Hazardous Waste Management System; Identification and Listing of Hazardous Waste: Inorganic Chemical Manufacturing Wastes; Land Disposal Restrictions for Newly Identified Wastes; and CERCLA Hazardous Substance Designation and Reportable Quantities

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 148, 261, 268, 271, and 302

[SWH-FRL-6864-5]

RIN 2050-AE49

Hazardous Waste Management System; Identification and Listing of Hazardous Waste: Inorganic Chemical Manufacturing Wastes; Land Disposal Restrictions for Newly Identified Wastes; and CERCLA Hazardous Substance Designation and Reportable Quantities

AGENCY:

Environmental Protection Agency.

ACTION:

Proposed rule.

SUMMARY:

The Environmental Protection Agency (EPA) proposes to amend the regulations for hazardous waste management under the Resource Conservation and Recovery Act (RCRA) by listing as hazardous three wastes generated from inorganic chemical manufacturing processes. We also propose not to list as hazardous various other process wastes. This action proposes to add the toxic constituents found in the wastes to the list of constituents that serves as the basis for classifying wastes as hazardous, and to establish treatment standards for the wastes.

The effect of this proposed regulation would be to subject the wastes to stringent management and treatment standards under Subtitle C of RCRA. Additionally, this action proposes to designate the wastes proposed for listing as hazardous substances subject to the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) and to adjust the one-pound statutory reportable quantities (RQs) for some of these substances.

DATES:

EPA will accept public comments on this proposed rule until November 13, 2000. Comments postmarked after this date will be marked “late” and may not be considered. Any person may request a public hearing on this proposal by filing a request with Mr. David Bussard by September 28, 2000.

ADDRESSES:

If you wish to comment on this proposed rule, you must send an original and two copies of the comments referencing docket number F-2000-ICMP-FFFFF to: RCRA Information Center, Office of Solid Waste (5305G), U.S. Environmental Protection Agency Headquarters, 1200 Pennsylvania Avenue, NW, Washington, D.C. 20460. Hand deliveries of comments should be made to RCRA Information Center, Crystal Gateway I, First Floor, 1235 Jefferson Davis Highway, Arlington, VA.

You also may submit comments electronically by sending electronic mail through the Internet to:

rcradocket@epamail.epa.gov.

You should identify comments in electronic format with the docket number F-2000-ICMP-FFFFF. You must submit all electronic comments as an ASCII (text) file, avoiding the use of special characters and any form of encryption.

Address requests for a hearing to Mr. David Bussard at: Office of Solid Waste, Hazardous Waste Identification Division (5304W), U.S. Environmental Protection Agency, 1200 Pennsylvania Avenue, NW, Washington, D.C. 20460, (703) 308-8880.

FOR FURTHER INFORMATION CONTACT:

For general information, contact the RCRA/Superfund Hotline at (800) 424-9346 or TDD (800) 553-7672 (hearing impaired). In the Washington, D.C., metropolitan area, call (703) 920-9810 or TDD (703) 412-3323. For specific aspects of the rule, contact Ms. Gwen DiPietro, Office of Solid Waste (5304W), U.S. Environmental Protection Agency, 1200 Pennsylvania Avenue, NW, Washington, D.C., 20460. [E-mail addressee and telephone number:

dipietro.gwen@epa.gov

(703-308-8285).] For technical information on the CERCLA aspects of this rule, contact Ms. Lynn Beasley, Office of Emergency and Remedial Response, Analytical Operations and Data Quality Center (5204G), U.S. Environmental Protection Agency, 1200 Pennsylvania Avenue, NW, Washington, D.C. 20460, [E-mail address and telephone number:

beasley.lynn@epa.gov

(703-603-9086).]

SUPPLEMENTARY INFORMATION:

If you do not submit comments electronically, we ask you to voluntarily submit one additional copy of your comments on labeled personal computer diskettes in ASCII (text) format or a word processing format that can be converted to ASCII (text). It is essential to specify on the disk label the word processing software and version/edition as well as your name. This will allow us to convert the comments into one of the word processing formats we utilize. Please use mailing envelopes designed to physically protect the submitted diskettes. We emphasize that submission of comments on diskettes is not mandatory nor will it result in any advantage or disadvantage to any commenter.

You should not submit electronically any confidential business information (CBI). You must submit an original and two copies of CBI under separate cover to: RCRA CBI Document Control Officer, Office of Solid Waste (5305W), U.S. EPA, 1200 Pennsylvania Avenue, NW, Washington, D.C. 20460. Any CBI data should be specifically and clearly marked. In addition, please submit a non-CBI version of your comments for inclusion in the public record.

Supporting documents in the docket for this proposal are also available in electronic format on the Internet: <

http://www.epa.gov/epaoswer/hazwaste/id/inorchem/pr2000.htm

>.

We will keep the official record for this action in paper form. Accordingly, we will transfer all comments received electronically into paper form and place them in the official record, which also will include all comments submitted directly in writing. The official record is the paper record maintained at the RCRA Information Center, also referred to as the Docket.

Our responses to comments, whether the comments are written or electronic, will be in a notice in the

Federal Register

or in a response to comments document placed in the official record for this rulemaking. We will not immediately reply to commenters electronically other than to seek clarification of electronic comments that may be corrupted in transmission or during conversion to paper form, as discussed above.

You may view public comments and supporting materials in the RCRA Information Center (RIC), located at Crystal Gateway I, First Floor, 1235 Jefferson Davis Highway, Arlington, VA. The RIC is open from 9 a.m. to 4 p.m., Monday through Friday, excluding federal holidays. To review docket materials, we recommend that you make an appointment by calling 703-603-9230. You may copy a maximum of 100 pages from any regulatory docket at no charge. Additional copies cost $0.15/page.

Customer Service

How Can I Influence EPA's Thinking on This Proposed Rule?

In developing this proposal, we tried to address the concerns of all our stakeholders. Your comments will help us improve this rule. We invite you to provide your comments on all data, assumptions and methodologies used to support our proposal, your views on options we have proposed, your ideas on new approaches we have not considered, any new data you may have, your views on how this rule may affect you, and other relevant information. Your comments must be submitted by the deadline in this proposal. Your

comments will be most effective if you follow the suggestions below:

• Explain your views as clearly as possible and provide a summary of the reasoning you used to arrive at your conclusions. Provide examples to illustrate your views wherever possible.

• Provide solid technical data to support your views.

• If you estimate potential costs, explain how you arrived at your estimate.

• Tell us which parts of this proposal you support, as well as which parts you disagree with.

• Offer specific alternatives.

• Reference your comments to specific sections of the proposal by using section titles or page numbers of the preamble or the regulatory citations.

• Clearly label any CBI submitted as part of your comments.

• Include your name, date, and docket number with your comments.

Contents of This Proposed Rule

I. Overview

A. Who Potentially Will be Affected by this Proposed Rule?

B. Why Does this Rule Read Differently from Other Listing Rules?

C. What are the Statutory Authorities for this Proposed Rule?

II. Background

A. How Does EPA Define a Hazardous Waste?

B. How Does EPA Regulate RCRA Hazardous Wastes?

C. What is the Consent Decree Schedule for and Scope of this Proposal?

III. Approach Used in this Proposed Listing

A. Summary of Today's Action

B. What Wastes Associated with the 14 Sectors Are Outside the Scope of the Consent Decree?

1. Mineral processing wastes exempt under the “Bevill” exemptions

2. Residuals used or reused in different industrial processes

3. Debris and other nonprocess wastes

C. What Information Did EPA Collect and Use?

1. The RCRA Section 3007 Survey

2. Field work: site visits, sampling and analysis

3. Other sources

D. How Did EPA Evaluate Wastes for Listing Determinations?

1. Listing policy

2. Characteristic hazardous waste

3. Evaluations of particular units and pathways of release

4. Evaluation of Secondary Materials

E. Description of Risk Assessment Approaches

1. What risk thresholds were used?

2. What leaching procedures were used?

3. How were wastes screened to determine if further assessment was needed?

4. How was the groundwater pathway evaluated?

5. How was the surface water pathway evaluated?

6. What are the limitations and uncertainties of the assessment?

F. Sector-specific Listing Determination Rationales

1. Antimony oxide

2. Barium carbonate

3. Boric acid

4. Cadmium pigments

5. Inorganic hydrogen cyanide

6. Phenyl mercuric acetate

7. Phosphoric acid from the dry process

8. Phosphorus pentasulfide

9. Phosphorus trichloride

10. Potassium dichromate

11. Sodium chlorate

12. Sodium dichromate

13. Sodium phosphate from wet process phosphoric acid

14. Titanium dioxide

G. What is the Status of Landfill Leachate from Previously Disposed Wastes?

IV. Proposed Treatment Standards Under RCRA's Land Disposal Restrictions

A. What are EPA's Land Disposal Restrictions (LDRs)?

B. What are the treatment standards for K176 (baghouse filters from production of antimony oxide)

C. What standards are the treatment standards for K177 (slag from the production of antimony oxide that is disposed of or speculatively accumulated)?

D. What are the treatment standards for K178 (nonwastewaters from the production of titanium dioxide by the chloride-ilmenite process)?

E. What Other LDR Provisions Are Proposed to Apply?

1. Debris

2. Soil

3. Underground Injection Wells that can be found in the administrative record for this rule

F. Is There Treatment Capacity for the Proposed Wastes?

1. What Is a Capacity Determination?

2. What are the Capacity Analysis Results?

V. Compliance Dates

A. Notification

B. Interim Status and Permitted Facilities

VI. State Authority

A. Applicability of Rule in Authorized States

B. Effect on State Authorizations

VII. Designation of Inorganic Chemical Wastes under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA)

A. Reporting Requirements

B. Basis for Proposed RQ Adjustment

VIII. Administrative Assessments

A. Executive Order 12866

1. Methodology Section

2. Results

B. Regulatory Flexibility Act (RFA), as amended by the Small Business Regulatory Enforcement Fairness Act of 1996 (SBREFA), 5 U.S.C. 601

et seq.

C. Paperwork Reduction Act

D. Unfunded Mandates Reform Act

E. Executive Order 12898: Environmental Justice

F. Executive Order 13045 : Protection of Children From Environmental Health Risks and Safety Risks

G. Executive Order 13084: Consultation and Coordination With Indian Tribal Governments

H. Executive Order 13132: Federalism

I. National Technology Transfer and Advancement Act

I. Overview

A. Who Potentially Will Be Affected by This Proposed Rule?

Beginning January 1, 1999 all documents related to EPA's regulatory, compliance and enforcement activities, including rules, policies, interpretive guidance, and site-specific determinations with broad application, should properly identify the regulated entities, including descriptions that correspond to the applicable SIC codes or NAICS codes (source: October 9, 1998 USEPA memo from Peter D. Robertson, Acting Deputy Administrator of USEPA). Today's action, if finalized, could potentially affect those who handle the wastes that we are proposing to add to EPA's list of hazardous wastes under the RCRA program. This action also may affect entities that may need to respond to releases of these wastes as CERCLA hazardous substances. These potentially-affected entities are described in the Economics Background Document placed in the docket in support of today's proposed rule. A summary is shown in the table below.

Summary of Facilities Potentially Affected by the USEPA's 2000 Inorganic Chemical Manufacturing Waste Listing Proposal

Item

SIC code

Industry sector name

Number of U.S. relevant inorganic mfg. facilities

1

2816

Inorganic Pigments

3

1

2819

Industrial Inorganic Chemicals, not elsewhere classified

3

The list of potentially affected entities in the above table may not be exhaustive. Our aim is to provide a guide for readers regarding entities likely to be regulated by this action. This table lists those entities that we are aware potentially could be affected by this action. However, this action may affect other entities not listed in the table. To determine whether your facility is regulated by this action, you should examine 40 CFR Parts 260 and 261 carefully in concert with the proposed rules amending RCRA that are found at the end of this

Federal Register

document. If you have questions regarding the applicability of this action to a particular entity, consult the person listed in the preceding section entitled

FOR FURTHER INFORMATION CONTACT.

B. Why Does This Rule Read Differently From Other Listing Rules?

Today's proposed listing determination preamble and regulations are written in “readable regulations” format. The authors tried to use active rather than passive voice, plain language, a question-and-answer format, the pronouns “we” for EPA and “you” for the owner/generator, and other techniques to make the information in today's rule easier to read and understand. This new format is part of our efforts toward regulatory re-invention, and it makes today's rule read differently from other listing rules. We believe that this new format will help readers understand the regulations, which should then increase compliance, make enforcement easier, and foster better relationships between EPA and the regulated community.

C. What Are the Statutory Authorities for This Proposed Rule?

These regulations are proposed under the authority of Sections 2002(a), 3001(b), 3001(e)(2), 3004(d)-(m) and 3007(a) of the Solid Waste Disposal Act, 42 U.S.C. 6912(a), 6921(b) and (e)(2), 6924(d)-(m)and 6927(a), as amended several times, most importantly by the Hazardous and Solid Waste Amendments of 1984 (HSWA). These statutes commonly are referred to as the Resource Conservation and Recovery Act (RCRA), and are codified at Volume 42 of the United States Code (U.S.C.), Sections 6901 to 6992(k) (42 U.S.C. 6901-6992(k)).

Section 102(a) of the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA), 42 U.S.C. 9602(a) is the authority under which the CERCLA aspects of this rule are proposed.

II. Background

A. How Does EPA Define a Hazardous Waste?

Section 3001 of RCRA and EPA's regulations establish two ways of identifying wastes as hazardous under RCRA. A waste may be hazardous either if it exhibits certain properties (called “characteristics”) which pose threats to human health and the environment, or if it is included on a specific list of wastes EPA has evaluated and found to pose unacceptable risks. EPA's regulations in the Code of Federal Regulations (CFR) define four hazardous characteristics: ignitability, corrosivity, reactivity, or toxicity. (See 40 CFR 261.21 through 261.24.) As a generator, you must determine whether or not a waste exhibits any of these characteristics by testing the material or by using your knowledge of the process that produced the waste. (See 40 CFR 262.11(c).)

EPA may also conduct a more specific assessment of a waste or category of wastes and “list” them if they meet criteria set out in 40 CFR 261.11. As described in 40 CFR 261.11, we may list a waste as hazardous if it:

—Exhibits any of the characteristics noted above ,

i.e.,

ignitability, corrosivity, reactivity, or toxicity (261.11(a)(1));

—Is “acutely” hazardous,

i.e.,

if they are fatal to humans or in animal studies at low doses, or otherwise capable of causing or significantly contributing to an increase in serious illness (261.11(a)(2)); or

—Is capable of posing a substantial present or potential hazard to human health or the environment when improperly managed (261.11(a)(3)).

Under the third criterion at 40 CFR 261.11(a)(3), we may decide to list a waste as hazardous (1) if it contains hazardous constituents identified in Appendix VIII to 40 CFR Part 261, and (2) if, after considering the factors noted in this section of the regulations, we “conclude that the waste is capable of posing a substantial present or potential hazard to human health or the environment when improperly treated, stored, transported, or disposed of, or otherwise managed.” We place a chemical on the list of hazardous constituents on Appendix VIII only if scientific studies have shown a chemical has toxic effects on humans or other life forms. When listing a waste, we also add the hazardous constituents that serve as the basis for listing to Appendix VII to part 261.

Residuals from the treatment, storage, or disposal of most listed hazardous wastes are also classified as hazardous wastes based on the “derived-from” rule (see 40 CFR 261.3(c)(2)(i)). For example, ash or other residuals from the treatment of a listed waste generally carries the original hazardous waste code and is subject to the hazardous waste regulations. Also, the “mixture” rule (see 40 CFR 261.3(a)(2)(iii) and (iv)) provides that, with certain limited exceptions, any mixture of a listed hazardous waste and a solid waste is itself a RCRA hazardous waste.

