EPA Document# EPA-740-R1-8014
United States October 2020
* * Environmental Protection Agency Office of Chemical Safety and
Pollution Prevention
Summary of External Peer Review and Public Comments
and Disposition for
Carbon Tetrachloride
(Methane, Tetrachloro-)
CASRN: 56-23-5
ci
cr i
vl
CI
October 2020
-------
Table of Contents
Environmental Fate and Exposure 8
Environmental Hazard and Risk Characterization 39
Occupational Exposure and Releases 64
Human Health Effects 95
Risk Characterization 133
Content and Organization 175
References 206
Page 2 of210
-------
This document summarizes the public and external peer review comments that the EPA's Office
of Pollution Prevention and Toxics (OPPT) received for the risk evaluation of carbon
tetrachloride (CC14). It also provides EPA/OPPT's response to the comments received from the
public and the peer review panel.
EPA/OPPT appreciates the valuable input provided by the public and peer review panel. The
input resulted in numerous revisions to the hazard summary.
Peer review charge questions1 are used to categorize the peer review and public comments into
specific issues related to the five main themes.
1. Environmental Fate and Exposure
2. Environmental Hazard and Risk Characterization
3. Occupational Exposure and Releases
4. Human Health Effects
5. Human Health Risk Characterization
6. Content and Organization
All peer review comments for the six charge questions are presented first, organized by charge
question in the following section. These are followed by the public comments. For each theme,
general comments pertaining to all chemicals are presented first, and then additional comments
pertaining to only one or several chemicals follows.
ABBREVIATIONS
ACC
ACR
AF
AEGL
AOP
APF
AT SDR
BMCLio
BMD
BMDL
BMDS
CAA
CASRN
CDR
CFR
CI
CNS
coc
CRED
American Chemistry Council
Acute to chronic ratio
Assessment factor
Acute Exposure Guideline Levels
Adverse outcome pathway
Assigned protection factor
Agency for Toxic Substances and Disease Registry
Benchmark concentration lower bound
Benchmark dose
Benchmark dose lower bound
Benchmark Dose Software
Clean Air Act
Chemical Abstracts Service Registry Number
Chemical Data Reporting
Code of Federal Regulations
Confidence Interval
Central Nervous System
Concentration of concern
Criteria for Reporting and Evaluating ecotoxicity Data
1 These are the questions that EPA/OPPT submitted to the panel to guide the peer review process.
Page 3 of210
-------
CWA
Clean Water Act
DMR
Discharge Monitoring Report
DNA
Deoxyribonucleic acid
DTso
Time at which the amount of compound is degraded by half
ECio
Effect Concentration at which 50% of test organisms exhibit the effect
EC50
Effect Concentration at which 50% of test organisms exhibit the effect
ECOTOX
EPA's ECOTOXicology knowledgebase
EDC
Ethylene dichloride
EDF
Environmental Defense Fund
E-FAST
Exposure and Fate Assessment Screening Tool
EPA
United States Environmental Protection Agency
EPI Suite
Estimation Programs Interface suite of models
EPN
Environmental Protection Network
GWAS
Genome-wide association studies
GWP
Global warming potential
HAP
Hazardous air pollutant
HBCD
hexabromocyclododecane, representing the cyclic aliphatic bromide cluster
HEC
Human equivalent concentration
HERO
Health & Environmental Research Online
HFC
Hydrofluorocarbons
HFO
Hydrofluoro-olefines
HSIA
Halogenated Solvents Industry Alliance
IARC
International Agency for Research on Cancer
IOM
Institute of Medicine
IPCS
International Programme on Chemical Safety
IUR
Inhalation unit risk
JBRC
Japan Bioassay Research Center
Koc
Soil Organic Carbon-Water Partitioning Coefficient
Kow
Octanol-Water Partitioning Coefficient
LC10
Lethal Concentration at which 10% of test organisms die
LC50
Lethal Concentration at which 50% of test organisms die
LD10
Lethal Concentration at which 10% of test organisms die
LMS
Linearized Multistage Model
LOAEL
Lowest Observed Adverse Effect Level
LOD
Limit of detection
MACT
Maximum Achievable Control Technology
MAK
Maximale Arbeitsplatzkonzentration, or the "maximum permissible concentration
of a substance as a gas, vapour or aerosol in the air at the workplace"
MCL
Maximum Contaminant Level
MCLG
Maximum Contaminant Level Goal
MOA
Mode of Action
MOE
Margin of Exposure
MP
Montreal Protocol
NAS
National Academies of Science
NASEM
National Academies of Sciences, Engineering, and Medicine
NATA
National Air Toxics Assessment
Page 4 of 210
-------
NEI
National Emissions Inventory
NHANES
National Health and Nutrition Examination Survey
NIH
National Institutes of Health
NIOSH
National Institute for Occupational Safety and Health
NOAEL
No Observed Adverse Effect Level
NOEC
No Observed Effect Concentration
NPDES
National Pollutant Discharge Elimination System
NPL
National Priorities List
NTTC
National Tribal Toxics Council
NTP
National Toxicology Program
OECD
Organisation for Economic Co-operation and Development
OES
Occupational exposure scenario
OHAT
Office of Health Assessment and Translation
OPPT
Office of Pollution Prevention and Toxics
ONU
Occupational non-user
OSHA
Occupational Safety and Health Administration
PBPK
Physiologically based pharmacokinetic
PDM
Probabilistic Dilution Model
PECO
Populations, Exposures, Comparators, and Outcomes
PEL
Permissible exposure limits
PESS
Potentially exposed or susceptible subpopulations
PF
Protection factor
POD
Point of departure
PPE
Personal protective equipment
ppm
Parts per million
PRISMA
Preferred Reporting Items for Systematic Reviews and Meta-Analyses
QSAR
Quantitative Structure-Activity Relationship
RCRA
Resource Conservation and Recovery Act
RE
Risk Evaluation
RESO
Receptor, Exposure, Setting (or Scenario), and Outcome
RQ
Risk quotient
SACC
Science Advisory Committee on Chemicals
SCHF
Safer Chemicals Healthy Families
SDS
Safety Data Sheet
SDWA
Safe Drinking Water Act
SEG
Similar Exposure Groups
SOP
Standard Operating Procedures
SR
Systematic Review
SSD
Species sensitivity distributions
STORET
STOrage and RETrieval database
TRI
Toxics Release Inventory
TSCA
Toxic Substances Control Act
TWA
Time-weighted average
UF
Uncertainty factor
UFa
Interspecies uncertainty/variability factor
UFh
Intraspecies uncertainty/variability factor
Page 5 of210
-------
USGS U.S. Geological Survey
VOC Volatile organic compound
WHO World Health Organization
WOE Weight-of-evidence
Page 6 of210
-------
l.isl of ( oinincnls
#
Docket l-'ilc
Siihinillcr
22
EP A-HO-OPPT-2019-0499-0022.
Christopher Bevan, Director, Scientific Programs, Halogenated Solvents Industry
Alliance, Inc. (HSIA)
23
EP A-HO-OPPT-2019-0499-0023
Jonathan Kalmuss-Katz, Staff Attorney, Earthjustice et al.
26
EP A-HO-OPPT-2019-0499-0026
Richard A. Denison, Lead Senior Scientist, Environmental Defense Fund (EDF)
27
EP A-HO-OPPT-2019-0499-C
Anonymous public comment
28
EP A-HO-OPPT-20
19-0499-0028
Environmental Investigation Agency (EIA)
29
EP A-HO-OPPT-20
19-0499-0029
Christopher Bevan, Director, Scientific Programs, HSIA
30
EP A-HO-OPPT-20
19-0499-0030
Michelle Roos, Environmental Protection Network (EPN)
31
EP A-HO-OPPT-20
19-0499-0031
Suzanne Hartigan, Senior Director, Regulatory and Technical Affairs, American
Chemistry Council (ACC)
32
EP A-HO-OPPT-20
19-0499-0032
Liz Hitchcock, Director, Safer Chemicals Healthy Families (SCHF) et al.
33
EP A-HO-OPPT-20
19-0499-0033
Jennifer Sass, Senior Scientist, Natural Resources Defense Council (NRDC)
37
EP A-HO-OPPT-20
19-0499-0037
Amy McCamphill, Senior Counsel, and Amy Chyao, Assistant Corporation
Counsel, Environmental Division, New York City Law Department
38
EP A-HO-OPPT-2019-0499-003 8
Randy Rabinowitz, Executive Director, Occupational Safety & Health Law
Project and Jonathan Kalmuss-Katz and Lakendra Barajas, Staff Attorneys,
Earthjustice
39
EP A-HO-OPPT-20
19-0499-0039
Christopher Bevan, Director, Scientific Programs, HSIA
40
EP A-HO-OPPT-20
19-0499-0040
J. Warshaw
41
EP A-HO-OPPT-20
19-0499-0041
Swati Rayasam, Science Associate, Program on Reproductive Health and the
Environment, Department of Obstetrics, Gynecology and Reproductive Sciences,
University of California, San Francisco (UCSF PRHE) et al.
42
EP A-HO-OPPT-20
19-0499-0042.
Dianne C. Barton, Chair, National Tribal Toxics Council (NTTC)
43
EP A-HO-OPPT-20
19-0499-0043
Liz Hitchcock, Director, SCHF et al.
44
EP A-HO-OPPT-20
19-0499-0044
Liz Hitchcock, Director, SCHF et al. (Attachments)
45
EP A-HO-OPPT-20
Suzanne Hartigan, Senior Director, Regulatory and Technical Affairs, ACC
SACC
N/A
Science Advisory Committee on Chemicals (SACC)
Page 7 of210
-------
Kiivironiiicnliil l-'aleaml Kxposure
Charge C
Question 1.1: Please comment on the data, approaches and/or methods used to characterize exposure to aquatic receptors.
#
Summary of Comments for Specific Issues Uclalcri to
Charge Question 1
KIW/OPPT Response
Fate assumptions/models
SACC
SACC COMMENTS:
Several Committee members suggested the need to
consistently use an appropriate environmental fate model
(e.g., similar to fugacity level 3) with realistic inputs
(emissions to water and air) to determine when a log Koc
value is low enough to ignore sorption to sediment and a
Henry's law constant is high enough to ignore all other fate
processes but volatilization.
Added to section 2.1.2 Fate and Transport:
"EPI Suite ( >012a) module that estimates
volatilization from lakes and rivers ("WVol") was run using
inputs to evaluate the volatilization half-lives of CCU in
various compartments. Given the measured vapor pressure of
115 mm Hg at 20ฐC and a calculated Henry's law constant of
2.76 x 10 "2 atm-m3/mol, these physical-chemical property
inputs to the WVol model in EPI Suite indicates that CCU
will volatilize from a model river with a half-life on the order
of 1.3 hours and from a model lake on the order of
approximately 5 days. Although volatilization is expected to
be rapid, a Level III Fugacity model predicted that when CCU
is continuously released to water, 80% of the mass will
partition to water, 19% to air, <1% to soil and < 1% to
sediment. Level III fugacity modeling results are impacted by
which compartments (air, water or soil) receive the chemical
releases so a second scenario was run assuming equal releases
of CCU to all three compartments. The model predicted that
when CCU is continuously released to air, water, and soil,
50% of the mass partitions to water, 47.3% to air, 2.5% to soil
and < P/o to sediment. Intermittent releases of CCU are not
expected to result in long-term presence in the aquatic
compartment."
SACC
SACC COMMENTS:
Several members indicated that complete
biodegradation (mineralization) was unlikely to occur
Added to section 2.1.2 Fate and Transport:
"Studies have shown the formation of degradation products
Page 8 of210
-------
under most environmental conditions. The potential for
the formation of products including chloroform,
methylene chloride, methyl chloride, and phosgene
should be discussed (see example flow diagram by
Tripp et al., 2020).
