment carbon are assumed
to be in equilibrium and the concentrations are related by a partition
coefficient, Koc. This assumption, described as equilibrium
partitioning (EqP), provides the rationale for the equality of wateronly
and sediment-exposure-effects concentrations on a pore water
basis: The sediment-pore water equilibrium system results in the same
effects as a water-only exposure. The Agency requests comment on the
use of this approach in support of today's proposal. In some cases,
protecting these
[[Page 66355]]
ecological receptors represents the critical pathway that limits the
projected exit level for management of a waste stream outside of the
Subtitle C hazardous waste program. These ecological receptors serve as
the basis for the proposed exit levels for 18 constituents, including 6
metals. To the extent that contaminants from these waste streams reach
off site areas, the Agency based its proposal on modeling the
ecological receptors on a neighboring land area of 500 acres or an
adjacent stream (with a total length of 12 miles). This approach as
currently modeled, may only serve as an indicator of a potential nearby
threat to ecological receptors (e.g., the soil fauna and plant life),
rather than serving as a measure or indicator of a broader threat to
the environment. The Agency solicits comment on the appropriateness and
relevance of these receptors as the basis for exit levels under the
HWIR program.
3. Sources of Data
a. Human
The two primary sources used to identify human health benchmarks
were the Integrated Risk Information System (IRIS) and the Health
Effects Assessment Summary Tables (HEAST). Both of these sources were
developed and are maintained by the USEPA. For a few constituents,
other Agency sources such as Carcinogen Assessment Group (CAG)
profiles, Health Effect Assessments (HEAs), and Health Assessment
Documents (HADs) were used to fill data gaps.
IRIS is the Agency's official repository of Agency-wide consensus
chronic human health risk information. IRIS evaluation are conducted by
the Agency's Work Group review process that leads to internal Agency
scientific consensus regarding risk assessment information on a
chemical. This information is recorded on IRIS and is considered to be
''Work Group Verified.''
The HEAST is prepared by EPA's Office of Research and Development.
They contain risk assessment information on chemicals that have
undergone a more limited review and have the concurrence of individual
Agency program offices; each is supported by an Agency reference. The
information has not, however, had enough review to be recognized as
Agency-wide consensus information.
b. Ecological
A thorough literature review was conducted to identify
toxicological data from laboratory and field studies for each of the
constituents of ecological concern. The review included secondary
sources such as the Synoptic Review Series published by the U.S. Fish
and Wildlife Service, the Ambient Water Quality Criteria documents, and
other Federal compendia of toxicity data (e.g. HEAs, the Derivation of
Proposed Human Health and Wildlife Bioaccumulation Factors for the
Great Lakes Initiative, Agency for Toxic Substances and Disease
Registry documents, PHYTOTOX, GRIN, TERRETOX, and AQUIRE). Toxicity
data on soil organisms were obtained for several constituents from van
de Meent et al. (1990). In addition to AQUIRE, the other primary data
source for toxicity data on aquatic plants were the Toxicological
Benchmarks for Screening Potential Contaminants of Concern for Effects
on Aquatic Biota:1994 Revision (Suter and Mabrey, 1994). On-line
literature searches were conducted to identify primary sources of
toxicity data on constituents lacking sufficient data in the secondary
sources. Additional studies were identified in conventional literature
reviews.
E. Risk Assessment
1. The Non-groundwater Risk Assessment
a. Introduction
The risk assessment underlying today's proposed rule is based upon
a comprehensive approach to evaluating the movement of many different
waste constituents from their waste management units, through different
routes of exposure or pathways, to the points where human and
ecological receptors are potentially exposed to these constituents.
This risk assessment is being used in today's proposed rule to
determine which listed hazardous wastes can be defined as ''low-risk''
wastes, able to exit the Subtitle C system and be managed in nonSubtitle
C units. The previous approach taken in the May 20, 1992,
proposed HWIR rule also addressed the risks associated with the
management of wastes containing hazardous constituents with very
diverse physical and chemical properties; however, only groundwater
ingestion exposures from landfill units were evaluated. That approach
led to a concern by the Agency, as well as commenters on the proposed
rule, that leachate from landfills contaminating groundwater and
subsequent consumption of the contaminated groundwater by humans may
not be the only exposure pathway important to evaluate. Although the
ingestion of contaminated groundwater pathway may be appropriate to
propose exit levels for some wastes and constituents, it may be underprotective
for others, depending on the physical and chemical
properties of each waste constituent. (For example, some constituents
have a high potential to bioaccumulate or bioconcentrate in living
organisms. Pathways in which these constituents come in contact with
fish, grazing livestock, wildlife, or edible plants would be important
to evaluate.) In addition, over the past 14 years of implementing the
RCRA program, the Agency has learned more about potential routes of
release to the environment from various management practices.