Some materials that would otherwise be classified as hazardous wastes under the rules described above are excluded from jurisdiction under RCRA if they are recycled in certain ways. The current definition of solid waste at 40 CFR 261.2 excludes secondary materials from the definition of solid waste that are used directly (

i.e.,

without reclamation) as ingredients in manufacturing processes to make new products, used directly as effective substitutes for commercial products, or returned directly to the original process from which they are generated as a substitute for raw material feedstock. (See 40 CFR 261.2(e).) As discussed in the January 4, 1985, rulemaking that promulgated this regulatory framework, these are activities which, as a general matter, resemble ongoing manufacturing operations more than conventional waste management and so are more appropriately classified as not involving solid wastes. (See 50 FR 637-640). Our approach to these issues is described in more detail below in section III.D.4.

B. How Does EPA Regulate RCRA Hazardous Wastes?

Wastes exhibiting any hazardous characteristic or listed as hazardous are subject to federal requirements under RCRA. These regulations affect persons who generate, transport, treat, store or dispose of such waste. Facilities that must meet the hazardous waste management requirements, including the need to obtain permits to operate, commonly are referred to as “Subtitle C” facilities. Subtitle C is Congress’ original statutory designation for that part of RCRA that directs EPA to issue those regulations for hazardous wastes as may be necessary to protect human health or the environment. EPA standards and procedural regulations implementing Subtitle C are found generally at 40 CFR Parts 260 through 272.

All RCRA hazardous wastes are also hazardous substances under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), as described in section 101(14)(C) of the CERCLA statute. This

applies to wastes listed in 40 CFR 261.31 through 261.33, as well as any wastes that exhibit a RCRA characteristic. Table 302.4 at 40 CFR 302.4 lists CERCLA hazardous substances along with their reportable quantities (RQs). Anyone spilling or releasing a substance at or above the RQ must report this to the National Response Center, as required in CERCLA Section 103. In addition, Section 304 of the Emergency Planning and Community Right-to-Know Act (EPCRA) requires facilities to report the release of a CERCLA hazardous substance at or above its RQ to State and local authorities. Today's rule proposes to establish RQs for some of the newly listed wastes.

C. What Is the Consent Decree Schedule for and Scope of This Proposal?

The 1984 Hazardous and Solid Waste Amendments (HSWA) to RCRA require EPA to make listing determinations for several specified categories of wastes, including “inorganic chemical industry wastes” (see RCRA section 3001(e)(2)). In 1989, the Environmental Defense Fund (EDF) filed a lawsuit to enforce the statutory deadlines for listing decisions in RCRA Section 3001(e)(2). (

EDF

v.

Browner;

D.D.C. Civ. No. 89-0598). To resolve most of the issues in the case, EDF and EPA entered into a consent decree, which has been amended several times to revise deadlines for EPA action. Paragraph 1.g (as amended) of the consent decree addresses the inorganic chemical industry:

EPA shall promulgate a final listing determination for inorganic chemical industry wastes on or before October 31, 2001. This listing determination shall be proposed for public comment on or before August 30, 2000. The listing determination shall include the following wastes: sodium dichromate production wastes, wastes from the dry process for manufacturing phosphoric acid, phosphorus trichloride production wastes, phosphorus pentasulfide production wastes, wastes from the production of sodium phosphate from wet process phosphoric acid, sodium chlorate production wastes, antimony oxide production wastes, cadmium pigments production wastes, barium carbonate production wastes, potassium dichromate production wastes, phenyl mercuric acetate production wastes, boric acid production wastes, inorganic hydrogen cyanide production wastes, and titanium dioxide production wastes (except for chloride process waste solids). However, such listing determinations need not include any wastes which are excluded from hazardous waste regulation under section 3001(b)(3)(A)(ii) of RCRA and for which EPA has determined that such regulation is unwarranted pursuant to section 3001(b)(3)(C) of RCRA.

Today's proposal satisfies EPA's duty under paragraph 1.g to propose determinations for inorganic chemical industry wastes.

As described above, the consent decree provides that EPA does not need to make listing determinations for certain wastes that it has exempted from hazardous waste regulations under the “Bevill amendments” to RCRA. See the discussion of “exempt mineral processing” wastes in section III.B.1 below.

III. Approach Used in This Proposed Listing

A. Summary of Today's Action

Manufacturers of the inorganic chemical products described above identified over 170 categories of residuals generated as part of their production processes. We first determined which of these residuals fell within the scope of our consent decree obligations. We then evaluated the risks posed by each of the remaining categories of residual materials. In some cases we used quantitative or qualitative screening methods. For 18 wastes we conducted full-scale modeling to predict risks.

As a result of this evaluation, we found that three wastes generated in the 14 inorganic chemicals manufacturing operations which we evaluated meet the criteria for listing set out in either 40 CFR 261.11(a)(1) or 261.11(a)(3). We conducted full-scale modeling of two of these wastes and propose to list them under 40 CFR 261.11(a)(3). We found that one waste warranted listing under 40 CFR 261.11(a)(1) because it exhibited hazardous waste characteristics. We did not model this waste. Since these are wastes from specific inorganic chemical industries, we propose to add them to Section 261.32 with K-waste codes. The three wastestreams we propose to list as hazardous, along with their corresponding hazard code and proposed EPA Hazardous Waste Numbers, are:

1

1

As required in 40 CFR 262.30, the listing description includes the hazard code. Wastes listed under 40 CFR 261.11(a)(1) due to the toxicity characteristic are designated “E,” and wastes listed under 40 CFR 261.11(a)(3) for toxicity are designated “T”.

K176 Baghouse filters from the production of antimony oxide. (E)

K177 Slag from the production of antimony oxide that is disposed of or speculatively accumulated. (T)

K178 Nonwastewaters from the production of titanium dioxide by the chloride-ilmenite process. (This listing does not apply to chloride process waste solids from titanium tetrachloride production exempt under 40 CFR 261.4(b)(7).) (T)

We found that all of the remaining wastes that we evaluated did not meet the criteria for listing in 40 CFR 261.11, and we are proposing not to list them as hazardous wastes. More information on our evaluations of particular wastes is set out in the background documents and the sector-specific discussions in section III.F of this preamble.

We have previously listed as hazardous a number of wastes in 40 CFR 261.32 from other inorganic chemicals industries, including wastes from the production of inorganic pigments (codes K002 through K008), and wastes from chlorine production (codes K071, K073, and K106). Today's proposal does not affect the scope of any existing hazardous waste listing, and we are not soliciting comments on those existing listing determinations.

We are also proposing other changes to the RCRA regulations as a result of the proposed listings. These changes include adding constituents to Appendices VII and VIII for Part 261, and setting new land disposal restrictions. We are proposing to add the following constituents to Appendix VII that serve as the basis for listing: K176—arsenic and lead, K177—antimony, and K178—manganese and thallium. We are also proposing to add manganese to the list of hazardous constituents in Appendix VIII, based on scientific studies that demonstrate manganese has toxic effects on humans and other life forms. Section IV of today's proposal describes the proposed changes to the land disposal restrictions, which would establish treatment standards for specific constituents in the wastes proposed for listing.

Also as a result of the proposed listings, these wastes would become hazardous substances under CERCLA. Therefore, in today's rule we are proposing to designate these wastestreams as CERCLA hazardous substances, and to adjust the one-pound statutory RQs for two of these wastestreams; this is described in section VII of today's proposal.

B. What Wastes Associated With the 14 Sectors Are Outside the Scope of the Consent Decree?

Determining the scope of our consent decree obligations was more complicated than usual for two reasons. First, Paragraph 1.g (quoted above in II.C) does not tell EPA which wastestreams it must evaluate. For most other listing obligations set out in the consent decree, the decree specifies particular wastestreams which EPA must evaluate for listing. See, for

example, paragraph 1.k identifying 14 specific petroleum wastestreams.

2

Second, paragraph 1.g contains an exemption for wastes found to be exempt from hazardous waste regulation in previous EPA actions implementing the so-called “Bevill exemptions” for mineral processing wastes. Thus, we needed to conduct some analysis to determine the scope of our obligations.

2

The revised consent decree is available in the docket for today's proposal.

We began by asking facilities to identify all of the residuals generated by their production processes. We then reviewed their lists. We found that some residuals were actually exempt “Bevill” wastes that we need not address. We found that other wastes were really associated with the manufacture of other materials. Also, we concluded that a few residuals were not “production” wastes and therefore were not covered by the decree. With the exceptions discussed below in our evaluation of the sodium dichromate and titanium dioxide sectors, we chose not to evaluate any of the wastes that we considered to be outside the scope of the decree. We concluded that evaluation was not possible under the time frame set out in the decree. In the following sections we provide an overview of the types of wastes that we consider outside the scope.

1. Mineral Processing Wastes Exempt Under the “Bevill” Exemptions

Many of the inorganic chemical manufacturing processes we address in this rule use ores and minerals as feedstocks. Some wastes derived from the processing of ores and minerals are exempt from regulation as RCRA hazardous wastes under decisions EPA made under statutory requirements known as the “Bevill” amendments. RCRA Sections 3001(b)(3) and 8002(p) required EPA to determine whether wastes from the extraction, beneficiation or processing of ores and minerals warranted regulation as hazardous wastes under Subtitle C of RCRA. Between 1989 and 1991 EPA completed a series of rules and studies establishing which mining wastes fit within the “extraction,” “beneficiation,” or “processing” definitions, and which of the wastes within each category were exempt from regulation as hazardous wastes. EPA concluded that all wastes produced during extraction and beneficiation are entitled to an exemption. EPA found that 20 categories of wastes from subsequent “mineral processing” operations met the “high volume/low toxicity” criteria and were exempt as well. See 54 FR 36592 (Sept. 1, 1989), 55 FR 2322 (Jan. 23, 1990), the July 31, 1990 Report to Congress on Wastes from Mineral Processing, and 56 FR 27300 (June 13, 1991).

EPA codified these “Bevill” exemptions at 40 CFR 261.4 (b)(7). EPA discussed some of these exemptions further in a 1998 final rule promulgating treatment standards for non-exempt mineral processing wastes that exhibit the toxicity characteristic. See the Land Disposal Restrictions Phase IV Final Rule at 63 FR 28598 (May 26, 1998).

Paragraph 1.g of the consent decree provides that EPA need not make listing determinations for wastes from any of the 14 inorganic chemical manufacturing processes which are “excluded from hazardous waste regulation under Section 3001(b)(3)(A)(ii) of RCRA and for which it has determined that such regulation is unwarranted pursuant to Section 3001(b)(3)(C) of RCRA.” In other words, the consent decree does not require us to make listing determinations for wastes which we exempted under the statute's “Bevill” provisions.

Paragraph 1.g. of the consent decree requires EPA to make a listing determination for “* * * titanium dioxide production wastes (except for chloride process waste solids).” EPA interprets the exception to refer to the chloride process waste solids from the production of titanium tetrachloride which are exempt under the Bevill rule, rather than all solids from the chloride process. Solids generated after titanium tetrachloride forms fall within the scope of the consent decree.

We reviewed the generators' lists of process residuals to determine whether they had included any Bevill exempt wastes which we need not assess. (In some cases, the generators had claimed that certain wastes were exempt under EPA's Bevill decisions.) This process was not always simple. We found it was sometimes difficult to determine whether a particular facility's waste fit within one of the exempt categories. For example, the mineral processing exemption for titanium dioxide covers only solid materials from an initial step in the production process. It was not always easy to tell whether particular waste solids were generated from the portion of the process that would make them exempt, or from later production steps. Sector-specific information regarding our conclusions appears in section III.F of this preamble for those sectors where we found this exemption had some relevance. We found that facilities in only three of the consent decree sectors generate Bevill exempt wastes: Boric acid, sodium dichromate, and titanium dioxide.

In other sectors, the facilities produce inorganic product chemicals from a mineral product. Under the Bevill exemption (54 FR 36620-21), chemical manufacturing begins if there is any further processing of a saleable mineral product. Since these facilities use saleable mineral products as feedstock, their processes are chemical manufacturing, and may not be classified as mineral processing. Therefore, none of the wastestreams generated by these facilities in the production of the other inorganic chemicals are Bevill exempt.

We emphasize that we are not re-opening any Bevill decisions made in earlier actions regarding the exemptions. We are not re-defining the boundaries between “extraction” and “beneficiation,” between “beneficiation” and “mineral processing,” or between “mineral processing” and non-exempt chemical manufacturing. Nor are we revisiting our decision that all wastes uniquely associated with the extraction and beneficiation of ores and minerals are exempt. Similarly, we are not re-opening any of our earlier decisions as to which categories of mineral processing wastes are exempt. Rather, we are determining whether particular wastestreams fall within any of the exempt categories. We are not requesting comment on, and do not intend to respond to comments relating to the earlier decisions.

We also found that some inorganic chemical processes generate composite wastestreams that contain both a Bevill exempt waste and one or more non-exempt wastes. We evaluated the non-exempt portions of such wastes to fulfill our consent decree requirements. We apportioned risks between the exempt and nonexempt portion of such commingled wastes, and made listing determinations for the non-exempt portions. We did not, however, assess the exempt portions of such streams. This assessment, therefore, does not re-open any earlier decision regarding exemptions for the “Bevill” component of the commingled streams.

2. Residuals Used or Reused in Different Industrial Processes

In some cases, facilities within the 14 inorganic chemicals sectors set out in the consent decree produce residuals that are used or reused in processes that are not among those listed in the decree. Those industries in turn produced residuals derived from the materials generated in the consent decree industries. We evaluated the

management of the original industry's material up to the point that the second industry inserts it into its production process. However, we generally considered the second production process and its associated wastes to be beyond the scope of the consent decree. We did not evaluate for listing purposes wastes generated from these non-consent decree industries. For example, in the titanium dioxide sector, one facility uses a residual from the production of titanium dioxide as an ingredient to make salt. We considered salt-making to be a separate production process outside the scope of the consent decree. We are not proposing any listing determinations for wastes generated in the salt plant.

However, in some cases, the reuse of the residual from a consent decree process involved an activity which we always regulate as waste management. In this situation, we considered the reuse to be waste management, and the waste to be within the scope of the consent decree. Consequently, we evaluated the residual for listing. For example, we found that one of the facilities which produces boric acid generates a waste which is used as a fuel. Under our recycling regulations, we regulate burning for energy recovery and so, we evaluated this waste. See 40 CFR 261.2(c)(2). We found that the waste is already being managed in a RCRA Subtitle C unit and decided not to list the waste. Others examples of reuse that we evaluated include land application of biological treatment solids from hydrogen cyanide production as a fertilizer or soil amendment, and land application of gypsum from the titanium dioxide sector. In two cases, however, we decided to make listing decisions for residuals generated during the production of non-consent decree products. In the titanium dioxide sector, the residuals are commingled with other wastes clearly within the scope of the decree. See the discussions of the sodium dichromate sector and the titanium dioxide sector in section III.F. In the sodium dichromate sector, residuals from the non-consent decree process are piped back to the consent decree process, making it difficult to determine whether the two processes are really separate.