Recommendation: Include a discussion of metabolic
pathways and environmental breakdown products.
such as chloroform, methylene chloride, methyl chloride, and
phosgene under various environmental conditions. Under
sulfate reducing conditions, partial complete dechlorination
of carbon tetrachloride has been observed (de Best et al..
1997). Carbon tetrachloride has been found to degrade under
anaerobic conditions to methane, carbon dioxide and carbon
monoxide through various metabolic pathways (Van Eekert et
al.. 1998). Additionally, abiotic transformation has been
observed to play an important role in degradation of carbon
tetrachloride to carbon disulfide, however substitutive and
oxidative dechlorination processes forming carbon dioxide
from degradants may pose as a potential pathway to
producing safe degradation products (Van Eekert et al..
1998)."
SACC
SACC COMMENTS:
One member found the discussion of whether CC14 wastes
are in the form of mixed liquids or as residues mixed into
solid wastes to be inadequate, as physical form affects
emissions and exposure estimates.
For Engineering: See Section 2.4.1.7.9. This section details
the disposal of carbon tetrachloride including information on
the form of the wastes for assessing emissions and exposures
of carbon tetrachloride.
23,26
PUBLIC COMMENTS:
EPA dismissed phosgene exposures because TRI data do
not show releases of CC14 and phosgene at the same
facility. At least one facility that reported releases of CC14
under the NEI also reported phosgene emissions under the
NEI and phosgene manufacture under the Chemical Data
Reporting (CDR). Other sources of data, such as the NEI,
should be considered before excluding a potential
exposure.
During problem formulation, EPA identified information on
the thermal decomposition of carbon tetrachloride into
phosgene, a highly toxic gas. However, thermal
decomposition of carbon tetrachloride is more likely to occur
in open environments and less likely in the type of closed
systems used during the manufacturing and processing of
carbon tetrachloride.
Because exposures to the general population from any
thermal decomposition of carbon tetrachloride would occur
via exposure pathways that fall under the jurisdiction of other
EPA-administered laws, such exposures are not within the
scope of the risk evaluation.
Page 9 of210
-------
EPA acknowledges that the SABIC Alabama facility reported
both carbon tetrachloride and phosgene to TRI. A facility
could have these chemicals (both phosgene and carbon
tetrachloride) onsite, but their co-existence does not imply
that the phosgene is present due to the decomposition of
carbon tetrachloride. Additional clarifications are indicated
below:
1. A site reported to TRI that the carbon tetrachloride is
manufactured as an impurity. It is not revealed which process
the carbon tetrachloride is manufactured as an impurity. This
facility involves a chlor-alkali process, that produces
chlorine. Chlorine is used in several other on-site processes.
This site also produces hydrochloric acid (HC1). SABIC, as a
company, produces ethylene dichloride (EDC, also known as
1,2-dichloroethane) and vinyl chloride monomer (VCM),
which are both sold as a product and used internally (EDC
used to make VCM, and VCM used to make PVC). Thus,
there could be a number of processes that use both chlorine
and carbon-based compounds to produce carbon tetrachloride
as an impurity (e.g., the production of phosgene, VCM,
EDC). EPA has also exercised its authority in TSCA Section
6(b)(4)(D) to exclude from the scope of this risk evaluation
conditions of use associated with carbon tetrachloride
generated as a byproduct. Carbon tetrachloride generated as a
byproduct during the manufacture of 1,2-dichloroethane will
be assessed in the risk evaluation for 1,2-dichloroethane (see
Final Scope of the Risk Evaluation for 1,2-Dichloroethane,
EPA-HQ-OPPT-2018-0427-0048).
2. The phosgene is typically manufactured to be used on-site
as a reactant due to its properties and toxicity. Phosgene is not
typically transported across the U.S. The specific site could
be producing the phosgene to use as a reactant to produce
Page 10 of 210
-------
polycarbonate (which is one of the polymer products of this
company). Phosgene is well known as a reactant that, with
bisphenol A, is used to produce polycarbonate.
Considering the above two items, there is no reason to think
the phosgene is present as a decomposition product of carbon
tetrachloride, especially when the CDR and TRI reports the
phosgene is intentionally manufactured as a site-limited
reactant.
Decomposition of carbon tetrachloride requires > 730ฐC, a
temperature at which phosgene could form from carbon
tetrachloride (Noweir et al 1973). However, phosgene,
typically formed otherwise, is not stable at temperatures
above 250ฐC, decomposes to form mixtures of carbon
monoxide, chlorine, carbon dioxide, and carbon tetrachloride
(ACC. 2018).
26
PUBLIC COMMENTS:
EPA cited no sources to demonstrate that decomposition of
CC14 is "more likely" to occur in open systems, which
EPA alleges will not happen because CC14 is only
manufactured and processed in closed systems. EPA does
not explain how releases to the environment of CC14
would not decompose and result in exposures to phosgene.
The SACC should call on EPA to address its failure to
consider CC14's decomposition into phosgene and any
resulting exposures to phosgene.
Carbon tetrachloride storage and handling are reported to be
performed in close and secure vessel (OxvChem. 2014). In
addition, samples could only be collected (potential release
source) from the closed systems that have built-in capabilities
to handle vents, provide nitrogen, process unused liquid
volume and results in a sample in a closed container
(OxvChem. 2.014). (OxvChem. 2018) reported closed loop
unloading systems are designed to minimize solvent vapor
emissions during transfer by exchanging the liquid solvent in
the trailer with the storage tank vapors. In addition, it was
also reported that the closed system cuts the water usage
(resource needs) and release of carbon tetrachloride
(Cheremisinoff and Rosenfeld. 2009). Carbon tetrachloride
has no flash point, it is not flammable.
Decomposition of carbon tetrachloride requires > 730ฐC, a
temperature at which phosgene could form from carbon
Page 11 of 210
-------
tetrachloride (ACC. 2018). However, phosgene, typically
formed otherwise, is not stable at temperatures above 250ฐC,
decomposes to form mixtures of carbon monoxide, chlorine,
carbon dioxide, and carbon tetrachloride (ACC, 2018).
Because exposures to the general population from any
decomposition of carbon tetrachloride would occur via
exposure pathways that fall under the jurisdiction of other
EPA-administered laws, such exposures are not within the
scope of the risk evaluation.
The revised risk evaluation document has also included the
following sentences:
"Carbon tetrachloride should be stored in labelled, airtight
containers in a well-ventilated place protected from light and
at a temperature below 30ฐC. It must be stored separated from
chemically active metals. Disposal of carbon tetrachloride
contaminated wastes via incineration is not recommended due
to the non-flammability of carbon tetrachloride and to the
formation of phosgene, hydrogen chloride and other toxic
gases on heating."
Presenta
tion of physical-chemical and fate properties
SACC
SACC COMMENTS:
Replaced section 1.1 Physical and Chemical Properties with:
"Physical-chemical properties influence the environmental
behavior and the toxic properties of a chemical, thereby
informing the potential conditions of use, exposure pathways
and routes and hazards being evaluated. A summary of the
physical and chemical properties of carbon tetrachloride are
listed in Table 1-1. Carbon tetrachloride is a colorless liquid
at room temperature with a sweet, aromatic and ethereal odor
resembling chloroform (Merck. 1996); (U.S. Coast Guard.
>). It is water miscible, has a melting point of -23 ฐC, a
Instances of incorrect terminology were noted:
CC14 is referred to as moderately miscible (p. 25, line
299). A compound is either miscible in water or not. It
cannot be partially miscible.
The risk evaluation states that CC14 is expected to
volatilize based on its high vapor pressure (p. 25, line
297). Vapor pressure is related to intermolecular
interactions, whereas volatilization depends on
interactions between CC14 molecules and the
Page 12 of 210
-------
environmental phases it is in contact with, along with
environmental conditions.
boiling point of 76.8 ฐC and its' density is 1.4601 g/cm3 at
20ฐC (Tide. 1999). Carbon tetrachloride has a Henrv's Law
Constant of 0.0276 atm m3/mole and a log Kow value of
2.8 3 (Lei eh ton and Calo. 1981); (Hansch et al.. 1995). Other
pertinent physical-chemicals properties are listed below in
Table 1-1."
The language regarding miscibility was changed to state that
carbon tetrachloride is "water miscible."
Volatilization is further discussed in section 2.1 Fate and
Transport. Additional detail was added on the level of
volatilization that is estimated to occur form different
environmental phases and under different environmental
conditions.
SACC
SACC COMMENTS:
Recommendation: Better define the quality and variability
associated with physical-chemical properties.
The discussion on data quality assessment and
variability for physical-chemical and fate properties,
including those obtained from EPA's EPI Suite
(both experimental and estimated values), should be
expanded. Several SACC members suggested
estimating confidence intervals (CIs) around each
property and conducting a sensitivity analysis to
determine if variability would change the outcome of
the quality pathway analysis.
Due to the differences among study conditions, generating
confidence intervals for each property would be very
complex. However, the range and quality of available data
was considered in the fate assessment of carbon tetrachloride.
The sources used to collect physical-chemical property data
for carbon tetrachloride were all subjected to data quality
evaluations based on metrics presented in the Application of
Systematic Review in TSCA Risk Evaluations document, and
the full data quality assessments are presented in a
supplemental file.
SACC
SACC COMMENTS:
When more than one estimation method is available in EPI
Suite, the estimation method that was used should be
specifically stated and the rationale for selecting one
estimation method over another should be provided. The
quality of the estimated value should be based on the
When multiple values are available, EPA presents the range
of values. Additional language regarding the use of EPI
Suite is provided in section 2.1 of the risk evaluation. EPA
employs guidance located in the EPI Suite User's Manual and
help files, along with scientific judgment to make decisions
on endpoint applicability. This suggestion will be considered
Page 13 of 210
-------
reliability of the estimation method.
further as we continue to develop our systematic review
process.
Current releases
28, 32,
43
PUBLIC COMMENTS:
The EPA draft risk evaluation reports that, according
to TRI data, U.S. air emissions for reporting facilities
totaled over 176,000 pounds in 2018. The 2016
Stratosphere-troposphere Processes And their Role in
Climate (SPARC) Report on the Mystery of Carbon
Tetrachloride concludes that the scale of emissions of
CC14 is several orders of magnitude higher than TRI
data suggest. The SPARC report estimated total CC14
emissions of 20ฑ5 Gg/year, narrowing, but not
eliminating, the gap with top down estimates.
CC14 is produced as a co-product of PCE
production or as a co-product of CM
production. In total, the combined emissions of
CC14 from PCE and CM plants, or unreported
non-feedstock uses, is 13 Gg.
CC14 is widely used as a feedstock to
manufacture hydrofluorocarbons (HFCs) and its
production and use is expected to expand
further to produce their replacements,
unsaturated HFCs or hydrofluoro-olefines
(HFOs). CC14 production for these so-called
'nondispersive' applications globally totaled
~200 Gg in 2012-2014 based on which bottom-
up emissions contributions of 2 Gg/year from
feedstock use have been derived.
The production and use of CC14, and thus
potential emissions of CC14 in chlor-alkali
The SPARC report is an important reference addressing
global sources and sinks of carbon tetrachloride. EPA
acknowledged in the final risk evaluation the global sources
of carbon tetrachloride in the atmosphere including feedstock
uses and non-feedstock emissions. Please see revised
paragraph in Section 1.2 (line 1187 - 1196). The revision
includes various carbon tetrachloride sources and their
emissions. The reasonably available information includes
citations of peer-reviewed articles used to inform global
sources of carbon tetrachloride. Assessing global emissions
of carbon tetrachloride is outside the scope of the risk
evaluation.