Therefore, for today's proposal the Agency undertook an extensive
risk assessment that examines numerous exposure pathways, rather than
just the groundwater ingestion pathway. In selecting the exposure
pathways, previous rulemakings were used as a guide, as well as other
special studies by the Agency that implement analyses examining
numerous pathways. (Tables A-1 and A-2 contain the human and ecological
pathways, respectively, evaluated in the assessment, and are presented
in appendix A to today's preamble.) With regard to waste management
units considered in the assessment, it is important to note that
because today's proposal establishes criteria for waste to exit the
Subtitle C system, the assessment evaluated exposures associated with
managing wastes in non-Subtitle C units. The human and ecological
receptors considered in the assessment were selected to represent a
range of behaviors, activities, dietary habits, and trophic levels that
influence exposure levels.
The risk assessment supporting this proposal is currently
undergoing review by the Science Advisory Board and EPA's Office of
Research and Development. As a result of these reviews, and of comments
received during the public comment period, it is likely that EPA would
make changes to the risk assessment or other parts of the rule. Topics
on which the Agency has received informal comment include the use of
ecological benchmarks for regulation and the overland transport of
waste constituents. The Agency, to the extent consistent with the
schedule negotiated in the consent decree for this rulemaking, would
publish a supplemental notice proposing any significant changes before
finalizing the rule.
b. How the Assessment is Structured
The non-groundwater assessment acknowledges that not all human and
ecological pathways arise from each
[[Page 66356]]
source; for example, movement of particles from an active surface
impoundment is not expected to occur. To account for this, the
assessment matched the environmental transport pathways with both the
releases from various types of waste management units and the various
receptors for the nearly 200 constituents examined. All constituents
were assessed in all pathways deemed plausible for a given waste
management unit, if the data permitted. Tables A-3, A-4, and A-5 of
appendix A show the pathways assessed for each waste management unit,
human receptors assessed for each pathway, and ecological receptors
assessed for each pathway, respectively. The assessment estimated the
constituent-specific concentrations in a waste at the management unit
that could be expected to result in an acceptable exposure for a human
or ecological receptor (determined through using the toxicity
benchmarks discussed in section V.B.), taking into account the various
pathways by which the constituent may move through the environment from
the waste management unit to the receptor.
The waste management units considered in the assessment are not
all-inclusive but were selected to reflect those that might be commonly
associated with the management of exited hazardous wastes (from
wastewaters to nonwastewaters) in non-Subtitle C waste management
units. These units were identified as commonly used in the management
of solid wastes in the 1988 Report to Congress entitled Solid Waste
Disposal in the United States Report. The Agency believes that risks
posed by other types of management of these exited wastes will be no
greater than those from the units assessed.
There is a high degree of variability in the physical and chemical
properties of the approximately 200 constituents evaluated. An
understanding of those properties and how they interact with the
physical and chemical properties that control persistence and mobility
in the environment is an essential element of the assessment. The
management units could potentially be located in the range of
environments that exist across the United States. These environments
have differing characteristics (e.g., meteorological conditions, soil
type) that are more conducive for the movement of certain constituents
in certain pathways than others. For example, an environment with a
high precipitation and high organic soil content may result in
significant exposures to fishers by constituents that readily adsorb to
soils (i.e., have a high log Kow) through erosion of contaminated
soil and uptake in the food chain. For other pathways, however, an
environment with these characteristics may result in relatively low
exposures. The assessment was designed to determine what conditions
would need to exist to cause higher exposures for each pathway rather
than developing a scenario and determining all the types of exposures
and receptors for that scenario. By determining the appropriate
conditions for which higher exposures from a given pathway will occur,
the Agency believes that environments where the conditions are not as
likely for a constituent to move through a pathway are protected.
