Federal Register / Vol. 56. No. 138 / Thursday, July 16, 1991 / Proposed Rules 32993
Consolidation of Dockets/Comment
Period
All previously docketed motions
dealing with the confidential treatment
of Form 41 Schedules B-7 and B-43 will
be considered and consolidated in this
rulemaking (Docket 46597). Previously
docketed motions include eight motions
filed by United Air Lines (Dockets 46597,
46868, 46869, 46933,47119, 47254,47485,
and 47486); and one motion filed by
American Airlines (Docket 47273).
Furthermore, all motions received to
date for the confidential treatment of the
first quarter 1991 Form 41 Schedule B-7
have also been filed in reference to '
Docket 46497. Finally all future
comments on any of the docketed
motions, and on this rulemaking, should
also be filed in this rulemaking docket.
Comments are due no later than 45
days after publication of this ANPRM in
the Federal Register.
Rulemakiag Analyses and Notices
DOT Regulatory Policies and Procedures
This ANPRM is not significant under
the Department's Regulatory Policies
and Procedures, dated February 26,
1979. Its economic impact should be
minimal and a full regulatory evaluation
is not required. There is no air carrier
reporting burden associated with this
rulemaking proposal. Any burden
associated with this rulemaking if
confidentiality is granted, would be
borne solely by the Department, which'
would need to establish and maintain
procedures for ensuring the
confidentiality of the affected data,
Executive Orders
Executive Order 12291
This action has been reviewed under
Executive Order 12291, and it has been
determined that this is not a major rule.
It will not result in an annual effect on
the economy of $100 million or more.
There will be no increase in production
costs or prices for consumers, individual
industries, Federal, state or local
governments, agencies or geographic
regions. Furthermore this ANPRM would
not adversely affect competition,
employment, investment, productivity,
innovation, or the ability of United
States-based enterprises to compete
with foreign-based enterprises in
domestic or export markets. These
proposed regulations would not affect
the reporting burden for large
certificated air carriers. Accordingly, a
regulatory impact analysis is not
required.
Executive Order 12612
This action has been analyzed in
accordance with the principles and
criteria contained in Executive Order
12612 and it has been determined that
this rulemaking action does not have
sufficient federalism implications to
warrant the preparation of a Federalism
Assessment.
Executive Order 12830
This action has been analyzed in
accordance with the principles and
criteria contained in Executive Order
12630 and it has been determined that
this rulemaking action does not pose the
risk of a taking of constitutionally
protected private property.
Legislative Acts
Regulatory Flexibility Act
I certify that this action will not have
a significant economic impact on a
substantial number of sinall entities. For
purposes of its aviation economic
regulations, Departmental policy
categorizes certificated air carriers
operating small aircraft (60 seats or less
or 18,000 pounds maximum payload or
less) hi strictly domestic service as
small entities for purposes of the
Regulatory Flexibility Act. This ANPRM
would affect only large certificated air
carriers.
Paperwork Reduction Act of 1980
An information copy of this notice is
being submitted to the Office of
Management and Budget (OMB). This
ANPRM would result in no change in air
carrier reporting burden; therefore there
is no need to submit a paperwork
package to OMB for their approval. If
confidentiality is granted the burden
impact of this proposal Would be borne
solely by the Department which would
be required to establish and maintain
procedures for ensuring the
confidentiality of the affected data. If a
notice ef proposed rulemaking is drafted
subsequent to the issuance of this
ANPRM, the rulemaking proposal will
be submitted to the Office of
Management and Budget for approval in
accordance with 44 U.S.C. chapter 35.
Regulatory Identification Number
A regulatory information number
(RIN) is assigned to each regulatory
action listed in the Unified Agenda of
Federal Regulations. The Regulatory
Information Service Center publishes
the Unified Agenda-in April and
October of each year. The RIN number
contained in the heading of this
document can be used to cross reference
this action with the Unified Agenda.
List of Subjects in 14 CFR Part 241
0 Air carriers and Uniform System of
Accounts and Reports.
Issued in Washington, DC on July 10,1991.
Robin A. CaldweU,
Director, Office of Airline Statistics.
[FR Doc. 91-17074 Filed 7-17-81; 8:45 am]
BILLING CODE 4910-62-U
ENVIRONMENTAL PROTECTION
AGENCY
40 CFR PART 261
[SW-FRL-3974-6]
Hazardous Waste Management
System; Identification and Listing of
Hazardous Waste; Proposed Use of
EPA's Composite Model for Landfills
(EPACML) and Proposed Exclusion
AGENCY: Environmental Protection
Agency (EPA).
ACTION: Proposed rule and request for
comment
SUMMARY: Today's notice announces the
Environmental Protection Agency's
(EPA or Agency) proposal to use the
EPA's Composite Model for Landfills
(EPACML) in the evaluation of a
delisting petition. Based on waste-
specific information providedvby the
petitioner, the Agency is proposing to
use the EPACML fate and transport
model to evaluate the impact of the
petitioned waste on human health and
the environment
Today's proposal provides
background information on the
mechanics of the EPACML, and the use
of the EPACML in delisting decision-
making. Based on its evaluation using
the EPACML, the Agency today is
proposing to grant a petition submitted
by Reynolds Metals Company.
(Reynolds), Bauxite, Arkansas, to
conditionally exclude certain solid
wastes to be generated by its thermal
treatment process from the lists of
hazardous wastes contained in 40 CFR •
261.31 and 261.32. This action responds
to a delisting petition submitted under
40 CFR 260.20, which allows any person
to petition the"' Administrator to modify
or revoke any provision of parts 260
through 265 and 268 of title 40 of the "
Code of Federal Regulations, and under
40 CFR 260.22, which specifically
provides generators the opportunity to
petition the Administrator to exclude a
waste on a "generator-specific" basis
from the hazardous waste lists. Today's'
proposed decision is based on an
evaluation of process and waste-specific.
information provided by the petitioner.
DATES: EPA is requesting public
comments on today's proposed
exclusion and on the applicability of the
EPACML to this proposal. Comments
Printed on Recycled Paper
-------
32994
Federal Register. / Vol. 58, No. 138 / Thursday. July 18, 1991 /Proposed Rules
will be accepted until September 3,1991.
Comments postmarked after the close of
the comment period will be stamped
"late".
Any person may request a hearing on
the proposed decision by filing a request
with the Director, Characterization and
Assessment Division, Office of Solid
Waste, whose address appears below,
by August 2,1991. The request must
contain the information prescribed in 40 •
CFR260.20fd).'
ADDRESSES: Send three copies of your
comments to EPA. Two copies should be
sent to the Docket Clerk, Office of Solid
Waste (05-305), U.S. Environmental
Protection Agency, 401M Street, SW.,
Washington, DC 20460. A third copy
should be sent to Jim Kent, Delisting ,
Section, Waste Identification Branch,
CAD/OSW (OS-333). U.S.
Environmental Protection Agency, 401M
Street, SW., Washington. DC 20460.
Identify your comments at the top with
this regulatory docket number: "F-91-
CML-FFFFF'.
Requests for a hearing should be
addressed to Director, Characterization
and Assessment Division, Office of
Solid Waste (OS-330), U.S.
Environntcntal Protection Agency, 401WL
Street, SW., Washington, DC 20460.
The RCRA regulatory docket for this
proposed rule is located at the U.S.
Environmental Protection Agency, 401M
Street, SW. (room M2427), Washington;
DC 20400, and is available for viewing
from 9 a an. to 4 p.m., Monday through
Friday, excluding Federal Holidays. Call
(202) 475-9327 for appointments. The
public may copy material from any •
regulatory docket at a cost of $0.15 per
page.
FOR FURTHER INFORMATION CONTACT:
For general information, contact the
RCRA Hotline, toll free at (800) 424-
9340. or at (703) 920-9810. For technical
information concerning this proposed
rule, contact Chichang Chen, Office of
Solid Waste (OS-333), U.S. . '?
Environmental Protection Agency, 401M
Street, SW., Washington, DC 20460,
(202) 382-7392.
SUPPLEMENTARY INFORMATION:
Preamble Outline
I. Proposed Use of EPACML for Delisting
Evaluations
A. Background
1. Authority
2. Use of Models for Delisting Evaluations
a. Reasonable Worst-Case Disposal
Scenarios
b.ThaVHS Model
c. Public Comments on the VHS Model
3. Development of the EPACML
a. The Final Toxicity Characteristics Rule
b. Public Comments on the Planned Use of
the EPACML for Delisting
4. Benefits of Replacing the VHS Model
with the EPACML for Delisting
B. Summary of the EPACML
1. Description of the EPACML , .
a. Disposal Scenario Assumptions
b. Fate end Transport Processes
c. Input Parameters and Data Sources
d. Monte Carlo/Uncertainty Analysis
2. Modifications of the EPACML for
Delisting
a. Conversion of Delisting Waste Volumes ,
to Disposal Unit Area
b. Use of a "Scaling Factor" for
Accumulation of Wastes Over Multiple
Years (and Possible Codisposal with *
Similar Wastes) ,
, 3. Dilution/Attenuation Factors (DAFs)
a. Selection of an Appropriate DAF
Percentile
b. Delisting DAFs for Landfills
c. Delisting DAFs for Surface
, Impoundments • • •
d. Comparison of Characteristics arid
Delisting DAFs
C. Use of the EPACML in Delisting
Evaluations • .
1. General Approach f
2. Selection of Landfill or Surface
Impoundment Disposal Scenario
.' 3,'Impact on Information Needed from
Delisting Petitioners
II. Proposed Exclusion ,
A. Background ....-•-..•
1. Approach Used to Evaluate This Petition
B. Disposition of Delisting Petition , .
1. Reynolds-Metals Company, Gum Springs,
Arkansas
, a. Petition for Exclusion
b. Background
c. Agency Analysis ;
d. Agency Evaluation
e. Conclusion
f. Verification Testing Conditions
m. Effective Date
IV. Regulatory Impact -
V. Regulatory Flexibility Act
VI. Paperwork Reduction Act . •' '
VH. List of Subjects in 40 CER Part 281
I. Proposed Use of EPACML for
Delisting Evaluations
A. Background
1. Authority
On January 16,1981, as part of its
final and interim final regulations
implementing section 3001 of RCRA,
EPA published an amended list of
hazardous wastes from non-specific and
specific sources. This list has been
amended several times, and is published
in 40 CFR 261.31 and 261.32i;These
wastes are listed as hazardous because
they typically and frequently exhibit one
or more of the characteristics of
hazardous wastes identified in subpart
C of part 261 (i.e., ignitability,
corrosivity, reactivity, and toxicity) or
meet the criteria for listing contained in
40 CFR 261.11 (a)(2) or (a)(3).
Individual waste .streams may vary, .
however, depending on raw materials,
industrial processes, and other factors.
Thus, while a waste that is described in
these regulations generally is hazardous,
a specific waste from an individual,
facility meeting the listing description
may not be. For this reason, 40 CFR
260.20 and 260.22 provide an exclusion
procedure, allowing persons to .
demonstrate that a specific waste from a
particular generating facility .should not
be regulated as a hazardous waste.
To have their wastes excluded,
petitioners must show that wastes
generated at their facilities do not meet
any of the criteria for which the wastes
were listed. See 40 CFR 260.22(a) and
the background documents for the listed
wastes. In addition, the Hazardous and
Solid Waste Amendments (HSWA) of
1984 require the Agency to consider any
factors (including additional
constituents) other than those for which
the waste was listed, if there is a
reasonable basis to believe that such
additional factors could cause the waste
to be hazardous. Accordingly, a
petitioner also must demonstrate that
the waste does not exhibit any of the
hazardous waste characteristics [Le., ,-
ignitability, reactivity, corrosivity, and
. toxicity)^ and must present sufficient
information for the Agency to determine
, whether the waste contains any other
toxicants at hazardous levels. See 40
CFR 260.22(a), 42 U.S.C. 6921(f), and. the
background documents for the .listed
wastes. Although wastes which are
"delisted" (j.e., excluded) have been
evaluated to determine whether or not
they exhibit any of the characteristics of
hazardous waste, generators remain
obligated under RCRA to determine,
whether or not their waste remains non-
hazardous based on the hazardous.
waste characteristics. . .
In addition to wastes listed as ,
hazardous in40 CFR 261.31 and 261,32,
residues from the treatment, storage, or
disposal of listed hazardous, wastes and
mixtures containing hazardous wastes
.are also considered hazardous wastes.
Such Wastes are also eligible fqr ,
exclusion and remain hazardous wastes
until excluded. See 40 CFR 261,3 (c) and
(d)(2). The substantive standards for
"delisting" a treatment residue or a
mixture are the; same as previously
described for listed wastes.
, If a petitioned waste is granted an
exclusion by the Agency (i.e., delisted),
the waste is no longer controlled under
federal hazardous waste regulations and
the petitioner must manage and dispose
of the waste according to local and state
requirements or specifications.
In some cases, the Agency grants an
exclusion conditioned upon the facility
. or waste meeting certain requirements.
For example, for wastes that are highly
variable in composition, the Agency
-------
Federal Register / Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules
32995
often imposes post-exclusion testing
requirements that the petitioner must
meet prior to waste disposal. Only those
batches of waste that have passed the
verification testing conditions provided
in the final exclusion could be managed
as non-hazardous wastes; failing
batches must be managed as hazardous.
2. Use of Models for Deh'sting
Evaluations
During a delisting determination, the
Agency often uses fate and transport
models to predict the concentration of
hazardous constituents that may be
released from petitioned wastes after
disposal to determine the potential
impact on human health and the
environment. An appropriate transport
model has been used to estimate the
potential impact of the leachable
hazardous constituents on the
underlying aquifer. Specifically, the
Agency has used the maximum
estimated waste volume and the
maximum reported leachate
concentrations as inputs to estimate the
constituent concentrations in the ground
water at a hypothetical receptor well
downgradient from the disposal site.The
calculated receptor well concentrations
(referred to as compliance-point
concentrations) were then compared
directly to the health-based levels used
in delisting decision-making for the
hazardous constituents of concern.
a. Reasonable Worst-Case Disposal
Scenarios. EPA's approach when
applying fate and transport models has
been to represent a reasonable worst-
case waste disposal scenario for the
petitioned waste rather than use site-
specific factors. The'Agency believes
that a reasonable worst-case scenario is
appropriate when determining whether
a waste should be relieved of the
protective management constraints of
RGRA subtitle C. The use of a
reasonable worst-case scenario results
in conservative values for the
compliance-point concentrations and
ensures that the waste, once removed
from hazardous waste regulation, will
not pose a threat to human health or the
environment if the petitioner chooses to
dispose of the waste in accordance with
subtitle D requirements. Site-specific
factors [e.g., site hydrogeology) are not
considered because a delisted waste is
no longer subject to hazardous waste
control, and therefore, the Agency is
generally unable to predict and does not
control how a waste will be managed
after delisting.
b. The VHS Model. Under a landfill or
surface impoundment disposal scenario,
the plausible exposure route of most
concern to the Agency is ingestion of
contaminated ground water. To evaluate
land-disposed wastes, the Agency has
used the Vertical and Horizontal Spread
(VHS) model {see 50 FR 7896, February
26,1985. and 50 FR 48898, November 27,
1985 for details), the VHS model
approximates the transport processes
likely to occur in an aquifer below a
waste disposal site. The model predicts
the dilution of the contaminants hi a
drinking water aquifer as a result of
dispersion in the vertical and horizontal
directions.
The waste-specific parameters used in
the VHS model are the leachate
concentrations of constituents of
concern and the annual volume of the
petitioned waste. The leachate
concentration is determined by
laboratory analysis, i.e., the Extraction
Procedure (EP) or the Toxicity
Characteristic Leaching Procedure
(TCLP) for constituents of concern.