3. Debris and Other Nonprocess Wastes

Some generators also identified debris and structural components of their production plants as intermittently-generated wastes. We concluded that these materials do not fall within the scope of the decree. Most of the wastes that fell in this category were refractory bricks which become wastes when facilities remove them to refurbish their furnaces. We consider this material to be a structural component of the plant where production takes place rather than a waste from the “production” of an inorganic chemical. Similarly, we consider a few analogous types of plant debris to fall outside the scope of the decree. This debris includes miscellaneous construction materials, insulation, reactor bed material, and piping. These wastes were reported for the following inorganic sectors: Phosphoric acid, barium carbonate, sodium dichromate, hydrogen cyanide, antimony oxide, sodium phosphate, and titanium dioxide.

We have never interpreted the decree to require us to consider listing tanks, pads, or other structural components housing production processes when they become wastes by being removed from use. Other paragraphs of the decree support this position. Paragraphs 1.c. (coke byproducts) and 1.k. (petroleum refining wastes) cover production processes involving reaction vessels lined with refractory or similar materials, and in neither case did the decree include wastes related to the reaction vessels themselves or related materials. Nor do any other provisions in the decree direct us to list any other type of structural components. We note that discarded refractory bricks and other debris would be regulated as hazardous wastes, if these materials were contaminated with a listed waste (including wastes listed as a result of today's rulemaking), or if they exhibited a hazardous waste characteristic.

A few facilities also reported environmental media (excavated soils or recovered groundwater) contaminated with process residuals as wastes from their production processes. We consider such contaminated media to be outside the scope of today's listing determinations, because these are not wastes generated during production processes, but rather wastes generated due to construction or remedial action. We note that none of the other consent decree provisions require us to evaluate contaminated media. See the specific listing background documents for the different sectors for a full listing of the wastes we considered to be out of scope of the decree.

C. What Information Did EPA Collect and Use?

Our investigation of the wastes generated by the inorganic chemicals manufacturing industry included two major information collection efforts: A survey of the industries and field investigations. The survey effort included the development, distribution, and assessment of an extensive questionnaire sent under the authority of RCRA section 3007 to all known facilities engaged in any of the 14 inorganic chemical manufacturing processes. During our field investigations we made site visits to familiarize ourselves with processes and residuals, and made additional visits to collect samples of residuals which we sent to laboratories for analysis. Finally, we collected data from other sources to help characterize the settings in which some of the wastes are managed. Each of these efforts is summarized below.

1. The RCRA Section 3007 Survey

We developed an extensive questionnaire under the authority of Section 3007 of RCRA for distribution to the inorganic chemicals manufacturing industry. The purpose of the survey was to gather information about solid and hazardous waste generation and management practices in the U.S. for the fourteen inorganic chemical industry sectors. The questionnaire collected information about the inorganic chemical products manufactured, the processes used, the wastes generated, the wastes characteristics, and how the wastes were managed. The questionnaire is included in the “General Background Document for the Inorganic Chemical Listing Determination.” which is in the docket for today's proposal. This document also provides more details on the producers identified in the inorganic sectors.

We distributed the survey in March of 1999 to all 124 facilities that we had identified as potential manufacturers of chemicals in the 14 targeted sectors. We developed the list of facilities from a review of the available literature, which included directories of chemical producers, reference works of chemical technology, chemical profile information, and previous work by EPA on these industries. From the surveys distributed, 57 facilities indicated that they manufacture chemicals from at least one of the 14 sectors. The other facilities notified us that they had either stopped operations or did not manufacture inorganic chemical products. From the survey, we confirmed that one product was no longer manufactured in this country (phenyl mercuric acetate).

We also conducted an exhaustive engineering review of the submitted surveys for accuracy and completeness. We conducted quality assurance reviews of the data to identify any inappropriate entries and missing data. The engineering review of each facility's

response resulted in follow-up letters and/or telephone calls to the facilities seeking clarifications, corrections, and additional data where needed.

Where we conducted sampling and analysis of the waste, we used this analytical data in our analysis (see the following section). Facilities also submitted data in their survey on the composition of some of their wastes. In the absence of our own analytical data, we used data provided by facilities in our evaluation. These cases are noted in the sector-specific discussions in section III.F. In some cases, these data consisted of results from testing to determine whether the wastes exhibited characteristics. We thought such data were reliable because of the consequences the facilities would face if their characteristic data were not accurate. In addition, survey respondents were required to certify the accuracy of their submittal.

2. Field Work: Site Visits, Sampling and Analysis

As part of our field work, we visited production facilities (engineering site visits), we took preliminary samples (familiarization sampling), and we obtained samples to fully characterize the waste for constituents of concern (record sampling). We initiated the sampling phase of this listing determination with the development of a Quality Assurance Project Plan (QAPP) for sampling and analysis. The QAPP describes the quality assurance and quality control requirements for the data collection. We also developed sampling and analysis plans (SAPs) for sampling at individual facilities. The QAPP and the SAPs are available in the public docket for this proposal.

The primary purpose of the engineering site visits was to gain first hand knowledge of the manufacturing processes, the waste generation and management, and to identify potential locations for waste sampling. We conducted site visits at 25 facilities in 12 of the sectors. We selected the facilities to visit based on logistics and to visit sites that represent the variety of process and wastes generated within industry sectors. Site visit reports are available in the docket for today's rule. During some of the engineering site visits, we collected 22 familiarization samples to help us identify potential sampling or analytical problems for the wastes of interest. For example, we used the familiarization samples to assess the effectiveness of the analytical methods that we planned to use during record sampling for a number of the targeted waste matrices.

During record sampling, we collected 69 waste samples from 13 different facilities. Additional samples were collected for QA/QC purposes. Largely due to the time constraints imposed by our consent decree schedule, we focused the sampling effort on the wastes that we most expected to present significant potential risks. Based on information obtained from the RCRA Section 3007 Surveys, we established sampling priorities by considering the reported management practices (e.g., wastes going to Subtitle D landfills and impoundments were of concern), and the likely presence of contaminants of concern.

We also found that we were able to make listing decisions on a variety of reported wastes without conducting sampling. In some cases, we were able to use information about the processes and the raw materials to conclude that a waste was not likely to present a significant risk. Also, we did not typically sample wastes that were reported to be characteristically hazardous waste and were already regulated as hazardous under RCRA. We felt that, for these wastes, we could make listing decisions without further information on waste constituents. In addition, we did not attempt to sample wastes that we found to be outside the scope of the consent decree, as described in Section III.B. Thus, for example, we did not sample a number of wastes that appeared to be exempt under the Bevill regulations.

We believe that the 69 record samples from 13 sites provide an adequate characterization of the wastes that we sampled. The 13 sites represent approximately 30% of the 42 identified production facilities within the specific sectors we chose for sampling. The wastes sampled also represent the major waste types of concern, e.g., specific process wastes/sludges, wastewater treatment sludges, wastewaters, and spent filter material. Section III.F of this proposal provides information on the specific wastes sampled in each sector. The docket for today's proposal also contains background documents for the specific sectors, which give details on which wastes we sampled and our evaluation of the need for sampling or modeling certain wastes.

For most sectors, we focused our analyses on metal constituents, because these are the constituents expected from the inorganic processes under evaluation. We analyzed for other constituents in those cases where we expected they might be present in the waste, or if other constituents showed up in the familiarization sampling. Thus, we analyzed wastes from the inorganic hydrogen cyanide industry for cyanide and volatile organics because of their potential to be present from the process. Similarly, in the titanium dioxide sector, we analyzed waste samples for semivolatile and chlorinated organics due to the use of coke and chlorine as raw materials in the production process for the titanium chloride intermediate. The overall list of target analytes are in the QAPP, which is in the docket for today's rule. The docket also contains the background documents for each sector and the corresponding waste characterization data reports, which show the chemical analyses performed and the analytes found in the waste samples.

In our analyses of wastes samples, we performed analyses to measure constituent concentrations in the wastes themselves (“total” analysis), as well as analyses for constituents that leach out of the wastes. We generally used the methods specified in OSW's methods manual (“Test Methods for Evaluating Solid Waste, Physical/Chemical Methods,” SW-846), as described in the QAPP, the SAPs, and the background documents for the specific sectors. We used two extraction methods to measure leaching, the Toxicity Characteristic Leaching Procedure (TCLP, SW-846 method 1311), and the Synthetic Precipitation Leaching Procedure (SPLP, SW-846 method 1312).

In general, we were able to measure the concentrations of constituents in waste samples at very low detection levels. However, for some constituents in some matrices the SPLP and/or TCLP analyses provided detection limits that were somewhat above health-based levels of concern. In such cases, we examined all of the analytical data to determine if the undetected constituent might possibly present a potential risk. Where we did not detect the constituent in the total analysis (

i.e.,

the analysis of a sample prior to any leaching), we assumed that the constituent was not present in the leachate. However, if the totals analysis showed the presence of a constituent that we did not detect in the leachate, then we assessed the risk that would be posed if the constituent were present at a concentration equal to one-half the detection limit. Section III.F shows the cases where we used this assumption in our evaluation of wastes for the different inorganic sectors, and further details are available in the background documents for each sector.

3. Other Sources

We also collected data from a variety of other sources to help characterize the settings in which these wastes are managed. For example, we contacted

several state and local authorities to collect information regarding the location of drinking water wells. We also obtained information and, in some cases analytical data, from state authorities and other sources to help in our evaluations. We note these sources in the sector-specific discussions in Section III.F when we relied on such data.

D. How Did EPA Evaluate Wastes for Listing Determinations?

1. Listing Policy

As discussed in section II.A. of this preamble, we consider the listing criteria set out in 40 CFR 261.11, in light of all the information we have relevant to the criteria, in making listing determinations. For decisions made under 40 CFR 261.11(a)(3), today's proposed listing determinations follow the elements of the EPA's hazardous waste listing policy presented the proposed listing for wastes generated by the dyes and pigments industry (see FR 66072, December 22, 1994). We have modified and adapted this policy in subsequent listings. See for example the recent Petroleum Refining proposal (60 FR 57747; November 20, 1995) and the Solvents waste proposal (61 FR 42318; August 14, 1996).

This policy uses a “weight-of-evidence” approach in which calculated risk information is a key factor to consider in making a listing determination under 40 CFR 261.11(a)(3). The criteria provided in 40 CFR 261.11(a)(3) include eleven factors for determining “substantial present or potential hazard to human health and the environment.” We incorporate nine of these factors, as described generally below, into our risk evaluation for the wastestreams of concern:

—Toxicity (261.11(a)(3)(i)) is considered in developing the health benchmarks used in the risk evaluation.

—Constituent concentrations and waste quantities (261.11(a)(3)(ii) and 261.11(a)(3)(viii)) are used to define the initial conditions for the risk evaluation.

—Potential to migrate, persistence, degradation, and bioaccumulation of the hazardous constituents and any degradation products (261.11(a)(3)(iii), 261.11(a)(3)(iv), 261.11(a)(3)(v), and 261.11(a)(3)(vi)) are all considered in our evaluation of constituent mobility (e.g., leaching from waste) and fate and transport models we used to project potential concentrations of the contaminants to which individuals might be exposed.

We considered two additional factors, plausible mismanagement and other regulatory actions (261.11(a)(3)(vii) and 261.11(a)(3)(x)) in selecting the waste management scenarios we evaluated in our risk assessments. For example, we used information that the waste generators submitted in their Section 3007 questionnaires to decide what types of waste management units are used. Using information about other federal environmental regulatory programs, we concluded that some units or some pathways did not pose risks requiring evaluation.

We separately considered the remaining factor, whether the available information indicated any impact on human health or the environment from improper management of the wastes of concern (261.11(a)(3)(ix)). Thus, we examined a variety of databases for information on damage incidents for the inorganic chemical production processes under investigation. For example, we examined databases for information on potential and actual Superfund sites (CERCLIS), releases reported under the Toxic Release Inventory System (TRIS), civil cases filed on behalf of EPA, and spills and releases reported to the National Response Center (NRC). A full description of our search is in the docket for this rule.

Most of the cases found for these industries typically resulted from spills or releases of products, and did not provide any useful information of possible risks presented by the wastes we evaluated for listing. In a few cases we found sites on the Superfund National Priority List (NPL) that included inorganic manufacturing processes. However these sites usually encompassed a variety of chemical manufacturing and mining industries, and it is difficult to attribute the damage reported to the specific inorganic manufacturing wastes under evaluation. Furthermore, contamination at these sites appears linked to historical management practices at closed or inactive manufacturing plants, and these were not useful in assessing current or potential hazards for the wastes at issue. In addition, Federal and State regulatory controls are now in place that would prevent mismanagement. For example, many of the wastes examined in today's proposal are regulated as characteristic waste, and releases or disposal to the land are addressed under the existing RCRA regulations. We did not find any evidence of actual damage cases.

We describe our decisions under 40 CFR 261.11(a)(3) in more detail in the sector-specific discussion in section III.F below, and in the background documents. Generally, we conducted full-scale risk modeling for 18 wastes in 5 sectors. We found that we could adequately address the risks of the remaining wastes with a variety of less time-consuming approaches. Some were qualitative; others were quantitative, but not as complex as full modeling.

We evaluated one waste using the single criterion set out in 40 CFR 261.11(a)(1) rather than the eleven factors referenced in 40 CFR 261.11(a)(3). This is the first time under this consent decree that we have proposed to make a listing decision based on this criterion. It relies on the existing characteristics to identify wastes posing significant risks and does not require the use of modeling. See the discussion of wastes from the production of antimony oxide in section III.F.1 of the preamble.

Our proposed listing determinations are based upon estimates of individual risk. We relied on individual risk estimates (HQs > 1), and not population risk estimates, because we are concerned about risks to individuals who are exposed to potential releases of hazardous constituents. We believe that using individual risk as a basis for our listing determinations (rather than population risk estimates) also is appropriate to protect against potential risks, as well as present risks that may arise due to the generation and management of particular wastestreams. EPA acknowledges that in cases where small populations are exposed to particular wastes and waste management practices, population risks may be very small. We did not attempt to calculate population risks for the proposed listings. In general, we expect population risks arising from contaminated groundwater due to waste management to be small, because often only a limited number of domestic wells will be near these facilities, and groundwater contamination often moves very slowly. Nevertheless, the increased risk for an exposed individual may be significant. In proposing the listing determinations for K176, K177, and K178, EPA is protecting against the potential risk for exposed individuals, regardless of how many individuals are exposed.

We set out below general observations about some of our approaches to risk assessment.

2. Characteristic Hazardous Waste

We describe in Section a. below our analysis for wastes which are “100% characteristic”—wastes which all generators report as characteristic and which all generators appear to manage in compliance with applicable hazardous waste regulations. We

describe our approach to wastes which are occasionally characteristic—but managed in compliance—in Section b. below. Finally, we discuss in Section c. one waste which appears to exhibit a characteristic frequently, but does not appear to be managed in compliance with hazardous waste regulations.

a. Wastes consistently exhibiting characteristics.

For wastes which these industries identified as characteristic and managed in compliance with hazardous waste regulations, we are proposing to find that there is no “plausible mismanagement” scenario to evaluate for listing. (See 40 CFR 261.11(a)(3)(vii).) The Subtitle C rules applying to characteristic wastes adequately protect human health and the environment, especially where waste generators are complying with them. 40 CFR 261.11(a)(3)(x) authorizes us to consider actions taken by other regulatory programs. We believe we can reasonably interpret this to include the rules for characteristic wastes under Subtitle C .

We acknowledge that the regulation of characteristic wastes differs in some ways from the regulation of listed wastes. For example, for characteristic wastes, residues from treatment required by the land disposal restrictions need not always be placed in hazardous waste disposal units. However, we do not regard the differences as “mismanagement.” Rather, we believe that both approaches protect human health and the environment. Consequently, for the purposes of this rule we decided that we would not propose to list a “100% characteristic” waste unless we found evidence of extraordinary risks under one or more of the other factors in 40 CFR 261.11(a)(3).