The assessment was structured using a deterministic approach. A
deterministic approach uses a single, point estimate of the value of
each input or parameter and calculates a single result based on those
point estimates. The assessment used the best data available to select
typical (i.e., approximately 50th percentile) and high-end (i.e.,
approximately 90th percentile) values for each parameter or parameter
group as discussed in Section E.2. below. Sometimes full distributions
were available but, more commonly, ranges of values or point values
were available with no description of distributions or variability. If
there was not a sufficient distribution for the parameter, best
professional judgement was used in determining typical and high-end
values (which sometimes would be the maximum).
The assessment is constructed as a set of calculations that begin
with an acceptable exposure level for a constituent at a receptor, and
back-calculates to a concentration in a waste in a management unit that
corresponds to the acceptable exposure level. For the human receptors,
the assessment was designed to determine constituent concentrations in
waste for each waste management unit that would correspond to
protecting receptors at the high-end of exposure (i.e., above the 90th
percentile of each of the receptor populations and types of exposures
being assessed). The Agency estimated waste concentrations
corresponding to the high-end exposure by identifying four critical or
sensitive parameters in the source/pathway/receptor equations and using
high-end input values for those parameters and using central tendency
values for the remaining parameters. The Agency also estimated central
tendency (approximately the 50th percentile) and bounding estimates
(worst-case) of constituent concentrations in waste for each of the
receptor populations and types of exposures being assessed. For
ecological receptors, the approximate percentile level of protection is
difficult to discern. The Agency believes the ecological analysis is
conservative with respect to the overall assessment endpoint (e.g.,
sustainability of the reproducing populations) because of the way the
source, fate and transport parameters are set, the dietary habits
assumed, and how the toxicity benchmarks are developed. However, the
degree to which this conservativeness transfers to ecosystems is not
known.
The steps of the assessment which provide estimates of acceptable
constituent-specific concentrations in waste include the following:
Step 1--Specify acceptable risk levels for each constituent and
each receptor. See Section V.B. in today's preamble for a discussion of
how benchmarks are set for both human and ecological receptors.
Step 2--Specify the exposure medium. Using the toxicity benchmarks
as a starting point and the exposure equations, the assessment backcalculates
the concentration of contaminant in the medium (e.g., beef,
milk, plant, air, water, soil) that corresponds to the ''acceptable''
exposure level. The exposure equations include a quantitative
description of how a receptor comes into contact with the contaminant
and how much the receptor takes in through specific mechanisms (e.g.,
ingestion, inhalation, dermal adsorption) over some specified period of
time. Thus, for the subsistence farmer eating contaminated beef, the
exposure specifies the amount of beef eaten on a daily basis, the
period of time over which the contaminated beef is eaten, and
descriptions for the individual such as body weight and lifetime. For
this example, the concentration in the beef is what is back-calculated.
Step 3--Calculate the point of release concentration from the
exposure concentration. Based on the back-calculated concentration in
the exposure medium (from Step 2), the concentration in the medium to
which the contaminant is released to the environment (i.e., air, soil,
groundwater) for each pathway/receptor was modeled. The end result of
this calculation is a medium concentration at the point of release from
the waste management unit.
Step 4--Calculate the concentration in the waste that corresponds
to the medium concentration at the point of release. This step depends
on the characteristics (e.g., area, cover practices, waste consistency)
of the waste management unit.
[[Page 66357]]
The output of the assessment is a range of constituent
concentrations, reflecting the range of pathway-receptor combinations
considered for each waste management unit. The lowest concentration
(per constituent) of this range represents the highest exposure
pathway-receptor combination for that waste management unit.
c. How Uncertainty is Addressed
Any analysis of the magnitude used in this rule-making will have
uncertainty associated with the outputs generated. The uncertainty can
be associated with the models or equations used and the data relied on
for the model parameters. In addition, policy assumptions, such as
waste management units assessed and receptors assessed, may also affect
the degree of representativeness of the assessment. In order to be
consistent with Agency policy on the characterization of risk,
stochastic and deterministic approaches were considered. A stochastic
approach, such as Monte Carlo analysis, which produces a distribution
of constituent concentrations, was initially considered due to the
tremendous interest in, and use of, these techniques in risk
assessment. However, after evaluation of the models and data available
for use, the Agency decided to use a deterministic approach for the
non-groundwater assessment.