In applying the VHS model, the
Agency made a variety of assumptions
to account for a reasonable worst-case
disposal scenario. The VHS model >
assumes that the waste is disposed in a
40-foot wide, 8-foot deep trench in an
unlined municipal landfill. The length of
'the trench is a function of the volume of ,
the petitioned waste. The model
assumes an infinite source of waste and
no aquifer recharge. The model
mathematically simulates the migration
of toxicant-bearing leachate from the
waste into the uppermost underlying
aquifer, and the subsequent dilution of
the toxicants due to dispersion within
the aquifer. The Agency used this model
to predict the maximum concentration of
the diluted toxicants at a hypothetical
receptor well (or compliance point)
located 500 feet from the disposal site.
The model does not consider
biodegradation, sorption, hydrolysis, or
unsaturated soil conditions.
c. Public Comments on the VHS
Model. The Agency has received
comments on the VHS model hi notices
announcing the use of the VHS and in
proposed petition denials and
exclusions. In general, the comments
expressed concern about ability of the
VHS to consider large waste volumes,
wastes stored in surface impoundments,
unsaturated soil conditions,
groundwater recharge, and longitudinal
contaminant dispersion in the aquifer.
The Agency acknowledged that the
VHS has some limitations, but
maintained that the use of the VHS was
appropriate until a more sophisticated
model was developed that would
account for these waste disposal
assumptions and transport processes,
3. Development of the EPACML
a. The Final Toxicity Characteristics
Rule. On June 13,1986, EPA proposed an
approach (see 51 FR 21648) for
•estimating regulatory levels of
hazardous constituents in a waste
leachate using chronic toxicity reference
levels, combined with constituent-
specific dilution/attenuation factors
(DAFs) derived from the application of a
subsurface fate and transport model
(EPASMOD). On August 1,1988 (53 FR
28892), EPA requested comments on
revisions to EPASMOD that were being
considered by the Agency. The revisions
were incorporated and EPASMOD was
subsequently referred to as EPACML. In
the TC Final Rule (55 FR 11798, March
29,1990), the EPACML was used to
support the choice of appropriate DAFs
for the development of TC regulatory
levels.
The EPACML simulates the movement
of contaminants from a subtitle D (i.e.,
municipal solid waste) waste
management unit and migration through
the subsurface environment to a
potential drinking-water well. The
EPACML accounts for one-dimensional
steady and uniform advective flow;
contaminant dispersion in the
longitudinal, lateral, and vertical
directions; sorption; and chemical
degradation from hydrolysis. The
EPACML estimates a DAF for
contaminants, which represents a
reduction ia the concentration expected
to occur during transport through soil
and ground water from the leachate
release point (bottom of.the waste
management unit) to an exposure point
(a receptor well serving as a drinking-
water supply, referred to as a
compliance point).
The Agency used the EPACML to
investigate the expected range of DAFs
associated with the mismanagement of
solid wastes. DAFs were used to
identify wastes whose leaching behavior
indicates that they pose a hazard to
human health unless they are controlled
under Subtitle C management standards.
Specifically, the regulatory levels
promulgated under the TC rule are
supported by DAFs generated from the
EPACML.
b. Public Comments on the Planned
Use of the EPACML for Delisting. In the
August 1,1988 proposal of the TC rule,
EPA solicited comments on the use of
the EPACML model hi the delisting
program. Although most commenters
agreed that there are fundamental
differences between the various
elements of the hazardous waste listing
and the corresponding elements in the
delisting program, all commenters
supported the use of the EPACML in the
delisting program. In response to these
comments, EPA stated in the final TC
rule that the EPACML would be used for
-------
32936
Federal Register / Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules
dclisling program in the future (see 55
FR11833; March 29,1990).
4. Benefits of Replacing the VHS Model
With the EPACML for Delisting
As stated above, the comments
received by the Agency supported the
use of EPACML In the delisting program.
The EPACML addresses comments
about the VHS model, including
consideration of an unsaturated zone,
ground-water recharge, evaluation of
large waste volumes and longitudinal
dispersion. Rather than using single
values for each input parameter, as is
the case with the VHS model, EPACML
uses a range of values for each input
parameter that are based on national
surveys of actual disposal facilities,
laboratory and field measurements, and
best available literature sources. The
uncertainty in the input parameters and
the range of possible combinations, of
different input values are quantified
using a Monte Carlo simulation
technique.
The use of EPACML by deh'sting
provides consistency between the
characteristics and deh'sting programs
and maintains the current delisting
approach for considering waste volume
and reasonable worst-case disposal
scenarios. In addition, the use of
EPACML by delisting does not add
complexity to or increase the time
required for a delisting evaluation.
B, Summary of the EPACML
1. Description of the EPACML
As discussed previously in today's
proposed rule, the EPACML was
developed to predict the transport of
hazardous constituents through soil and
ground water from a waste management
unit to a compliance point (a receptor
xvell serving as a drinking-water supply).
The output of the EPACML, the dilution/
attenuation factor (DAF), represents the
reduction in contaminant concentration
resulting from subsurface processes
such as adsorption, three-dimensional
dispersion, and dilution from ground-
water recharge.
The DAF is obtained by dividing the
contaminant concentration in leachate
leaving the disposal unit by the
concentration in the receptor well
predicted by the EPACML. For example,
if the leachate concentration of the
waste in the disposal unit is 10 ppm and
the predicted contaminant concentration
in the receptor well is 1 ppm, then the
DAF for the contaminant would be 10.
Actual use of the DAF in the evaluation
of petitioned wastes is discussed in
section I.C.
The EPACML uses a Monte Carlo
simulation technique to account for the
wide range of hydrogeolpgic settings,
found at municipal waste landfills, and
to account for the uncertainty in the
input data. The Monte Carlo procedure
randomly selects values from frequency
distributions developed for each input
parameter and results in a cumulative
frequency distribution of DAFs. Details
of the Monte Carlo method are
described in detail in section LB.l.d.
The method the Agency used to apply
EPACML to delisting is similar to its
application for the TC rule, except for
some modifications made to reflect the
criteria currently used for deh'sting
evaluations. Specifically, petitioned
wastes are evaluated based on the
volume of the waste and the leachable
concentrations of the constituents of
concern found in the waste. However,
the EPACML requires as an input
parameter the area of a disposal unit.
Therefore, the Agency developed a
procedure to convert the volume of a
petitioned waste to the disposal unit
area required by the EPACML. This
procedure is described in detail in the
docket in today's notice. Otherwise, the
fate and transport assumptions and •
input data are the same as those used
for the TC rule and are summarized
here.
For additional information about the
development and the use of the
EPACML, the reader is referred to the
TC rule {55 FR 11798, March 29,1990)
and to the Background Document for the
EPACML contained in the docket in
today's notice. This background
document presents hi detail each of the
technical issues raised in the public
comments on the model and the
Agency's response to these issues.
a. Disposal Scenario Assumptions.
The Agency developed the EPACML
model to evaluate the fate and transport
of hazardous constituents released from
waste disposed in landfills. During the
development of the TC rule, the Agency
received comments about the need to
consider the fate and transport of
wastes disposed in surface
impoundments. In response, the Agency
used data relevant to a surface
impoundment scenario with the
EPACML to simulate migration from an
impoundment The assumptions
associated with both the landfill and
surface impoundment disposal scenario
and the input data necessary to evaluate
surface impoundments using the '
EPACML are discussed below.
The selection of the disposal scenario
assumption determines the method of
selecting input parameters that are
subsequently used to compute leachate
infiltration rate. For landfills, the
Hydrologic Evaluation of Landfill
Performance (HELP) model (Ref. 1) was
used to determine the leaching rate from
landfills. The HELP model calculates
leachate flux as a function of landfill
cap, climatologic data, surface runoff,
and percolation. Leaching rates from
surface impoundments were estimated
by considering the relationship between
vertical groundwater velocity, the,
porosity and permeability of subsurface •
materials and the solution of the
nonlinear steady state flow problem.
Leachate rates from surface .
impoundments also consider the depth
of ponding in the impoundment and the
thickness and permeability of the sludge
layer at the bottom of the impoundment.
The reader is referred to the background
documents in the docket for the details
of the assumptions! associated with each
disposal scenario..
b. Fate and Transport Processes. The
EPACML fate and transport model
consists of an unsaturated zone module
and a saturated zone module, both of
which were reviewed and endorsed by
EPA's Science Advisory Board (SAB).
Each module is based on a number of
key assumptions pertaining to ground-
water flow, subsurface characteristics,
and the behavior of hazardous
constituents. These assumptions were
necessary to develop mathematical
representations of the fate and transport
processes considered. Selection of the
assumptions was based on their relative
significance in modeling the transport of
contaminants, the availability of data
required for the assumption, and the
ability to develop mathematical
solutions for the models.
The unsaturated and saturated zone
modules are best described by
summarizing the assumptions each uses
to model unsaturated zone flow and
contaminant transport. The major
assumptions for flow in the unsaturated
zone include:
• Flow is steady in the vertical
direction, and negligible in the lateral
and transverse directions;
• No vapor phase or immiscible liquid
flow occurs, and the water phase is the
only flowing material;
• Flow is isothermal;
• Effects of variations in the
unsaturated zone hydraulic properties
caused by alternating moisture
conditions (hysteresis) are negligible;
• The flow field is uniform and
continuous in direction and velocity;
and
• The unsaturated zone is
homogeneous and isotropic.
The major assumptions used to model
contaminant trans port in the
unsaturated zone are:
-------
Federal Register /Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules 32997
• Chemical transport is vertical;
lateral and transverse transport is
negligible;
• Chemical sorption is modeled as a
reversible, linear equilibrium process;
• Vapor phase transport of chemicals
is negligible; and '
• Unsaturated zone transport of
chemicals is solved for the steady-state
condition.
Ground-water flow and contaminant
transport in the saturated zone are
based on the following assumptions:
'• Ground-water flow regions are of
infinite extent in the longitudinal
direction, semi-infinite extent in the :
lateral direction, and finite in the
vertical'direction;
• The aquifer is homogeneous and
isotropic with a constant thickness;
• Ground-water flow is uniform and
continuous in direction and velocity;
• Recharge, as a result of
precipitation, supplies water to the
disposal unit and the aquifer;
• The ground water upstream of the
disposal site is initially free of
contamination;
• An infinite source supplies a
constant mass flux of chemical into the
aquifer;
• Contaminants follow a linear
equilibrium adsorption isotherm;
• Receptor well locations are
anywhere within the areal extent of the
contaminant plume.
These assumptions and the
mathematical methods used to model
them are described in greater detail in
the Background Document for EPACML
(Ref. 2).
c. Input Parameters and Data Sources.
The modeling assumptions and the
Monte Carlo simulation method
(described in detail in section I.B.l.d)
define the data requirements for
applying the EPACML. Specifically, the
Monte Carlo method randomly selects
values from frequency distributions of
each input parameter. Therefore, the
information required for each input
parameter includes; the type of
frequency distribution [e.g., normal,
exponential) and summary statistics
such as mean, standard deviation and
range of values (minimum and
maximum). The data sources and
analyses performed to obtain the
frequency distributions are described in
detail in the background documents for
the EPACML (Ref. 2) and are
summarized here.
The input data required by the
EPACML consists of several types of
data: source (disposal unit) data,
chemical constituent data, unsaturated
zone data, and aquifer data. Values for
each of these types of data were
obtained from Agency surveys,
laboratory and field measurements, and
best available literature sources.
The Agency used source and site-
specific data from the OSW Survey of
Solid Waste Landfills (Ref. 3) to obtain
values for disposal unit area, distance to
the closest downgradient drinking water
wells, and thickness of the unsaturated
zone. The Agency used this survey data
to develop the method to convert waste
volume to landfill disposal unit area, as
described in I.B.2.a below and in the
background document (Ref. 2). In
addition, the Agency has obtained a set
of data on Subtitle D surface
impoundments, including disposal unit
volumes and depths (Ref. 4).
The OSW survey contains data for a
representative range of aquifer types in
the country, including values for
saturated zone thickness, ground-water
pH, particle diameter and fraction of
organic carbon. Dispersivity values are
computed as a function of distance to
the receptor well based on a detailed
analysis of data gathered from field
tests. Seepage velocity, hydraulic
conductivity, and porosity are derived
from particle diameter and other
parameters using analytical equations.
Climatologic and soils data were used
to compute frequency distributions of
leachate flux from the disposal unit. The
Agency obtained national climatologic
data from 100 stations for precipitation
and evaporation data that are based on
20-year climatic records. Soils data were
acquired from the US Soil Conservation
Service (SCS) to determine landfill cover
and unsaturated zone soil properties.
A complete list of all input parameters
and then1 frequency distributions is
presented in the background document
(Ref. 2). It is to be noted that a recent
statistical analysis (based on
comparison with a hydrogeologic
database developed from a national
survey of National Water Well
Association members (Ref. 5)} supported
the Agency's use of the OSW landfill
survey data. In particular, this
publication concluded that the
statistical distributions used by EPA for
the two most important hydrogeologic
parameters used in the model, seepage
velocity-and hydraulic conductivity, are
sound.
d. Monte Carlo/Uncertainty Analysis.
Data are incorporated into the EPACML
using a Monte Carlo procedure. The
Monte Carlo procedure randomly selects
values from a frequency distribution for
each parameter. The model is run a
sufficient number of times (typically
several thousand) to produce a
frequency distribution of the model's
output (DAF's). The output frequency
distribution represents a full range of
interactions of each individual
parameter and can be viewed as a
ranked order of increasingly higher
downgradient concentrations, from the
"best-case" situations at a low
cumulative frequency (large DAFs) to
the "worst-case" situations at high
cumulative frequencies (small DAFs) for
the scenario being investigated (55 FR
11826; March 29,1990).
Monte Carlo simulation was chosen
as the preferred method to evaluate a
variety of land disposal scenarios and to
account for the uncertainty inherent in
the input data. The wide range of
environmental conditions {e.g., ground-
water velocities, ground-water pH,
exposure point/receptor well locations)
found at disposal sites across the nation
makes the selection of a reasonable
worst-case for each input parameter
difficult.
Further justification for the use of the
Monte Carlo method is provided by the
very complex manner in which the many
model parameters interact Unless many
(hundreds to thousands) combinations -
of variables are investigated, it is simply
not possible to account for the wide
range of physical settings found at.
various disposal sites. Therefore, the
Monte Carlo method was utilized to
ensure a conservative yet physically
reasonable estimation of contaminant
fate and transport
2. Modifications of the EPACML for
Delisting •
a. Conversion of Delisting Waste
Volumes to Disposal Unit Areas. As
described previously in today's
proposed rule, the waste-specific data
used in delisting evaluations are the
maximum waste volume and the
maximum leachable concentrations of
constituents in the waste. To use the •
EPACML for delisting evaluations, the
area of the waste management unit,
rather than the maximum waste volume,
is used as an input parameter to the
EPACML.
The conversion of waste volume to
disposal unit area was part of the pre-
processing calculation for the EPACML.
The preprocessing calculation correlates
waste volume with disposal unit area ,
via a regression equation developed
from national data on disposal unit
dimensions contained in the OSW
Survey. This calculation uses the
regression equation to develop
frequency distributions of disposal unit
areas, taking into consideration the
regression variance to account for the
uncertainty in the volume/area
relationship. Methods to convert Waste
volume to disposal unit area were
developed for both landfills and surface
impoundments (Ref. 6).