For a few of the 100% characteristic wastes in this rule, we found factors warranting further consideration. For example, we found that the sole generator of cadmium pigment wastes codes them as hazardous, arranges for treatment to comply with the land disposal restrictions, and then disposes of treated residues in a Subtitle D landfill. At the same time, we found that the waste contains very high levels of cadmium. We decided to investigate further to ensure that the treatment residues did not present significant risks. We examined data relating to the treatment process and leachate monitoring data from the landfill receiving the residues. Based on these data, we concluded that the residues did not pose risks warranting listing.

b. Wastes which sometimes exhibit characteristics.

Information submitted in responses to the Section 3007 questionnaires also showed that there are a number of wastes that exhibit characteristics at some facilities, but not others. Consistent with previous listing decisions (see for example, the most recent petroleum refining listing at 63 FR 42137), we focused on the volumes of waste that did not exhibit characteristics in our listing evaluation. For wastestreams identified as exhibiting characteristics and apparently managed in compliance with applicable regulations, we relied on the “no plausible mismanagement” and “other regulations” analysis described above. A hypothetical example follows. If one facility generated 40 tons per year of a properly-managed characteristic waste, and a second facility generated 60 tons per year of a non-characteristic waste, we would not evaluate the total of 100 tons of waste under a single approach. Rather, we would evaluate the characteristic waste under the approach described above. For the waste that did not exhibit a characteristic, we would conduct the type of risk assessment described below in section III.E.

c. Characteristic wastes not managed in compliance with Subtitle C.

In one case, we found a characteristic waste where we believe that existing Subtitle C rules do not adequately prevent mismanagement. Four facilities generate a baghouse filter waste from the production of antimony oxide. Data from our sampling and analysis at 2 facilities showed exceedences of the toxicity characteristic. Two facilities recycle these wastes in a manner that may comply with applicable regulations. Two other facilities, however, did not identify their waste as characteristic wastes, and appear to manage them in ways which do not comply with Subtitle C rules. Because of this apparent noncompliance, we concluded that it would be appropriate to disregard the characteristic rules in an analysis of the factors in 40 CFR 261.(a)(3). However, we also concluded that it was not necessary to conduct such an analysis. Since this waste fails the toxicity characteristic, it clearly contains levels of constituents which could pose threats to human health via groundwater when placed in a municipal landfill, if leachate were to migrate to a drinking water well at sufficient concentrations. Since the generators are not managing the wastes in compliance with applicable Subtitle C regulations, we assume that this type of mismanagement could occur at other sites. Accordingly, we exercised our authority to propose to list this waste under 40 CFR 261.11(a)(1). As noted above, this provision authorizes (but does not require) EPA to list wastes that exhibit characteristics without the analysis required under 40 CFR 261.11(a)(3). We believe that noncompliance is an appropriate reason to use this authority to list a characteristic waste.

d. Non-characteristic wastes disposed of in hazardous waste units.

We identified nine wastes which do not appear to exhibit any characteristic, but which are disposed of in Subtitle C management units. Four of these wastes are sent to combustion unit regulated under Subtitle C of RCRA. The remaining 5 wastes are sent to Subtitle C landfills. We found that all of these wastes receive some treatment before land disposal. In one case available data indicates that the waste meets applicable LDR treatment standards as generated.

In general, these wastes have very limited potential for mismanagement under 40 CFR 261.11(a)(3)(vii). This is particularly true for wastes which generators place in on-site, Subtitle C units with ample capacity. Also, in some cases, some of the wastes are generated in very small quantities (less than 1 metric ton per year). These wastes are distinguishable from a non-characteristic organobromine waste sent to a hazardous waste unit that we decided to list in 1998. That waste had extremely high concentrations of a constituent posing significant risks, and received no treatment (see May 4, 1998; 63 FR 24596).

3

We request comment on the individual rationales set out in the sector-specific discussions and the background documents.

3

On April 9, 1999, the D.C. Circuit in Great Lakes Chemical Corporation v. EPA ordered that the organobromine listing determinations be vacated. Accordingly, EPA removed the listings from CFR (see 65 FR 14472: March 17, 2000).

3. Evaluations of Particular Units and Pathways of Release

We are proposing to find that some pathways of release from some units present low risks because they are adequately controlled under other Federal environmental regulations that minimize the likelihood of releases. We are also proposing to find that other risk pathways present low risks due to physical or chemical attributes of the wastes. In some cases, we evaluated all release pathways at a single unit under a combination of these approaches.

a. Wastewater management.

Facilities in these industries generally treat wastewaters in on-site wastewater treatment plants and discharge to surface waters, or pretreat the waste and discharge to an off-site wastewater

treatment facility,

e.g.,

a Publicly Owned Treatment Works (POTW). Under the Clean Water Act (CWA), discharges to surface waters are controlled under the National Pollutant Discharge Elimination System (NPDES) and require an NPDES permit, while discharges to a POTW are subject to State and national pretreatment standards.

4

Point source discharges for the various sectors in the inorganics listing are regulated under the CWA by the effluent guidelines and pretreatment standards in 40 CFR Parts 415 (Inorganic chemical manufacturing) and 422 (Phosphate manufacturing). Therefore, we did not evaluate NPDES effluent or discharges to POTWs in today's proposal. This approach is consistent with other listing rules. See, for example, 60 FR 57759 (November 20, 1995, petroleum refining wastes proposal). In a few cases, facilities reported disposal of wastewaters by deep well injection in a permitted Class I UIC hazardous waste injection well. In these cases, the wells were units regulated by the Underground Injection Control (UIC) program under the Safe Drinking Water Act (40 CFR Part 144). These wells also had no migration exemptions under Section 148.20 to allow disposal of untreated hazardous waste. Therefore, we did not evaluate this scenario further.

4

In fact, 40 CFR 261.4 excludes “any mixture of domestic sewage and other wastes that passes through a sewer system to a POTW for treatment” (40 CFR 261.4(a)(1)(ii), and industrial wastewater discharges that are point source discharges subject to regulation under Section 402 of the CWA (40 CFR 261.4(a)(2)).

For surface impoundments, we concluded that releases to air were not likely to present concerns. For most sectors, the constituents of concern are nonvolatile metals, and this makes volatilization a highly unlikely pathway for constituents from normal wastewater treatment practices. We recognize that releases of volatile organic chemicals from impoundments may be a potential route of concern for one sector, inorganic hydrogen cyanide production. EPA is developing maximum achievable control technology (MACT) standards for cyanide manufacturing under the Clean Air Act (CAA), which may address these emissions.

5

EPA is evaluating possible air releases from wastewaters in impoundments as part of the MACT rulemaking. Therefore, we did not do any further evaluation of these emissions as part of today's listing determination. We assessed the potential for groundwater releases from the impoundments.

5

Clean Air Act—Title III: Upcoming MACT Standards—Cyanide Chemical Manufacturing; Unified Air Toxics Website:

http://www.epa.gov/ttn/uatw/mactupd.html:

The hydrogen cynaide industry would also be subject to regulations under 40 CFR Part 60, Subpart YYY under the CAA for volatile organic compound (VOC) emissions from wastewater treatment at facilities in the synthetic organic chemical manufacturing industry (SOCMI), which was proposed September 19, 1994 (59 FR 46780).

For sectors and wastes where facilities did not use surface impoundments for wastewater management, we determined that “plausible mismanagement” would be continued management in existing tank-based treatment systems. We do not view abandonment of existing treatment systems for surface impoundments as “plausible,” because the manufacturers have already made a considerable investment in wastewater treatment systems using tanks and will continue to use them. Further, we assumed that wastewater treatment tanks retain sufficient structural integrity to prevent wastewater releases to the subsurface (and therefore to groundwater), and that overflow and spill controls prevent significant wastewater releases. Thus, based on the lack of any significant likelihood of release of the constituents to groundwater, we did not project significant risks to groundwater from these wastes in the tank-based wastewater treatment scenario. We did not model any releases to groundwater from tanks. This is consistent with our approach in other listing rules (see, for example, the proposed rule for chlorinated aliphatics production wastes at 64 FR 46476; August 25, 1999). We also considered the possibility of air releases from tanks. For most wastes, the constituents of concern are nonvolatile metals, making volatilization a very unlikely pathway of release from tanks. For the hydrogen cyanide sector, where volatile compounds are likely and tanks are used in wastewater treatment systems, the tanks will also be covered by other CAA regulations as described above. In addition, in many cases facilities have installed tank covers, further reducing the likelihood of release to the air. As a result, we have not modeled releases to air from tanks for any wastes in this listing determination.

b. Waste solids management.

We concluded that we did not need to model any releases of volatile constituents from solids for the same reasons set out above. The management practices of concern for waste solids were landfills, including disposal in on-site and off-site landfills, and in a few cases, waste piles. We evaluated the potential for groundwater releases from all landfills and piles. We also considered the possibility of releases of airborne particulates by a multistep process where we compared the total concentrations of the constituents of concern to a series of soil screening levels (see section III.E.3).

4. Evaluation of Secondary Materials

RCRA gives EPA jurisdiction only over materials that are discarded. EPA's current definition of discard is set out in the definition of solid waste at 40 CFR 261.2. Under this approach, process residuals (or “secondary materials”) destined for recycling are solid wastes within our jurisdiction if the recycling closely resembles waste management. Conversely, if the materials are recycled as part of an ongoing manufacturing process, they are not solid wastes. The existing rules specifically exclude secondary materials from jurisdiction that are used directly (without reclamation), as ingredients in manufacturing processes to make new products, used directly as effective substitutes for commercial products, or returned directly to the original process from which they are generated as a substitute for raw material feedstock. 40 CFR 261.2(e). In addition, the existing rules allow for closed loop reclamation where secondary materials can be reclaimed and returned to the original production process provided that the entire process is closed, the reclamation does not involve controlled flame combustion, and the reclaimed material is not used to produce a fuel or a material that is used in a manner constituting disposal. (40 CFR 261.4(a)(8)) As discussed in the January 4, 1985, rulemaking, these are activities which, as a general matter, resemble ongoing manufacturing operations more than conventional waste management and so are more appropriately classified as not involving solid wastes. However, materials which would otherwise qualify for exclusion under these provisions are not excluded if EPA finds that the recycling is not legitimate. EPA considers a variety of economic and chemical factors when it determines whether or not a specific recycling practice is legitimate. (See Memorandum from Sylvia K. Lowrance, Director Office of Solid Waste, concerning F006 Recycling, dated April 26, 1989). These determinations are very site-specific and tend to be very time consuming. EPA typically makes them in the context of site-specific enforcement or permitting actions.

The existing rules, however, do not exclude materials that are either contained in or used to produce fuels or that are directly used, or incorporated into a product that is used, in a manner constituting disposal. EPA asserts RCRA jurisdiction for these types of use/reuse

circumstances as they more closely resemble conventional waste management rather than ongoing manufacturing. (See 50 FR 637-640, January 4, 1985).

A series of court decisions also address the issue of our jurisdiction over recycled materials. In general, they hold that EPA lacks authority to regulate materials that are immediately reused in an ongoing manufacturing or industrial process.

American Mining Congress

v.

EPA

(824 F. 2d 1177 (D.C. Cir. 1987) (AMC I));

American Mining Congress

v.

EPA

(907 F. 2d 1179, 1186 (D.C. Cir. 1990) (AMC II));

American Petroleum Institute

v.

EPA

(216 F. 3d 50 (D.C. Cir. 2000)). The most recent decision,

Association of Battery Recyclers, Inc.

v.

EPA

(208 F. 3d 1047 (D.C. Cir 2000)), remanded a rule regulating the reuse of some closely related materials. We are still evaluating the impacts of this decision. However, the remand does not affect this rule because we are not relying on the exemptions in the remanded rule.

6

6

On May 26, 1998, we promulgated a conditional exclusion from the definition of solid waste for secondary materials (other than listed wastes) generated within the primary mineral processing industry from which minerals, acids, cyanide, water, or other values are recovered by mineral processing, with certain provisions. Because this conditional exclusion only applied to non-listed wastes, and we were making listing determinations, we did not use this exclusion as a basis to not evaluate wastes for listing purposes. On April 21, 2000, the D.C. Circuit Court issued a decision vacating a portion of this conditional exclusion. [See

Association of Battery Recyclers, Inc.

v.

EPA

. 208 F.3d 1047 (D.C. Cir. 2000)].

For almost all of the residual materials from these manufacturing processes which are re-used or recycled in some way, we decided not to attempt to determine whether the recycling practice is not subject to regulation under the court decisions and regulations described above. Such determinations can be very time-consuming, particularly where we find recycling practices that appear not to be regulated, and then need to determine whether or not such practices are legitimate. Consequently, we decided that it would be more efficient to examine first the potential risks posed by the reported recycling practices. If we found no significant risks, we would decide not to list the material. If, on the other hand, we found risks, we evaluate the recycling practice prior to making a listing decision.

To assess the risks of materials recycled on-site by reusing them in one of the consent decree manufacturing processes, we first evaluated the management of the materials prior to their re-use. We looked for closed piping, covers on containers, or similar barriers to releases to the environment. Where we found such management practices, we determined that there was no significant potential for releases. We then evaluated the potential for releases from the consent decree process itself. We found that the only points at which releases were expected were either those where we were already evaluating solid wastes for the purposes of this listing or points where the facility released uncontained gases outside of RCRA jurisdiction. Consequently, we felt that we were evaluating all of the potential risks (within our jurisdiction) associated with the recycling of these materials. In the antimony oxide sector, however, we found one residual that was being held in containers for several years for potential reuse. Our rules identify this practice as “speculative accumulation” and classify the materials held in such a manner as solid wastes. Accordingly, we assessed the risks posed by these accumulated wastes.

We found that a few materials are inserted into separate manufacturing processes co-located on-site with consent decree processes. We evaluated the potential for releases prior to reinsertion into that separate process. However, as explained above in section III.B, we did not evaluate any risks posed by use of residuals in processes that are not subject to our consent decree deadline.

We also considered the risks of materials recycled off-site. We considered the potential for release before the materials were transferred off-site. We did not assess the off-site uses which involved non-consent decree manufacturing processes. In a few cases, however, we found that the reuse involved land placement or burning for energy recovery. These activities are always regulated as waste management under the rules and court decisions described above. In those cases, we concluded that the materials were wastes from the consent decree process where they were generated, and we evaluated risks posed by the use. For example, we evaluated the risks posed by use of residual materials from the production of boric acid as fuels for cement kilns. In one case involving antimony oxide residuals, we found that the residuals were sent off-site to another smelter producing antimony oxide. This smelter happens to be located outside of the country. We did not evaluate risks from its residuals, as we have no legal jurisdiction to regulate them. We have evaluated the production of antimony oxide within the U.S. in this rulemaking, so we have evaluated the risks that would be posed if this generator changed its practice and sent the materials to an antimony oxide smelter located within the U.S.

For purposes of convenience, in the sector specific discussions below (and in the various background documents) we describe all of the residuals as wastes. We emphasize, however, that we have not determined whether any of the residuals that are recycled are solid wastes as defined in 40 CFR 260.2. We believe it is more appropriate to leave such site-specific determinations to other decision-making processes.

E. Description of Risk Assessment Approaches

Before turning to the details of the risk assessment approaches used, we want to highlight two general issues. First, we note that for this proposal we used a variety of screening methodologies to assess a large number of wastes. Due to time constraints imposed by the consent decree schedule, we chose —where appropriate— to use these methodologies rather than conducting more time-consuming, full-scale, risk assessment modeling. In general, however, we believe that these screening methodologies conservatively assessed risks, so that wastes that we “screened out” are unlikely to present significant risks.