The Agency's deterministic approach used for this assessment, like
most such approaches, uses point values in all calculations and
produced point estimates of constituent concentrations for waste in
each management unit-exposure pathway-receptor combination. However, in
selecting and developing point values for parameters, EPA considered
all available data. Wherever possible, the Agency developed both a
central-tendency and high-end value for each parameter used in the
assessment. This was not possible in all cases because some parameters
were a property, such as density of water, and because some values were
fixed by Agency-wide policy decisions. (For example, EPA used standard
Agency-wide human toxicity benchmarks and body weights.) EPA then
calculated constituent concentrations based on a mixture of centraltendency
and high-end values.
EPA believes that the deterministic approach described above (based
on identifying critical parameters and using higher-end values only for
those parameters and central-tendency values for the other parameters)
allowed it to derive constituent concentrations in waste for each waste
management unit that are reasonably protective across a range of
conditions and for a range of receptors. EPA also believes that this
approach is consistent with EPA's risk assessment policy.
EPA further believes that the approach chosen allows both the
Agency and the public to determine more easily which parameters played
the most critical roles in determining the constituent concentrations
in waste for each waste management unit. This furthers general
understanding of the assessment and helps commenters effectively target
their resources for reviewing what EPA is proposing. It has also helped
EPA target its own data collection and input selection efforts. It is
often more difficult to identify critical parameters in a stochastic
assessment because of the greater number of iterations and because
results are reported as probability distributions. This is particularly
true for an analysis with a large number of parameters such as the
assessment used for this proposed rule.
EPA notes that stochastic approaches are also consistent with
Agency risk assessment policy. In fact, EPA applied a stochastic
''Monte Carlo'' approach to the separate analysis of dilution and
attenuation of groundwater performed for this proposal. That analysis,
however, has been under development for many years and EPA is more
familiar with the underlying data and the relationships between various
parameters. In addition, the public has had a chance to comment on
aspects of that analysis in previous rule-makings. EPA was more
comfortable applying a stochastic analysis for the groundwater analysis
than a stochastic approach to the non-groundwater analysis.
EPA believes that it is not necessary to resolve all issues
relating to the relative merits of the two approaches or to determine
which approach would be ideal for each of the assessments described
above. Rather, the debate should focus on whether the approaches chosen
allowed EPA to reach reasonable regulatory decisions.
The Agency solicits comment on the use of a deterministic approach
as described above. Specifically, the Agency seeks comment on whether
the approach proposed is a reasonable approach for setting protective
levels across a set of types of management units and exposure pathways.
d. Linkage of the Non-groundwater Risk Assessment to the Groundwater
Risk Assessment
In the non-groundwater risk assessment, the pathways involving
potentially contaminated groundwater (e.g., bathing) are backcalculated
from the receptor to the wellhead (i.e., the assessment
provides constituent concentrations in the groundwater at the well). In
order to determine the concentration of a constituent in leachate
coming from a waste management unit that would result in the estimated
constituent concentration at the water well, the Agency used a separate
groundwater fate and transport risk analysis. That analysis is
described in detail in Section D.8. elsewhere in today's proposal. The
well concentrations estimated from the pathways involving bathing are
used as input to the groundwater fate and transport modeling from which
a leachate concentration is determined.
e. Risk Targets Used
As previously discussed in Section V.B. of today's proposed rule,
the Agency used existing toxicity benchmarks when available. However,
many ecological benchmarks were developed for this rule-making, as
discussed in Section V.B. of today's proposed rule. As described in
that section, the Agency used a cancer risk target of 1 x 10-6,
and a hazard quotient equal to 1 for non-carcinogens. For ecological
benchmarks, a hazard quotient equal to 1 was used. The Agency solicits
comment on the risk targets being used for today's proposed rule.
2. Detailed Overview of the Non-groundwater Risk Analysis
The assessment can be broken down into six components:
Constituents; toxicity benchmarks; receptors; exposure; fate and
transport; and waste management units. Each of these components is
discussed in turn below, except the constituents and toxicity
benchmarks which were discussed earlier in section V.A and V.B. It is
important to recognize that the assessment was not able to evaluate all
constituents in all receptor-pathway-waste management unit combinations
because of data gaps in either toxicity or chemical properties, or
inadequate methodologies. Many of these gaps have been identified in
different sections of the Technical Support Document for the Hazardous
Waste Identification Rule: Risk Assessment for Human and Ecological
Receptors'' (denoted ''Uncertainties and Issues of Concern''). The
Agency requests additional data or other information that would assist
in filling these gaps.