-------
32998
Re^8tar / Vo1' 56» No' *38 I Thursday, July 18. 1991 / '
b. USB of a "Scaling Factor" for
Accumulation of Wastes Over Multiple
Years (and Possible Co-disposal with
Similar Wastes}. The delisting program
primarily receives petitions for wastes
generated on an on-going basis (/.a, a
petition is based on an annual waste
generation rate}. To account for the total
amount of waste generated and
ultimately placed in a landfill, the
Agency has chosen to multiply the
annual waste volume by a factor of 20,
based on the assumption of a 20-year
active lifetime of a municipal subtitle D
unit This conservative assumption is
based on the average active life of
Subtitle D disposal facilities in the U.S.
(Ref.7).
Petitions are also occasionally
submitted for wastes that have been
placed In a waste •management unit, but
are no longer being generated (i.e., a
petition for a one-time exclusion). In
order to maintain a consistent delisting
evaluation methodology for the landfill
scenario, the Agency believes it is
generally appropriate to also use the
same scaling factor of 20 for these one-
time exclusions. In this way, the Agency
would discourage petitions for
intermittently generated batches of
waste, which could not be delisted
based on total volume, but may pass the
volume-based delisting evaluation when
evaluated separately. However, in some
cases the Agency believes the use of a
factor of 20 may not be appropriate for
one-time exclusions. That is, if the
petitioner can conclusively demonstrate
that the waste is no longer being
generated, it will not be generated in the
future, and no other similar waste is
stored or disposed of on-site, then the
factor of 20- seems unwarranted.
Therefore, the Agency intends to
evaluate petitions for one-time
exclusions on a case-by-case basis.
For wastes that are likely to be
disposed In surface impoundments,, the
Agency will determine the size of the
unit based on the maximum annual
waste generation rate. The Agency is
assuming that most active surface
impoundments generally utilize a
continual flow-through of liquid, i.e.,
liquid waste is discharged to a surface
impoundment, solids are allowed to
settle and the remaining liquid is
discharged via an NPDES permit.
Retention times for liquids in surface
impoundments range from hours to
hundreds of days (Surface Impoundment
Survey. Ref. 8). In addition, sludge that
accumulates at the bottom of a surface
impoundment may, in some cases, be
removed to maintain the active life of
the surface impoundment Therefore, the
Agency has chosen to use the annual
volume of liquid wastes to determine the
size of surface impoundments. The
Agency believes that using the volume
of liquid waste generated over 20 years
to determine the size of the
impoundment is unnecessary and
unreasonable given the continual flow
through of liquid. One-time surface
impoundment wastes-will also be sized
based on the reported volume of stored
liquid waste.
3. Dilution/Attenuation Factors (DAFs)
a. Selection of an Appropriate DAF •
Percentile. The EPACML Monte Carlo
simulation yields a probability
distribution of DAFs. Therefore, the
selection of the DAF percentile is
essentially a selection of the level of
confidence in the DAF value. In other
words, the DAF percentile expresses the
probability that a toxic constituent
disposed of in. a municipal solid waste
landfill will undergo certain dilution/
attenuation as it moves through a
subsurface environment to an exposure
point (55 FR11828, March 29,1990). The
DAF value depends on the selected
probability or DAF percentile.
The characteristics program used the
85th percentile DAF to set regulatory
levels. The characteristics program is
designed to identify those wastes that
should be subject to hazardous waste
regulations; characteristics levels are
those equal to or above which a waste is
clearly hazardous. However, wastes
may still be hazardous even if they
leach hazardous constituents at levels
less than the TC regulatory levels.
The delisting program is meant to
define those wastes that can be let out
of the hazardous waste system; delisting
regulatory levels are those below which
a waste is clearly honhazardous.
Therefore, the Agency believes it is
appropriate that the delisting program
be more stringent than the
characteristics program.
The Agency suggested in the final TC ,
rule that more stringent DAF percentiles
may be used when input parameters are
more narrowly defined. The disposal
unit area (as a function of petitioned
waste volume) has been more narrowly
defined for delisting, and the delisting
application of EPACML produces a
distribution of DAFs for fixed volumes
of waste. The TC rale application of the
EPACML randomly selected disposal
unit size (/.a, waste volume) from the
entire range, of landfill sizes represented
in the OSW landfill survey.
The Agency believes it is justified in
selecting a confidence level for the
EPACML as adapted for delisting that is
more stringent than the 85th percentile
chosen to define the TC. Therefore, the
Agency is proposing to use the 95th
percentile DAFs for delisting. The
Agency believes the 95th percentile
represents a conservative and
reasonable worst-case disposal scenario
for delisted wastes. . . .
The 95th percentile confidence level
also represents a conventional level of
significance that is frequently used in
statistical evaluations. For example,
when evaluating ground-water
monitoring data for hazardous waste
management facilities, as required by 40
CFR 264.97, the Agency uses a
confidence level of 95% for determining
constituent concentrations in multiple-
well comparisons (53 FR 39728; October
11,1988). Since the goals of the delisting
and RCRA ground-water monitoring
programs are similar,, Le., determining if
wastes are releasing hazardous
constituents to ground water at levels
above health-based thresholds, the
Agency believes the 'use of the same
statistical significance level is
appropriate.
b. Delisting DAFs for Landfills.
EPACML analyses for landfills produced
DAFs for a range of petitioned waste
volumes that might be considered by the
delisting program. The resulting DAF
curve for landfills is shown in Exhibit 1,.
and representative DAF values for
selected volumes are given in Table 1.
The results are consistent with the
expected degree of dilution as a function
.of waste volumes; i.e., leachate from
small waste volumes undergoes a
greater degree of dilution than large
waste volumes. It can also be noted that
the highly non-linear nature of the curve
(rapid decrease in DAF as waste volume
increases) may provide an additional
disincentive for dilution of wastes to
decrease the concentration of hazardous
constituents (/.e., dilution of a waste will
result in higher volumes and, therefore*
lower DAFs). The Agency believes that
the use of a "sliding scale" based on
volume to evaluate delisting petitions
may also provide Incentives to
petitioners to minimiae waste •
generation.
BILUNQCODE 6S60-SMI
-------
Exhibit I
180
EPACML Landfill and Surface Impoundment
OAFS for Delisting
LF 95th Percent He
—SI 95th Percentile
100
§•
I
00
oo
I
n,
s
CO
CO
I
o
CO
CD
o-
1
CD
CO
so too iso soo
Waste Volume (Thousands of Cubic Yards)
250
300
w
-------
33000
Federal Register / Vol. 56, No. 138 / Thursday, July 18» 1991 / Propose j
TABLE L—DAFs FOR LANDFILLS AND
SURFACE IMPOUNDMENTS
Wu!a volume (cubic
yanis p«r you)
1,000 ...„„......,.„..,_..,....,
1,250 «.^,M.^,.™» ,
1 500 .**»•«•»•..(...»«»..,....
1,760 —..,._.„.,.,.._....,-
2,000 ..._„.„,.....„....,....
2,500 .— .,~™..~_
3,000 ........ .-..,......_
4,000 «_™.__..«_™™...
5,000 ...............................
6,000 —.„,«—«...,.,.-.....„
7,000 „..„„..,_.............,....
8,000 „«,,..„, ™
9,000 .......,...,_........._..,..,
10,000 _w,..__.™™..
1 2,500 ™«,.™,«.w_,.,.....
16,000 ......I........... ...... „..,,
20,000 „„.,„„.__,„,„:,„„„
25,000 .M^^™..... _,„..
30.000 ~-M._.™,»_^,,
40,000 ™.«, ,.,»™»
50,000 .«,,,,,™....,.,.™......
eo.ooo.™™™.™
80,000 ,™™™«,..,.,.™,...,.
00,000 ,„....„...,....,.,.,._....
100.000_™~™«
160,000.™,™
200,000 ™™ .„„
gf^OQQ,,,-,,,, -,„„„„„,„„,
300,000 __.____.„.._,..
95th percentila
Landfill
MOO
86
90
84
79
74
68
57
54
48
45
43
40
36
33
29
27
24
23
20
19
17
17
16
15
14
13
12
12
Surface
impound-
ments
MOO
MOO
93
85
1 78
68
63
54
48
43
41
36
35
34
29
26
23
20
18
15
13
12
10
10
9
7
6
6
6
characteristic levels that define wastes
that are clearly hazardous. The Agency
believes that the use of waste volume as
a determining factor in delisting is
consistent with the criteria used to
define waste as hazardous through the
listing of hazardous waste (see 40 CFR
1DAF cutoff Is 100 corresponding to the Toxiclty
Characteristic nuto.
c. Deli'sting DAFa for Surface
Impoundments. EPACML analyses for
surface Impoundments also produced
DAFs for a range of petitioned waste
volumes. Exhibit 1 shows that the
surface impoundment DAFs for the 95th
pcrcentile are somewhat lower than the
landfill DAFs despite the use of a
scaling factor of 20 for the landfill DAFs,
compared to a factor of one for the
impoundment scenario. This is a
function of the relatively higher
infiltration rate for surface
impoundments, as a result of the liquid
in the surface impoundment and
subsequent increase in hydraulic head.
The difference between landfills and
surface impoundments becomes more
pronounced as volume increases. The
Agency believes that the use of lower
DAFa for surface impoundments than
landfills is representative of the higher
'eacbale flux that may potentially occur
at surface impoundments.
d. Comparison of Characteristics and
DcJisting DAFs. The range of disposal
unit areas (petitioned waste volumes)
and the use of a higher DAFpercentile
for delisting have resulted in DAF
values that are different than those
promulgated for characteristic wastes.
As noted previously, the Agency
believes it is appropriate for delisting
levels to be more conservative than
The Agency believes that the
EPACML provides a "continuum" for the
definition of hazardous waste, from the
TC level (DAF of 100) to a DAF of
approximately 10 based on the volume
of waste. Note that this is an incentive
to minimize waste and a disincentive to
dilute wastes to meet delisting levels.
The delisting regulations (40 CFR •
260.22(d)(3)) preclude delisting any
waste that exhibits a Characteristic,
including the TC. Therefore, DAFs used
in delisting will not exceed the DAF of
100 that serves as the basis for the
regulatory TC levels, and delisting DAFs
for waste volumes below 1,000 cubic
yards will not increase above 100.
C. Use of the EPACML in Delisting
Evaluations
1. General Approach
The Agency intends to use the
EPACML (modified for delisting
purposes), when appropriate, as a tool
to estimate the concentration of
contaminants in ground water that could
result from a potential contaminant
release from a waste disposal unit The
EPACML allows the Agency to
determine the effect the petitioned
waste could have on human health if it
is not managed as a hazardous waste.
The Agency considered the results of
the model evaluation as a factor in
proposing to exclude waste from the
lists of hazardous waste in subpart D of
40 CFR part 261 in today's notice (see
section n).
The typical input parameters for the
EPACML in the delisting process are the
constituent leachate concentration, the
type of waste unit [i.e., landfill or
surface impoundment) and the volume
of the petitioned waste. The EPACML
allows a "sliding-scale" approach to be
used to determine the DAF value based
on a specific petitioned waste volume.
This approach involves identifying the
most likely disposal scenario (typically
landfills and surface impoundments)
and selecting the DAF values based on
the reported volume of waste.
2. Selection of Landfill or Surface
Impoundment Disposal Scenario
The Agency will select the
appropriate model and DAF value, for
either landfills or surface
impoundments, depending on the
expected disposal scenario for the
petitioned waste. The assessment of
whether or not a petitioned waste is
disposed hi a landfill or surface
impoundment will be determined on a
case-by-case basis. The Agency will
determine which type of model to use
(landfill or surface impoundment) based
on the disposal method, the waste
composition [e.g., percent solids), and '
current methods used to manage the
waste.
3. Impact on Information Needed from
Delisting Petitioners
The Agency does not anticipate that
use of the EPACML for delisting
decisions, as appropriate, will have a
significant impact on the information
needed for facilities submitting a
delisting petition. As before, petitioners
will be required to submit the
information and data required by 40
CFR 260.20 and 260.22. Waste sampling
procedures for the collection of input
leachate data remain the same as those
used for other fate and transport models
and are not likely to impose any
additional burden on the petitioner (see
"Petitions to Delist Hazardous Wastes:
A Guidance Manual", EPA publication
EPA/530-SW-85-003;; April 1985).
EPA solicits comments on all aspects
of this model and its use in the .context
of this delisting petition, which is
discussed in more detail below.
II. Proposed Exclusion
A. Background
1. Approach Used to Evaluate This
Petition
This .petition requests a delisting for a
Listed hazardous was te. In making the
initial delisting determination, the .
Agency evaluated the petitioned waste
against the listing criteria and factors
cited hi 40 CFR 261.11 (a)(2) and £a)(3).
Based on this review, the Agency agrees
With the petitioner that the waste is non-
hazardous with respect to the original
listing criteria. (If the Agency had found,
based on this review, that the waste
remained hazardous based on the
factors for which the waste was
originally listed, EPA would have
proposed to deny the petition.) EPA then
evaluated the waste with respect to
other factprs or criteria to assess
whether there is a reasonable basis to
believe that such additional factors
could cause the waste; to be hazardous.
The Agency considered whether the
waste is acutely toxic, and considered
the toxicity of the constituents, the
concentration of the constituents in the
waste, their tendency to migrate and to
bioaccumulate, their persistence hi the
environment once -eleased from the
-------
Federal Register / Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules
33001
waste, plausible and specific types of
management of the petitioned waste,
and the quantities of waste generated.
For this delisting determination, 'the
Agency identified plausible exposure
.routes for hazardous constituents
present in the waste. The Agency is
proposing the use of an organic leachate
model in lieu of analytical results
quantifying the mobility of hazardous
organic constituents in the petitioned
waste. The Agency also used the
EPACML described in Section I of this
notice'to predict fee concentration of
hazardous constituents that may be
released from the petitioned waste after
disposal and to determine the potential
impact of the unregulated disposal of
Reynolds' petitioned waste on human
health and the environment.
Specifically, the EPACML was used to:
(1) Predict compliance-point
concentrations which were then
compared directly to the levels of
regulatory concern for particular
hazardous constituents, and (2)
calculate proposed maximum allowable
leachable concentrations (I.e., delisting
levelsj for fee hazardous constituents of
concern in the waste.
EPA believes that this fate and
transport model represents a reasonable
worst-case waste disposal scenario for
the petitioned waste, and that a
reasonable worst-case scenario is
appropriate when evaluating whether a
waste should be relieved of the
protective management constraints of
RCRA subtitle C. Because a delisted
waste is no longer subject to hazardous
waste control, the Agency is generally
unable to predict and does not control
how a waste will be managed after
delisting. Therefore, EPA currently
believes that it is inappropriate to
consider extensive site-specific factors
when applying the fate and transport
model. For example, a generator may
petition the Agency for delisting of a
metal hydroxide sludge which is
currently being managed in an on-site
. landfill and provide data on the nearest
drinking water well, permeability of the
aquifer, dispersivities, etc. If the Agency
were to base its evaluation solely on
these site-specific factors, the Agency
might conclude that the waste, at that
specific location, cannot affect the
closest well, and the Agency might grant
the petition. Upon promulgation of the
exclusion, however, the generator is
under no obligation to continue to
manage the waste at the on-site landfill.
In fact, it is likely that the generator will
either choose to send the delisted waste
off site immediately, or will eventually
reach the capacity of the on-site facility
and subsequently send the waste off site
to a facility which may have very
different hydrogeological and exposure, •
conditions.
The Agency also considers the
applicability of ground-water monitoring
data during the evaluation of delisting
petitions. During the development of its
thermal treatment process using a rotary
kiln, Reynolds disposed of the kiln
residue in an on-site non-hazardous
landfill at its Bauxite, Arkansas facility.