EPA did not include the emission pathways to ambient air
from commercial and industrial stationary sources, because
stationary source releases of carbon tetrachloride to ambient
air are under the jurisdiction of Section 112 of CAA. In
addition, carbon tetrachloride production and use are
controlled under the 1987 Montreal Protocol. Resulting
exposures were out of scope as described in section 1.4.3 of
the final risk evaluation for carbon tetrachloride.
Under TSCA section 6(b), EPA is required to determine
whether a chemical substance presents unreasonable risks
without consideration of costs or other non-risk factors.
Consideration of technically and economically feasible
alternative substances is a step that may occur as part of a
potential risk management action developed pursuant to
TSCA section 6(c)(2)(C). This type of analysis could be
considered as part of a subsequent risk management action if
Page 14 of 210
-------
facilities, constitutes up to -10 Gg of emissions
each year. There is no recognized alternative to
CC14 in chlor-alkali production, and no
foreseeable end to the use of CC14 as a
feedstock or chemical intermediary.
While production of CC14 continues, illegal trade and
use of CC14 is expected to persist. Recent use of CC14
as a feedstock has been linked to unexpected
emissions of CFC-11, and its widespread illegal
production and use in China, and with observed
concentrations of CC14 emissions in the same region
where the increased emissions of CFC-11 were
observed. In eastern Europe, Georgia and Armenia
have seized illegal CC14 entering the European Union
(EU) from Russia. While these incidences were, in
theory, nondispersive, illegal dispersive uses of CC14
production have also been recorded.
EPA must consider all available scientific information
regarding observed global and regional emission trends
and concentrations of CC14 when considering these risks,
and not rely solely on industry reported data. We urge
EPA to evaluate and subsequently further regulate CC14
production and intermediate uses under TSCA to avoid
unreasonable risk to human health and the environment.
unreasonable risk is determined and regulatory
considerations are pursued.
The illegal production, trade, and use of carbon tetrachloride
in Asia and Europe are not conditions of use because these
activities are not known, intended, or reasonably foreseen to
occur in the United States. EPA assumes compliance with
existing laws and regulations, including those applicable to
the production, trade, and use of carbon tetrachloride, and
EPA has no evidence that these illegal materials are being
manufactured (including imported) here.
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation. Assessing
global emissions of carbon tetrachloride is outside the scope
of the risk evaluation
SACC,
26, 28,
30, 32,
28
SACC COMMENTS
Table E-l indicates a >300-pound release from one
facility in 2014 and a 14-pound spill from another facility
in 2015. How many spills per year occur in the population
of 49 facilities? If the average is 1 per year, then analysis
of releases should factor in these occurrences.
PUBLIC COMMENTS:
Spills and leaks generally are not included within the scope
of TSCA risk evaluations because in general they are not
considered to be circumstances under which a chemical
substance is intended, known or reasonably foreseen to be
manufactured, processed, distributed, used, or disposed of.
To the extent there may be potential exposure from spills and
leaks, EPA is also declining to evaluate environmental
Page 15 of 210
-------
EPA states that environmental spills are not within the
scope of the risk evaluation and were thus not evaluated.
This exclusion is contrary to TSCA's mandate that EPA
evaluate the conditions of use of a chemical substance.
"Conditions of use" under TSCA mean "the
circumstances, ... under which a chemical substance
is intended, known, or reasonably foreseen to be
manufactured, processed, distributed in commerce,
used, or disposed of." Spills are a "reasonably
foreseen" aspect of the circumstances under which
CC14 is manufactured, processed, distributed, used, or
disposed of.
Spills and leaks are undoubtedly reasonably
foreseeable, and indeed, when preparing
environmental impact statements (EISs) for federal
projects, the federal government regularly analyzes the
potential for spills and leaks because they are
reasonably foreseen aspects of such projects.
EPA cites two instances of known spills: a San Diego
spill that exceeded permit limits and a Dover
Chemical Corp. spill in 2014. These spills are known
conditions of use that result in actual exposures to
people and the environment.
In 2016 and 2017, a 200% increase in CC14
emissions above 2015 levels was reported, coming
from a facility owned by Dover Chemical
Corporation. This is the same facility where a large
accidental spill of 'chlorinated wax material'
containing CC14 byproduct occurred from a reactor
in 2014, leading to concerns about EPA's
voluntary reporting program.
Table 4-2 in the EPA draft reports that "San Diego
Sea World facility (CA0107336) was not included
exposure pathways addressed by other EPA-administered
statutes and associated regulatory programs.
First, EPA does not identify carbon tetrachloride spills or
leaks as "conditions of use." EPA does not consider carbon
tetrachloride spills or leaks to constitute circumstances under
which carbon tetrachloride is manufactured, processed,
distributed, used, or disposed of, within TSCA's definition
of "conditions of use." Congress specifically listed discrete,
routine chemical lifecycle stages within the statutory
definition of "conditions of use" and EPA does not believe it
is reasonable to interpret "circumstances" under which
carbon tetrachloride is manufactured, processed, distributed,
used, or disposed of to include uncommon and unconfined
spills or leaks for purposes of the statutory definition.
Further, EPA does not generally consider spills and leaks to
constitute "disposal" of a chemical for purposes of
identifying a condition of use in the conduct of a risk
evaluation.
In addition, even if spills or leaks of carbon tetrachloride
could be considered part of the listed lifecycle stages of
carbon tetrachloride, EPA has "determined" that spills and
leaks are not circumstances under which carbon tetrachloride
is intended, known or reasonably foreseen to be
manufactured, processed, distributed, used, or disposed of, as
provided by TSCA's definition of "conditions of use," and
EPA is therefore exercising its discretionary authority under
TSCA section 3(4) to exclude carbon tetrachloride spills and
leaks from the scope of the carbon tetrachloride risk
evaluation. The exercise of that authority is informed by
EPA's experience in developing scoping documents and risk
evaluations, and on various TSCA provisions indicating the
Page 16 of 210
-------
in the analysis since the reported level is above
permit discharge limits; noncompliance and spills
are not in the scope of this risk evaluation." Given
the relevance of the 2016 Lautenberg TSCA
amendment and Ninth Circuit finding that EPA
should no longer be ignoring spills, it might be
worthwhile to inquire whether those
understandings also apply to NPDES permit
discharge limits.
EPA does not evaluate occupational exposures from
spills and other accidental releases of CC14. The
SACC should comment on EPA's failure to consider
this condition of use.
intent for EPA to have some discretion on how best to
address the demands associated with implementation of the
full TSCA risk evaluation process. Specifically, since the
publication of the Risk Evaluation Rule, EPA has gained
experience by conducting ten risk evaluations and
designating forty chemical substances as low- and high-
priority substances. These processes have required EPA to
determine whether the case-specific facts and the reasonably
available information justify identifying a particular activity
as a "condition of use."
With the experience EPA has gained, it is better situated to
discern circumstances that are appropriately considered to be
outside the bounds of "circumstances... under which a
chemical substance is intended, known, or reasonably
foreseen to be manufactured, processed, distributed in
commerce, used, or disposed of' and to thereby
meaningfully limit circumstances subject to evaluation.
Because of the expansive and potentially boundless impacts
that could result from including spills and leaks as part of the
risk evaluation (e.g., due to the unpredictable and irregular
scenarios that would need to be accounted for, including
variability in volume, frequency, and geographic location of
spills and leaks; potential application across multiple
exposure routes and pathways affecting myriad ecological
and human receptors; and far-reaching analyses that would
be needed to support assessments that account for
uncertainties but are based on best available science), which
could make the conduct of the risk evaluation untenable
within the applicable deadlines, spills and leaks are
determined not to be circumstances under which carbon
tetrachloride is intended, known or reasonably foreseen to be
manufactured, processed, distributed, used, or disposed of, as
Page 17 of 210
-------
provided by TSCA's definition of "conditions of use."
Exercising the discretion to not identify spills and leaks of
carbon tetrachloride as a condition of use is consistent with
the discretion Congress provided in a variety of provisions to
manage the challenges presented in implementing TSCA risk
evaluation. See e.g., TSCA sections 3(4), 3(12), 6(b)(4)(D),
6(b)(4)(F). In particular, TSCA section 6(b)(4)(F)(iv)
instructs EPA to factor into TSCA risk evaluations "the
likely duration, intensity, frequency, and number of
exposures under the conditions of use....," suggesting that
activities for which duration, intensity, frequency, and
number of exposures cannot be accurately predicted or
calculated based on reasonably available information,
including spills and leaks, were not intended to be the focus
of TSCA risk evaluations. And, as noted in the preamble to
the Risk Evaluation Rule, EPA believes that Congress
intended there to be some reasonable limitation on TSCA
risk evaluations, expressly indicated by the direction in
TSCA section 2(c) to "carry out [TSCA] in a reasonable and
prudent manner."
For these reasons, EPA is exercising this discretion to not
consider spills and leaks of carbon tetrachloride to be
conditions of use.
Second, even if carbon tetrachloride spills or leaks could be
identified as exposures from a condition of use in some
cases, these are generally not forms of exposure that EPA
expects to consider in risk evaluation. TSCA section
6(b)(4)(D) requires EPA, in developing the scope of a risk
evaluation, to identify the hazards, exposures, conditions of
use, and potentially exposed or susceptible subpopulations
Page 18 of 210
-------
the Agency "expects to consider" in a risk evaluation. This
language suggests that EPA is not required to consider all
conditions of use, hazards, or exposure pathways in risk
evaluations. EPA has chosen to tailor the scope of the risk
evaluation to exclude spills and leaks in order to focus
analytical efforts on those exposures that present the greatest
potential for risk.
In the problem formulation documents for many of the first
10 chemicals undergoing risk evaluation, EPA applied the
same authority and rationale to certain exposure pathways,
explaining that "EPA is planning to exercise its discretion
under TSCA 6(b)(4)(D) to focus its analytical efforts on
exposures that are likely to present the greatest concern and
consequently merit a risk evaluation under TSCA...." This
approach is informed by the legislative history of the
amended TSCA, which supports the Agency's exercise of
discretion to focus the risk evaluation on areas that raise the
greatest potential for risk. See June 7, 2016 Cong. Rec.,
S3519-S3520.
In addition to TSCA section 6(b)(4)(D), the Agency also has
discretionary authority under the first sentence of TSCA
section 9(b)(1) to "coordinate actions taken under [TSCA]
with actions taken under other Federal laws administered in
whole or in part by the Administrator." TSCA section
9(b)(1) provides EPA authority to coordinate actions with
other EPA offices, including coordination on tailoring the
scope of TSCA risk evaluations to focus on areas of greatest
concern rather than exposure pathways addressed by other
EPA-administered statutes and regulatory programs, which
does not involve a risk determination or public interest
finding under TSCA section 9(b)(2). EPA has already
Page 19 of 210
-------
tailored the scope of this risk evaluation using such
discretionary authorities with respect to exposure pathways
covered under the jurisdiction of other EPA-administered
statutes and associated regulatory programs (see section
1.4.3).
Following coordination with EPA's Office of Land and
Emergency Management (OLEM), EPA has found that
exposures of carbon tetrachloride from spills and leaks fall
under the jurisdiction of RCRA. See 40 CFR 261.33(d)
(defining in part a hazardous waste as "any residue or
contaminated soil, water or other debris resulting from the
cleanup of a spill into or on any land or water of any
commercial chemical product or manufacturing chemical
intermediate having the generic name listed [40 CFR
261.33(e) or (f)], or any residue or contaminated soil, water
or other debris resulting from the cleanup of a spill, into or
on any land or water, of any off-specification chemical
product and manufacturing chemical intermediate which, if it
met specifications, would have the generic name listed in [40
CFR 261.33(e) or (f)]"); 40 CFR 261.33(f) (listing carbon
tetrachloride as hazardous waste no. U211). As a result, EPA
believes it is both reasonable and prudent to tailor the TSCA
risk evaluation for carbon tetrachloride by declining to
evaluate potential exposures from spills and leaks, rather
than attempt to evaluate and regulate potential exposures
from spills and leaks under TSCA.