Second, we want to describe our selection of plausible mismanagement practices for both screening and full modeling assessments. In general, we assessed the types of management units which, according to data available to us, facilities have actually used or contemplated using. Frequently, we found that facilities had made economic investments that would make them likely to continue to use the same types of units. For example, where facilities had paid to install tanks to store or treat wastes, we assumed that they would continue to use tanks rather than place wastes in pits or surface impoundments. Furthermore, we found that some waste quantities were so large that it would be prohibitively expensive to transport wastes off-site. Similarly, where facilities had installed piping to return residual materials to their production processes, we assumed that they would continue to use these systems to recycle those residuals. We also assumed that such facilities had found it more economical to return those residuals to their processes, and were thus not likely to send them to landfills or other types of disposal units.

We seek comment on all data, assumptions and methodologies used in our risk assessment for this proposal.

1. What Risk Thresholds Were Used?

EPA's listing program generally defines risk levels of concern for carcinogens as risks within or above a range of 1×10

-6

to 1×10

-4

(from 1 in 1,000,000 to 1 in 10,000) at the upper end of the risk distribution (

e.g.,

90th or 95th percentile risk for a particular exposure scenario). The level of concern for non-cancer effects is generally indicated by a hazard quotient (HQ) of 1 or greater at the upper end of the distribution. Consistent with the listing policy described in the dyes and pigments proposal (59 FR 66075-66078) we used a 1×10

-5

risk level and/or HQs of one to identify which wastes are candidates for listing. To make a listing determination, we then used a weight-of-evidence approach that considers the risk estimates along with other information related to the factors described in 40 CFR 261.11(a)(3). For cancer, a risk threshold of one in 1,000,000 represents the probability that an individual will develop cancer over a lifetime as a result of exposure to a chemical contaminant. When we estimate the lifetime excess cancer risk, we use an upper bound estimate of the carcinogenic slope factor (CSF) as derived from laboratory studies in animals or from human epidemiological studies. In addition, because the CSF typically relies on a number of extrapolations (

e.g.,

from animals to humans and from high doses to low doses) there is some uncertainty in the value of the CSF.

For non-cancer effects, which include a wide variety of health effects, we used EPA's reference dose (RfD) as a risk threshold. A reference dose is an estimate of an oral exposure that is likely to be without an appreciable risk of adverse effects in the general population, including sensitive individuals, over a lifetime. The RfD can be derived from a NOAEL, LOAEL, or benchmark dose. Uncertainty factors are applied to address limitations of the available toxicological data and are necessary to ensure the RfD is protective of individuals in the general population. The use of uncertainty factors is based on long-standing scientific practice. Uncertainty factors when combined commonly range from 10 to 1000 depending on the nature and quality of the underlying data. The RfD methodology is expected to have an uncertainty spanning perhaps an order of magnitude. To assess risks associated with non-cancer effects, we used a hazard quotient (HQ), which is defined as the ratio of the estimated dose of a given chemical to an individual to the reference dose for that chemical. A hazard quotient (HQ) of one (1) indicates that the estimated dose is equal to the reference dose (RfD) and, therefore an HQ of 1 is EPA's threshold of concern for non-cancer effects. Usually, doses less than the RfD (HQ<1) are not likely to be associated with adverse health risks and, therefore, are less likely to be of regulatory concern. As the frequency and/or magnitude of the exposures exceeding the RfD increase (HQ>1), the probability of adverse effects in a human population increases. However, it should not be categorically concluded that all doses below the RfD are “acceptable” (or will be risk-free) and that all doses in excess of the RfD are “unacceptable” (or will result in adverse effects).

The values of the CSF and RfD that we use for assessing risks are generally taken from EPA's on-line toxicity data base called IRIS. However, in some cases we used EPA's compilation of toxicity benchmarks known as HEAST or other sources, such as toxicological issue papers prepared by EPA's National Center for Environmental Assessment (NCEA).

7

7

EPA's Integrated Risk Information System (IRIS) may be found at http://www.epa.gov/iris. See also “Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wastes” (August 2000) for a discussion of the toxicity benchmark values used in today's rule.

2. What Leaching Procedures Were Used?

As noted in III.C, we used the TCLP and SPLP leaching procedures to evaluate the wastes in today's rule. EPA developed the TCLP as a tool to predict the leaching of constituents from the waste in a municipal solid waste landfill, and the TC regulations use this method to determine if a waste is hazardous under 261.24 (see the Toxicity Characteristic rule, 55 FR 46369; November 2, 1990). We have also used the TCLP in the listing program to estimate leaching concentrations for use in groundwater modeling (for example, see the recent petroleum listing, 63 FR 42110, August 6, 1998). We believe the TCLP is the most appropriate leaching procedure to use for wastes in municipal landfills, because the leaching solution is similar to the type of leachate generated from the decomposition of municipal waste. The TCLP leaching solution is a solution containing acetic acid that is adjusted to a pH of 4.93 or 2.88, depending on the acidity of the waste sample.

EPA developed the SPLP as a method to predict leaching from wastes or soils under exposure to the slightly acidic, dilute solution generated by normal rainfall. The SPLP test uses a leach solution which mimics acid rain, while the TCLP uses a leach solution which mimics acids formed in municipal landfills. In past actions, EPA has recognized that the TCLP's use of organic acids may not be appropriate for disposal scenarios that do not involve municipal landfills. For example, in the proposed rule for management and disposal of lead-based paint debris, EPA used the SPLP to assess leaching from landfills that do not accept municipal wastes (see 63 FR 70189; December 18, 1998). Similarly, EPA utilized the SPLP in screening low hazard wastes as part of its 1989 Bevill determination (see 54 FR 36592; September 1, 1989).

In the context of EPA's more recent mineral processing sector actions, we considered the relative merits of both the TCLP and the SPLP for various wastes in the mineral processing industries; EPA decided to continue to rely on the TCLP for defining characteristically hazardous Bevill wastes, in part because we found that disposal in municipal landfills did occur for some sectors. See the Land Disposal Restrictions Phase IV Final Rule at 63 FR 28598 (May 26, 1998). For today's rule, however, we have specific data showing that some wastes do not go to municipal landfills and are unlikely to be disposed of in municipal landfills. We used the SPLP sampling results for wastes that were not likely to go to municipal landfills, and we used the TCLP results for wastes going to municipal landfills.

3. How Were Wastes Screened To Determine If Further Assessment Was Needed?

We used a number of approaches to eliminate from further consideration those wastes that could not plausibly pose unacceptable risks. This served to identify those wastes and chemical constituents that required further assessment. Different screening approaches were used depending on the type of waste management practices employed in the industry and, in some instances, the waste volume and the location of the waste management units.

For wastes that are managed in landfills, groundwater contamination is the primary source of human exposures, particularly for certain metals and other inorganic compounds that are nonvolatile, such as those present in the wastes that are the subject of today's rule. We compared leachate concentrations derived from the TCLP or SPLP test measurements to levels in drinking water that are protective of human health. These levels, referred to as health-based levels (HBLs), are designed to be protective of both

children and adults. Health-based levels (HBLs) are levels in environmental media that would not exceed EPA's risk thresholds given conservative assumptions regarding exposure (

e.g.

, a level in drinking water that would not exceed a risk threshold for an individual whose drinking water intake was at the high end of the distribution for the general population).

8

Although an HBLs represents a concentration level at the point of exposure, we conservatively assumed direct contact with the wastes (

i.e.,

no dilution) for the purpose of screening out wastes and chemical constituents that could not pose unacceptable risks and, therefore, do not merit further analysis. As explained previously, we used SPLP measurements for wastes that are managed in landfills containing only industrial wastes and TCLP measurements for wastes that are managed in landfills which also contain municipal wastes. For wastewaters that are managed in surface impoundments, we used the concentration in the filtered liquid (

i.e.

, the SPLP filtrate) because the filtrate is more representative of the fraction of the waste that could infiltrate into the subsurface environment. Regardless of the type of measurement, if the result of the chemical analysis for a particular compound was below the limit of detection but the compound was detected in the waste, then we used

1/2

the value reported by the laboratory as the limit of detection for that compound. Any chemical contaminant in a waste that did not screen out against HBLs (

i.e.,

the waste concentration was a factor of 2 or less times the HBLs

9

) we identified as a constituent of concern (CoC) requiring further assessment. However, very low volume wastes were subject to further screening, as described below.

8

Details on how HBLs are derived may be found in the risk assessment background document for today's proposal, “Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wastes” (August 2000).

9

We used professional judgment to screen out constituents with concentrations within a factor of two of the HBLs. Historically, our models have suggested that the dilution and attenuation of constituents in the subsurface will generally result in dilution and attenuation factors (DAFs) of 2 or greater. (See, for example, the DAFs estimated for the petroleum refining listing determination, 63 FR 42110, and the docket for today's rulemaking for DAFs calculated to support today's proposal.) If our consideration of site-specific factors had indicated that a very low DAF were likely for actual exposure (

e.g.

known drinking water wells placed very close to the management until boundary), we would have modeled that waste rather than screening it out using professional judgment.

For very low volume wastes that did not screen out against HBLs, we performed an additional conservative screen to determine if the waste could plausibly pose a risk to human health when disposed of in a landfill. Typically wastes generated in volumes of less than 1 or 2 metric tons per year were considered as candidates for this

de minimis

analysis. This analysis assumed that the entire mass of the chemical contaminant in a volume of waste that is generated in a year's time would leach out of the waste and infiltrate into groundwater in the same year. The only dilution that was assumed to occur was with the volume of water that infiltrated into the landfill. To minimize the amount of dilution we chose a conservative infiltration rate based on the infiltration that could occur for a relatively low permeability soil underlying a relatively small landfill (corresponding to the 10th percentile of the distribution of municipal landfill areas nationwide). However, in some cases the resulting infiltration was less than the amount of water that would be withdrawn from a well by a household for domestic usage. In these instances, we diluted the infiltrate into the minimum volume of water needed to support a household well, which we estimated from data on U.S. per capita water consumption assuming a family of four. The concentration derived using this procedure was then compared to the HBLs. Any chemical contaminant that did not screen out as a result of this analysis we identified as a constituent of concern (CoC) requiring further assessment. While we do expect the

de minimis

screen to be conservative overall, the degree to which it is conservative depends on many waste and site-specific factors. (For example, our sampling and analysis data indicate that in some cases essentially all of the chemical constituent leached out of the sample over the duration of the leach test.)

For wastes managed in waste piles and landfills, we performed a multi-level screening analysis to determine if further assessment of the air pathway was needed. Wind blown dust from wastes managed in piles is a potential source of human exposures. This pathway is also possible for landfills, but likely to result in much lower releases due to the common usage of daily and longer-term cover at landfills. In the first level screen we compared the waste contaminant total concentrations to background levels in soils. Background soil levels were taken from published compilations of levels in native soils nationwide and were generally characterized using a geometric mean or (in a few instances) an arithmetic mean concentration of the available data.

10

If the waste concentrations exceeded background levels in soils, we performed a second level screen by comparing the waste concentrations with soil ingestion HBLs. Soil ingestion HBLs assume direct contact with the waste and, therefore, are more conservative than HBLs based on inhalation exposures. In those instances when the waste concentrations exceeded both background levels and soil ingestion HBLs, we performed a third level screen using the results of EPA's air characteristics study. This study developed levels of chemical contaminants in wastes that are protective of human health with respect to inhalation exposures when managed in a variety of ways.

11

In particular, air characteristic levels were developed for waste piles at several different distances from a potential receptor. We used the air characteristic levels corresponding to a downwind distance of 25 or 150 meters (80 or 500 feet). Because the air characteristic levels include the effect of atmospheric dilution, they are significantly higher than soil ingestion HBLs. In most cases waste concentrations are either below background or below soil ingestion HBLs for the wastes EPA evaluated. Moreover, we found no instances in which air characteristic levels are exceeded. In the cases where waste concentrations exceeded the soil ingestion levels, the exceedence was typically less than a factor of 2 to 3. We believe it is highly unlikely that off-site exceedences due to windblown dust from piles or landfills would actually exceed the soil ingestion levels given this low level of exceedence in the waste. Therefore, we conclude that risks associated with particulates from piles and landfills transported by an air pathway are not significant and no further assessment is needed.

10

Different statistics may be used for characterizing background levels depending on the data available. The mathematical properties of the arithmetic mean allow it to be used when only average values rather than the original data are available. However, if the original data are available, the data can be pooled and a geometric mean can be calculated. If the data are positively skewed, as is often the case, the arithmetic mean will be higher than the geometric mean. We consider either statistic to be a central tendency measure of background levels. However, background levels are highly variable and may be considerably higher or lower than the national average at any given location. See, for example, the U.S. Geological Survey paper “Elemental Concentrations in Soils and Other Surficial Materials in the Conterminous United States,” paper 1270, U.S. Government Printing Office, 1984.

11

See U.S. EPA, “Revised Risk Assessment for the Air Characteristic Study,” Office of Solid Waste, EPA 530-R-99-019, November 1999.

EPA derived HBLs for chemical contaminants from toxicity benchmarks and a set of exposure assumptions that differ depending on the type of health effect and exposure pathway. For carcinogenic effects, HBLs were derived from a cancer slope factor (CSF) for the oral route of exposure. For non-cancer effects, HBLs were derived from EPA's oral reference dose (RfD) for the compound. Risk thresholds were as described previously. Drinking water and soil ingestion HBLs for individual chemical contaminants are presented elsewhere.

12

The exposure assumptions we used for deriving the HBLs are described as follows.

12

See the risk assessment background document for today's proposal, “Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wasters” (August 2000).

For drinking water exposures, we derived HBLs for carcinogenic effects for an adult exposed for 30 years and having a tap water intake of 1.4 liters per day. This represents 21 milliliters per day on a per kilogram body weight basis, which is the mean tap water intake for adults. A duration of exposure of 30 years represents the 95th percentile of the distribution of residential occupancy periods for adults nationwide. We derived HBLs for non-cancer effects for a child having a tap water intake of 1.3 liters per day. This represents 64 milliliters per day on a per kilogram body weight basis and corresponds to the 90th percentile of the distribution of tap water intakes in children that are 1 to 10 years of age.

13

Because the drinking water HBLs incorporate conservative exposure assumptions, we consider them to be appropriate for screening purposes.

13

See EPA's “Exposure Factors Handbook” (EPA/600/P-95/002Fa), August 1997, for additional details on human exposure factors.

Soil ingestion HBLs were derived from either the CSF or the RfD assuming a soil ingestion rate of 200 milligrams per day and an exposure duration of 8 years. A soil ingestion rate of 200 milligrams per day (about 3/100th of a teaspoon) is a conservative estimate of the mean intake rates for children in the age range of 1 to 7. An exposure duration of 8 years is an estimate of the mean residential occupancy period for a 6 year old child. In selecting these values for use in deriving soil ingestion HBLs, we considered the likelihood that children would actually come into direct contact with the wastes.

In cases where wastes are known to be managed in on-site landfills or surface impoundments that are located adjacent to or in close proximity to surface waters, we used additional screening criteria to identify wastes that could have the potential to adversely impact surface waters before eliminating the wastes from further consideration. We used EPA's national water quality criteria for this purpose. Specifically, we compared waste concentrations (

i.e.