In this case, the Agency determined that
it would be inappropriate to request
ground-water monitoring data from
Reynolds' on-site landfill for several
reasons. First, the waste disposed of in
the on-site landfill was generated over a
period of several years during the
development of the thermal treatment
process. Therefore, the landfill contains
wastes treated using various process .
operating conditions and the wastes are
not fully representative of the final
treatment process or the petitioned
waste. Second, the petitioned waste is
not currently generated or disposed of in
the on-site landfill. For these reasons,
any ground-water monitoring data
collected from the on-site landfill would
not characterize the effects of the
petitioned waste on the underlying
aquifer at the eventual disposal site and
thus would serve no purpose.
Furthermore, the petitioned waste;
presently classified as EPA Hazardous
Waste No. K088, was not a listed
hazardous waste in the State of
Arkansas during the period that
Reynolds disposed of the waste in its
on-site landfill. Under 40 CFR
271.21{e)(2), States, such as Arkansas,
with final authorization for their
hazardous waste management programs
must modify their programs to reflect
Federal program changes within
specified time frames. The deadline by
which an authorized State had to modify
its program to adopt the Federal listing
of K088 was July 1,1990 (provided no
statutory change was needed). See 53
FR 3S412, September 13,1988. Because
Reynolds did not place any of its
thermally-treated K088 waste in the on-
site landfill after June 30,1990, and the
landfill never received any other
hazardous waste, the landfill is not
subject to the ground-water monitoring
requirements of 40 CFR part 264 or 265
or the equivalent authorized state
requirements. Therefore, groundwater
monitoring data is not available for the
petitioned waste. EPA has proposed a
rule clarifying die Agency's use of
ground-water data in delisting decisions
(see 54 FR 41930, October 12,1989}.
Reynolds petitioned the Agency for an
upfront exclusion (for a waste that has
not yet been generated) based on
descriptions of a full-scale process used
to treat spent potliners, characterization
of untreated spent potliners, and results
from the analysis of kiln residue
generated at Reynolds' Bauxite,
Arkansas facility during the treatment of
spent potliners from four Reynolds
aluminum reduction facilities.
Specifically, Reynolds requested an
upfront exclusion for. kiln residue
generated from the treatment of spent
potliners received from the four
• Reynolds facilities and any other
aluminum reduction facility. In addition,
Reynolds plans to move its thermal
treatment process from Bauxite,
Arkansas to another Reynolds facility
located in Gum Springs, Arkansas, and
requested that the upfront exclusion
apply to kiln residue generated at the
new facility location. Moreover,
Reynolds requested that the exclusion
also apply to future waste generated by
one additional rotary kiln, hi order for
Reynolds to expand its spent potliner
treatment capacity. The second kiln
would be established in conjunction
with the first kiln, in Gum Springs,
Arkansas, and similarly treat spent
potliners.
Similar to other facilities seeking
upfront exclusions, this upfront
exclusion would be contingent upon
Reynolds conducting analytical testing
of representative samples of the .
petitioned waste once the treatment unit
is on-line at the new facility location.
Specifically, Reynolds will be required
to collect representative samples from
each of the rotary kilns once they are
operational in Gum Springs. Arkansas,
to verify that the rotary kilns are on-line
and operating as described in the
petition. The verification testing requires
Reynolds to demonstrate mat the rotary
kilns, once on-line at its R.P. Patterson
facility in Gum Springs, Arkansas can
render spent potliners non-hazardous
" [i.e., meeting die Agency's verification ,
testing conditions).
From the evaluation of Reynolds'
delisting petition, a list of constituents
was developed for fee verification
testing conditions. Proposed maximum
allowable leachable concentrations for
these constituents then were derived by
back calculating from fee delisting
health-based levels through fee
EPACML adapted for use in delisting
(see Part I of this notice). These
concentrations {ie., "delisting levels")
are fee proposed verification testing
conditions of fee exclusion.
The Agency encourages fee use of
upfront delisting petitions because they
have fee advantage of allowing fee
applicant to know what treatment levels
for constituents will be sufficient to
-------
33002
Federal Register / Vol. 56, No. 138 /Thursday, July 18,. 1991 / Proposed Rules
render specific wastes non-hazardous,
before investing in new or modified
waste treatment systems. Therefore,
upfront delistings will allow new.
facilities to receive exclusions prior to
generating wastes, which, without
upfront exclusions, would unnecessarily
have been considered hazardous.
Upfront delistings for existing facilities
can be processed concurrently during
construction or permitting activities;
therefore, new or modified treatment
systems should be capable of producing
wastes that are considered non- ,
hazardous sooner than otherwise would
be possible. At the same time,
conditional testing requirements to
verify that the delisting levels are
achieved by the fully operational
treatment systems will maintain the
Integrity of the delisting program and
will ensure that only non-hazardous
wastes are removed from subtitle C
control
Finally, the Hazardous and Solid
Waste Amendments of 1934 specifically.
require the Agency to provide notice
and an opportunity for comment before
granting or denying a final exclusion.
Thus, a final decision will not be made
until all public comments on today's
proposal to exclude Reynolds' waste are
addressed and the proposal to use the
EPACML in this delisting decision is
finalized. ,
B. Disposition ofDelisting Petition
1. Reynolds Metals Company. Gum
Springs, Arkansas
a. Petition for Exclusion. Reynolds
Metals Company's [Reynolds) Hurricane
Creek facility, located in Bauxite,.
Arkansas, is a closed bauxite mine and
alumina refining processing plant.
Reynolds petitioned the Agency to
exclude kiln residue derived from
processing spent potliners using its
rotary kiln treatment process. The kiln
residue is presently listed, in accordance
with 40 CFR 261.3[c)(2)(i) (i.e., the
"derived from" rule), as EPA Hazardous
Waste No. K088—''Spent potliners from
primary aluminum reduction". The listed
constituent of concern for K08S waste is
cyanide (complexes) (see 40 CFR part
201, appendix VH). Reynolds plans to
move its spent potliner treatment
equipment from Bauxite, Arkansas to its
RJP. Patterson facility located at Gum
Springs, Arkansas, if a final exclusion
for the treatment residue is granted.
Reynolds petitioned to exclude its kiln
residue because it does not believe that
the waste will meet the criteria of the
listing. Reynolds also believes that its
treatment process will generate a non-
hazardous waste because the
constituent of concern is present at low
levels. Reynolds further believes that
the waste will not be hazardous for any
other reason (i.e., there are no additional
constituents or factors that could cause
the waste to be hazardous). Review of
this petition included consideration of
the original listing criteria, as well as the
additional factors required by the
Hazardous and Solid Waste
Amendments (HSWA) of 1984. See
section 222 of HSWA, 42 U.S.C. 6921(f),
and 40 CFR 260.22(d)(2H4). Today's
proposal to grant this petition for
delisting is the result of the Agency's
evaluation of Reynolds' petition.
b. Background. On August 14,1989,
Reynolds petitioned the Agency to
exclude its kiln residue generated from
the treatment of spent potliner wastes
by its rotary kiln process, and
subsequently provided additional,
information to complete its petition.
Specifically, Reynolds requested that
the Agency grant an upfront exclusion
(i.e., an exclusion that applies to waste
not presently generated) for kiln residue
generated by the rotary kiln process at
its future location in Gum Springs,
Arkansas. .
In support of its petition, Reynolds
submitted (1) detailed descriptions of its
xvaste treatment process; (2) a
description of the processes generating
spent potliners' that were treated by the
rotary kiln process; (3) total constituent
analysis results for the eight EP metals
listed in 40 CFR 261.24 (EPA recently
adopted the Toxicity Characteristic
Leaching Procedure (TCLP) in the
Toxicity Characteristic (TC) rulemaking
(55 FR 11798; March 29,1990) as a
replacementto the EP for the
establishment of the TC regulatory
levels and these eight metals are now
referred to as the TC metals.); (4) total
constituent analysis results for
antimony, beryllium, nickel, cyanide,
and fluoride from representative
samples of both the kiln residue and the
untreated spent potliners; (5) EP
leachate analysis results for the.eight
metals listed in 40 CFR 261.24, antimony,
beryllium, nickel, cyanide, and fluoride
from representative samples of the kim
residue; (6) TCLP leachate analyses for
the TC metals (except mercury),
antimony, beryllium, nickel, cyanide,
and fluoride from representative
samples of the kiln residue; (7) total
constituent analysis results for volatile
and semi-volatile organic compounds,
dioxins, and furans from representative
samples of the kiln residue; and (8) test
results and information regarding the
hazardous waste characteristics of
ignitability, corrosivity, and reactivity.
Similar to other facilities seeking
upfront exclusions, once an operational
rotary kiln is present at the new facility
location in Gum Springs, Arkansas,
Reynolds would be required to submit .
additional analytical data for the
petitioned waste to verify that the rotary
kiln, once on-line at the new location, '.
.- meets the treatment capability of the
Baukite rotary kiln as described hi the
petition and the verification testing
conditions specified in the exclusion
(see section ILB.l.f.-—Verification
Testing Conditions |.
Reynolds has developed a process to
treat spent potliners that have been
generated during the primary reduction
of aluminum. In support of its delisting
demonstration, Reynolds treated
potliners generated at its four primary
aluminum reduction plants in North
America (located at Longview,
Washington; Massena, New York;
Troutdale, Oregon; and Baie Comeau,
Quebec).
Approximately 23 facilities presently
generate spent potliners in the United
States at a rate of approximately 150,000
to 200,000 metric tons of spent potliner
per year. (See "Summary of Generation.
Disposal, and Treatment Practices for
Spent Potlinera from the Primary
Reduction of Aluminum", March 12,
1990, in the RCRA public docket for this
notice.) All primary aluminum produced •
in the United States is manufactured by
the Hatt-Heroult process. Aluminum is
refined by dissolving alumina
(aluminum oxide) in a molten cryolite
(NasAlFe) bath and then introducing a
direct electric current to reduce the
alumina to aluminum. The reduction
takes place in carbon-lined, cast iron
electrolytic cells or pots. These pots
consist of a steel shell lined with
refractory brick with an inner lining of
carbon. .
The cathode of the aluminum
reduction cell is a carbon liner on which
the pool of cryolite/molten aluminum
rests. Alumina is added to the.bath
periodically to maintain the
concentration of dissolved alumina •
within the desired range. The aluminum
is withdrawn intermittently from the
bottom of the molten bath. The molten
aluminum is collected in ladles and then
cast as the final product into ingots or
pigs at a separate casthouse facility. In
order to retain purity of the aluminum
product and structural integrity of the
cell, the molten aluminum must be kept
isolated from the iron shell of the cell.
Over the life of the cathode, the carbon
lining materials become impregnated
with the cryolite electrolytic solution. As
the cryolite is absorbed into the
cathode, the integrity of the lining can
be reduced and cracks or heaving of the
lining can occur. A service life of three
-------
Federal Register / Vol. 56, No. 138 / Thursday July 18, 1991 / Proposed Rifles
33003
to seven years for a potliner is common.
• Upon failure of a liner, the cell is
emptied and cooled. The steel shell is
stripped of the carbon lining by
mechanical drilling. This carbon lining,
or spent potliner, is the subject of the
K088 listing.
Reynolds originally constructed
several Traylor rotary kilns at its
Bauxite, Arkansas facility for the
purpose of calcining alumina.- The
calcining of alumina is a process that
was used as a final step in the refining
of bauxite to alumina. Specifically,
alummum trihydroxide from the
processing of bauxite was transformed
to anhydrous alumina in a rotary kiln at
approximately 1,450° F. The kilns no
longer perform this function at the
Bauxite facility. Reynolds used one of
these kirns to treat spent potliner and
generate data in support of its delisting
petition.
In the rotary kiln treatment process,
spent potliner is first crushed and milled
to a %-inch particle size. Brown sand
and limestone are also ground to %-
inch particles. Brown sand is an alkaline
mud generated in the two-stage process
of extracting alumina from bauxite. In
the past, Reynolds stored this waste on-
site atrthe Bauxite facility and has built
iap a significant stockpile of this ;
material. Reynolds mines the brown
sand from the dry lake beds at their
Hurricane Creek facility and crushes it
, for use in their treatment process. The
crushed'spent potliner, brown sand, and
limestone can then be blended in
varying ratios depending on the results
of initial spent potliner characterization
(i.e., the greater the cyanide and fluoride
levels in the spent potliner, the more
brown sand and limestone is added). In
general, the spent potliner can
contribute between 30 to 45 percent of
the influent to the treatment process
generating the petitioned waste.
Reynolds adds the brown sand to help
prevent the mixture being treated from
agglomerating in the kiln. Limestone
reacts with the soluble fluoride salts
{sodium fluoride and cryolite} in spent
potliner to form stable, relatively
insoluble calcium fluoride, thereby
reducing the leaching potential of
fluorides in the kiln residue.
The rotary kiln at the Bauxite facility
is approximately 250 feet in length and
9.5 feet in diameter and operates
counter-currently. Natural gas is used to
heat the kiln to the 1,200° F operating
temperature at the burner end. The flue
gas is sent through cyclones and an
electrostatic precipitator (ESP) to
remove solids. Reynolds plans to recycle
the solids from the cyclones to the kiln,
*vhilessolids generated from the ESP
(which are not fee subject of this
petition) will be handled as hazardous
waste. The kiln residue, the subject of
the petition, is cooled by contact
spraying with lake water and stored in
waste piles. Reynolds plans to either
dispose of the kiln residue at an pn-site
or off-site non-hazardous waste landfill,
or send the waste to an offrsite
* materials recovery facility, if the
exclusion is granted.
To collect representative samples
from a single waste treatment unit (e.g.,
rotary kiln) like Reynolds", petitioners .
are normally requested to collect a
minimum of four composite samples -
composed of independent grab samples
collected over time (e.g., grab samples
collected every hour and composited by
shift). See "Test Methods for Evaluating
Solid Waste: Physical/Chemical
Methods," U.S. EPA, Office of Solid
Waste and Emergency Response,
Publication SW-848 (third edition),
November 1986, and "Petitions to Delist
Hazardous Wastes—A Guidance
Manual," U.S. EPA, Office of Solid
Waste (EPA/530-SW-85-003), April
1985.
Reynolds petitioned for an upfront
delisting for kiln residue generated from
its rotary kiln treatment process which
will treat spent potliners from various
facilities. Therefore, Reynolds initiated a
sampling program to characterize
untreated spent potliners and treated
kiln residue. Reynolds initially collected
four composite samples of spent potliner
generated from each of three Reynolds
primary aluminum reduction facilities
(i.e., Massena, New York; Longview,
Washington; and Baie Comeau, Quebec}
for a total of 12 composite samples).
Four composite samples of untreated
material that had been crushed and
milled were collected from one railroad
car of spent potliner from each facility,
Reynolds collected grab samples of the
crushed spent potliner every five
minutes following the milling process to
form the composite samples. Reynolds
claims that the railcars containing spent
potliner from each primary aluminum
production facility are representative of
spent potliner likely to be generated at
each facility. Each of the 12 composite
samples of the untreated spent potliner
was analyzed for total constituent
concentrations (i.e., mass of a particular
constituent per mass of waste) of the TC
metals,.antimony, beryllium, nickel,
cyanide, and fluoride; total constituent :
concentrations of volatile organic
compounds; and total constituent
concentrations of semivolatile organic
compounds,
In order to characterize the variability
of the untreated spent potliners,
Reynolds also collected grab samples to
be analyzed in addition to the
composited samples. Reynolds collected
16 grab samples of untreated spent
potliner from each of the three facilities.
Each of these 48 grab-samples of
untreated waste was analyzed for total
cyanide and total fluoride.