EPA has evaluated disposal as a condition of use of carbon
tetrachloride with respect to occupational exposures from
disposal activities.
Regarding regulatory action, EPA must evaluate all the
Page 20 of 210
-------
conditions of use it expects to consider under TSCA in the
risk evaluation and propose risk management for any
condition of use which the Agency determines presents
unreasonable risk in the final risk evaluation. Risk
management activities are outside the scope of the risk
evaluation. As the commenter indicated, for any condition of
use determined to have unreasonable risk, EPA will consider
this and other comments during risk management.
EPA has clarified Sea World carbon tetrachloride discharges.
EPA has estimated the surface water concentration from Sea
World carbon tetrachloride releases using a proxy facility in
San Diego since Sea World permit data was not available in
E-FAST2014. The Risk Evaluation has been revised to
include these surface water carbon tetrachloride estimate
results. The Risk Evaluation has also been revised to include
a greater explanation of the E-FAST 2014 modeling
approach, model calculations, inputs and results.
26
PUBLIC COMMENTS: 26
EPA "expects insignificant or unmeasurable
concentrations of CC14 in the manufactured
chlorinated substances in the commercially available
products." The only corroborating sources that it
provided were qualified comments, with no data, from
representatives of the chemical industry asserting that
levels are low.
EPA has no reasonably available information indicating the
presence of carbon tetrachloride in commercially available
products in concentrations at significant or measurable
levels. In addition, the high volatility of carbon tetrachloride
and the extent of reaction and efficacy of the
separation/purification process for purifying final products
supports EPA's assumption that there are insignificant or
unmeasurable concentrations in these products.
While carbon tetrachloride is used in the manufacturing of
other chlorinated compounds that may be subsequently
added to commercially available products, EPA expects that
consumer use of such products would present only de
minimis exposure to, or otherwise insignificant risk from,
carbon tetrachloride given the high volatility of carbon
Page 21 of 210
-------
tetrachloride and the extent of reaction and efficacy of the
separation/ purification process for purifying final products.
No additional information was received by EPA following
the publication of the problem formulation that would update
the problem formulation conclusion that carbon tetrachloride
is expected to be present in consumer products at trace levels
resulting in de minimis exposures or otherwise insignificant
risks and therefore that consumer uses do not warrant
inclusion in the risk evaluation. For that reason, EPA
exercised its discretionary scoping authority under TSCA
sec. 6(b)(4)(D) to exclude this use from the scope of the risk
evaluation in order to focus the Agency's analytical efforts
on those exposures that are likely to present the greatest
concern. See section 1.4.2.2 of the Risk Evaluation; sections
2.2.2 and 2.2.2.1 of the Problem Formulation of the Risk
Evaluation for Carbon Tetrachloride (May 2018); 82 FR
33736, 33729 (July 20, 2017).
Legacy releases
23, 26,
30, 32,
42, 43
PUBLIC COMMENTS:
In the 2017 Scoping document, EPA stated, "In the case of
CC14, legacy uses and associated legacy disposals will be
excluded from the scope of the risk evaluation."
Disposal is a condition of use that must be considered
in a TSCA risk evaluation. A decision in the U.S.
Court of Appeals for the Ninth Circuit issued in late
2019 found that legacy activities should NOT be
excluded from the definition of conditions of use and
should be analyzed during risk evaluations.
EPA's SACC noted that EPA failed to consider
releases associated with disposal.
EPA has determined that general population exposures due to
drinking water contamination, ambient-water contamination,
and disposal pathways are regulated under other statutes and
are outside the scope of this risk evaluation. See section 1.4.3
of the risk evaluation.
EPA did not identify any "legacy uses" or "associated
disposals" of carbon tetrachloride, as those terms are
described in EPA's Risk Evaluation Rule, 82 FR 33726 (July
20, 2017). Therefore, no such uses or disposals were added to
the scope of the risk evaluation for carbon tetrachloride
following the issuance of the opinion in Safer Chemicals,
Healthy Families v. EPA, 943 F.3d 397 (9th Cir. 2019).
Page 22 of 210
-------
The Agency further stated that "As a result of this phase-
out and ban, it is highly unlikely that there are any ongoing
uses of CC14 that could be considered legacy uses, and no
such uses have been evaluated."
The SPARC Report estimated that up to 10 Gg/year of
global CC14 emissions is likely from legacy emissions
from contaminated soils and toxic waste treatment
facilities.
According to the latest TRI data, in 2018, >73,000
pounds of CC14 were released to land through
underground injection, disposal in hazardous waste
landfills, and "other land disposal." According to 2017
TRI data, total CC14 production-related waste totaled
36,838,580 pounds, of which 26,838,850 underwent
treatment. Landfills and other waste-treatment
operations reported environmental releases accounting
for 34% of total CC14 releases.
ATSDR indicates that CC14 was detected in soil at 103
National Priorities List (NPL) hazardous waste sites, in
sediment at 23 NPL hazardous waste sites, in
groundwater at 310 NPL hazardous waste sites, and in
surface water at 53 NPL sites.
EPA has detected CC14 inside homes above or around
Superfund sites where CC14 was found in the
groundwater, indicating a potential vapor intrusion
pathway.
The Agency is obligated to revise this draft risk
evaluation to incorporate the assessment of any
identified legacy uses and then re-issue updated
assessment for further peer review and public
comment. In particular, The National Tribal Toxics
Council (NTTC) strongly urges that environmental
release from waste management sites, including
The use of carbon tetrachloride in the past are not "legacy"
uses. As described in EPA's Risk Evaluation Rule (82 FR
33726 (July 20, 2017)), a legacy use is an ongoing use of a
chemical substance in a particular application where the
chemical substance is no longer being manufactured,
processed, or distributed in commerce for that application.
The example provided in the Rule is insulation, which may be
present in buildings after a chemical substance is no longer
being made for that use. In contrast, the uses of carbon
tetrachloride phased out as a result of the Montreal Protocol
and CAA Amendments of 1990 as well as the uses banned by
CPSC in 1970 (excluding unavoidable residues not exceeding
10 ppm atmospheric concentration) are no longer being
manufactured, processed, distributed in commerce, used, or
disposed of, to the best of EPA's knowledge, which is based
on EPA's research and outreach. Specifically, EPA received
no information from any commenters or otherwise indicating
that products in the United States had been stockpiled or that
use or disposal was ongoing. Therefore, carbon tetrachloride
uses that have been phased out or banned are not conditions
of use because they are not known, intended, or reasonably
foreseen.
Page 23 of 210
-------
transfer sites, construction and demolition sites,
materials recovery facilities, and landfills be evaluated
with consideration of unlined facilities with resulting
leachate subsurface flow, ponded water, direct surface
water and snowmelt runoff, ambient emissions from
uncovered disposal areas, and untreated waste burning
emissions.
Future releases
SACC,
28, 32,
43
SACC COMMENTS:
The SACC is concerned about the trends in increasing
CC14 releases. The SACC report details several points of
concern:
Data indicate that water releases are increasing in both
quantity and fraction of total releases.
The National Air Toxics Inventory (2015) shows an
increase in atmospheric CC14 over a 10-year period.
The number of facilities with water releases is
increasing.
The pattern of water releases is variable, but most
facilities show an increasing trend.
Accidental releases are not considered in TSCA
evaluations.
Smaller companies can manufacture/import/use
slightly less than 10,000 pounds of CC14 and dispose
all of this without reporting to TRI.
Removal mechanisms {i.e., biodegradation, photolysis
in the troposphere) are likely too slow to prevent
environmental concentrations of CC14 from increasing.
Overall, future releases, while uncertain, are expected to
increase from current levels unless regulatory action is
taken.
PUBLIC COMMENTS:
CC14 production in the U.S. is increasing due to
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes, including
pathways involving air and water releases, has been added to
Section 1.4.3 of the Risk Evaluation.
Water releases vary significantly between 2014 and 2018.
EPA has revised Appendix E to also include surface water
releases for 2019.
The National Air Toxic Inventory (2015) data presents
ambient air monitoring data for a number of chemicals
including carbon tetrachloride. The risk evaluation did not
consider the ambient air pathway due to its coverage under
the of the Clean Air Act.
Though some facilities show an increasing trend between the
2016 and 2018, there is considerable variability among the
number of facilities discharging carbon tetrachloride and the
amount of these releases over five years. Forty-nine facilities
discharged carbon tetrachloride in 2017 whereas 42 in 2018
and 39 in 2019. In addition, many facilities discharge one or
two years, then have zero releases other years. Therefore, two
years of upward trend is not necessarily a predictor of future
releases. It appears that total 2019 carbon tetrachloride
Page 24 of 210
-------
growing demand for CC14 as a feedstock in the
manufacture of HFO refrigerants. Unregulated
feedstock and intermediate uses of CC14 are
expected to increase by >50% in the near future.
U.S. production and import of CC14 has already
increased 10% from 129.1 million pounds in 2012
to 142.6 million pounds in 2015 according the
CDR database.
releases decreased from 2018 to levels similar to 2017
confirming the variable nature of releases. EPA therefore
averaged carbon tetrachloride releases over 5 years to capture
this variability.
EPA agrees with the SACC comment regarding limitations on
the population of facilities manufacturing/importing/using
and releasing carbon tetrachloride reflected in TRI. EPA
therefore relied on the DMR data in EPA's ECHO database to
capture releases to surface water.
Please see comment response under "Current Conditions of
Use and Emissions" for discussion of accidental releases."
Carbon tetrachloride shows minimal susceptibility to indirect
photolysis by hydroxyl radicals in the troposphere, where its
estimated tropospheric half-life exceeds 330 years.
Ultimately, carbon tetrachloride diffuses upward into the
stratosphere where it is photodegraded to form the
trichloromethyl radical and chlorine atoms (( ).
Carbon tetrachloride is efficiently degraded by direct
photolysis under stratospheric conditions and the DT50
(Dissipation Time for 50% of the compound to dissipate)
value is in the order of minutes. However, the troposphere to
the stratosphere migration of carbon tetrachloride is very long
and this migration time limits the dissipation. The rate of
photodegradation increases at altitudes >20 km and beyond.
Carbon tetrachloride dissolved in water does not
photodegrade or oxidize in any measurable amounts, with a
calculated hydrolysis half-life of 7,000 years based on
experimental data at a concentration of 1 ppm (OEC ).
Removal mechanisms from water could include volatilization
Page 25 of 210
-------
due to the Henry's Law constant and anaerobic degradation in
subsurface environment.
Domestic production and importation of carbon tetrachloride
is currently prohibited under regulations implementing the
Montreal Protocol (MP) and CAA Title VI, except when
transformed (used and entirely consumed, except for trace
quantities, in the manufacture of other chemicals for
commercial purposes), destroyed (including destruction after
use as a catalyst or stabilizer), or used for essential laboratory
and analytical uses. (40 CFR Part 82, 60 FR 24970, 24971
(May 10, 1995).) Carbon tetrachloride is used and entirely
consumed in feedstock and intermediate uses, and EPA does
not believe rising emissions from these uses are likely.
In any event, EPA determined that both the manufacture of
carbon tetrachloride and the processing of that chemical as a
reactant in the production of HFOs present an unreasonable
risk of injury to the health of workers and ONUs, and will
initiate TSCA section 6(a) risk management actions on these
conditions of use as required under TSCA section 6(c)(1).
Mass balance assessment of releases
SACC,
26, 43
SACC COMMENTS:
Recommendation: Include a mass balance assessment of
CC14 released to the environment.