, SPLP measurements for wastes managed in on-site landfills and SPLP filtrate measurements for wastes managed in surface impoundments) directly to ambient water quality criteria that have been established for the protection of both human health and aquatic life. Any chemical contaminant in a waste managed under these circumstances that did not screen out against ambient water quality criteria (within a factor of 2) we identified as a constituent of concern (CoC) requiring further assessment.

EPA recently republished ambient water quality criteria for a large number of chemical contaminants (see 63 FR 68354; December 10, 1998). Separate criteria for the protection of aquatic life have been established for fresh water and salt water. In a number of instances waste management units are located adjacent to estuarine environments. In these cases, for screening purposes, we used the lower of the fresh water and salt water criteria.

4. How Was the Groundwater Pathway Evaluated?

We conducted modeling analyses to assess possible risks to human health from wastes managed in land-based units such as landfills and surface impoundments. We used fate and transport models to estimate contaminant concentrations that might occur in a residential drinking water well from migration of uncontrolled releases of leachate from a waste management unit through the subsurface environment. We assessed human exposures to these contaminants from information on the amount of tap water an individual drinks and the length of time an individual might reside at a residence and utilize water from a residential well. We then assessed what the human health risks would be as a consequence of such exposures.

We took a probabilistic approach to the assessment of human exposures. In this approach, we used Monte Carlo simulation techniques to determine the distribution of groundwater concentrations to which an individual could be exposed and combined this with distributional data for the general population on the intake rates of tap water and the duration of exposure. We then assessed the risks to human health from both the middle (central tendency) and upper (high end) portions of the distribution of human exposures. EPA defines high end as the 90th percentile and greater of the distribution of exposures in the population. Central tendency generally refers to the mean or 50th percentile of the distribution. Central tendency and high end estimates may be generated using either probabilistic or deterministic approaches.

14

14

We relied upon the probabilistic risk estimates for today's proposal. However, both deterministic and probabilistic approaches are presented in the risk assessment background document.

We evaluated potential groundwater exposures over a 10,000 year time period. Evaluating peak doses over this time horizon allows the model to capture the slow movement of some chemicals through the subsurface. While exposure assumptions (

e.g.,

land use patterns, climate, environmental and other exposure assumptions) are expected to change over 10,000 years, such changes are difficult to predict. We believe such a time period is appropriate to ensure human health is protected. Even with long time periods, we are still concerned with the risk that would result once contamination reaches potential drinking water wells. Given that the metals of concern do not degrade in the environment, we believe a long modeling time period is necessary. Further, there is uncertainty in when peak concentrations at the receptor well may occur, and using the 10,000 year time frame makes it more likely that we will capture the peak risk in our evaluation. EPA has used similar time horizons for groundwater modeling in past hazardous waste rules.

15

15

See HWIR proposal at 64 FR 63429, November 19, 1999, and the final rule for the recent listing of wastes from petroleum refining at 63 FR 42157, August 6, 1998.

For modeling chemical concentrations in ground water, many input parameters were varied. These included waste characterization data (

e.g.,

chemical concentrations and waste volumes), waste management practices (

e.g.,

waste management unit size and infiltration rates), hydrogeological parameters (

e.g.,

depth to water table, hydraulic conductivity, and aquifer thickness), and chemical parameters (

e.g.,

soil-water partition coefficient). We conducted extensive sensitivity analyses to determine which of these parameters had the greatest influence on the risk results. For a detailed discussion of the ground water analysis, including parameter distributions, input assumptions, and sensitivity analyses, see the risk assessment background document for today's proposal, “Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wastes” (August 2000).

In assessing groundwater exposures for wastes managed in off-site landfills, we considered the locations of every industrial and municipal landfill known to receive the wastes and the volume of wastes managed at each of these sites. In so doing, we considered only that volume of waste that is currently not being managed as hazardous waste. For wastes managed on-site by multiple facilities, we generally considered only those facilities where groundwater exposures are expected to be the highest. These locations were identified by considering the concentration levels of chemical constituents in the waste managed at the site and the proximity of on-site waste management units, namely landfills and surface impoundments, to potential off-site receptors. Our rationale for selecting particular locations for conducting modeling analyses is discussed in section III.F for the specific inorganic sectors.

a.

How were contaminant concentrations in groundwater modeled?

For modeling fate and transport in the subsurface environment, we used the groundwater model EPACMTP (EPA's Composite Model for Leachate Migration with Transformation Products). The model consists of two coupled modules: (1) A one-dimensional module that simulates infiltration and dissolved contaminant transport through the unsaturated zone, and (2) a three-dimensional saturated zone flow and transport module. Fate and transport processes accounted for in the model are advection, hydrodynamic dispersion, sorption equilibria, hydrolysis, and dilution from recharge to the saturated zone. The model assumes that the soil and aquifer are uniform porous media. EPACMTP (as used in this analysis) does not account for heterogeneity of the aquifer or for preferential migration pathways such as fractures and macro-pores or for colloidal transport, any or all of which could be important at a particular site. Although EPACMTP simulates steady-state groundwater flow in both the unsaturated zone and the saturated zone, the model (as used in this analysis) simulates contaminant transport from a finite source and predicts the peak contaminant concentration arriving at a down-gradient groundwater well. Only migration of chemical contaminants within the surficial aquifer is modeled by EPACMTP. We did not model migration of contaminants to deeper aquifers but, instead, based our assessment on exposures that might occur from groundwater withdrawn from the uppermost aquifer where contaminant concentrations are expected to be the highest.

Equilibrium sorption of chemical contaminants onto soil and aquifer materials is parameterized in the EPACMTP model using a soil-water partition coefficient (Kd). For today's proposed rule, we used values for Kd that have been derived from field studies and have been published in the scientific literature. An empirical distribution was used to characterize the variability of Kd for chemical contaminants for which sufficient published data were available. However, for several chemical contaminants having relatively few published values (

e.g.,

antimony and thallium), a log uniform distribution was used.

16

Our use of empirically derived partition coefficients assumes that sorption is linear with respect to groundwater concentration (

i.e.,

the Kd isotherm is linear). However, sorption is not unlimited and will tend to level off as groundwater concentrations increase beyond the linear range (

i.e.,

the Kd isotherm becomes non-linear). This condition is most likely to occur in the unsaturated zone where dilution is limited, if leachate concentrations are sufficiently high.

16

A log uniform distribution is a distribution that has equal probabilities at all percentiles when the parameter is transformed into logarithms. For these chemical constituents, we used a log uniform distribution that was centered on the geometric mean of the available data and had a width of 3 logs. This was done to better account for the variability normally seen in measurements of Kd.

EPA has sometimes used the MINTEQA2 equilibrium speciation model to estimate Kd's for a variety of metals rather than relying solely on field measurements. However, recently a number of technical issues have been raised concerning the model and its application. EPA is in the process of evaluating the model to address those issues. Therefore, we have decided not to use MINTEQA2 for today's proposed rule. Once the evaluation is completed and the issues are satisfactorily resolved, EPA may again choose to use the model in an appropriate form in future rulemakings.

Infiltration of leachate from landfills into the subsurface is modeled using the HELP model (Hydrologic Evaluation for Landfill Performance), a quasi-two-dimensional hydrologic model used to compute water balances for landfills. We assumed that landfills have a final earthen cover but no liner or leachate collection system. The net infiltration rate that is calculated by the model considers, among other factors, precipitation, evapotranspiration, and surface runoff and depends on the type of soil and the climate where the landfill is located. For surface impoundments, the infiltration rate is estimated from the liquid depth in the impoundment and from the hydraulic conductivities and thicknesses of the sediments and the underlying soil. We assumed that surface impoundments have no liner or leachate collection system. Unconsolidated or loose sediments are treated as free liquid so that the pressure head on the underlying, consolidated sediments is determined by the depth of the liquid in the impoundment and the depth of the unconsolidated sediments. As sediment accumulates at the base of the impoundment, the weight of the liquid and upper sediments acts to compress (or consolidate) the lower sediments. The result is the formation of a consolidated sediment layer having a hydraulic conductivity that is much lower than the previously unconsolidated sediment.

We assumed that landfills have an operational life of 30 years.

17

In landfills, leaching of contaminants from the waste leads to an exponential decrease in the leachate concentration with time. The rate at which this occurs depends on the volume of waste disposed of in the landfill and the total concentration of chemical contaminants in the waste. We used the measured TCLP concentration (for disposal in a municipal landfill) or SPLP concentration (for disposal in an industrial landfill) as the initial leachate concentration for modeling. In contrast, we assumed that surface impoundments have an operational life of 50 years.

18

Many surface impoundments are periodically dredged and, therefore, can be maintained in service for longer periods of time. With surface impoundments, leachate concentrations are not expected to decrease over time and, therefore, leachate concentrations are assumed to remain constant during their operational life. We used the total concentration of chemical contaminant measured in the wastewater or (for wastewaters with high levels of solids) the concentration measured in the SPLP filtrate as the leachate concentration for modeling.

17

U.S. Environmental Protection Agency, “Draft National Survey of Solid Waste (Municipal) Landfill Facilities,” Office of Solid Waste, Washington, D.C., 1988 (EPA/530-SW-88-034).

18

See assumptions made for the recent proposed hazardous waste identification rule at 64 FR 63382; November 19, 1999.

The fate and transport simulation modules in EPACMTP are linked to a Monte Carlo module to allow quantitative consideration of variability

and uncertainty in groundwater concentrations due to variability and uncertainty in model input parameters. We use a regional site-based methodology to associate the appropriate regional climatic and hydrogeologic conditions to the location of actual waste management sites. This methodology accommodates dependencies between the various model input parameters. In this approach, a site location is assigned to one of 13 hydrogeologic regions and one of 97 climatic regions that are linked to databases of climatic and hydrogeologic parameters. A climatic data set provides infiltration and recharge values for three soil textures at each of 97 climatic centers in the contiguous United States. The soil textures are based on a Soil Conservation Service soil mapping database and U.S. Department of Agriculture definitions of coarse, medium, and fine soil textures. (These textures are represented in EPACMTP by sandy loam, silt loam, and silty clay loam, respectively.) Infiltration rates for the waste management unit and the recharge rate for the surrounding region were determined for each soil type and climatic center using the HELP model. A site location is generally assigned to the climatic center that is geographically closest to the site.

Each site location is also located on a groundwater resource map (from a U.S. Geological Survey inventory of State groundwater resource maps) and a hydrogeologic region is assigned to the site based on the primary aquifer type at that location. A hydrogeologic database provides a distribution of values for depth to groundwater, aquifer thickness, hydraulic gradient, and hydraulic conductivity for each of 13 hydrogeologic regions. The hydrogeologic data base (HGDB) was developed from a survey of hydrogeologic parameters for approximately 400 hazardous waste sites nationwide. These site-specific data were then regrouped according to hydrogeologic classifications, and a distribution of parameter values was generated for each of the 13 hydrogeologic regions (made up of 12 specific hydrogeologic environments and one miscellaneous category). In the analysis for today's rule, we modified the above approach for on-site waste management units to enable available site-specific information on depth to groundwater to be used in place of the values found in the database.

We also used a regional site-based methodology to associate the appropriate soil characteristics to a given site location. In this approach, a distribution of soil textures at a site is determined by associating the site location with a soils classification region. We defined soil classification regions from information on the soil types found within a 100 mile radius of the site location. The distribution of soil textures for the region was determined by identifying the soil texture classifications from data contained in the U.S. Department of Agriculture (Natural Resources Conservation Service) STATSGO (State Soil Geographic) data base. The predominant soil textures within each mapping unit (which represents a collection of soils) were identified and the fraction of the three soil textures used in the EPACMTP model were determined (i.e, sandy loam, silt loam, and silty clay loam). These soil classification regions were used for modeling off-site municipal and industrial landfill sites. A similar approach was taken for on-site landfills and surface impoundments except that the predominate soil textures from mapping units that correspond to the site location itself were identified. These were compared for consistency with other soils information available for the site. Once the fraction of the three soils textures is determined for a given site location, a distribution of soil parameter values is generated from information on the distribution of soil parameter values for the three soil textures and the fraction of each soil texture for the site. These parameters are used for modeling groundwater flow and contaminant transport in the unsaturated zone and include saturated conductivity, moisture retention properties, water content, and organic matter content.

A full description of the groundwater modeling analyses conducted for today's proposed rule may be found in the background document, “Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wastes” (August, 2000).

b. How were human exposures assessed?

Our assessment of human exposures to contaminated ground water is based on a residential drinking water scenario. A different approach was used for determining the location of exposure depending on whether the wastes are managed on-site or are shipped off-site for disposal. For waste shipped to off-site municipal landfills, we used EPA's National Survey of Municipal Landfills

19

to determine the distance from the landfill to the receptor well. We also used these same data for off-site industrial landfills. For wastes managed on-site in either landfills or surface impoundments, we attempted to determine the closest point at which a residential well could be located and, therefore, the point at which human exposures could plausibly occur. We considered the location of the facility property boundary, the type of land use adjacent to the property boundary, the presence of surface waters that could intercept ground water flow, utilization of ground water for residential or agriculture uses, and the existence of residential drinking water wells in the direction of ground water flow. For both on-site and off-site waste management, we assumed the receptor well was located down-gradient from the waste management unit and that ground water is withdrawn from the top ten meters of the aquifer and within the lateral extent of the contaminant plume. Exposures were further assumed to occur out to a distance of a mile from the waste management unit.

19

Ibid.

Our assessment of human exposures did not consider naturally occurring background levels in ground water. Background levels in ground water are not a significant source of human exposure for several of the more important chemical constituents in the wastes that are the subject of today's proposal (

e.g.

, antimony and thallium). However, for manganese, dietary exposures are a significant source of background exposures. We did not attempt to quantify the cumulative risks from both dietary and drinking water exposures combined and, therefore, this is a source of uncertainty in our assessment of risks from manganese in these wastes.

Human exposures were characterized in terms of lifetime average daily dose (LADD) and average daily dose (ADD) for both children and adults. We used the LADD for assessing cancer risks and the ADD for assessing risks from non-cancer effects (including reproductive, developmental, neurological, cardiovascular, hematologic, metabolic, and a wide variety of other physiologic effects). Exposures to children of age one to six years and adults of age 20 to 64 years were assessed. We used information from EPA's Exposure Factors Handbook

20

to characterize tap water intake rates for individuals and residential occupancy periods of households (and, therefore, the length of time an individual could be exposed to contaminated ground water). Distributional data on tap water intake rates for individuals and residential occupancy periods for households were

used to generate both the ADD and LADD exposure estimates. For assessing lifetime exposures, we averaged the well water concentrations over the duration of exposure (

i.e.

, the residential occupancy period). We also averaged the tap water intake rates over the duration of exposure to account for the changes in tap water intake rates with age that are seen among children. For estimating the ADD, we used the peak 9-year average well water concentration but did not further average the estimated exposure (which we believe would be inappropriate given the range of possible health effects we want to protect against). Previous work with the EPACMTP ground water model has shown that the peak 9-year concentration and the maximum predicted concentration are nearly identical.

20

U.S. Environmental Protection Agency, “Exposure Factors handbook”, Volumes I and III, Office of Research and Development, National Center for Enironmental Assessment, Washington, DC., August 1997 (EPA/600/P-95/002Fa and c).

A full description of the methods and data used in the exposure assessment for today's proposed rule may be found in the background document, “Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wastes” (August, 2000).