. Reynolds treated.the spent potliner
from each facility in discrete runs. The
test period for each facility consisted of
four runs of material from each spent
potliner source. During the ransKsamples,
of the kiln residue were collected every
15"minutes and composited to form one
composite per run. Each run. consisted of
approximately 3.5 to 4-hour processing
time with a kirn residence time of
approximately 90 minutes. Thus, four
composite samples were obtained for
each of the three spent potliner sources
for a total of 12 composite samples. Each
of the 12 composite samples of kiln
residue were analyzed for total
constituent concentrations and
extraction procedure (EP) leachate
concentrations (i.e., mass of a particular
constituent per unit volume of extract)
of the TC metals, antimony, beryllium,
nickel, cyanide, and fluoride; total
constituent concentrations of volatile
organic compounds; and total
constituent concentrations of semi-
volatile organic compounds.
In order to study the variability of the
kiln residue, Reynolds took grab
samples of kiln residue at fifteen minute
intervals during one run from each
facility. Each of the 44 grab samples was
analyzed for levels of total and
leachable cyanide and fluoride.
Reynolds in its initial demonstration
collected and analyzed kiln residue
generated from a treatment process
using approximately 45 percent spent
potliner, 30 percent brown sand, and 2?
percent limestone by weight. An
evaluation of the data submitted by
Reynolds revealed that fluoride levels in
the kiln residue varied significantly from
run to run and in some cases exceeded
the maximum allowable delisting level
for fluoride established for Reynolds'
petitioned waste. At the request of the
Agency, Reynolds evaluated the results
of the treatment of the spent potliner
from fee three facilities and then
modified its treatment process to
stabilize the fluoride more effectively in
the waste. Reynolds modified its
treatment process by increasing the
ratio of brown sand and limestone to
spent potliner. Subsequently, to treat
spent potliner generated at its Troutdale,
Oregon facility, Reynolds used an
average ratio of 30 percent spent
potliner, 35 percent brown sand, and 35
percent limestone. The spent potliner
-------
from the Troutdale facility is considered
by Reynolds to contain "worst-ca.se"
levels of hazardous constituents, •
principally cyanida and fluoride. The
presence of contaminants in spent
notliners varies depending on cell design
and cell life. The Troutdale facility
utilizes pots of older design which
generate greater levels of cyanide and
fluoride in the spent potliner. Reynolds
collected four composite samples of
spent potliner generated front the ' ..•••
Troutdale facility. Reynolds sampled the
spent poUmerinfluent by .collecting
grab samples of the crushed, spent
potliner every five minutes following the
milling process to form the composite
samples. Curing Jhe run, samples of the
kiln residue were collected every 15
minutes and composited to form one
composite per run. Each run consisted of
approximately 3.5 to 4-hour processing
time with a kfln residence time of
approximately 90 minutes. Thus,
Reynolds collected a total of four
composite samples of Mln residue
generated from the treatment of spent
potliner from Reynolds' Troutdale,
Oregon facility.
Each of the four composite samples of
tha untreated spent potliner was,
analyzed for total constituent
concentrations of the metals listed in 40
CFR 2B1J24, antimony, beryllium, nickel,
cyanide, and fluoride; total constituent
concentrations of volatile organic
compounds; and total constituent
concentrations of semi-volatile organic'
compounds.
Reynolds analyzed each of the four
composite samples of kiln residue
collected for the total constituent and EP
leachate concentrations of the TC
metals, antimony, beryllium, nickel,
cyanide, and fluoride; TCLP leachate
concentrations of the TC metals (except
mercury), "antimony, beryllium, nickel,
cyanide, and fluoride; total constituent
concentrations of volatile organic
compounds; total constituent
concentrations of semivolatile organic
compounds; and total constituent
concentrations of dioxins and furans.
Reynolds claims that analytical
results generated from this sampling
scheme provided data representative of
tho Influent spent potliner, kiln residue,
and variability of the thermal treatment
process. The Agency believes that the
material generated from the initial
treatment process (i.e., treatment using
45 percent spent potliner, 30 percent
brown sand, and 25 percent limestone)
is not fully representative of the waste
generatedfrom Reynolds' modified
process. Thus, the Agency did not use
the data provided by Reynolds using its
initial treatment process in the
evaluation of this petition. While these .
data were not used in the evaluation of
tte petition, the Agency notes that
levels of all constituents except for
several fluoride levels met delisting
levels. A summary of the earlier data
and the Agency's evaluation of that,data
is presented in the RCRA public docket
for this notice. The Agency believes that
the waste generated from the modified
process is representative of the type of
waste that will be generated by
Reynolds. Since the initial Reynolds
process was not completely successful
in treating the spent potliner and the
modified process appears to have been
successful, the Agency is proposing to
limit the exclusion to spent potliners
treated using the modified Reynolds
treatment process.
c. Agency Analysis. Reynolds used
SW-846 Methods 6010 and 7041 through
7740 to quantify the total constituent
concentrations of the TC metals, nickel,
antimony, and beryllium in both the
untreated spent potliners and kiln
residue. Reynolds used Method 340.2 in
"Methods for Chemical Analysis of
Water and Wastes" to quantify the total
constituent concentrations of fluoride in
both the untreated spent potliners and
kiln residue. Reynolds used SW-846
Method 9010 to quantify the total
constituent concentrations of cyanide in
the spent potliners and kiln residue.
Reynolds used SW-848 Method 1310
(standard EP) to quantify the leachable
concentrations of the TC metals,
antimony, beryllium, nickel, cyanide,
and fluoride in the kiln residue.
Reynolds used the TCLP (SW-848
Method 1311 as described in 40 CFR Part
261, Appendix H) to quantify the
leachable concentrations of the TC
metals (except mercury), antimony,
beryllium, nickel, cyanide, and fluoride
in the kiln residue..
Table 2 presents the maximum total
constituent concentrations of the TC
metals, antimony, beryllium, nickel,
cyanide, and fluoride for the untreated
spent potliners and kiln residue from the
Troutdale facility. Table 3 presents the
maximum EP and TCLP leachate
concentrations of the TC metals,
antimony, beryllium, nickel, cyanide,
and fluoride in the kiln residue from the
treatment of this spent potliner.
Detection limits in Tables 2 and 3
represent the lowest concentration
quantifiable by Reynolds, when using
the appropriate SW-848 analytical
method to analyze its waste. (Detection
limits may vary according to the waste
and waste matrix being analyzed, Le.,
the "cleanliness" of waste matrices
varies and "dirty" waste matrices may
cause interferences, thus raising the
detection limits.)
TABLE 2.—MAXIMUM TOTAL CONSTITUENT
CONCENTRATIONS (PPM) * UNTREATED
SPENT POTLINER AND KILN RESIDUE
Constituents
Antimony.........™.....-
Arsenic'.... ".. ••
Barium.. — ..
Cadmium
Chromium -»
Lead.:
Mercury -
Nickel ....-,
Selenium
Silver. .........
Fluoride
Untreated
i&pent potliner
<3.3
<40.0
180.0
19.0
0.44
26.0
26.0
<0.1
51.0
<2.0
0.99
5,800.0
113,000.0
Kiln
residue
16.0
<40.0
110.0
7.2
1.9
66.0
13.0
<0.1
35.0
<2.0
4.4
16.0
35,000.0
< Denotes that tha constituent was not detected
at the detection limit specified in the table.
«These levels represent the highest concentra-
tions of the constituents found in Troutdale samples
of untreated spent potlinar and kiln residue collected
by Reynolds. The maximum level of a specific con-
stituent in the untreated spent potliner does not
necessarily correspond to the maximum level of the
constituent in the kiln residua. I" addition, these
levels dp not necessarily represent the specific
levels found in one sample.
Using SW-846 Method 9070, Reynolds
determined that its kiln residue had a
maximum oil and grease content of
0.0133 percent; therefore, the EP
analyses did not have to be modified in
accordance with the Oily Waste EP
methodology (i.e., wastes having more
than one percent total oil and grease
may either have significant
concentrations of constituents of
concern in the oil phase, which may not
be assessed using fthe standard EP .
leachate procedure*, or tie concentration
of oil and grease may be sufficient to
coat the solid phase of the sample and
interfere with the leaching of metals
from the sample). See SW-846 Method
1330.
TABLE 3.—MAXIMUM EP AND TCLP
LEACHATE CONCENTRATIONS (PPM)
TROUTDALE KILN RESIDUE »
Constituents
Antimony »«
Arsenic .. ;
Barium — •
Beryllium.......... — ••
Cadmium — ........
Chromium .
Mercury ..
Nickel ...•——....— ~.
Selanlum .......——.—...
Silver............-..........——""
Cyanide'™..- — .............
Fluoride -
TCLP
leachate
analyses
<0.007
0.01S
0.68
<0.008
<0.02
<0.01
0.0091
»NA
0.042
<0.002
' 0.046
0.014
29.0
EP
leachate
analyses.
<0.20
0.031
0.44-
0.0091
0.0066
-------
Federal Register / Vol. 56, No. 138 /Thursday, July 18, 1991 / Proposed Rules
33005
'These levels represent the highest concentra-
tions of the constituents" found in the Troutdale kiln
residue samples collected by Reynolds. These levels
do not necessarily represent the specific levels
found in one sample.
* Not analyzed.
3 Extraction done with distilled water instead of
acetate buffer,
Reynolds used SW-846 Method 9010
and 9030, following acidification, to
quantify the levels of reactive cyanide
and sulfide, respectively, in the kiln
residue. The maximum reported levels
of reactive sulfide and reactive cyanide
in the waste were 53 ppm and <0.25
ppm, respectively. Reynolds provided
information, pursuant to 40 CFR 260.22,
indicating that the kiln residue is not
expected to demonstrate the
characteristics of ignitability or
corrosivity. See 40 CFR 261.21 and
261.22, respectively.
Reynolds used SW-846 Method 8240
to quantify the total constituent
concentrations of volatile organic
compounds hi the untreated spent
potliner and kiln residue. Reynolds used
SW-846 Method 8270 to quantify the
total constituent concentrations of semi-
volatile organic compounds in the
untreated spent potliner and kiln
residue, following extraction by SW-846
Method 3540. Reynolds used SW-846
Method 8290 to quantify the total
constituent concentrations of dioxins
and furans in the kiln residue, A list of
the compounds analyzed by sample, and
corresponding detection limits, may be
found in the RCRA public docket for this
notice. Table 4 presents the maximum
reported concentrations for hazardous
organic constituents detected in the
Troutdale untreated spent potliner and
in the kiln residue. As in Table 2, the
detection limits in this table represent
the lowest concentrations quantifiable
by Reynolds.
TABLE 4.—MAXIMUM TOTAL CONSTITUENT
CONCENTRATIONS (PPM) » OF ORGANIC
, COMPOUNDS IN UNTREATED SPENT
POTLINER AND KILN RESIDUE
.Constituents
Acenaphthene „
Benz(a)anthracene .
Benzo(a)pyrene .
Benzo(b)fluoranthene
Benzo(k)fluoranthena
Benzo(g,h,i)pery!ene ....„
Bis(2-ethylhexyl)phtha!ate
Chrysene .„
Flubranthene ..
lndeno(1 ,2,3-c,d)pyrene
Phenanthrene .......
Pyrene............ .....
2,3,7.8-Tetrachlorodibenzo-
p-furan
Untreated
spent
potliner
15.0
40.0
53.0
'10.0
110.0
27.0
108.0
490
520
26.0
28.0
390
"NA
Kiln
residue ,
qqqqqqqc
V V VV VVV\
•£"\ Q
<1.0
<1.0
4X10-6
< Denotes that the constituent was not.detected
at the detection limit specified in the table.
•These levels represent the highest concentra-
tions of the constituents found in Trputdala samples
of untreated spent potilner and kiln residue collected
by Reynolds. The maximum level "of a specific con-
stituent in the untreated spent potliner does not
necessarily correspond to the maximum level of the
constituent In the kiln residue. In addition,, these
levels do not necessarily represent the specific
levels found in one sample.
. "Not analyzed.
Reynolds submitted a signed
certification stating that, based on
projected annual waste generation, the
maximum annual generation rate of kiln
residue from the treatment of spent
potliner produced by the four Reynolds
primary aluminum reduction facilities
will be approximately 50,000 tons per
year. The Agency reviews a petitioner's
estimates and, on occasion, has
requested a petitioner to re-evaluate ,
estimated waste volume. EPA accepts
Reynolds' certified maximum estimate
of 50,000 tons (approximately 50,000
cubic yards) per year for the four
Reynolds sites. If Reynolds uses two
kilns as proposed in their petition and
treats wastes generated by other
aluminum producers, then Reynolds
calculates that its maximum annual
treatment capacity would be 378,000
tons of kiln residue. This calculation is
based on operation of the two treatment
kilns with a maximum feed rate of 24
tons per hour, operation 24 hours per
day, 365 days per year with
approximately 10 percent kiln
maintenance downtime per year. The
Agency believes that the estimate of
378,000 tons of kiln residue generated
per year is based on assumptions that
reflect absolute maximum throughput of
the kiln and may not reflect potential
operational problems that are normally
associated with operating a kiln of this
size. Operational problems may
preclude the kilns from being operated
at their maximum capacity at all times.
Therefore, the Agency felt that a more
reasonable maximum annual volume of
waste generated by the Reynolds
process should be calculated. Reynolds
provided a probable range of the
influent materials of between 20 arid 24
tons per hour, Reynolds' demonstration
during the Troutdale test run used a feed
rate of between 21.2 and 21.5 tons per
hour. Therefore, the Agency based its
calculation on operation of two kilns
with a feed rate of 20 tons per hour,
operation 24 hours per day, operation
365 days per year with 15 percent kiln
downtime. This results hi a calculated
maximum annual volume of
approximately 300,000 tons per year.
Therefore, the Agency has chosen to cap
the volume generated at 150,000 tons per
year per kiln as a more realistic
maximum annual generation rate.
EPA does not generally verify
submitted test data before proposing
delisting decisions, and has not/verified
the data upon which it proposes to grant
Reynolds' exclusion. The sworn
affidavit submitted with this petition
binds the petitioner to'preseht truthful
and accurate results. The Agency,
however, has initiated a spot-check
sampling and analysis program to verify
the representative nature of tie data for
some percentage of the submitted
petitions. A spot-check visit to a
selected facility may be initiated before
finalizing a delisting petition or after
granting an exclusion.
d. Agency Evaluation. The Agency
considered the appropriateness of
alternative waste management
scenarios for Reynolds' kiln residue and
decided that disposal in a landfill is the
most reasonable, worst-case disposal
scenario for this waste. Under this
disposal scenario, the major exposure
route of concern for any hazardous
constituents would be ingestioh of
contaminated ground water. The
. Agency, therefore, evaluated the
petitioned waste using the modified
EPA's Composite Model for Landfills
(EPACML). See Section I of this notice
and the RCRA public docket for this
notice for a detailed description of the
EPACML and the modifications made
for delisting.
In addition, the Agency used its
Organic Leachate Model (OLM) to
estimate the leachable portion of the
organic constituents hi the petitioned
waste. See 50 FR 48953 (November 27,
1985), 51 FR 41084 (November 13.1986),
and the RCRA public docket for these
notices for a detailed description of the
OLM and its parameters. The results of
the OLM analysis were used in
conjunction with the EPACML model to
estimate the potential impact of organic
constituents on the underlying aquifer^
The Agency requests comments on the
use of the EPACML and the OLM as
applied to the evaluation of the
petitioned waste.
Specifically, the Agency used the
EPACML to evaluate the mobility of the
hazardous inorganic constituents .
detected in the EP and TCLP extract of
Reynolds' kiln residue. The Agency's
evaluation, using an estimate of 300,000
cubic yards per year and the.maximum
reported leachate concentrations (the
maximum concentrations, whether EP or
TCLP, were used, see Table 3), yielded
the compliance-point concentrations
shown in Table 5.