Several Committee members recommended using a
table of the amounts of CC14 manufactured/imported in
the U.S., and the amounts used in processes/products,
released to the environment, or recycled. This approach
allows for better estimation of CC14 discharges to the
environment that are not captured in databases such as
EPA does not have reasonably available mass balance data to
conduct such an analysis for carbon tetrachloride. EPA's
analysis uses TRI and DMR to estimate the highest local per
site water releases of carbon tetrachloride. The NEI, which is
compiled every 3 years for the purpose of supporting residual
risk evaluations as required by Section 112 of the CAA. NEI
contains air emission estimates, which sites estimate using a
variety of methods, such as emission factors, mass balance,
stack monitoring. Purchase and disposal records are not
Page 26 of 210
-------
TRI.
reported to NEI. However, NEI could not be used to
reasonably estimate all media releases as it only includes air
releases from larger facilities and would not include releases
from many smaller shops that use carbon tetrachloride. EPA
acknowledged in the revised Risk Evaluation the global
sources of carbon tetrachloride in the atmosphere including
feedstock uses and non-feedstock emissions (see responses
below against #23, 30, 32, 43). Please see revised paragraph
in Section 1.2 (line 1187 - 1196). The revision includes
various carbon tetrachloride sources, their emissions and
citations of peer-reviewed articles.
SACC
SACC COMMENTS:
The draft risk evaluation notes that CC14 is used as a
feedstock in the production of
hydrochlorofluorocarbons, HFCs, HFOs, and
perchloroethylene (multiple locations), and that
production of HFC-245fa and HFC-365mfc accounted
for 71% and 23%, respectively, of total CC14
consumption in 2016 (p. 73). HFC-245fa and HFC-
365mfc are being phased out as part of the EPA's
Significant New Alternatives Policy (SNAP) program
and usage of CC14 for this condition of use is expected
to decrease significantly.
A mass-balance accounting of the condition of use
should be incorporated to better account for existing
feedstock usages.
Mass balance estimated discharges could be used along
with environmental fate models (e.g., fugacity level 3
model) to supplement limited monitoring data.
Given the relatively long aerobic half-life of CC14, if
continual discharge is occurring, exposure to aquatic
life would be ongoing and not require trophic transfer
or bioaccumulation.
See above response regarding the hydrochlorofluorocarbons,
HFCs, HFOs, and perchloroethylene, HFC-245fa and HFC-
365mfc.
Please refer response to mass-balance approach described
earlier. EPA's evaluation of the conditions of use accounted
for the existing feedstock usages and other published
information as cited in the risk evaluation document.
Level II fugacity model discussion included in Fate section of
revised RE (section 2.1).
Mass balance of releases of carbon tetrachloride, as reported
by various researchers, has been discussed in the revised risk
evaluation document. Appropriate citations are also included.
EPA addressed exposure to aquatic life: environmental
monitoring data were used to assess ambient water exposure
to aquatic organisms. Details of these exposure estimates as
compared to the aquatic toxicity benchmarks (concentrations
of concern) are available in Section 4.1.2.
Page 27 of 210
-------
Uncertainty associated with modeled estimates
SACC
SACC COMMENTS:
Recommendation: Discuss the uncertainty associated with
estimated exposures to aquatic organisms by the lack of
monitoring data.
Monitoring data from EPA (1977) provide CC14
concentrations in water downstream of five industrial
facilities. The 92nd and 75th percentiles are higher than
any 20-day estimate from the E-FAST and over 10
times higher than the 95th percentile of the 20-day
predictions reported in the draft risk evaluation. These
monitoring data should be included in this risk
evaluation as a justification for using higher percentile
E-FAST estimates, rather than the average.
EPA assessed facilities reporting monitoring data (DMRs) of
carbon tetrachloride discharges and presents data over five
years (2014 - 2018). These data are more representative of
the environmental concentrations than monitoring values that
predate many of the regulations placed on carbon
tetrachloride (e.g., CWA and CAA).
23, 43
PUBLIC COMMENTS:
The draft risk evaluation states that "the literature search
results for environmental exposures yielded 393 data
sources. Of these data sources, none were determined to be
relevant to the draft risk evaluation." EPA thus disregards
all of the environmental exposure data in its possession,
and instead calculates environmental risks based solely on
modeling, as opposed to actual surface water, soil, and air
concentrations. If EPA truly has no usable environmental
exposure data, then it has the authority under TSCA to
compel companies that manufacture, import, or use CC14
to produce or generate such data. EPA's exclusive reliance
on modeling, with no data to validate the results, does not
provide a sufficient basis for the evaluation of CC14's
environmental risk.
EPA had sufficient information to complete the carbon
tetrachloride risk evaluation using a weight of scientific
evidence approach. EPA selected the first 10 chemicals for
risk evaluation based in part on its assessment that these
chemicals could be assessed without the need for regulatory
information collection or development.
The TSCA risk evaluation process does not require EPA to
compel the generation of new data. In fact, in conducting a
risk evaluation, EPA must "take into consideration . . . hazard
and exposure information, under the conditions of use, that is
reasonably available" (TSCA ง 26(k)). When preparing this
risk evaluation, EPA obtained and considered reasonably
available information, defined in 40 CFR 720.33 as
information that EPA possesses or can reasonably generate,
obtain, and synthesize for use in risk evaluations, considering
the deadlines for completing the evaluation. EPA has
explained in its regulations that "EPA will use [information
gathering] authorities on a case-by-case basis during the
Page 28 of 210
-------
performance of a risk evaluation to obtain information as
needed to ensure that EPA has adequate, reasonably available
information to perform the evaluation" (40 CFR
702.41(b)(2)). As explained at 40 CFR 702.41(a)(7), "To the
extent a determination as to the level of risk presented by a
condition of use can be made using models or screening
methodologies, EPA may determine that no further
information or analysis is needed to complete its risk
evaluation of the condition(s) of use." In this case, consistent
with EPA's approach of conducting fit-for-purpose risk
evaluations, described in greater detail in 82 FR 33726 at
33739-40 (July 20, 2017), EPA determined that a technically
sound risk determination could be made, consistent with the
best available science, without the generation of additional
data (which, in any event, likely would not have been
possible to produce and incorporate in the risk evaluation
within the timeframe specified in TSCA section 6(b)(4)(G)).
EPA used the peer-reviewed E-FAST model to estimate
carbon tetrachloride surface water concentrations using
facility monitoring and loadings data as reported to EPA in
Discharge Monitoring Reports. EPA has high confidence in
the model and the estimates of surface water concentrations
given location-specific flow data and carbon tetrachloride
discharge data.
EPA will continue to improve on its method and data
collection for the next round of chemicals to be assessed
under TSCA.
45
PUBLIC COMMENTS:
EPA applies a number of conservatisms to its
environmental exposures estimates. While these
approaches may suffice for screening level assessments,
they do not represent real world exposures. For example,
EPA used PDM within E-FAST 2014 to estimate surface
A refined analysis for the five sites that indicated risk to
aquatic organisms has been added to section 4.1.2 and in the
appendix (Table F-2). Briefly, EPA calculated surface water
concentrations using E-FAST and associated, site-specific
RQs to determine whether risk was or was not indicated at the
five facilities that indicated risk during time periods relevant
Page 29 of 210
-------
water concentrations. For situations where environmental
exposures determined by E-FAST lead to a RQ >1.0,
additional investigation about the site should be pursued.
Worst-case assumptions in the model, such as no dilution
during 7Q10 receiving stream flows or for the "still water
body" scenario, may be unlikely. EPA should conduct a
higher tier analysis of any facility for which it has concerns
about exceedances based on the current approach.
to amphibian development. Risk was not indicated during
time periods relevant to amphibian development at Eco
Services Operations Facility (RQs < 1 for the three years
where monitoring information was available). At the other
four facilities for which a refined analysis was conducted, risk
was indicated (RQs >1) during the time periods relevant to
amphibian development for at least 2 separate reporting
periods.
Justification of exclusion of exposures regulated under other environmental statutes
SACC
SACC COMMENTS:
EPA should provide additional documentation (i.e.,
links to the specific environmental programs and
statutes) that shows how other regulations will address
terrestrial risk.
Releases to non-aqueous phases should be considered.
At a minimum, a rationale should be added for
exclusion of non-aqueous media.
Recommendation: Improve the
justifications/documentation for excluding non-
aqueous media from consideration.
In the problem formulation, EPA indicated that CC14
was identified in biosolids. This indicates that it will
sorb to environmental solids and suggests that if CC14
is discharged into streams, it is likely to be found in
sediments. Therefore, stating that CC14 discharged into
streams rapidly distributes into air cannot be supported
without monitoring data or a dynamic stream
contaminant model that can predict distribution to
water, air, and sediment.
Recommendation: Be consistent and better define
how physical-chemical properties and terminology
are used to justify the exclusion of various
Section 1.4.3 in the final risk evaluation contains information
on EPA administered regulatory programs and statutes with
jurisdiction over exposure pathways to terrestrial ecological
receptors from carbon tetrachloride emissions.
A Level III fugacity model was assessed to investigate
sorption to sediment. See section 2.1.2 Fate and Transport for
narrative indicating the following:
"Although volatilization is expected to be rapid, a Level III
Fugacity model predicted that when carbon tetrachloride is
continuously released to water, 80% of the mass will partition
to water, 19% to air, <1% to soil and < 1% to sediment."
Page 30 of 210
-------
environmental fate processes and distributions.
26, 30,
32,41,
42, 43
PUBLIC COMMENTS:
The CC14 draft lacks any assessment of risks to the general
population or to the environment from CC14's presence in
air, water, and soil. EPA has excluded all general
population risks from exposures due to releases of CC14 to
land, air, and water, based on the assumption that other
statutes adequately address these exposures. Yet, no
analyses or data have been presented to show that these
other statutes are protective of the general population.
Established scientific principles for exposure assessment
require that known exposures (including from air, water,
land, and all other pathways) be included in the
assessment, or exposure will not be accurately quantified,
and risk will be underestimated. The incorrect
determination that emissions are not in scope is deeply
concerning. Under TSCA, EPA must conduct a
comprehensive assessment of exposures, and by failing to
consider this pathway, EPA will miss potentially exposed
or susceptible subpopulations (PESS) within the general
population. The SACC has faulted EPA risk evaluations
(1,4-dioxane, methylene chloride) for excluding
environmental pathways of exposure.
To justify this exclusion, EPA claims that it need not
address "exposure pathways under programs of other
environmental statutes" because they "adequately assess
and effectively manage exposures" using "long-standing
regulatory and analytical processes."
Risk evaluations under section 6(b)(4)(A) must
determine "whether a chemical substance presents
As part of the Problem Formulation for carbon tetrachloride
( 2018b), EPA found that exposures to the general
population may occur from the conditions of use due to
releases to air, water or land. The exposures to the general
population via surface water, drinking water, ambient air and
sediment pathways fall under the jurisdiction of other
environmental statutes administered by EPA, i.e., CAA,
SDWA, CWA, and RCRA. As explained in more detail in
section 1.4.3 of the final risk evaluation, EPA believes it is
both reasonable and prudent to tailor TSCA risk evaluations
when other EPA offices have expertise and experience to
address specific environmental media, rather than attempt to
evaluate and regulate potential exposures and risks from those
media under TSCA. EPA believes that coordinated action on
exposure pathways and risks addressed by other EPA-
administered statutes and regulatory programs is consistent
with the statutory text and legislative history, particularly as
they pertain to TSCA's function as a "gap-filling" statute, and
also furthers EPA aims to efficiently use Agency resources,
avoid duplicating efforts taken pursuant to other Agency
programs, and meet the statutory deadlines for completing
risk evaluations. EPA has therefore tailored the scope of the
risk evaluations for carbon tetrachloride using authorities in
TSCA sections 6(b) and 9(b)(1). See section 1.4.3 of the Risk
Evaluation.