5. How Was the Surface Water Pathway Evaluated?

A number of facilities that generate wastes covered by today's proposed rule are located adjacent to rivers or bays. As a consequence, the potential exists for subsurface releases of chemical contaminants from on-site management of the wastes to enter these river and bay systems through ground water inflow. In instances where no direct contact with ground water is likely to occur (as there would otherwise be if, for example, ground water was being used for residential drinking water), it becomes important to evaluate the potential water quality impacts of these releases on surface waters in the context of hazardous waste listings. However, we wish to emphasize that the surface water impacts considered in today's proposed rule are due to subsurface releases to ground water only. Direct discharges to surface waters are already regulated by the Clean Water Act under the NPDES permit system and are not considered further in today's proposal.

21

21

Industrial wastewater discharges that are regulated under the National Pollutant Discharge Elimination System (NPDES) Permit Program are specifically excluded from regulation as hazardous wastes under 40 CFR 261.4(a)(2).

We conducted a screening level analysis to evaluate potential surface water impacts. In this analysis, we estimated the volume of leachate that would infiltrate into ground water and assumed that this entire volume would be intercepted by surface water. Because this is a screening analysis, we made conservative assumptions that are likely to overstate the infiltration of leachate and, therefore, the potential release to surface water. For example, for on-site landfills, we assumed a soil type (sandy loam) that is likely to overstate the infiltration rate even in the absence of liners or leachate collection systems. Similarly, for surface impoundments we assumed a sludge thickness (8 inches) and soil type (sandy loam) that is likely to overstate the infiltration rate. In addition, we assumed no retardation in the migration of chemical contaminants in ground water due to sorption or other processes.

Due to the nature of these releases, which are likely to occur over a wide area, we assumed that the inflow of contaminated ground water was rapidly diluted into surface water and that there was little or no mixing zone. We followed EPA's Office of Water guidance for determining the design flows for rivers as regards water quality criteria. The appropriate design flow depends on the particular water quality impact being evaluated. For assessing potential impacts on aquatic life, we used the “7Q10” as the design flow. The 7Q10 is the seven day low flow with a return frequency of once every 10 years and is the recommended design flow for use with chronic water quality criteria for the protection of aquatic life. We believe that chronic water quality criteria are the appropriate criteria for evaluating the potential impact of continuing steady releases, such as those that would result from subsurface discharge of contaminated ground water. On the other hand, EPA generally uses the “30Q5” as the design flow for assessing potential impacts on human health. The 30Q5 is the thirty day low flow with a return frequency of once every 5 years and is the recommended design flow for use with water quality criteria for the protection of human health as regards non-cancer effects. However, a 30Q5 design flow was not available in all cases. In these instances, we estimated the 30Q5 based on the 7Q10 design flow.

22

For carcinogens (

e.g.,

arsenic), lifetime exposures are the primary concern and a design flow that corresponds to a longer averaging time is appropriate. For this reason, EPA recommends the long-term harmonic mean be used as the design flow.

23

The harmonic mean is always less than the arithmetic mean and is used in place of it because low flow conditions drive long-term average water quality. However, because this flow statistic was not available, we estimated the harmonic mean flow from the arithmetic mean flow and the 7Q10.

22

EPA guidance provides a simple rule of thumb for estimating the 30Q5 from the 7Q10 depending on the size of the river. For smaller rivers (defined as those with a low flow of 50 cfs or less), the 30Q5 is 1.1 times the 7Q10. For larger rivers (low flow of 600 cfs or greater), the 30Q5 is 1.4 times the 7Q10. See “Technical Support Document for Water Quality-Based Toxics Control,” EPA/505/2-90-001, March 1991.

23

The harmonic mean is defind as the inverse of the average of the sum of the inverses of the recorded flows.

As a result of the screening level analysis, all wastes screened out for which the ground water to surface water pathway was a concern. Therefore, no additional analysis of this pathway was conducted.

6. What Are the Limitations and Uncertainties of the Assessment?

Our assessment of exposures and risks is subject to a variety of limitations and uncertainties. These are discussed in some detail in the background document for today's proposed rule. A number of these are highlighted here.

We assumed our sampling and analysis data are fully representative of the range of wastes generated in the effected industries. However, our own data show that there are significant variations in waste concentrations across facilities in a given industry. Variability in waste concentration that is unaccounted for could lead to an over- or under-estimation of risks. However, any tendency toward under-estimation is likely to be mitigated to some extent by our selection of wastes and exposure scenarios that are intended to capture the highest risks.

We also assumed that our methods for measuring the leaching behavior of wastes (

i.e.,

the TCLP and SPLP test procedures) are both representative of the range of leaching conditions that exist under real world conditions and accurately quantify the concentrations of contaminants that leach into the subsurface environment from a given waste management unit. However, we know that many metals exhibit varying (or amphoteric) behavior with respect to pH and that any one test procedure is capable of characterizing leaching behavior only under a particular set of conditions.

The ground water model we used in our analysis (

i.e.,

EPACMTP) is designed to characterize dilution and attenuation in the subsurface environment under homogeneous conditions. The model does not account for subsurface heterogeneities, nor does it account for fractured flow or colloidal transport. These conditions, if present at

a site, can lead to less dilution and attenuation of contaminant levels than predicted by the model, causing ground water concentrations to be under estimated. In addition, sorption of metal species onto soil and aquifer materials exhibits considerable variability depending on geochemical conditions and the total concentration of the metal present at a given location. Although our use of empirically derived Kd values captures some of this variability, the available published data are fairly limited for certain metals (

e.g.

, antimony). We have accounted for the uncertainty associated with the small number of data points explicitly for these metals by expanding the range of Kd values used for modeling (to three orders of magnitude). Even for metals that have abundant data (e.g., arsenic), it is unlikely that the range of variability apparent in the data could exist at a given site.

Uncertainty associated with the specification of Kd as noted above could lead to an over- or under-estimation of risk. However, a tendency toward over-estimation is likely to be mitigated by the fact that under near steady-state conditions (when ground water impacts are the greatest), concentrations in ground water are little influenced by Kd. Under non-steady conditions, any tendency toward over- or under-estimation is limited by the variability inherent in the empirical distributions of Kd used in the analysis, which include both relatively high and relatively low values of Kd. Nevertheless, in general the risk estimates are sensitive to the specification of Kd and, therefore, this is an important source of uncertainty in our analysis.

As indicated previously, for wastes managed on-site we based our assessment of human exposures on the plausibility of ground water being used for drinking water. While some information was available on utilization of ground water for drinking water, very limited information was available from which to determine the location of exposure at a given site. For wastes managed off-site we assumed that ground water is used for drinking water (or will be in the future) and we used national data on the distribution of distances to residential wells to assess human exposures and risk. Our analysis did not consider possible changes in the location of on-site waste management operations in the future. These exposure assumptions (about which there is considerable uncertainty) may have an impact on the estimated risks and, therefore, the outcome of the risk assessment.

Other important uncertainties include those related to the health effects of chemical contaminants in humans (hazard identification), absorption and metabolism of ingested contaminates (pharmacokinetics), and biological response (dose-response relationships). These and other limitations and uncertainties are discussed in the background document, “

Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wastes

” (August, 2000).

F. Sector-Specific Listing Determination Rationales

We seek comments on all proposed listing decisions in this section, and the underlying rationales used to support our proposals.

1. Antimony Oxide

a. Summary.

We have evaluated antimony oxide production wastes and propose to list two wastes from this process as hazardous: (1) Baghouse filters and (2) slag that is disposed of or speculatively accumulated. We propose to list the baghouse filter waste under the criterion in 40 CFR 261.11(a)(1) because it routinely exhibits one or more of the characteristics of hazardous waste, but the waste is not consistently managed in compliance with Subtitle C regulations. We propose to list the slag under the criteria in 40 CFR 261.11(a)(3) because of risks associated with land disposal.

K176 Baghouse filters from the production of antimony oxide. (E)

K177 Slag from the production of antimony oxide that is disposed of or speculatively accumulated (T).

Other wastes generated by the antimony oxide industry do not meet the criteria set out at 40 CFR 261.11(a)(3) for listing a waste as hazardous. They do not pose a substantial present or potential threat to human health or the environment. We identified no risks of concern associated with the current management of these other wastes.

b. Description of the antimony oxide industry.

Antimony oxide was produced by four facilities in the United States in 1998. Antimony oxide is used as a flame retardant in plastics and textiles, a smoke suppressant, a stabilizer for plastics, an opacifier in glass, ceramics and vitreous enamels, and a coating for titanium dioxide pigments and chromate pigment.

The manufacturers use two different processes to produce antimony oxide. In the first process, antimony metal is roasted in the presence of air. The antimony oxide forms as a fume, cools and condenses in a baghouse. In the second process, crude (low grade) antimony oxide is roasted in the presence of air to produce higher grade antimony oxide. The antimony oxide cools and condenses in a baghouse. The crude antimony oxide comes either from off-site or is recycled from within the facility.

c. How does the Bevill Exclusion apply to wastes from the antimony oxide manufacturing processes?

Antimony oxide producers use a range of raw materials to produce antimony oxide, including antimony metal ingots, sodium antimonate, and antimony ore concentrate, and some facilities have claimed that wastes generated from the production of antimony oxide are Bevill exempt. Wastes generated from processes using either antimony ingots or sodium antimonate (both of which are saleable mineral products) are considered chemical manufacturing wastes rather than mineral processing wastes and are not eligible for the Bevill exemption. The September 1, 1989 Bevill final rule states at 54 FR 36620-21 that chemical manufacturing begins if there is any further processing of mineral product.

Two of the facilities also purchase an antimony ore concentrate as a raw material and place this material in kilns to produce antimony oxide. The smelting of a ore concentrate above the fusion point is defined as mineral processing (See 54 FR 36618). At these antimony oxide facilities, since mineral processing has begun, wastes from the process are not eligible for the Bevill exemption as beneficiation wastes (See 40 CFR 261.4(b)(7)(i)). In addition, although there is a Bevill exemption for 20 specific mineral processing wastes form various mineral processing sectors, the wastes generated from antimony oxide mineral processing are not included as one of these 20 wastes and are not excluded. (See 40 CFR 261.4(b)(7)(ii)). Thus there are no antimony oxide wastes that qualify for the Bevill exemption.

d. Wastes generated by these processes.

Table III-1 summarizes our information about the wastes generated from the production of antimony oxide:

Table III-1.—Antimony Oxide Production Wastes

Waste category

Number of

reported

generators

1998 volume

(MT)

Reported waste

hazard codes

Management practices

Antimony slag not recycled in process

3

113

D008

Sent to lead smelters for lead and/or antimony recovery; or on-site drum storage prior to future on-site land disposal.

Baghouse filters

4

9

No code reported

In-process antimony recovery; off-site antimony recovery; industrial Subtitle D landfill; or non-hazardous waste incinerator.

Empty supersacks

1

15

No code reported

Disposal in off-site Subtitle D landfill or recycled.

In addition to these wastes, there are other materials produced that are reused in the antimony oxide production process. Antimony oxide and antimony slag are captured at various points in the facility and reinserted into a furnace to produce antimony oxide, either on-site or off-site. Because these materials are managed prior to reuse in ways that present low potential for release, and because we evaluated process waste generated after the secondary materials are reinserted into the process, we do not believe that these secondary materials present significant risks.

e. Agency evaluation.

(1) Antimony slag not recycled in antimony oxide process.

How Are These Wastes Currently Managed?

Three facilities produced antimony slag that is not recycled in the antimony oxide process. Two of these facilities send the slag to lead smelters. One of the two facilities reported its slag to be TC hazardous because of its lead content (D008). The third facility, however, has historically stored a portion of its slag on-site in drums, reporting that they plan to reclaim antimony when antimony prices are more favorable. Recent revisions to the facility's Operating Permit,

24

however, require that the slag be placed in an on-site engineered “slag storage pit” to be constructed in the next two to three years.

24

“United States Antimony Corp. Stibnite Hill Mine Project Operating Permit 00045”, 6th review draft, January 1999. This draft permit is issued under the Metal Mine Reclamation Act, 82-4-301 MCA. It was prepared by the facility, approved by the State of Montana on August 12, 1999 (with a number of stipulations), and subsequently approved by the Forest Service.

We assessed the on-site disposal scenario, reflecting the projected management practice for this waste. For a number of years, the facility has been placing approximately 20 MT/yr in steel drums on pallets on the ground. The facility reported that they intend to reclaim the antimony from this slag when antimony prices are favorable. We consider storage on-site for more than one year to be speculative accumulation and consider these materials to be solid wastes. We believe that the length of time secondary materials are accumulated before being recycled is an important indicator of whether or not they are wastes. This is supported by the large number of recycling damage cases where secondary materials that were overaccumulated over time caused extensive harm. (See 50 FR 614) “Under RCRA and the implementing regulations, permanent placement of hazardous waste, including perpetual “storage” falls into the regulatory category of land disposal.”

25

(See also

American Petroleum Institute

v.

EPA,

216 F. 3d 50 (D.C. Cir. 2000).) Since the Operating Permit requires the facility to build and use an on-site, land-based unit for this waste, we assessed the on-site landfill scenario for this waste.

25

“Above Ground Land Emplacement Facilities, N.J. Law,” Letter to Honorable James J. Florio, Chairman, Subcommittee on Commerce, Transportation, and Tourism, Committee on Energy and Commerce, House of Representatives, from J. Winston Porter, Administrator, EPA, dated March 26, 1986.

How Was This Waste Category Characterized?

We selected two of the three facilities for sampling and analysis. At the site which stores the slag indefinitely, we collected one sample of “reduction furnace slag” that was designated as containing less than 5 percent antimony (AC-1-AO-01) and one sample of “reduction furnace slag” that was designated as containing between 5 and 10 percent antimony (AC-1-AO-06). Based on characterization information provided by the facility in its RCRA Section 3007 Survey response, we believe these samples are representative of all of the slags generated at the facility. We conducted total, TCLP and SPLP analyses of these slags. The analytical results for the constituents found to be present in the leachates at levels exceeding the HBLs are presented in Table III-2.

We collected a third sample (LI-1-AO-01) at a facility that reclaims its slag for lead. This sample failed the TC for lead, as the facility reported in its RCRA Section 3007 Survey response. The results are available in “Waste Characterization Report, Laurel Industries Inc., La Porte, Texas” in the docket for today's proposal.

Table III-2.—Characterization of Speculatively Accumulated Antimony Slag

Constituent of concern

AC-1-AC-01

Total

mg/kg

TCLP

mg/L

SPLP

mg/L

AC-1-AO-06

Total

mg/kg

TCLP

mg/L

SPLP

mg/L

HBL

mg/L

Antimony

11,500

55.8

114

127,000

110

211

0.006

Arsenic

301

2.0

2.9

478

3.1

3.8

0.0007

Boron

<500

9.8

9.3

<2,500

8.5

8.1

1.4

Selenium

<50

0.6

0.6

<250

0.6

0.3

0.08

Vanadium

<50

1.3

1.1

<250

0.6

1.0

0.14

How Was the Groundwater Ingestion Risk Assessment Established?

We modeled the annual volume reported to be stored on-site indefinitely (20 MT). (The facility reports that it processes sodium antimonate from two facilities and returns the resultant slag to the process for further processing. We did not include these recycled slag volumes in our modeling.)

We used the total and SPLP results as model inputs, reflecting the industrial nature of the on-site unit. We used only the analytical results for the facility that stores the slag indefinitely. Both of the samples for this facility are relevant because they represent the material stored on-site and destined for the on-site slag pit. We did not model the sample from the other sampled facility because they acknowledged that their waste exhibited the TC. Both this facility and the facility that was not sampled reclaim these wastes in a manner that is excluded from regulation under Subtitle C. We believe that it is reasonable to assume that they will continue to manage their slags in ways that do not violate Subtitle C regulations. Also, in this case, the SPLP results are higher than the TCLP results, making the industrial landfill the worst case scenario.