-------
33006
Federal Register / Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules
TABLE 5.—EPACMb COMPLIANCE-POINT
CONCENTRATIONS (PPM) TROUTDALE
KILN RESIDUE
Constituents
A»sontc~..
Sotonlum.,
Silver..
Fluoride.
Compliance-
point
concentra-
tions
0.0026
0.057
0.00076
0.00055
0.0014
0.0035
0.00051
0.0033
0.021
2.42
Levels of
regulatory
concern*
0.05
1.0
0.001
0.005
0.015
0.1
0.05
0.05
0.2
4.0
1 Sea "Docket Report on Health-based Levels and
SoJubiKtfes Ifsod In tho Evaluation of Delisting Peti-
tions," May 1991, located In tha RCRA public docket
(or today's notice.
The Agency did not evaluate the
mobility of the remaining inorganic
constituents (i.e., antimony, chromium,
and mercury) from Reynolds' waste
because they were not detected in the
EP and TCLP extracts using the
appropriate SW-848 analytical test
methods (see Table 3). The Agency
believes'that it is inappropriate to
evaluate non-detectable concentrations
of a constituent of concern in its
modeling* efforts if the nondetectable
value was obtained using the
appropriate analytical method.
Specifically, if a constituent cannot be
detected (when using the appropriate
analytical method with an adequate
detection limit), the Agency assumes
that the constituent is not present and
therefore does not present a threat to
either human health or the environment
The kiln residue exhibited arsenic,
barium, beryllium, cadmium, lead,
nickel, selenium, silver, cyanide, and
fluoride levels at the compliance point
below the health-based levels used in
delisting decision-making. Additionally,
the maximum reported levels of reactive
sulfide and cyanide in the waste (i.e., 53
ppm and <0.25 ppm, respectively) are ,
below the Agency's interim standards of
500 and 250 ppm, respectively. See
"Interim Agency Thresholds for Toxic
Gas Generation," July 12,1985, Internal
Agency Memorandum in the RCRA
public docket.
The Agency also evaluated the
mobility of the hazardous organic
constituent detected in Reynolds' waste
using the OLM and EPACML. The only
organic constituent detected was 2,3,7,8-
tetrachlorodibenzofuran (2,3,7,8-TCDF)
at 4 parts per trillion (ppt) in one
sample. The Agency believes that this
reported quantification level may be
somewhat suspect because (1) the level
detected is very close to the detection
limit (1 ppt), and (2) no other chlorinated
dioxins or furans were detected.
Typically these dioxins and furans occur
as mixtures of varying chlorine content
and 2,3,7,8 isomers are a small fraction
of the total dioxin/furan content. In any
case, the Agency evaluated the detected
concentration of 2,3,7,8-TCDF by
applying the applicable 2,3,7,8-
tetrachlorodibenzo-p-dioxin (2,3,7,8-
TCDD) toxicity equivalent factor (0.1 for
2,3,7,8-TCDF) and evaluating the
mobility of the resultant equivalent (0.4
ppt) using the OLM. A TCDD equivalent
is calculated by multiplying all detected
concentrations of tetra-, penta-, and •
hexa-chlorinated dioxins and furans by
weighting factors and summing them to
estimate a 2,3,7,8-TCDD equivalent
concentration. The calculation of TCDD
toxicity equivalents, equivalent factors,
and their derivation are described in
"1989 Update to the Interim Procedures
for Estimating Risk Associated with
Exposures to Mixtures of Chlorinated
• Dibenzo-p-Dioxins and Dibenzofurans"
U.S. EPA, Risk Assessment Forum,
March 1989.
The resulting leachable concentration
was used as an input into the EPACML
in order to assess the potential impact of
the constituent upon the ground water.
The calculated compliancepoint
concentration for this constituent (0.0001
ppt) would be below the level of •
regulatory concern (0.05 ppt; see 55 FR
30370, July 25,19SO).
The Agency does not believe that
verification testing for dioxin and furans
is necessary because (1) the Agency
believes that the influent to the kiln is
unlikely to contain dioxin precursors (2)
the one detected occurrence of 2,3,7,8-
TCDF may have been an analytical
anomaly, and (3) the level apparently
detected would pass the delisting
evaluation.
The Agency did not evaluate the
mobility of the remaining hazardous
organic constituents from Reynolds kiln
residue because they were not detected
in the kiln residue using appropriate
analytical methods. As stated.
previously, the Agency will not evaluate
non-detectable concentrations of a
constituent of concern in its modeling
efforts if the non-detectable, value was
obtained using the appropriate
analytical method.
During its evaluation of Reynolds'
petition, the Agency considered the
potential impact of the petitioned waste
via nonground water routes, specifically,
with regard to airborne and waterborne
dispersal of waste contaminants. The
Agency believes that direct contact from
airborne exposure to hazardous
contaminants from Reynolds' waste is
unlikely due to the physical and
chemical nature of the petitioned waste
[Le., Reynolds' waste "sets up" as a
result of the lime content following its
removal from the kiln and exposure to
weathering). However, due to the
significant volume of waste that
Reynolds' estimates it will generate, the
Agency evaluated the potential hazards
resulting from airborne exposure to
waste contaminants from the petitioned
waste. In this evaluation, the Agency
assumed that the petitioned kiln residue,
under some conditions, could be ground
to form loose particles. The results of
this conservative, worst-case evaluation
indicated that there is no substantial
potential hazard to human health from
airborne exposure to constituents from
Reynolds' waste. A complete description
of the Agency's assessment of the
potential impact of Reynolds' waste,
with regard to airborne dispersal of
waste contaminants., is presented in the
docket for today's proposed rale.
With regard to waterborne dispersal
of waste contaminants, the Agency
believes that it may be possible for
runoff (ie., rainwater, leachate, or other
liquid) to transport contaminants from a
waste disposal area to a nearby surface
water body. As described in today's
proposal, the Agency believes that
landfill disposal is & reasonable worst-
case management scenario for Reynolds'
petitioned waste. While contamination
of surface water might occur through
runoff from the waste disposal area
including both contaminants leached
from the waste as well as suspended
partieulate matter, the Agency believes
that the concentrations of any
hazardous constituents in this runoff
will tend to be lower than the extraction
procedure test results reported in
today's proposal due to the aggressive
acidic medium used for extraction. In
addition, the Agency believes that any
transported contaminants would be
further diluted in the receiving surface
water body. Finally, the Agency believes
that, in general, leachate derived from
the waste will hot directly enter a
surface water body without first
traveling through the saturated.
(subsurface) zone where dilution of
hazardous constituents may occur; the
EPACML accounts for the presence of
this saturated zone. As a result, the
Agency does not believe Reynolds'
treated wastes will pose a threat to
human health or the environment
through the waterbome dispersal of
waste constituents.
The Agency concluded, after
reviewing Reynolds' processes and raw
materials list, that no other hazardous
constituents, other than those tested for,
are likely to enter into the thermal
-------
Federal Register / Vol. 56. No. 138 / Thursday. July 18. 1991 /Proposed Rules
treatment process or be generated by
the process, and that no other
constituents of concern are likely to be
present in Reynolds' waste.
In addition, based on test results and
information provided by Reynolds,
pursuant to 40 CFR 260.22, the Agency
concludes that the kiln residue will not
exhibit any of the characteristics of
ignitability, corrosivity, or reactivity.
See 40 CFR 261.21, 261.22, and 261.23,
respectively.
e. Conclusion, The Agency believes
that the descriptions of Reynolds'
thermal treatment process and
analytical characterizations, in
conjunction with the proposed delisting
testing requirements, provide a
reasonable basis to grant Reynolds'
petition for an upfront conditional
exclusion. The Agency also believes
that Reynolds' sampling plan adequately
represents the variations in raw
materials and processing. Furthermore,
the Agency concludes that the data
submitted in support of the petition
show that Reynolds' process can render
spent potliners non-hazardous. The
Agency believes that, in general,
Reynolds can treat spent potliner to
reduce fluoride to levels below delisting
levels of concern. Specifically, using a
kiln influent consisting of 30 percent by
weight spent potliner, with
approximately equal parts of brown
sand and limestone by weight, the
fluoride in the spent potliner is expected
to be effectively immobilized. To
address the potential concerns regarding
fluoride, the Agency is proposing to limit
the exclusion to kiln residue generated
from influent composed of no more than
35 percent spent potliner with
approximately equal parts of brown
sand and limestone by weight That is,
the maximum ratio of spent potliner
cannot exceed the demonstrated
effective treatment ratio {30 percent) by
more than 5 percent. This will allow
Reynolds some flexibility in tailoring its
process to the treatment of a particular
spent potliner but will be similar to the
process demonstrated by Reynolds to be
effective in treating the spent potliner. In
addition, under the continuous testing
provisions of a conditional exclusion,
Reynolds will be required to retreat or
dispose as hazardous any batch
exhibiting fluoride extract levels above
a specified level (i.e., "delisting level")
(see section ILB.l.f.—Verification
Testing Conditions.)
However, the Agency is concerned
that the concentrations of the
constituents of concern in the kiln
residue may vary somewhat depending
on the quality of spent potliners
generated at various facilities.
Therefore, the Agency is proposing to
require initial and subsequent testing of
the petitioned kiln residue, prior to
disposal, to ensure that the rotary kiln
effectively handles the potential
variation in constituent concentrations
(see section H.B.l.f.—Verification
Testing Conditions).
The Agency proposes to grant a
conditional exclusion to Reynolds
Metals Company, located in Gum
Springs, Arkansas, for the kiln residue
described in its petition as EPA
Hazardous Waste No. K088. The
Agency's decision to exclude this waste
is based on process descriptions,
characterization of untreated spent
potliner waste, and results from the
analysis of kiln residue generated by the
rotary kiln located at its Bauxite,
Arkansas facility. This exclusion does
not apply to electrostatic precipitator
(ESP) dust generated by the rotary kiln.
If the proposed rule becomes effective,
the petitioned kiln residue, provided the
conditions of the exclusion are met, will
no longer be subject to regulation under
40 CFR parts 262 through 268 and the
permitting standards of 40 CFR part 270.
Reynolds requested that the exclusion
be applicable to a future kiln facility to
be established in Gum Springs,
Arkansas. Reynolds plans to move its
kiln, presently located in Bauxite,
Arkansas, to Gum Springs, if the
exclusion is granted. Reynolds plans to
construct and operate the new facility in
the same manner as the Bauxite facility
and to close the Bauxite facility. The
Agency is proposing conditional testing
requirements for Reynolds' new rotary
kiln facility location. Because the same
rotary kiln-treatment process and
influent ratio (i.e., ratio of spent potliner.
limestone, and brown sand) will be
used, the generated waste is expected to
be similar to the kiln residue generated
at the Bauxite, Arkansas facility.
As part of its petition, Reynolds
requested that the exclusion be applied
to all possible spent potliner sources in
addition to their own facilities identified
in Reynolds' petition {ia, Massena,
New York; Longview, Washington;
Troutdale, Oregon! and Baie Comeau,
Quebec). The Agency believes that the
Reynolds treatment process has the
potential to effectively treat a variety of
spent potliner material from other
aluminum producers. However, total
constituent concentrations of certain
compounds (e.g., cyanide and fluoride)
in untreated spent potliners generated at
other facilities can be somewhat higher
than those reported as detected in
Reynolds' spent potliner. (See
"Summary of Generation, Disposal, and
Treatment Practices for Spent Potliners
33007
from the Primary Reduction of
Aluminum," March 12,1990, in the
RCRA public docket for this notice.)
Therefore, the Agency is proposing to
exclude the kiln residue from the
treatment of spent potliners from other
generators only if Reynolds can
demonstrate, through extensive
verification testing, that the new waste
(i.e., spent potliners from other sources)
can be effectively treated. This
condition is described in more detail
below. '
As part of its petition, Reynolds also
requested that the exclusion be applied
to kiln residue generated from one
additional rotary kiln which Reynolds
plans to establish at the same'locatioh
in Gum Springs. Reynolds proposed to
establish this additional kiln hi order to
have the capacity to treat spent
potliners generated from primary
aluminum facilities other than the
Reynolds facilities. This exclusion will
not initially include kiln residue
generated from spent potliner produced
by facilities other than Reynolds' four
facilities. However, Reynolds may add
an additional kiln if it can demonstrate
that spent potliners from other
generators can be successfully treated.
Therefore, the Agency may grant
Reynolds' request for the scope of the
exclusion to cover one additional rotary
kiln (without further notice and
comment) if Reynolds can demonstrate
that the new kiln can meet the
verification testing conditions specified.
However, the proposed conditional
exclusion initially covers only one:kiln.
f. Verification Testing Conditions. As
stated earlier, the proposed exclusion
contains verification testing
requirements. These testing
requirements are to be conducted in two
phases, initial and subsequent. The
initial testing requirements apply to the
first 20 days that the rotary kiln, once
established at the new facility location
in Gum Springs, Arkansas, is operated
as an on-line, full-scale unit at typical
operating conditions (£e,,>similar to
those residence times, temperatures'
described in the petition, using no more
than 35 percent spent potliner by weight
and approximately equal percentages of
brown sand and limestone, and .other .
conditions specified in the initial
verification testing requirements),. The
subsequent testing requirements for the
rotary kiln apply to the period of time
following the initial 20-day period.
If the final exclusion is granted as
proposed, Reynolds will be required to:
(1) Submit information on the operating
parameters of the newly located rotary
kiln, (2) collect and analyze daily
composite samples (over a 20-day •
-------
33008
Federal Register / Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules
period) to verify that the new facility,
once on-line, meets the treatment
capability of the rotary kiln described in
the petition, and (3) continue to collect
and analyze daily and weekly samples
of the petitioned waste to verify that the
kiln residue continues to meet the
Agency's verification testing limitations
(i.e., "delisting levels"). These proposed
conditions are specific to the upfront
exclusion petitioned for by Reynolds.
The Agency may choose to modify these
proposed conditions based on comments
that may be received during the public
comment period for this proposed rule.
The proposed exclusion for Reynolds'
rotary kiln in Gum Springs, Arkansas is
conditional upon the following
requirements: •
(I) Operating Conditions:
(A) Initial Verification Testing: During the
first 20 days of full-scale operation of the
rotary kiln, at typical operating conditions,
Reynolds must monitor and submit to EPA
the rotary kiln operating conditions
(Including, but not limited to: temperature
ranga of the kiln (hot and cold end), kiln
residua exit temperature, spent potliner feed
rate, brown sand feed rate, limestone feed
rate, natural gas feed rate, oxygen/air feed
rate, and rotary kiln residence time of thfe
raw materials). The ratio.of the spent potliner
feed rate to the combined feed rates of the
spent potliner, brown sand, and limestone
mutt be no more than 0.35. Information on all
other operating conditions should encompass
all conditions used for preliminary testing
runs and those anticipated for subsequent
waste processing. During initial verification
testing, the petitioner must also demonstrate
to EPA how the range of operating conditions
could affect the process (i.e., submit analyses
of representative grab samples, as specified
under Condition (2), of the kiln residue
generated under the expected range of
operating conditions). The source of the
brown sand must be from Reynolds' dry lake
beds at the Bauxite, Arkansas facility.
Reynolds must submit the information
specified in this condition and obtained
during this Initial period no later than 90 days
after the treatment of the first full-scale batch
of spent potliner.