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation.
Page 31 of 210
-------
an unreasonable risk of injury to health or the
environment." This requirement cannot be met
without examining all sources of exposure that
contribute to health and environmental risk.
Section 6(b)(4)(A) provides that a risk evaluation
must determine the substance's risks under "the
conditions of use," defined as "the circumstances . .
. under which a chemical substance is intended,
known or reasonably foreseen to be manufactured,
processed, distributed in commerce, used or
disposed of." These "circumstances" clearly
include environmental releases that result in
pathways of human exposure, whether or not they
might be controlled under other environmental
laws.
If Congress had intended a blanket exemption for
environmental releases from risk evaluations under
section 6(b), it would have said so explicitly. But
not only is there no such exemption in the law, its
legislative history and structure demonstrate that
Congress intended TSCA to provide a
comprehensive framework for identifying and
managing chemical risks, including those that
derive from environmental exposure pathways
subject to other environmental laws.
Additional points:
EPA's position that other environmental laws
should displace TSCA risk evaluations for all
chemicals arbitrarily assumes that these laws
provide equivalent protection of public health and
the environment and that there is no added benefit
in addressing environmental pathways of exposure
Page 32 of 210
-------
under TSCA. But these other laws vary greatly in
the degree of protection that they afford against
chemical risks and the extent of their application to
unsafe chemicals. Many other laws do not regulate
the entire universe of polluting sources. Other laws
may impose controls based not on risk but on other
considerations like cost or available technology.
The CAA, SDWA, CWA, and RCRA are specific
to individual media; they do not authorize an
examination of exposure and risk across media.
Other EPA authorities may lack the bandwidth to
tackle serious chemical risks that do not represent
immediate priorities if they are not mandated to do
so. These limitations are why Congress gave EPA
comprehensive authority over chemical risks under
TSCA in 1976 and strengthened that authority in
2016.
EPA relies on the CAA to dismiss the need to
assess exposures to CC14 from air emissions. The
standards under the CAA for HAPs are set for
individual source categories, meaning that the
exposures to CC14 from all sources in combination
are never considered.
In a recent proposed rule for a source category,
EPA stated: "Although we are interested in placing
source category and facility-wide HAP risk in the
context of total HAP risk from all sources
combined in the vicinity of each source, we are
concerned about the uncertainties of doing so" (84
Fed. Reg. 58,268, 58,273). Thus, it is clear that
EPA does not look at overall risk from a chemical
substance in those assessments.
Page 33 of 210
-------
Under the CAA, the first step is setting the
Maximum Achievable Control Technology
(MACT) standard, which does not require a risk
evaluation. The mandate for the standard is to
achieve the reduction in emissions possible,
considering technology, costs, and energy
requirements.
After the promulgation of the MACT standard,
under the legal requirements for the CAA, it would
take EPA 8 years to evaluate residual risk to the
population and, if necessary, create a stricter
standard; during the 8 years, people will continue
to be exposed to harmful chemical levels.
Many of these other statutes require EPA or other
agencies to consider factors such as cost and
feasibility when setting standards - factors that
TSCA explicitly forbids EPA from taking into
account when assessing risks (Section 6(b)(4)(A)):
"The Administrator shall .. .determine whether a
chemical substance presents an unreasonable risk
of injury ..., without consideration of costs or other
nonrisk factors."
Extensive monitoring required by EPA showed
exceedances of the EPA maximum contaminant level
(MCL) and widespread contamination at levels that pose a
cancer risk of >1 in one million and exceed the California
public health goal (PHG).
In 1987, EPA set a maximum contaminant level
goal (MCLG) of zero and an MCL of 5 (J,g/L. The
MCL was based on the LOD for CC14 in drinking
water at the time. Subsequently developed
Page 34 of 210
-------
analytical methods can detect CC14 at lower
concentrations.
Some states recognize that the MCL should be
lowered to assure health protection. California's
Office of Environmental Health Hazard
Assessment (OEHHA) set a PHG of 0.1 [j,g/L for
CC14 in drinking water in 2000.
The 2010 Integrated Risk Information System
(IRIS) assessment for CC14 determines that
drinking water exposures over a lifetime to
0.5 [j,g/L - a tenth of the MCL - pose a cancer risk
of 1 in a million.
The CC14 problem formulation notes that: "Internal
analysis for SYR3 (2006-2011) data... show that
118 of 55,735 systems (0.212%) have mean
[drinking water] concentrations greater than the
Minimum Reporting Level of 0.5 (J,g/L. SYR 2
(1998-2005) data showed 650 systems or 1.289%
of 50,446 systems had detects greater than 0.5
(J,g/L... Only 57 (0.113%) systems had detects of
CC14 greater than the MCL of 5 (J,g/L."
In monitoring of public water systems, the USGS
detected CC14 in source water and finished water at
levels above the PHG.
The 2019 Update of the Environmental Working
Group (EWG) Tap Water Database reports that
CC14 was detected in drinking water of 256 water
suppliers in 34 states, serving a total population of
3.1 million people, and that 167 drinking water
utilities serving 1.1 million people had CC14
concentrations above the California PHG.
The ATSDR notes that some studies show drinking
water concentrations well above the MCL (i.e., at
Page 35 of 210
-------
16 and 29 (J,g/L) and that "based on the STORET
database, CC14 was detectable in 12% of 8,858
ambient water samples," with a median
concentration of 0.1 (J,g/L.
The EPA drinking water program has not
conducted an assessment of cancer and noncancer
risk from CC14-contaminated drinking water based
on current science and has no plans to do so despite
extensive evidence that CC14 levels in drinking
water exceed EPA's threshold for acceptable
cancer risk. EPA's exclusion of drinking water
from its TSCA evaluation creates a serious and
unjustified gap in health protection of exactly the
type Congress intended for TSCA to address.
45
PUBLIC COMMENTS:
The ACC agrees that existing EPA regulatory programs
addressing environmental media pathways (air, water,
land) can and do adequately assess and manage exposures
to these media. EPA has deviated from this assessment by
choosing to address the ambient water pathway despite the
existence of CWA regulations. EPA's consideration of the
ambient water pathway did not uncover unreasonable risk,
nor did it even produce recommendations to other program
offices to pursue additional regulation under the statutes
for which they have jurisdiction. Is OPPT's attempt to
address environmental pathways that are already subject to
significant EPA regulation under other environmental
statutes in these draft TSCA risk evaluations a good use of
EPA's resources?
To address TSCA Section 9 and transparency concerns,
Clarifying language on exposure pathways and risks under
the jurisdiction of other EPA-administered statutes have been
added to section 1.4.3 of the final risk evaluation document.
General population exposures from the ambient water
pathway are excluded from the scope of the risk evaluation
based on coverage under CWA section 304(a) and
implementing regulations.
OPPT worked closely with other EPA program offices during
the course of the risk evaluation process and will continue to
engage intra-agency coordination for future TSCA risk
evaluations. This is consistent with TSCA section 9(b)(1),
which directs EPA to "coordinate actions taken under
[TSCA] with actions taken under other Federal laws
administered in whole or in part by the Administrator."
Page 36 of 210
-------
ACC recommends that EPA should seek: (a) for OPPT to
better understand the regulatory requirements and
processes of the various environmental statutes under
EPA's purview; (b) for OPPT to reach agreement with the
other program offices on what criteria should drive the use
of TSCA risk evaluations to address air, water, and other
waste pathways under the conditions of use of a TSCA
high priority chemical; (c) for other program offices to
understand the potential value of TSCA risk evaluations to
these other EPA programs; and (d) to establish better
approaches for coordinating what each program office
(including EPA OPPT) can provide the others to improve
environmental protection under their respective statutory
authorities more efficiently and without duplication.
TSCA was never intended to replace regulation by other
EPA environmental programs, each of which has different
requirements and standards and approaches for regulatory
decision-making.
The purpose of risk evaluation under TSCA is to determine
whether a chemical substance presents an unreasonable risk
to health or the environment, under TSCA conditions of use.
Clarifying language on exposure pathways and risks under
the jurisdiction of other EPA-administered statutes have been
added to section 1.4.3 of the final risk evaluation document.
Impacts of CCI4 on climate change
23, 30,
32, 43
PUBLIC COMMENTS:
CC14 has a significant global warming potential (GWP),
which makes it 1,730 times more potent than carbon
dioxide (C02). Assuming U.S. emissions of CC14 are
nearly 9 million pounds per year as estimated by SPARC,
C02 equivalent emissions would be 6.9 million metric
tons. This amount is higher than the C02 emissions of
most coal-fired power plants and equals the annual C02
emissions from over 1.5 million cars. The well-known
consequences of global warming include far-reaching
impacts on human health and the environment that should
be addressed in a comprehensive risk evaluation. Yet there
is no mention of CC14's GWP in the draft evaluation, let
Clarified the following after Table 1-2 in the final risk
evaluation document:
"Carbon tetrachloride had several uses in the past, primarily
as a feedstock for the production of chlorofluorocarbons.
Current uses are now confined by the Montreal Protocol to be
in contained processes. Sherry et al. (2018) reported global
industrial production of carbon tetrachloride in 2014 was
consumed in: (i) incineration (29 Gg); (ii) as a
perchloroethylene feedstock (64 Gg); (iii) as
hydrofluorocarbon feedstock (58 Gg); in (iv) methyl chloride
production (26Gg); (v) in divinyl acid chloride production (23
Gg); and (vi) for use as process agents and laboratory
Page 37 of 210
-------
alone any analysis of the significance of its emissions in
contributing to climate change.
purposes (3 Gg). Sherry et al. (2018) estimated 13 Gg year"1
of global emissions from unreported non-feedstock emissions
from chloromethane and perchloroethylene plants as the key
carbon tetrachloride source. Additionally, 2 Gg year"1 are
estimated as fugitive emissions from the usage of carbon
tetrachloride as feedstock and possibly up to 10 Gg year"1
from legacy emissions and chlor-alkali plants."
Impacts of CCI4 on stratospheric ozone
SACC,
23, 28,
30, 32,
43
SACC COMMENTS:
The impact of CC14 as an ODS in the stratosphere should
be further discussed.
The draft risk evaluation indicates that CC14 is released
into the atmosphere and rapidly degrades in the
stratosphere (p. 137, line 4437).
However, EPA verbally presented that CC14 is very
stable in the troposphere and that the movement to the
stratosphere is an extremely slow process and is
unlikely to significantly reduce tropospheric
concentrations over the short term. This was supported
by information in the problem formulation indicating
an extremely long half-life in the troposphere.
This contradicts statements in the draft risk evaluation
stating that CC14 diffusion into the stratosphere is an
important removal mechanism.
Recommendation: Add more discussion on the impact of
more atmospheric input and long tropospheric half-lives on
ozone depletion.
One Committee member cited the SPARC report on CC14
as a source for more information on impacts.
PUBLIC COMMENTS:
CC14 is a significant contributor to ozone depletion,
accounting for about 12% of the globally averaged
chlorine and bromine in the stratosphere, compared to
Assessing ozone depletion is out of scope for this Risk
Evaluation. EPA did not include the emission pathways to
ambient air from commercial and industrial stationary
sources, because stationary source releases of carbon
tetrachloride to ambient air are under the jurisdiction of
Section 112 of the CAA. Resulting exposure were out of
scope as described in section 1.4.3 of the final risk evaluation
for carbon tetrachloride.
Carbon tetrachloride is regulated under the CAA as a
Hazardous Air Pollutant (HAP) and an ozone depleting
substance (CAA Sections 112 and 604). HAP provisions
already account for ozone depletion and climate change in
accordance with Montreal Protocol.