a. Waste Management Units
The manner in which constituents are released to environmental
media and the relative quantity released to each
[[Page 66358]]
medium will affect the pathways of most concern for a particular
constituent. The pathway presenting the highest risk to human or
ecological receptors is not always easily determined because of the
complex interactions of the waste management unit and its types of
releases, the physical and chemical properties of the constituent, and
the properties that control mobility and persistence in a particular
environmental medium. For some constituents, the management practice
will determine which exposure pathway is of most concern. For example,
benzene tends to migrate to both air and groundwater. Upon examining
the risks from exposure to these two media arising from releases from a
quiescent surface impoundment, the groundwater ingestion pathway may
pose the highest risks. But, when examining the risks from these two
media for releases from an aerated tank, the air inhalation pathway may
pose the higher risks. Further, the air inhalation risks may even be
higher than groundwater ingestion risks from the quiescent surface
impoundment.
Therefore, and as stated earlier, the selection of non-Subtitle C
waste management units examined in the assessment attempted to reflect
both the influence of the type of unit on pathways and those that might
be commonly associated with the management of exited hazardous wastes
in non-Subtitle C waste management units. Again, the Agency believes
that risks posed by other types of management of these exited wastes
will be no greater than those from the units assessed. The management
units examined include the following:
• Aerated treatment tanks. Relative to all other types of
management, aerated tanks containing wastewaters can potentially have
the most significant releases of volatile organics to air.
• Quiescent surface impoundments. This type of unit
containing wastewaters also can potentially result in significant
releases of volatile organic constituents to air. These units also have
a potential to affect surface water bodies if the unit is not well
maintained or constructed. The sludges generated, which may contain
high concentrations of metals and hydrophobic constituents, may impact
groundwater. (As discussed above, the groundwater fate and transport
analysis was conducted in a separate analysis.)
• Land application. This type of unit, when used for nonwastewaters
can potentially have significant releases of certain
constituents to nearby land and surface water bodies through erosion
and runoff, particularly if run-on and run-off control measures are not
practiced. In addition, significant releases of volatile organics
constituents to air are possible. Further, after the unit is closed,
significant on-site exposures to some persistent and relatively
immobile constituents may occur as well as continued long-term releases
to the nearby land and surface water bodies. The Agency believes such
units will pose higher exposures relative to landfills in all pathways
except those arising from groundwater. Therefore, the non-groundwater
assessment did not examine landfills, but they were examined in the
groundwater fate and transport analysis.
• Ash monofill. This type of unit used for ash disposal can
potentially have significant releases of particulates to air which may
be inhaled or may deposit on land and plants, and result in exposure
through food and soil ingestion.
• Wastepiles. This type of unit used for nonwastewaters can
have significant releases of particulates to air as well as significant
releases of particulates through erosion and runoff.
Each of the pathways that evaluates a receptor using contaminated
groundwater other than as a source of drinking water (i.e., bathing)
are back-calculated to a concentration in a drinking water well. The
pathways are applicable to all of the waste management units modeled
(except tanks). All of the waste management unit and chemical-specific
portions of the groundwater fate and transport analysis and subsequent
estimated leachate concentrations are contained in the Agency's
separate groundwater fate and transport analysis (see Section E.3
below).
One exception to the above discussion of the types of waste
management units evaluated involves the combustion of wastes. Although
the Agency attempted to include this type of management in the
assessment, it became clear that the emissions from combustion are not
easily predicted from the waste inputs to the units. The combustion
process both destroys and creates constituents. Although destruction of
constituents can be predicted based on certain operating
characteristics of combustion units, the creation of other
constituents, referred to as products of incomplete combustion (PICs),
is not easy to predict. It may be possible to make such predictions for
a specific waste and a specific combustion unit; however, the extensive
data (e.g., on the variety of combustion units, waste types,
constituent combinations) needed for the assessment used in this
rulemaking relating wastes with emissions are not available. Therefore,
acceptable constituent levels in waste going to a combustion unit could
not be established. However, the Agency is developing emission
standards for various types of combustion units and those emission
standards may be a more appropriate vehicle for addressing combustion.
In addition, the assessment does not address accidental or
catastrophic releases, such as transportation accidents or tank
failures. The Agency determined that, although such releases are
possible, they are of low probability and non-routine and, therefore,
are not appropriate for developing exit criteria that apply to all
wastes.