(B) Subsequent Verification Testing: During
subsequent verification testing, Reynolds
must monitor the performance of the rotary
kiln at all times to ensure that it falls within
the range of operating conditions
demonstrated, during initial verification
testing, to be adequate to maintain the levels
of hazardous constituents below the delisting
levels specified in Condition (4). The feed
rates of spent potliner, lime and brown sand
are to be as that described in Condition
(1)(A). Records of the operating conditions of
the rotary kiln (including, but not limited to:
temperature range of the kiln, kiln residue
exit temperature, spent potliner feed rate,
brown sand feed rate, limestone feed rate,
natural gas feed rate, oxygen/air feed rate,
and rotary kiln residence time of the raw
materials) should be maintained on site for a
minimum of fiva years. This information must
be furnished upon request and made
available for inspection by any employee or
representative of EPA or the State of
Arkansas.
The purpose of this condition is to
ensure efficient treatment of the spent
potliners. The Agency is proposing
limitations on the spent potliner feed
ratio because analytical data revealed
that the spent potliner was effectively
treated using a feed rate of
approximately 30 percent spent potliner.
Treatment of the spent potliner was not
demonstrated to be completely effective
using a feed rate of 45 percent spent
potliner. The Agency, however, would
like to allow Reynolds some flexibility
hi optimizing then1 process. Since the
kiln residue generated by the Reynolds
process is subject to verification testing,
the Agency is proposing to limit
Reynolds to using up to 35 percent spent
potliner by weight, and approximately
equal amounts by weight of brown sand
and limestone.
(2) Testing: Sample collection and analyses
(including quality control (QC) procedures)
must be performed according to SW-846
methodologies. For fluoride, samples must be
analyzed using Method 340.2 from "Methods
for Chemical Analysis of Water and Waste".
If the EPA judges the treatment process to be
effective under the operating conditions used
during the initial verification testing,
Reynolds may replace the testing required In
Condition (2}(A) with the testing required in
Condition (2)(B). Reynolds must continue to
test daily composites of kiln residue
generated beyond the time period specified in
Condition (2)(A) until and unless notified by
EPA in writing that testing in Condition (2) (A)
may be replaced by Condition (2)(B) (to the
extent directed by EPA).
(A) Initial Verification Testing: During the
first 20 operating days of full-scale operation
of the new on-line rotary kiln, Reynolds must
collect and analyze daily composites of kiln
residue. Daily composites must be composed
of representative grab samples collected
every 6 hours during each 24-hour kiln
operating cycle. The kiln residue samples
must be analyzed, prior to the disposal of the
kiln residue, for all constituents listed in
Condition (4); Reynolds must report the
analytical test data, including quality control
information, obtained during this initial
period no later than 90 days after the
treatment of the first full-scale batch of
untreated spent potliner.
The Agency has determined, through
its review of similar petitions, that
approximately four weeks are required
for a facility to tram operators and to
collect sufficient data to verify that a
full-scale treatment process is operating
correctly. Because Reynolds has already
generated data from the full-scale
process,, the Agency believes that
approximately three weeks or 20
operating days are sufficient in this
case. The initial verification testing
conditions, if promulgated as proposed,
will require daily composite samples of
kiln residue to be collected during the
first 20 operating days of full-scale
operation of the rotary kiln at the new
facility location in Gum Springs,
Arkansas. The Agency proposes this
initial verification testing condition both
to gather data obtained from the rotary
kiln and to ensure that the rotary kiln is
closely monitored during the start-up
period. If the Agency determines that
the data collected under this condition
reveal that the rotary kiln is not being
operated as described in Reynolds'
petition, the exclusion will not cover the
generated kiln residue. If the Agency
determines that the data from the initial
verification period demonstrates the
treatment process is effective, EPA will
notify Reynolds hi writing that the
testing conditions in 2(A) may be
replaced with the testing conditions hi
2(B).
As stated in section H.B.l.e., the
Agency believes that the concentrations
of the constituents of concern in the kiln
residue can vary over time depending on
the source and quality of spent potliner
treated by the kiln. As a result, in order
to ensure that Reynolds' treatment
process effectively handles the likely
variation in constituent concentrations
in spent potliner, the Agency is
proposing a subsequent verification
testing condition. The proposed
subsequent testing is expected to verify
and demonstrate that the kiln is
operated hi a manner similar to its
operation during the initial verification
testing and that the kiln residue does not
exhibit unacceptable levels of toxic
constituents even though the
composition of the feedstock (e.g., spent
potliner} may change somewhat over
tune. Therefore, the Agency is proposing
to require the Reynolds to analyze daily
and weekly composites of the kiln
residue as described in Condition (2)(B).
(B) Subsequent Verification Testing:
Following notification by EPA, Reynolds may
substitute the testing conditions in (2)(B) for
(2)(A). Reynolds must collect and analyze
both daily and weekly composites of kiln
residue. Daily composites must be composed
of representative grab samples collected
every 6 hours during a 24-hour kiln operating
cycle and these samples must be analyzed,
prior to the disposal of the kiln residue, for
leachable concentrations of cyanide and
fluoride. Weekly composites must be
composed of representative grab samples
collected every 6 hoars during a 24-hour kiln
operating cycle for each day in the week that
the kiln is operating. The weekly samples
must be analyzed, prior to the disposal of the
kiln residue, for the teachable concentrations
of the inorganics listed in Condition (4)(A)
and leachable levels of the semi-volatile
organic compounds listed in Condition (4)(B).
Analyses of both dally and weekly samples
-------
Federal Register / Vol. 56, No. 138 / Thursday* July 18, 1991 / Proposed Rules
33009
must be completed prior to the disposal of
waste generated during that week as set forth
in Condition (3). The analytical data,
including quality control information, must be
compiled, summarized, and maintained on
site for a-minimum of five years. These data
must be furnished upon request and made
available for inspection by any employee or
representative of EPA or the State of
Arkansas.
The Agency believes that collecting
daily and weekly composite samples
will ensure that Reynolds' treatment
process is able to handle the potential
variability hi concentrations of those
constituents of most concern. However,
the Agency is seeking comments as to.
whether the daily testing for leachable
concentrations of cyanide and fluoride
required in Condition (2)(B) is necessary
or if weekly testing for these parameters
(similar to the testing requirement for
the other constituents in Condition (4))
would be sufficient. In addition, the
Agency is seeking comments on whether
it would be appropriate to reduce the
required testing frequency once
Reynolds has established a significant
database on constituent concentrations
in the kiln residue.
Future delisting proposals and
decisions issued by the Agency may
include different testing and reporting
requirements based on an evaluation of
the manufacturing and treatment
processes, the waste, the volume of
waste (including whether there is a
fixed volume of waste or an infinite
source), and other factors normally
considered in the petition review
process. For example, wastes with
variable constituent concentrations,
discussed in previous delisting decisions
(see e.g., 51FR 41323, November 14,
1986), may require continuous batch
testing.
(3) Waste Holding and Handling: Reynolds
must store, as hazardous, all kiln residue
generated until verification testing (as
specified in Condition (2)(A) and (2)(B)) is
completed and compared, by the petitioner,
with the delisting levels set forth in Condition
(4). If the levels of hazardous constituents
measured in the samples of kiln residue
generated do not exceed any of the levels set
forth in Condition (4), then the kiln residue is
non-hazardous and may be managed and
disposed of in accordance with all applicable
solid waste regulations. If hazardous
constituent levels in any daily or weekly
sample exceed any of the delisting levels set
in Condition [4), the kiln residue generated
sample must be retreated until it meets these
levels (analyses must be repeated] or
managed and disposed of in accordance with
Subtitle C of RCRA. Kiln residue which is
generated but for which the required analysis
is not complete or valid must be managed
and disposed of in accordance with Subtitle
C of RCRA, until valid analysis demonstrates
that Condition (4) is satisfied.
The purpose of this condition is to
ensure that kiln residue which contains
hazardous levels of specific inorganic or
organic constituents is managed and
disposed of in accordance with subtitle
C of RCRA. Holding the kiln residue
until characterization is complete will
protect against improper handling of
hazardous material. Both the daily
composite sample and its corresponding
weekly composite sample must be
analyzed for the appropriate
parameters, and must meet the
appropriate delisting levels, in order for
the waste to be considered non-
hazardous.
(4] Delisting Levels: All concentrations
must be measured in the waste leachate by
the method specified in 40 CER 261.24.
(A) The leachable concentrations for
inorganics may not exceed the following
levels (ppm): arsenic, selenium, or silver—
0.60; barium—12.0; antimony—0.12;
cadmium—0.05; lead—0.18; chromium or
nickel—1.2; mercury—0.024; beryllium—0.012;
fluoride—48.0; and cyanide—2.4 (cyanide
extraction must be conducted using deionized
water).
(B) The leachable constituent
concentrations for organics may not exceed
the levels listed below (ppm):
Acenapthene .•„.... 24
Benz(a)anthracene 1.2X10"4
Benzo(b)fluoranthene 2.4X10"4
Benzo(a)pyrene 2.4 X 10~3
Chrysene 2.4X10~3
Fluoranthene .,.., 12
Indeno(l,2,3-cd)pyrene 2.4X10"3
Pyrene : 12
The Agency established the delisting
levels for Condition (4) by back-
calculating from the health-based levels
for the constituents of concern using the
DAF of 12 derived from the EPACML.
These delisting levels correspond to the
allowable levels measured hi the TCLP
leachate of the waste. The Agency did
not use the OLM to calculate delisting
levels for these consituents in the kiln
residue itself, because the Agency
believes that the TCLP leachate values
provide a more direct indication of
leachable levels for the kiln residue.
The Agency selected the set of organic
constituents specified in Condition (4)(B)
after reviewing information about the «•
composition of spent potliners,
descriptions of Reynolds' treatment
process, and the health-based levels
used in delisting decision-making. Most
of the these constituents {Le., PAHs) are
also products of incomplete combustion.
Condition (4)(B) as listed above
provides the list of organic constituents
for which Reynolds must test the
leachate from the kiln residue, as well
as the levels at which (or below which)
the wastes will be considered non-
hazardous. The constituents hi
Condition (4)(B) reflectall of the organir
constituents that were found in the
untreated spent potliner from Reynolds'
Troutdale facility (none were detected
in the kiln residue). The PAHs in the
verification list also serve as excellent ,
indicators of the efficiency of the
treatment process, because most of
these substances are among the most
difficult to destroy through incineration.
One phthalate ester (bis(2-
ethylhexyl)phthalate) found in the
Troutdale spent potliner was not
included hi the conditional testing list
\si a delisting level of 0.036 ppm),
because (1) it was only detected in one
out of five Troutdale spent potliner
samples; (2) it was not quantified in any
kiln residue sample; (3) it is a common
laboratory contaminant due to its use as
a plasticizer and may be a laboratory -
artifact; and (4) it is unlikely that this
compound would exist in the kiln
residue if the PAHs being monitored are
not present, because this compound is
easier to destroy by incineration than
the PAHs (see Appendix D of "Guidance
on Setting Permit Conditions and
Reporting Trial Burn Results", EPA
Publication No. EPA/625/6-89/019,
January 1989; a copy is enclosed in the
public docket for today's notice).
Five additional constituents were also
detected hi either the untreated spent
potliner or kiln residue during the
treatment of material from the first three
Reynolds facilities (Le., Massena, New
York; Longview, Washington; and Baie
Comeau, Quebec). These constituents,
(and their corresponding delisting
levels), are: dichlorodiQuoromethane
(8.4 ppm), chloroform (0.078 ppm)
methylene chloride (0.06'ppm), methyl
ethyl ketone (24 ppm), and di-n-octyl
phthalate (8.4 ppm). These organic
compounds detected in Reynolds'
samples are likely to be analytical
artifacts (most are commonly used
laboratory solvents). All but di-n-octyl
phthalate (DNOP) are volatile
compounds that clearly are not expected
to survive the high temperature (1200° F)
of the kiln. The Agency believes that it
is highly unlikely that any of these
constituents could be present at
significant levels in the treated spent
potliners. In the case of DNOP, the
compound was not found in any spent
potliner samples, and only in one out of
19 kiln residue samples at a level (1.5
ppm) barely above detection limit (1.0
ppm) and below the practical
quantification limit (5 ppm).
Furthermore, similar to the argument
made above for the other phthalate [bis
(2-ethylhexyl) phthalate), DNOP is a
common laboratory contaminant (due to
-------
33010
Federal Register / Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules
its use as a plastlcizer), and is easier to
destroy by incineration than the PAHs
chosen as verification testing
parameters. Therefore, the Agency is
confident that if the treatment process is
successful in meeting the levels for the
other difficult to destroy constituents in
Condition (4)(B), it is highly unlikely that
any of these other constituents could
exist hi the kiln residue at levels of
concern.
(6) Changes in Operating Conditions and
Waste Sources: If after completing the initial
verification tost period in Conditions (l)(A)
and (2](A), Reynolds decides to treat spent
potliner from any other primary aluminum
reduction facility; or use a new source for
brown aand; or otherwise significantly
change the operating conditions developed
under condition (11; then Reynolds must
notify EPA in writing prior to instituting the
change. Reynolds must also re-institute the
reporting and testing required in Conditions
(1){A) and (2){A), and fulfill all other
requirements in Conditions (1) and (2), as
appropriate. Reynolds may also add one
additional kiln at its R.P. Patterson facility in
Cum Springs, Arkansas if it can demonstrate
that the new kiln can successfully treat spent
potllncra. Reynolds must fulfill all
requirements contained in Conditions (1) and
(2) for the second kiln. Reynolds must
continue to test any kiln residue generated
beyond tha time period specified in Condition
(2](A) until and unless notified in writing by
EPA that testing Condition (2)(A) may be
replaced by Condition (2)(B) to the extent
directed by EPA.
The Agency is proposing that the
exclusion initially only apply to kiln
residue generated from spent potliner
generated from Reynolds' four primary
aluminum reduction facilities Massena,
New York; Longview, Washington;
Troutdale, Oregon; and Baie Comeau,
Quebec). However, the Agency believes
that the Reynolds treatment process has
the potential to effectively treat a
variety of spent potliner material from
other aluminum producers. Therefore,
the Agency is proposing to allow
Reynolds to accept spent potliners from
other generators if Reynolds can
demonstrate through verification testing
that the new waste can be effectively
treated.
Reynolds requested that the exclusion
be applied to kiln residue generated
from one additional rotary Win at the
Gum Springs facility. As discussed
above, Reynolds may add an additional
kiln if it can demonstrate that spent
potliners can be successfully treated by
the new kiln. Reynolds must fulfill all
testing and reporting requirements of
conditions [1) and (2) for the exclusion
to be hi effect for the second kiln.
(6] Data Submittals: Reynolds must notify
in writing the Section Chief, DelisUng Section
(see address below] when the rotary kiln is
on-line and two weeks prior to when waste :
treatment will begin. The data 'obtained
through Conditions (1}(A) and (2)(A) must be
submitted to the Section Chief, Delisting
Section, OSW (05-333), U.S. EPA. 401M
Street, SW., Washington. DC 20460 within the
time period specified. At the Section Chiefs
request, Reynolds must submit any other
analytical data obtained through Conditions
(1)(B) and (2)(B) within the time period
specified by the Section Chief. Failure to
submit the required data within the specified
time period or maintain the required records
on site for the specified time will be
considered by the Agency, at its discretion,
sufficient basis to revoke the exclusion to the
extent directed by EPA. All data must be
accompanied by a signed copy of the
following certification statement to attest to
the truth and accuracy of the data submitted:
"Under civil and criminal penalty of law
for the making or submission of false or
fraudulent statements or representations
(pursuant to the applicable provisions of the
Federal Code, which include, but may not be
limited to, 18 USC1001 and 42 USC 6928), I
certify that the information contained in or
accompanying this document is true, accurate
and complete.