See additional language in Section 2.1.2 of the final risk
evaluation:
"Carbon tetrachloride shows minimal susceptibility to
indirect photolysis by hydroxyl radicals in the troposphere,
where its estimated tropospheric half-life exceeds 330 years.
Ultimately, carbon tetrachloride diffuses upward into the
stratosphere where it is photodegraded to form the
trichloromethyl radical and chlorine atoms ( ).
Carbon tetrachloride is efficiently degraded by direct
Page 38 of 210
-------
14% for CFC-12 in 2012. CC14 has an ozone depletion
potential (ODP) of 0.82, which makes it nearly as
potent as several of the CFCs.
CC14 is a Class I ODS under the 1987 Montreal
Protocol (MP) and is subject to the stratospheric ozone
protection provisions of Title VI of the CAA.
Feedstock and process agent uses are considered
'nondispersive' by the MP and CAA. CC14 continues
to be legally produced and used under the CAA for
'non-dispersive' uses as feedstocks, despite evidence
that chemical manufacturing and feedstock use is
dispersive.
In spite of the MP controls, "there are large ongoing
emissions of [CC14] into the atmosphere." According
to SPARC, "atmospheric levels of [CC14] are currently
declining at a rate slightly faster than 1% per year," 2-3
times slower than would be expected in the absence of
significant emissions.
Global emissions of CC14 are substantial when
compared with other ODSs, accounting for 11-17% of
all ozone depletion-weighted emissions.
Knvironinonlsil ll;i/:ii'(l ;iihI Risk (liurncleriziilion
Charge Question 2.1: Please comment on whether the information presented supports the hazard and risk findings in the draft
environmental hazard section (Section 3.1) and draft risk characterization section (Section 4.1).
#
Sn 111111:1 r\ of Comments lor Specific Issues Uchilcri lo
Chsirge Question 2
KIW/OPPT Response
photolysis under stratospheric conditions and the DT50
(Dissipation Time for 50% of the compound to dissipate)
value is in the order of minutes. However, the troposphere to
the stratosphere migration of carbon tetrachloride is very long
and this migration time limits the dissipation. The rate of
photodegradation increases at altitudes >20 km and beyond."
Page 39 of 210
-------
Selection of species for inclusion in risk evaluation
SACC,
30, 43,
45
SACC COMMENTS:
Recommendation: Include an evaluation of risk for
terrestrial receptors or provide convincing logic why risk
to terrestrial receptors would be negligible.
Terrestrial receptors should have been assessed given
the large amount of waste disposed in this manner.
Terrestrial organisms are briefly mentioned but were
excluded from evaluation since they were considered
to be covered under other EPA programs. Some
Committee members expressed concern that the
rationale for ignoring pathways for terrestrial
organisms was cursory. Is it the existence of the other
environmental regulatory statutes or the stated
adequacy of those programs in addressing these
pathways that justify the exclusion? If it were
demonstrated that the other environmental statutes
administered by EPA do not adequately assess or
effectively manage these specific exposures, would
terrestrial species exposure pathways then be covered
under TSCA?
EPA could provide more information on how other
EPA statutes are relevant to those hazards, perhaps in a
flowchart.
The Agency should cite the specific documents that
have examined terrestrial exposures and associated
risks from CC14.
PUBLIC COMMENTS:
EPA did not consider any environmental release data or
any data on toxicity to terrestrial or sediment-dwelling
species. Other governments have classified CC14 as
"ecotoxic to terrestrial vertebrates" and the draft risk
evaluation acknowledges that "terrestrial species
As explained in section 2.5.3.2 of the problem formulation
(U.S. EPA. 2.018b), exposure to terrestrial organisms was
removed from the scope of the evaluation. However, in the
final risk evaluation, EPA qualitatively evaluated the soil and
land-applied biosolid pathways leading to exposure to
terrestrial organisms. Exposures to terrestrial organisms from
air were considered out of scope due to its coverage under the
jurisdiction of the Clean Air Act. Section 1.4.3 in the final
risk evaluation contains information on EPA administered
regulatory programs and statutes with jurisdiction over
exposure pathways to terrestrial ecological receptors from
carbon tetrachloride emissions.
With respect to sediment-dwelling aquatic species, carbon
tetrachloride is not expected to partition to or be retained in
sediment and is expected to remain in aqueous phase due to
its water solubility (793 mg/L) and low partitioning to
organic matter (log Koc= 0.79 - 1.93 in aquifer sediments
and 1.67 in marine and estuary sediments) (see section 2.1).
According to the reasonably available information, carbon
tetrachloride is likely to be in pore water and not adsorbed to
the sediment organic matter. Thus, qualitatively, sediment-
bound carbon tetrachloride exposure concentrations are
expected to be low.
EPA also added a quantitative assessment of exposure to
sediment-dwelling aquatic organisms, which is available in
Table 4-2 and Section 4.1.3 in the final risk evaluation.
Briefly, the COCs were calculated based on toxicity
information available in one study conducted on Chironomus
tentans (Lee et aL. 2006), and were based on bodv drv
weight, and an AF of 5 for the acute COC, and an ACR of 10
Page 40 of 210
-------
populations living near industrial/ commercial facilities
... may be exposed via multiple routes such as ingestion
of surface waters and inhalation of outdoor air." We
believe this exclusion is unjustified under TSCA, which
requires a comprehensive assessment of risks to the
environment, and recommend that EPA revise the
evaluation to address hazards and exposures to
terrestrial organisms and make a risk determination for
these organisms.
For sediment-dwelling species, EPA writes that "CC14
is not expected to partition to or be retained in sediment
and is expected to remain in aqueous phase due to its
water solubility and low partitioning to organic matter."
However, CC14 has been detected in sediment
throughout the United States, including at more than 20
federal Superfund sites. Because EPA does not measure
or estimate the levels of CC14 in that sediment or
compare it to concentrations of concern for sediment-
dwelling organisms, it cannot determine whether the
risks to those organisms are unreasonable.
EPA OPPT decided to update its analysis of releases of
CC14 to surface waters and resulting concentrations of
CC14, based on "additional data" about ecological
hazards that came to the Agency's attention after
completing the CC14 problem formulation. EPA has not
explained what "additional data" drove this decision
and what role the EPA Office of Water played in
OPPT's reaching this decision. The mere absence of an
EPA-developed water quality criteria on aquatic life (or
human health) should not in and of itself trigger OPPT
to include ambient water pathways in TSCA risk
for chronic COC. Because only one study was available for
sediment dwelling organisms, EPA also generated acute and
chronic COCs using aquatic invertebrates (e.g., Gammarus
pseudolimnaeus and Daphnia magna) as a surrogate species
to provide an additional line of evidence to estimate toxicity
to sediment-dwelling organisms in the final risk evaluation.
Based on the COCs generated both from (Lee et at.. 2.006)
and from the use of aquatic invertebrates as a surrogate, risk
to sediment dwelling organisms was not indicated for acute
(RQs < 1) or chronic exposures to carbon tetrachloride (RQ <
1 or RQ > 1 and less than 20 days of exceedance).
As a result of a screening-level comparison of the reasonably
available environmental aquatic hazard data with aquatic
exposure concentrations, it was determined that no further
hazard analyses were necessary (see section 2.5.3.1. of the
problem formulation document) (U.S. EPA, 2018b). Upon
further evaluation of the reasonably available hazard data of
carbon tetrachloride after the problem formulation phase,
EPA decreased the environmental hazard chronic COC from
7 |ig/L to 3 |ig/L. Consequently, EPA assessed the risk of
aquatic organisms in the risk evaluation. The derived acute
COC (90 |ig/L) and chronic COC (3 |ig/L) are based on
environmental toxicity endpoint values (e.g., ECso) from
Brack and Rottler (Brack and Rottler, 1994) and (Black et al.,
1982; Birge et al., 1980), respectively. The data were based
on high quality studies and represent the lowest bound of
carbon tetrachloride data available in the public domain.
Further details about the environmental hazards of carbon
tetrachloride are available in Table 3-1.
Page 41 of 210
-------
evaluations.
Chronic ecological COC
SACC,
45
SACC COMMENTS:
It was unclear why the COC was changed from 7 to 3
[j,g/L. The calculations based on amphibians and algae are
included in Table G.6 of the draft risk evaluation.
Recommendation: Justify the change in COC for
environmental risk from 7 to 3 (J,g/L.
PUBLIC COMMENTS:
EPA decreased the environmental hazard chronic COC
from 7 to 3 (J,g/L. For clarity, EPA should reproduce the
process for developing the COCs from the problem
formulation document and discuss why the value was
changed from 7 to 3 (J,g/L.
Further, a summary table of the results used to calculate
the COCs should be provided in this section, rather than
leaving the reader to recreate it from Appendix G.
The chronic COC was initially determined to be 7 [j,g/L as a
result of a screening-level comparison of the reasonably
available environmental hazard data (see section 2.5.3.1. of
the problem formulation document) ( ). Upon
further evaluation of the reasonably available hazard data of
carbon tetrachloride after the problem formulation phase,
EPA decreased the environmental hazard chronic COC from
7 |ig/L to 3 |ig/L. Consequently, EPA assessed the risk of
aquatic organisms in this draft risk evaluation. The derived
acute COC (90 |ig/L) and chronic COC (3 |ig/L) are based on
environmental toxicity endpoint values (e.g., ECso) from
Brack and Rottler (Brack and Rottler. 1994) and (Black et aL,
1982; Birge et aL 1980), respectively. The data represent the
lowest bound of all carbon tetrachloride data available in the
public domain and provide conservative hazard values.
Further details about the environmental hazards of carbon
tetrachloride are available in Table 3-1.
EPA used hazard data from the most sensitive species to
estimate lethality and overall effects to aquatic organisms.
The chronic COC, 0.003 mg/L, was based on the LCio for the
European common frog (Rana temporaria). The COC for
algae, 0.007 mg/L, was calculated separately, and was based
on the ECio for green algae (Chlamydomonas reinhardtii).
EPA used an AF of 10 for the chronic and algal COC
calculations to account for species that may be more sensitive
but were not represented in the available data.
EPA used the lowest LCio (0.03 mg/L, chosen from LCios
from four amphibian species ranging from 0.025 to 0.436
Page 42 of 210
-------
mg/L) to calculate the chronic COC because, as both Birge, et
al. ( )) and Black et al., (1982) noted, it delineates the
concentration at which substantial reproductive impairment
could occur, resulting in population-level effects.
EPA incorporated this suggestion. The summary table of
aquatic toxicity studies and hazard ranges used to determine
the COCs has been moved into the environmental hazard
section in the main document (Table 3-1).
SACC
SACC COMMENTS:
The Committee suggested the use of LD10 for chronic
exposures, even when using an AF of 10 would be
insufficient to be protective of amphibian populations.
The assumption that the larval life stage is particularly
sensitive may not be accurate (Appendix G, p. 272, line
7023). Studies have shown that metamorphosis can be
a more sensitive life stage for some compounds (e.g.,
thyroid disrupting substances; Johnson et al., 2017).
Recommendation: Consider using benchmark dose (BMD)
methods to determine a POD for amphibians and either
defend the application of the AF of 10, or use an AF of 100
from the LC10, which is considered in many publications
to be protective of lethal effects in aquatic organisms
(Kienzler et al., 2017). Alternatively, EPA should consider
using an AF of 100 instead of 10, which would incorporate
additional uncertainty into risk characterization for
developmental effects.
Development and metamorphosis are both sensitive endpoints
for amphibians, and EPA acknowledges uncertainty due to
lack of data encompassing amphibian metamorphosis.