The Agency has identified several specific areas giving rise to
uncertainty in the characterization of the waste management units and
for which the Agency seeks comment:
(1) Use of Subtitle D Survey.
• The Agency relied upon data from a 1987 survey of Subtitle
D facilities to characterize waste management units. That survey, used
in the 1988 Report to Congress on Solid Waste Disposal in the United
States, was designed primarily to collect estimates of the following
parameters:
• Number of establishments that manage Subtitle D wastes on
site;
• Number of establishments that manage Subtitle D wastes on
site in land application units, wastepiles, surface impoundments, or
landfills;
• Number of land application units, wastepiles, surface
impoundments, or landfills used to manage Subtitle D wastes;
• Amount of Subtitle D wastes managed on site in land
application units, wastepiles, surface impoundments, or landfills.
In addition to these parameters, data were also collected for some
other parameters, such as the area of the waste management units.
Although the survey was not designed to collect accurate estimates for
these other parameters, it is the most comprehensive date available to
characterize these other parameters. One difficulty encountered in
using these data is that the survey requested information on total area
or waste quantity for all of each type of units at a facility. The
total area or waste quantity was divided by the number of each type of
unit at the facility (number of each unit being one of the primary
parameters the survey was designed to estimate) to estimate average
unit area. Further, it is not certain how well the on-site units (which
are used routinely for wastes generated on-site) reflect the
characteristics of off-site units. Uncertainty related to the
[[Page 66359]]
representativeness of the data is important because exited wastes could
be managed in units off-site as well as on-site. The Agency seeks
comment on the use of the Subtitle D survey to characterize the waste
management units.
In evaluating the waste management unit components of the risk
assessment, the Agency made certain assumptions when data were not
available or were incomplete. A description of the waste management
unit parameters for which there was little to no data is described
below. The rationale behind these assumptions is presented (e.g.,
results of any sensitivity analyses, references to other work, etc.).
The Agency requests comment on the specific issues raised for each
management unit.
(2) Fate and Transport
Fate processes, particularly biodegradation and hydrolysis, were
accounted for only in the land application unit since that unit had
wastes applied intermittently and that unit was being examined for onsite
risks after closure (assuming human occupation of the site begins
10 years after closure occurs). Because waste is continuously applied
to the other waste management units, biodegradation and hydrolysis were
presumed to have minimal influence on the subsequent availability of
constituents to the above ground pathways. The Agency requests comment
on not considering biodegradation and hydrolysis in waste management
units other than the land application unit. The Agency also requests
comment on the appropriateness of the data and methods used to account
for the fate and transport of constituents in waste management units,
with particular emphasis on data and methods of determining
biodegradation and hydrolysis of constituents in land application
units.
(3) Ash Monofill
(i) Particle Size Distribution for Air Dispersion Modeling
A size distribution of ash particles that become airborne from an
ash monofill was not available. Therefore, a sensitivity analysis was
performed to assess the importance of the particle size distribution in
the calculation of air concentrations and deposition rates. Different
distributions were modeled reflecting a variety of assumptions for
particle size distributions between PM10 and PM30 classes. The greatest
deviation among the modeled conditions in the estimated air
concentration of PM10 was 12 percent; for the estimated deposition rate
for PM30 the greatest deviation was 59 percent. Given the uncertainties
and variabilities inherent in the assessment, these variations were
considered minor, therefore, the Agency assumed an equal distribution
of particle sizes between the two size classes used in the assessment.
(ii) Monofill Characterization
Because limited data were available to characterize hazardous waste
ash monofills, data from municipal waste ash monofills were used.
However, because ash generation rates for municipal waste incinerators
ere more than 100 times greater than ash generation rates for hazardous
waste incinerators and reuse-as-fuel combustors resulting in
significantly larger municipal monofills, EPA calculated an ash
monofill volume for this analysis based on generation rates reported in
the 1988 National Survey of Hazardous Waste Treatment, Storage,
Disposal, and Recycling Facilities, assumed bulk density of the ash,
and assumed lifetime of the monofill. The Agency is not certain that
hazardous waste monofills should be sized in the same manner as
municipal waste monofills. The Agency also assumed that each waste
monofill would accept ash from only a single combustor. Accepting
wastes from more than one combustor may underestimate monofill size.