As to the (those) identified section(s) of
this document for which I cannot personally
verify its (their) truth and accuracy, I certify
as the company official having supervisory
responsibility for the persons who, acting
under my direct instructions, made the
verification that this information is true,
accurate and complete.
In the event that any of this information is
determined by EPA in its sole discretion to be
false, inaccurate or incomplete, and upon
conveyance of this fact to the company, I
recognize and agree that this exclusion of
wastes will be void as if it never had effect or
to the extent directed by EPA and that the
company will be liable for any actions taken
in contravention of the company's RCRA and
CERCLA obligations premised upon the
company's reliance on the void exclusion."
If made final, the proposed exclusion
will initially apply only to the kiln
residue generated by one rotary kiln at
Gum Springs, Arkansas, during the
treatment of spent potliner produced by
Reynolds' four primary aluminum
reduction facilities [i.e., Massena, New
York; Longview, Washington; Troutdale,
Oregon; and Baie Comeau, Quebec). The
proposed exclusion would apply to kiln
residues generated from a second kiln at
the site, or residues from the treatment
of spent potliners from other primary
aluminum production facilities, only if
the requirements in Condition (5) are
satisfied. The maximum annual volume
of kiln residues covered by this
exclusion is a total of 300,000 cubic
yards for all treatment kirns operated by
Reynolds.
Although management of the waste
covered by this petition would be
relieved from subtitle C jurisdiction
upon final promulgation of an exclusion,
the generator of a delisted waste must
either treat, store, or dispose of the
waste hi an on-site facility, or ensure
that the waste is delivered to an off-site
storage, treatment, or disposal facility,
either of which is permitted, licensed, or
registered by a State to manage
municipal or industrial solid waste.
Alternatively, the delisted waste may be
delivered to a facility that beneficially
uses or reuses, or legitimately recycles
or reclaims the waste, or treats the
waste prior to such beneficial use, reuse,
recycling, or reclamation.
m. Effective Date
This rule, if finally promulgated, will
become effective immediately upon such
final promulgation. The Hazardous and
Solid Waste Amendments of 1984
amended Section 3010 of RCRA to allow
rules to become effective in less than six
months when the regulated community
does not need the six-month period to
come into compliance. That is the case
here, because this rule, if finalized,
would reduce the existing requirements
for persons generating hazardous
wastes. In light of the unnecessary
hardship and expense that would be
imposed on this petitioner by an
effective date six months after
promulgation and the fact that a six-
month deadline is not necessary to
achieve the purpose of Section 3010,
EPA believes that this exclusion should
be effective immediately upon final
promulgation. These reasons also
provide a basis for making this rule
effective immediately, upon
promulgation, under the Administrative
Procedure Act, pursuant to 5 U.S.C.
553(d).
IV. Regulatory Impact
Under Executive Order 12291, EPA
must judge whether a regulation is
"major" and therefore subject to the
requirement of a Regulatory Impact
Analysis. The proposal to grant an
exclusion is not major, since its effect, if
promulgated, would be to reduce the
overall costs and economic impact of
EPA's hazardous waste management
regulations. This reduction would be
achieved by excluding waste generated
at a specific facility from EPA's lists of
hazardous wastes, thereby enabling this
facility to treat its waste as non-
hazardous. There is no additional
impact, therefore, due to today's rule.
This proposal is not a major regulation;
therefore, no Regulatory Impact
Analysis is required.
V. Regulatory Flexibility Act
Pursuant to the Regulatory Flexibility
Act, 5 U.S.C. 601-612, whenever an
agency is required to publish a genpral
-------
Federal Register / Vol. 56, No. 138 / Thursday, July 18, 1991 / Proposed Rules
33011
notice of rulemaking for any proposed or
final rule; it must prepare and make
available for public comment a
regulatory flexibility analysis which
describes the impact of the rule on small
entities (i.e., small businesses, small
organizations, and small governmental
jurisdictions). The Administrator or
delegated representative may certify,
however, that the rule will not have a
significant economic impact on a
substantial number of small entities.
This rule, if promulgated, will not
have an adverse economic impact on
small entities since its effect would be to
reduce the overall costs of EPA's
hazardous waste regulations.
Accordingly, I hereby certify that this
proposed regulation, if promulgated, will
not have a significant economic impact
on a substantial number of small
entities. This regulation, therefore, does
not require a* regulatory flexibility
analysis.
VI. Paperwork Reduction Act
Information collection and
recordkeeping requirements associated
With this proposed rule have been
approved by the Office of Management
and Budget (OMB) under the provisions
of the Paperwork Reduction Act of 1980
(P.L. 98-511,44 U.S.C. 3501 etseq.) and
have been assigned OMB Control
Number 2050-0053.
VII. List of Subjects in 40 CFR Part 261
Hazardous Waste, Recycling, and
Reporting and recordkeeping
requirements. ;
"Dated: July 2.1991.
Don R. Clay,
Assistant Administrator, Off ice of Solid
Waste and Emergency Response.
For the reasons set out, in the
preamble, 40 CFR Part 261 is proposed
to be amended as follows:
PART 261—IDENTIFICATION AND
LISTING OF HAZARDOUS WASTE
1. The authority citation for Part 261
continues to read as follows:
Authority: 42 U.S.C. 6905, 6912{a), 6921,
6922, and 6938. ...'.'
2. In Table 2 of appendix IX of part
261, add the following wastestream in
alphabetical order by facility to read as
follows:.
Appendix IX—Wastes Excluded Under
§ 260.20 and § 260.22.
TABLE 2.—WASTES EXCLUDED FROM SPECIFIC SOURCES
Facility
Address
Waste description
Reynolds Metals Gum Springs,
• Company. Arkansas.
Kiln residue (generated at a maximum annual volume of 300,000 cubic yards per year} from rotary kiln treatment of spent
potliners (EPA Hazardous Waste No. K088). This exclusion does not apply to electrostatic precipitator dust generated by
the rotary kiln. This exclusion initially applies only to the treatment by one rotary kiln of potliners generated by Reynolds
Metals' four primary aluminum facilities (Massena, New York; Lortgview, Washington; Troutdale, Oregon; and Baie Comeau,
Quebec) described In the petition. Reynolds may only accept spent potliners from other sources, or modify Its treatment
process, or add an additional rotary kiln in accordance with Condition (5). This exclusion is conditional upon the submission
of data obtained from each rotary kiln after ft is established at the R.P. Patterson facility in Gum Springs, Arkansas. To
ensure that hazardous constituents are not present in the waste at levels of regulatory concern while the treatment facility
Is in operation, Reynolds must implement a testing program. This testing program must meet the following conditidns for the -
exclusion to be valid:
(1) Operating Conditions: (A) Initial Verification Testing: During the first 20 days of full-scale operation of the rotary kiln, at
- typical operating conditions, Reynolds must monitor and submit to EPA the rotaiy .kiln operating conditions (including, but
not limited to: temperature range of the kiln (hot and cold end), kiln residue exit temperature, spent pptliner feed rate, brown
sand feed rate, limestone feed rate, natural gas feed rate, oxygen/air feed rate, and rotary kiln residence time of the raw
materials). The ratio of the spent potliner feed rate to the combined feed rates of the spent potiiner, brown sand, and
limestone must be no more than 0.35. Information on all other operating conditions should encompass all conditions used
for preliminary testing runs and those anticipated for subsequent waste processing. During initial verification testing, the1
petitioner must also demonstrate to EPA how the range of operating conditions could affect the process (I.0., submit
analyses of representative grab samples, as specified under Condition (2), of the kiln residue generated under the expected
range of operating conditions). The source of the brown sand must be from Reynolds' dry lake beds at the Bauxite,
Arkansas facility. Reynolds must submit the information specified in this condition and obtained during this initial period no
later than 90 days after the treatment of the first full-scale batch of spent potliner. . .
(B) Subsequent Verification Testing: During subsequent verification testing, Reynolds must monitor the performance of .the
rotary kiln at all times to ensure that it falls within the range of operating conditions demonstrated, during initial verification
testing, to be adequate to maintain the levels of-hazardous constituents below the delisting levels specified in Condition (4).
The feed rates of spent potiiner, lime and brown sand are to be as that described in Condition (1)(A). Records of the
operating conditions of'the. rotary kiln (including, but not-limited to: temperature range of the kiln, kiln residue exit
temperature,- spent potiiner feed rate, brown sand feed rate, limestone feed rate, natural gas feed rate, oxygen/air feed
rate, and rotary kiln residence time of the raw materials) should be maintained on site for a minimum of five years. This
Information must be furnished upon request and made available for Inspection by any employee or representative of EPA or
the State of Arkansas.
(2) Testing: Sample collection and analyses (including quality control (QC) procedures) must be performed according to SW
846 methodologies. For fluoride, samples must be analyzed using Method 340.2 from "Methods for Chemical Anatysis of
Water and Waste". If the EPA judges the treatment process to be effective under the operating conditions used during the
initial verification testing, Reynolds may replace the testing required in Condition (2)(A) with the testing required in Condition
(2)(B). Reynolds must continue to test daily composites of kiln residue generated beyond the time period specified in
Condition (2)(A) until and unless notified by EPA In writing that testing in Condition (2)(A) may be replaced by Condition
(2)(B) (to the extent directed by EPA).
(A) Initial Verification Testing: During the first 20 operating days of full-scale operation of the new on-line rotary kiln, Reynolds
, must collect and analyze daily composites of kiln residue. Daily composites must be composed of representative grab
samples collected every 6 hours during each 24-hour kiln operating cycle. The kiln residue samples must be analyzed, prior
to the disposal of the kiln residue, for all constituents listed in Condition (4). Reynolds must report the analytical test data,
Including quality control information, obtained during this initial period no later than 90 days after the treatment of the fjrst
full-scale batch of untreated spent potiiner.
-------
33012
Federal Register / Vol 56, No. 138 / Thursday. July .13, 1991 / Proposed Rules
TABCS 2.— WASTES EXCLUDED FROM SPECIFIC SOURCES— Continued
FacRty
'Address
Waste description
(B) Subsequent Verification Testing: Following notification by EPA, Reynolds may substitute ths testing conditions in (2)(B) for
{2)(A). Reynolds must collect and analyze both daily and weekly composites of kiln resklue. Daily composites must be
composed of representative grab samples collected every 6 hours during a 24-hour kiln operating cycle and these samples
must be analyzed, prior to ths disposal of the kiln residue, for teachable concentrations of cyanide and fluoride. Weekly
composites must be composed of representative grab samples collected every 6 hours during a 24-hour kiln operating cyde
for each day in the week that the kiln Is operating. The weekly samples must be analyzed, prior to the disposal of »he Win
residua. Jor the teachable concentrations of the Inorganics listed in Condition (4)(A) and teachable levels of the semi-volatile
organic compounds flsted In Condition (4)(B). Analyses of both daily and weekly samples must be completed prior to the
disposal of waste generated during that week as set forth In Condition (3). The analytica'l data, including quality control
information, must be compiled, summarized, and maintained on site for a minimum of fee years. These data must be
furnished upon request and made available for inspection by any employee or representative of EPA or the State of
Arkansas.
(3) Waste Holding and Handling: Reynolds must store, as hazardous, all kiln residue generated until verification testing (as
specified in Condition (2)(A) and (2)(B)) is completed and compared, by the petitioner, with the delisting levels set forth in
Condition (4). If the lavels of hazardous constituents measured In the samples of kiln residua generated do not exceed any
of the levels set forth in Condition (4), then the kiln residue is non-hazardous and may bo managed and disposed of Jn
accordance with all applicable solid waste regulations, tf hazardous constituent levels in any daily or weekly sample exceed
any of ttie delisting levels set In Condition (4), the kiln residue generated during the tirna period coiresponding to this
sample must be retreated until it moats these levels (analyses must be repeated) or managed and disposed of in
accordance with subtitle C of RCRA. Kiln residue which is generated but for which the required analysis is not complete or
valid must be managed and disposed of In accordance with Subtitle C of RCRA, until valid analysis demonstrates that
Condition (4) is satisfied.
(4) Del/sting Levels: All concentrations must be measured in the waste leachate by the method specified in 40 CFR part
(A) Tha teachable concentrations for metals may not exceed the following levels (ppm): arsenic, selenium, or silver—0.60;
barium—12.0; antimony—0.12; lead—0.18; cadmium—0.06; chromium or nickel—1.2; mercury—0.024; beryllium—0.012;
fluoride—48.0; and cyanide—2.4 (cyanide extraction must be conducted using detonlzed water).
(B) The leachable constituent concentrations for organics may not exceed the levels listed below (ppm)-
Acenapthene—24 • . •
6enz(a)anthracene—13. x 10~4
Benzo(b)fluoranthenB—2.4X 10-«
Banzo(a)pyrens—2.4x10-»
Chrysene—2.4x10-'
Fluoranthene—12
Indeno (1,2,3-cd)pyrene—2.4x10-'
Pyrene—12 ,
(5) Change? in Operating Conditions and Waste Sources: If after completing the initial verification test period in Conditions
(1)(A) and (2){A). Reynolds decides to treat spent potliner from any other primary aluminum reduction facility; or use a new '
source for brown sand; or otherwise significantly change the operating_conditions developed under condition (1); than
Reynolds must notify EPA in writing prior to instituting the change. Reynolds must also reinsBtute the reporting and testing
required in Conditions (1)(A) and (2)(A). and fulfill all other requirements in Conditions (1) ami (2), as appropriate. Reynolds
may also add one additional kiln at Us R.P. Patterson facility in Gum Springs, Arkansas if it can demonstrate that the new
win can successfully treat spent potliners. Reynolds must fulfill all requirements contained in Conditions (1) and (2) for the
second kiln. Reynolds must continue to test any kiln residue generated beyond the time period specified in Condition (2)(A)
until and unless notified in writing by EPA that testing Condition (2)(A) may be replaced by Condition (2)(B) to the extent
directed by EPA.
(6)i Data Submittalx Reynolds must notify in writing the Section Chief. Delisting Section (see aiidress below) when the rotary
kiln Is on-line and-two weeks prior to when waste treatment will begin. The data obtained through Conditions (1)(A) and
(2){A) must be submitted to the Section Chief. Deiisfing Section, OSW (OS-333). U.S. EPA, 401 M Street, SW., Washington,
DC 20460 within the time period ispecified. At the Section Chief's request, Reynolds must submit any other analytical data
obtained through Conditions (1)(B) and (2)(B) within the time period specified by the Section Chief. Failure to submit the
required data within the specified time period or maintain the required records on site for the specified time will be
considered by the Agency, at its discretion, sufficient basis to revoke the exclusion to the extent directed by EPA. All data
must be accompanied by a signed copy of the following certification statement to attest to the truth and accuracy of the
data submitted:
"Under civil and criminal penalty of law for the making or submission of false or fraudulent statements or representations
(pursuant to the applicable provisions of the Federal Code, which Include, but may not be limited to, 18 U.S.C. 1001 and 42
U.S.C. 6928), I certify that the information contained in or accompanying this document Is true, accurate and complete
As to the (those) identified sect!on(s) of this document for which I cannot personally verify its (their) truth and accuracy, I
certify as the company official having supervisory responsibility for the persons who, acting under my direct instructions
made the verification that this information is true, accurate and complete.
In the event that any of this information is determined by EPA in its sole discretion to be false, Inaccurate or incomplete, and
upon conveyance of this fact to the company, I recognize and agree that this exclusion of wastes will be void as if it never
had effect or to ttie extent directed by EPA and that the company will be liable for any actions taken in contravention of the
company's RCRA and CERCLA obligations premised upon the company's reliance on the void exclusion."
* * * * • > *
IFR Doc. 91-10971 Filed 7-17-91; 8:45 am]
BrLUNQ CODE eMO-60-M
------- |