We examined records available from the State where the slag is stored to determine the appropriate distance-to-well to model. We identified four residential wells within several miles of the facility. These data demonstrate that groundwater is a viable and actively used resource in this area. One well is located 1.4 miles directly downgradient. Based on local topography and groundwater information, we do not believe the other identified wells could be affected by releases to groundwater from the facility. We modeled potential releases to a downgradient residential well. Given that our groundwater model is not configured at this time to model releases further than one mile, we did not assess the full distance to the known well. In our probabilistic analysis, we varied the well distance from the closest property boundary that appeared to be potentially downgradient to the limit of the model (one mile). Our results therefore are conservative with respect to this particular well, but otherwise reflect the fact that future residences and wells may be placed closer to the facility and any potential groundwater plumes associated with its operation. Specifically, we modeled potential distances to wells from the facility's southern boundary to one mile.

We used a regional site-based approach in modeling this unit, as described in section III.E.4. We modified this to enable us to use available depth to groundwater information at this particular site.

What Is EPA's Listing Rationale for This Waste?

Where these slags are reused and present no exposure route of concern, we did not evaluate these secondary materials further.

The results of the risk assessment for the on-site disposal scenario for boron, selenium, and vanadium were very low. In the 90th to the 95th percentile range, the highest hazard quotient for these three constituents was in the range of 0.001. For this reason, the full results for these three constituents are not presented here. The results of the risk assessment for the on-site disposal scenario for antimony and arsenic are presented in Table III-3:

Table III-3.—Probabilistic Risk Assessment Results for Speculatively Accumulated Antimony Slag

Percentile

Adult risk

Child risk

Adult risk

Child risk

1

Antimony hazard quotient

1

Arsenic—cancer risk

90

th

%

2.2

4.6

4 E-07

3 E-07

95

th

%

4.5

9.4

1 E-06

9 E-07

For a more complete description of this analysis, see “Risk Assessment for the Listing Determinations for Inorganic Chemical Manufacturing Wastes” (August 2000) in the docket for today's proposal.

In our modeling results, the dilution and attenuation factors (DAFs) were relatively high. For example, high end antimony DAFs were as high as 8,000. This is the result of the hydrogeological setting of the site evaluated. Due to the high hydraulic conductivity we used in modeling, the landfill leachate is readily diluted into a large volume of groundwater. Given the uncertainty about the actual ultimate management practice and the site-specific nature of the modeling, DAFs could be considerably lower in other disposal scenarios, resulting in much higher hazard quotients and, therefore, higher potential risks.

Our modeling approach assumes that the slag will be placed in an unlined unit. Information from the facility, however, indicates that they plan to place the waste in an on-site lined storage pit, upon completion of construction, that will be governed by a state mining permit. We considered whether our decisionmaking should account for the added protection provided by a liner system. Our first consideration is the current uncertainty regarding this waste's disposition. While the facility has stated its intended placement of this waste in a lined unit, our most recent information indicates that construction had not yet begun. The facility may in fact choose to place this waste in an off-site commercial landfill that would not necessarily be lined. This uncertainty is greater than in most waste management scenarios that we have assessed in this rulemaking, where there is a long term history of management in a particular type of management unit (

e.g.

, an operating on-site landfill, a local off-site landfill). Because of this uncertainty, we are hesitant to give much weight to a liner system that may be constructed in the future.

More generally, we considered the efficacy of landfills (and any liners) over the modeled risk assessment period, which covers 10,000 years. Landfills are used actively until their capacity is reached (our models assume an active life of 30 years), and at the end of their active life, we assume landfills are closed and the wastes remain in the unit indefinitely. The effectiveness of liner systems depends on how they are designed. Composite and double liners that combine two or more layers of liner material with leachate collection and leak detection will no doubt minimize leakage to the subsurface during the period when the leachate collection system is actively managed. However, depending on the regulatory controls relevant for a particular unit, monitoring would continue for a limited post-closure period. There is also uncertainty associated with liner performance, in the near term as well as in the long term. There are a variety of factors that may influence longevity and performance, such as poor construction, installation

or facility operation, or geologic movement below the liner that can cause holes, tears or larger failures. Some defects may have a significant effect. Because of our uncertainty regarding the efficacy of the liner system over long periods of time, and the uncertainty over the ultimate disposal for this waste, we believe our use of the modeling results for an unlined landfill is appropriate.

In deciding whether to list this waste as hazardous, we also considered other factors in addition to the risk results noted above. First, we considered the very high levels of toxic constituents present in the waste and in test leachate (which is one of the criterion cited in 261.11(a)(3)(ii)). The levels of antimony and arsenic are quite high. The antimony level exceeds 10% in the waste (up to 127,000 mg/kg), and the SPLP antimony concentration exceeds the drinking water HBL by a factor of >35,000. Another key factor is the lack of any appreciable degradation expected for these metals (a constituent's degradation or persistence is also a criterion for listing given in 261.11(a)(3)). Unlike some organic compounds, metals such as antimony will not degrade over time. Thus, even if the loss in effectiveness of a liner system only occurs over the very long term, the metals would still be present for leaching. It is difficult to assess the impact of the long-term effectiveness of the liner system in question for today's proposal. However, we note that the effectiveness of the liner system would have to be sufficient to reduce the antimony concentration at the well by close to 90% in order to keep the risks below an HQ of 1.

Therefore, given the reasons cited above, we propose to list these slags as hazardous:

K177 Slag from the production of antimony oxide that is disposed of or speculatively accumulated.

It is important to note that this listing has been developed to capture only those wastes that are not recycled. Thus, this listing, as proposed, would not apply to generators that recycle or reclaim this material as long as it is not speculatively accumulated. If slags have been speculatively accumulated (

i.e.,

held for more than a calendar year without recycling) at the time of the effective date of this final rule, these slags would meet the listing immediately.

26

26

A material is not accumulatively speculatively, however, if the person accumulating it can show that the material is potentially recyclable and has a feasible means of being recycled and that—during the calendar year (commencing on January 1)—the amount of material that is recycled or transferred to a different site for recycling equals at least 75 percent by weight or volume of the amount of that material accumulated at the beginning of the period. (40 CFR 261.1(c)(8))

We also propose to add antimony to Appendix VII to Part 261, which designates the hazardous constituents for which K177 would be listed.

(2) Baghouse filters.

How Are These Wastes Currently Managed?

These filters capture product or off-specification product. Two facilities place antimony laden baghouse filters in their on-site production furnaces. One of these facilities also sends a portion of its baghouse filters to Mexico for antimony recovery. Two other facilities dispose of these wastes in a non-hazardous waste incinerator and an industrial Subtitle D landfill. None of these wastes are handled as hazardous, although our sampling efforts showed this waste to exhibit the toxicity characteristic for lead and arsenic.

How Was This Waste Category Characterized?

We collected a total of three samples of this waste category from two facilities. At one facility we collected one sample of the “oxidation furnace” baghouse filters (AC-1-AO-03) and one sample of the “reduction furnace” baghouse filters (AC-1-AO-07). At the other facility, we collected a sample of the baghouse associated with its kiln (LI-1-AO-03). Because the facilities sampled represented the range of production practices within the industry, we believe these samples are representative of all of the baghouse filters generated by this industry. We conducted total, TCLP and SPLP analyses of these baghouse filters. The analytical results for the constituents found to be present in the leachates at levels exceeding the HBLs are presented in Table III-4. Two of the three samples of the waste, one from each facility that generate this waste, exceed the toxicity characteristic for either lead or arsenic. (The third sample exhibits TCLP lead levels close to the TC standard).

Table III-4.—

Characterization of Baghouse Filters From Antimony Oxide Production (mg/kg or mg/L)

Constituent of Concern

LI-1-AO-03

Total

TCLP

SPLP

AC-1-AC-03

Total

TCLP

SPLP

AC-1-AO-07

Total

TCLP

SPLP

HBL

TC

Limit

Antimony

91,400

9.3

6.2

150,000

9.9

4.3

145,000

68.7

287

0.006

Arsenic

114

<0.5

0.6

<250

<0.5

0.09

<250

1

6.9

6.9

0.0007

5.0

Boron

24.0

6.5

1.0

<2500

<2

0.2

<2500

<2

0.7

1.4

Cadmium

5.3

0.3

0.5

<250

0.3

0.3

411

<0.05

0.9

0.0078

1.0

Lead

3.1

1

8.5

16.9

<2500

2.8

1.0

<250

<0.5

<0.05

0.015

5.0

Mercury

0.9

<0.002

0.001

0.1

<0.002

<0.0002

95.2

0.03

0.4

0.0047

0.2

Thallium

<2

<2

0.06

<1000

<2

0.06

<1000

<2

0.1

0.0013

1

Exceeds Toxicity Characteristic level.

What Is EPA's Listing Rationale for This Waste?

We propose to list the baghouse filters waste because our data show it routinely exhibits one or more of the characteristics of hazardous waste (

i.e.,

TC lead or arsenic), yet the generators do not identify their wastes as exhibiting the toxicity characteristic and the generators that dispose of this waste do not comply with Subtitle C regulations. We propose to list this waste under the 261.11(a)(1) criteria:

K176 Baghouse filters from the production of antimony oxide (E).

Because we believe we have sufficient reason to list this waste under 261.11(a)(1) based on the TC exceedences and lack of compliance with hazardous waste regulation, we chose to conserve our time and resources and did not conduct formal risk assessment modeling of the off-site landfill scenario, as we would traditionally do to support a 261.11(a)(3) listing. Such modeling would reflect reported management practices. Antimony is not a TC constituent and, therefore, was not considered in the 261.11(a)(1) listing decision. However, antimony levels are high and would likely result in risk if modeled. Leach results for the waste exceed the HBLs by

a wide margin,

e.g.,

the SPLP results for antimony are up to 48,000 times the HBL. The high levels of antimony in the waste (up to 15%) would provide a long-term source of the metal for leaching into the groundwater. Thus, we expect that modeling an off-site Subtitle D scenario would yield significant drinking water risk.

Note that, when facilities process the antimony oxide product captured in these filters by reinserting the product-containing filters back into the furnace where the antimony oxide originated, without reclamation, these materials would not be solid wastes.

27

27

As noted above, these filters capture product materials. EPA does not regulate reclamation of these products. See 50 FR 14216, April 11, 1985: “Under the final rules, commercial chemical products and intermediates, off-specification variants, spill residues, and container residues listed in 40 CFR 261.33 are not considered solid wastes when recycled except when they are recycled in ways that differ from their normal use—namely, when they are burned for energy recovery or used to produce a fuel”

We also propose to add arsenic and lead to Appendix VII to Part 261, which designates the hazardous constituents for which K176 would be listed.

The “mixture” rule for listed wastes currently provides an exemption for wastes listed solely because they exhibit characteristics (see 40 CFR 261.3(a)(2)(iii)). Mixtures of such listed wastes lose their listed waste status when they cease to exhibit characteristics for which they were listed. (However, they would still need to comply with Land Disposal Restriction requirements.) In the both of the last two Hazardous Waste Identification Rule (HWIR) proposals (60 FR 66344, December 21, 1995) and (64 FR 63382, November 19, 1999), we proposed to narrow the exemption to only include wastes listed for ignitability, corrosivity, and reactivity.

This narrowing would make any waste listed for the toxicity characteristic (TC) (including the waste proposed today for listing under the (a)(1) criteria) ineligible for the current exemption. In other words, under current mixture rule regulations, mixtures containing these baghouse filters would become nonhazardous wastes once they ceased exhibiting the characteristic. Under the HWIR proposal, however, such mixtures would remain hazardous wastes even after they cease to exhibit the TC. As we state in the TC rule, chemicals can still pose hazardous at levels below the TC (see 55 FR 11799, March 29, 1990). Under an amended consent decree (

Environmental Technology Council

v.

Browner, C.A.

No. 94-2119 (TFH), April 11, 1997), EPA is required to sign a notice taking final action with respect to the proposed revisions to the mixture rule by April 30, 2001.

(3) Empty supersacks. One facility ships crude antimony oxide in supersacks and then reuses them to store intermediate materials until they wear out. The facility then sends these empty supersacks either to an off-site industrial Subtitle D landfill or to an off-site plastic recycler. The facility claims that the supersacks are empty and would meet the standard in 40 CFR 261.7 (which exempts “empty” containers formerly used to manage hazardous waste). Although 40 CFR 261.7 does not literally apply to these sacks, we think it is reasonable to take a similar approach here. We believe that the levels of crude antimony oxide in worn-out supersacks would be low because the material is the primary feedstock (raw material) used in this process. We do not believe it follows that these supersacks should be regulated, when other similarly empty containers would be exempt. Therefore, we propose not to list this waste as hazardous.

2. Barium Carbonate

a. Summary.

We have evaluated the wastes, waste management practices, and potential risk exposure pathways associated with the barium carbonate production processes and propose not to list any wastes from this industry as hazardous under Subtitle C of RCRA. Some wastes in this industry are D002 or D005 characteristic hazardous wastes, which are both currently subject to RCRA Subtitle C regulation and managed in compliance with those regulations. For other wastes, not identified as characteristic hazardous wastes, we have identified no risks of concern associated with the current management of these wastes that would warrant listing. These wastes do not meet the criteria listed under 40 CFR 261.11(a)(3) for listing a waste as hazardous.

b. How is barium carbonate produced?

There are two facilities in the United States that produce significant quantities of barium carbonate. A Georgia facility produces barium carbonate for commercial sale. A Pennsylvania facility produces barium carbonate only for use as a feedstock in its own internal manufacturing processes. A third facility is a specialty manufacturer that produces extremely small amounts of barium carbonate (approximately 10 kg in 1998).

28

For more detailed information concerning this industry, see “Barium Carbonate Listing Background Document for the Inorganic Chemical Listing Determination” in the docket for today's proposal.

28

Since, as explained below, we find no significant risks from the larger volume wastes we assessed, we conclude that any low volume wastes from this third facility also would not pose any risks warranting listing.

Barium carbonate (BaCO

3

) has a wide range of uses, including feedstock for production of other barium chemicals, an additive in various glasses, ceramics, bricks, and other construction materials, an additive in oil-drilling suspensions, and a brine purification chemical in the chlor-alkali industry.

The two primary barium carbonate production facilities use different manufacturing processes to make barium carbonate. The Georgia facility uses locally mined barite ore, containing barium in the form of barium sulfate, as the primary feedstock. The ore is crushed and milled, thermally reduced in a roasting kiln, and leached with water to dissolve the barium. The resulting barium sulfide solution is filtered and reacted with carbon dioxide gas to produce a barium carbonate precipitate. This precipitate is then dried, and sized for sale.

The Pennsylvania facility uses a commercially purchased high purity barium chloride solid as the primary feedstock. The facility dissolves the barium chloride in water, heats and filters the resulting solution, and precipitates barium carbonate by reacting the barium chloride solution with ammonium bicarbonate. The resulting barium carbonate precipitate is washed, filtered, dried and sized before the facility utilizes it as a feedstock in other manufacturing processes on-site.

c. What wastes are generated? Table III-5 below briefly lists the facility-reported residuals from the barium carbonate manufacturing processes, residual volumes generated in 1998, reported RCRA hazard codes, and residual management practices.

Table III-5.

—Barium Carbonate Production Wastes

Waste category

1998 volume (MT)

Reported RCRA hazard codes

Sequential residual management practices

Barite Ore Feedstock Process—Georgia Facility

Treated barium wastes (D005 barium wastes include barite ore leaching waste, barium

This text is long and has been trimmed here. Open the source document for the complete record.