(iii) Vehicle Traffic
The estimates of number of ash trucks per day are dependent on the
size of truck. Limited data were available on the sizes of trucks
hauling ash. These data were used to characterize a range of truck
sizes. The truck sizes may either under- or overestimate the size of
trucks actually used at hazardous waste ash monofills depending on the
representativeness of municipal waste ash truck sizes.
No data were available on other vehicular traffic; therefore, these
values were estimated, introducing additional uncertainty into the
overall amount of traffic at the ash monofill.
(iv) Emission Equations for Ash Blown From Trucks and During Spreading
and Compacting
The emission equation used for ash blown from trucks was developed
for windblown emissions from storage piles. This was adapted to trucks
by using the truck speed to estimate frequency of wind greater than 5.4
m/s. Because this equation was not derived for windblown emissions from
moving trucks, the results of its application to such emissions are
uncertain. It may over- or underestimate actual emissions of
particulates blown from trucks.
Similarly, the emission equation used for spreading and compacting
was developed for agricultural tilling. Agricultural tilling was
thought to approximate the process of spreading and compacting;
however, the use of this equation may under- or overestimate emissions
due to spreading and compacting.
(4) Land Application Unit
(i) Particle Size Distribution for Air Dispersion Modeling
A size distribution of soil particles that become airborne was not
available. The same assumption was made for soil particles as was done
for ash particles when modelling the monofill (see above). As described
above for ash particles, the Agency assumed an equal distribution of
particle sizes between the two size classes.
(ii) Area of Land Application Unit Relative to Agricultural Field
The assessment examined the impact of subsistence farming on the
land application unit beginning 10 years after closure. Based on the
distribution of sizes for land application units and agricultural
fields, the Agency selected a combination of fields such that the
central tendency land application unit (61,000 m2) is smaller in
area than the central tendency agricultural field (2,000,000 m2).
The significantly larger size of the agricultural field suggests that
the model may inappropriately average the constituent concentration
over the agricultural field. However, the Agency does not believe this
to be a significant impact on the analysis because: (1) The area of the
agricultural field is not an explicit input to the model; (2) the size
of the land application unit is large enough to support a subsistence
farmer; and (3) this pathway is driven by the assumptions for the highend
analysis. The Agency requests comment on the relationship between
the land application unit and the agricultural field.
(iii) Application Rate
The waste application rate is an important parameter in determining
the constituent's soil concentration after application. In practice,
this rate is a function of the characteristics of the waste being
applied, the characteristics of the receiving soil, the environmental
conditions, and the purposes for which the waste is being applied.
Information from the Subtitle D survey was used to calculate the rates,
since those rates
[[Page 66360]]
were not expressly requested in the survey. The rates were calculated
from the area receiving the wastes and the waste quantity applied. This
introduces uncertainty for it combines rates applicable to both
treatment of wastes and rates for specific uses (e.g., farming, mine
reclamation). To account for the potential of having application rates
be much too high for the site they are being applied to, the data on
receiving area and waste quantity applied were linked.
(iv) Waste Characteristics
Limited data were available on the characteristics of wastes being
land applied. As a result, soil values for most parameters (e.g.,
hydraulic conductivity, moisture retention index) were used to
characterize nonwastewaters. It is not known to what extent these soil
values differ from the waste properties.
(v) Depth of Contamination
Depth of contamination affects the amount of constituent available
for exposure. For the non-groundwater pathways, only constituents at
the soil surface were assumed available for each exposure pathway. The
Agency selected tilling depth as the depth of contamination available
to the non-groundwater pathways as over time, the depth of the waste
layer would increase and a portion of the mass of waste would move out
of the zone available for the surface pathways. The model kept the
depth of contaminated soil constant that was available for the surface
pathways. The Agency recognizes that the use of the tilling depth may
underestimate the depth of contamination in some cases and overestimate
it in others. Thus, the Agency requests comment on the use of tilling
depth as a surrogate for depth of contamination.
(vi) Partitioning
Releases from the land application unit were partitioned among
volatilization, evaporative losses, hydrolysis, erosion, runoff, and
leaching. Periodic application of waste was factored into the
partitioning model during the active life of the unit. Biodegradation
was factored in during both the active life and closed period. The
finite source Jury model was used to estimate volatilization emissions.
The Jury model, which models the convection of constituents caused by
the flux of water in soil, was used for evaporative losses. Runoff and
leaching losses were calculated using the soil-water partition
coefficient (Kd) to determine const ![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}