However, metamorphosis is not anticipated to be a more
sensitive life stage than early amphibian development. While
amphibians can be particularly vulnerable to thyroid
endocrine disruption at low concentrations during
metamorphosis, EPA does not have evidence that carbon
tetrachloride is a thyroid endocrine disruptor. EPA is also
considering (Johnson, et al.. 2017) in an ongoing analysis
examining amphibian variation in sensitivity to inform the
use of amphibian data in future risk assessment (described
below).
EPA examined whether BMD modeling could be applied to
the toxicitv data from Birge, et al. (1980) and Black et al.,
(1982) used to derive the acute and chronic concentrations of
concern using the EPA peer reviewed BMDS
(https://www.epa.gov/bmds/about-benchmark-dose-software-
bmds). This methodologv has been added to the Appendix. In
brief, because the BMDS requires a measure of error
(STD/STE) for model calculation, EPA was not able to apply
these methods with the data provided by the Birge, et al.
Page 43 of 210
-------
(1980) and Black et al., (1982) papers. However, EPA has
high confidence in the toxicity values provided by both
papers because the study authors applied an appropriate
modeling technique (log-probit analysis) to generate LC10
and LC50 POD estimates for fish and amphibian species.
EPA used OPPT methodology as cited in the risk evaluation
( 2013. 2012b) and applied an AF of 10 for chronic
data. EPA is considering the Keinzler et al. ( ) study,
referred to by the SACC, in its assessment. EPA has
developed a data driven approach to deriving AFs for a case
study with amphibian data relevant to the carbon tetrachloride
risk evaluation and results are summarized below:
Because amphibian species are typically under-represented in
chemical risk assessments relative to other taxonomic groups,
little is known about the amount of variation observed across
amphibian species. EPA tested whether an AF of 10, typically
applied to the lowest chronic toxicity value for fish, daphnia,
and algae to account for species-level variation in sensitivity,
is protective of amphibians. Single chemical toxicity effects
for growth, development, or mortality specific to amphibian
larva were obtained from EPA's ECOTOX knowledgebase.
Chemicals were characterized as having specific-acting or
narcotic MO As as predicted from chemotype (ToxPrints) and
bioassay activity (ToxCast and Tox21 hits) features, and
species sensitivity distributions (SSDs) were used to
characterize variation in sensitivity. Based on the available
data, which included 1071 ECso and LCso endpoints spanning
202 chemicals and 41 amphibian species, an AF of 10 was
protective of 95% of the amphibian species, on average, when
toxicity data for at least 5 and 10 species were available for
Page 44 of 210
-------
narcotic and specific-acting chemicals, respectively. For
carbon tetrachloride, this suggests that the AF of 10 that EPA
applied to the lowest 9-day LCso chosen from 7 amphibian
species to calculate the acute COC, could be protective of
amphibians, as they are currently represented within the
ECOTOX database.
For chronic exposures, the paucity of long-term data for
amphibians and other taxonomic groups will make it difficult
to generate data-derived AFs. However, for amphibians, short
exposures during development and metamorphosis can
produce effects that are relevant through the lifespan of an
organism (e.g., developmental abnormalities that affect
growth and reproduction later in life). Initial analysis based
on metamorphic and developmental endpoints, but without
longer exposure chronic data, suggests that a larger AF could
be warranted to generate a chronic COC that is protective of
amphibians for carbon tetrachloride. However, EPA is still in
the process of evaluating the body of available literature data
to determine whether to revise standards for application of
AFs under TSCA.
30
PUBLIC COMMENTS:
The EPN is inclined to accept the approach used in the
2020 EPA draft for assessing CC14 risk to algae. This view
recognizes that 72- or 96-hour algal testing can be
appropriately described as both an acute and a chronic
exposure to a test substance because exposure takes place
in a relatively short duration, but it also occurs during the
reproduction of populations of individual algal cells, and
it's those developing and changing cell populations that are
measured. The fairly well-defined and easily measured
endpoint of death in individual organisms (e.g., fish) is
quite different from the measuring of inhibition of growth
The approach is carried through in the final risk evaluation.
Page 45 of 210
-------
in large populations of photosynthetic algal cells. Those
endpoints are clearly quite different.
Ecological risk characterization/interpretation
SACC
SACC COMMENTS:
Overall, the information presented did not support the
conclusions that expected environmental
concentrations were below hazard thresholds for
aquatic species.
EPA did not use conservative values to assess either
exposure or hazard threshold for aquatic species.
Mean values for exposures were compared to rounded
values of higher concentration threshold and did not
include adequate safety factors given the uncertainty of
the estimates.
Recommendation: The Agency should evaluate
degradation products of CC14 and conduct risk evaluations
in terrestrial organisms as well as aquatic and endangered
species.
In the absence of chronic amphibian studies, EPA viewed the
amphibian study 4-days post-hatch (8-9 days total) as sub-
chronic and applied an AF of 10 to derive a chronic hazard
value per current OPPT methodology (U.S. EPA. 2013,
2012bY
EPA chose the most conservative hazard values from data
available in the public domain to calculate acute and chronic
COCs relevant to aquatic ecosystems. In addition, an AF of
10 was applied to the most conservative acute and chronic
hazard values to account for species that may be more
sensitive but were not represented in the available data. This
AF was higher than the factor of 5 normally used to calculate
acute COCs for aquatic invertebrates and fish, because EPA
wanted to incorporate the added uncertainty around
amphibians into the COC.
The amphibian chronic COC of 0.003 mg/L used in this risk
evaluation is two orders of magnitude more protective than if
the chronic COC were derived from fish (0.2 mg/L), and
more protective than if the chronic COC were derived by
applying an ACR of 5 to the lowest amphibian acute hazard
value.
The TSCA risk evaluation focuses on exposures to particular
species and environmental receptors, and appropriately
considered impacts to affected species.
During problem formulation, terrestrial species exposure
pathways were determined to be covered under other
Page 46 of 210
-------
environmental statues administered by EPA (e.g., RCRA and
CAA). However, in the final risk evaluation exposure to
terrestrial organisms from the soils and biosolids pathway
was evaluated qualitatively. Clarifying language about what
pathways are addressed under the jurisdiction of other EPA-
administered statutes has been added to Section 1.4.3 of the
Risk Evaluation.
SACC
SACC COMMENTS:
Amphibians were more sensitive than fish to carbon
tetrachloride in acute exposure scenarios by two orders of
magnitude (amphibian acute hazard value of 0.9 mg/L versus
fish acute hazard value of 10.4 mg/L). Thus, they were used
to generate the acute COC to assess risk to aquatic organisms
(excluding algae). EPA also applied a larger AF (10 versus
the traditional 5 applied to acute fish data) to allow for
uncertainties in the use of amphibian data. If EPA were to use
the lowest toxicity value derived for fish divided by an AF of
10 versus 5, as recommended by the SACC to account for
uncertainty in MOA for fish, the acute COC would still be
less conservative than the COC generated using amphibian
data (1.04 mg/L versus 0.09 mg/L). The use of amphibian
toxicity data yields a COC most protective of aquatic life in
acute exposure scenarios.
A similar argument for additional safety factors can be
made for the acute COC of 7 [j,g/L in fish. Table G-l in
the Evaluation clearly shows the MOA (hepatoxicity)
of carbon tetrachloride was consistent between fish and
mammals. The liver serves an important role in fish
reproduction. Consequently, since it appears to be a
target organ in fish as well as rodents, the WOE
indicates reproduction may also be impaired and
indicates additional uncertainty for the risk
characterization statement of "no unreasonable risk."
SACC
SACC COMMENTS:
Recommendation: A 9-day exposure value should be
compared to the value of 2.5 [j,g/L instead of the rounded-
up value of 3 (J,g/L.
The use of 20 days of exceedance to determine risk
was based on chronic invertebrate and fish assays
normally taking 21 or 28 days. However, if a
developmental assay is used as a threshold of effect,
days of exceedance are not a relevant comparison, as
development can be altered by exposure at even hourly
After the application of an AF of 10, the chronic COC was
rounded to the nearest 1 ppb. The COC was rounded from 2.5
to 3 [j,g/L due to lack of precision in the reported experimental
data (where the LOD was 5 ppb) and uncertainty in the
extrapolation of data from a few organisms to represent
hazard for entire trophic levels.
EPA considered the recommendation that shorter exceedance
values (< 20 days) may be relevant to determine risk to
aquatic organisms when hazard is derived from
Page 47 of 210
-------
exposure durations.
The rationale to consider RQ exceedances for up to 20
days as acceptable for a lethal endpoint was not
provided. Ambystomidae, a species of salamanders (not
the species used in this draft risk evaluation, but the
lifespan is expected to be similar), live for 30 years.
Therefore, by definition, a chronic exposure would
exceed 3 years, although only ~1 month is typically
spent in any one water body.
One Committee member suggested that any time point
can be used for RQ evaluations given the uncertainties
of the data. If a developmental value is to be used, the
designation "Non-applicable" should be placed in
columns for acute days of exceedance in Tables E2 and
E3 of the risk evaluation.
The risk characterization was not straightforward, and
uncertainties should have been explicitly stated and an
attempt should have been made to account for them.
developmental endpoints. Exposure for short durations during
development can cause permanent adverse effects in
vertebrates. Because, for carbon tetrachloride, the chronic
COC was based on mortality observed during amphibian
development in a 9-day exposure, EPA added calculations of
chronic risk for amphibians where RQs were > 1 (and
exceedance was 0 days or greater). This scenario was
compared to the traditional assessment methodology (where
chronic risk = RQ >1 and > 20 days exceedance) to provide a
range that considered the biological relevance of short
exposures during development. This risk calculation has been
added in section 4.1.1.
Although there were no data reasonably available for long
lived salamander species, EPA expects the chronic COC
should be protective of amphibians. EPA used the most
sensitive toxicity value from a 9-day exposure during
development (a sensitive life history stage), applied an AF of
10 to account for uncertainty surrounding differences across
life stages and species, and added a risk bracket for
conservative scenarios where RQ > 1.
EPA explicitly stated uncertainties in Section 4.4.3 and
accounted for them by applying AFs in its risk calculations.
SACC
SACC COMMENTS:
There was no consideration of effects in the aquatic prey
base, which were not evaluated.
Hazard data for algae and aquatic invertebrates were
evaluated and were found to be less sensitive than
amphibians.
SACC
SACC COMMENTS:
If Table 4.2 is evaluated in the light of any exceedance
of the predicted E-FAST value, the occurrence of RQ
>1 for 5 out of 21 sites for the 20-day exposure
estimates and 4 out of 21 for the 250-day exposure
estimates would indicate that a more refined risk
characterization is needed, perhaps with measured
There were five facilities that indicated risk for aquatic
organisms from chronic exposure to carbon tetrachloride (RQ
> 1 for the chronic COC based on a developmental endpoint).
EPA subsequently refined the assessment to examining when
released occurred at each of these five facilities to determine
if amphibian development could realistically be affected.
Page 48 of 210
-------
values in surface water. The Committee suggested
these values could be obtained from the NPDES
monitoring reports from the same dischargers used to
estimate surface water values.
Recommendation: If the RQ is >1 in multiple sites, a more
refined risk characterization with better uncertainty
estimates is needed.
Timing of exposure is important to consider because
amphibian development is constrained seasonally throughout
the U.S., and typically spans only 2-4 months out of any
given year. Where releases occurred and data were available,
EPA calculated surface water concentrations using E-FAST
and associated, site-specific RQs to determine whether risk
was or was not indicated at the facilities during these key
time periods. Risk was not indicated during time periods
relevant to amphibian development at Eco Services
Operations Facility (RQs< 1 for the three years where
monitoring information was available). At the other four
facilities, risk was indicated (RQs > 1) during the time
periods relevant to amphibian development for at least 2
separate reporting periods. However, risk was not consistent
or predictable across years or facilities (e.g., some years no
releases of carbon tetrachloride occurred, or RQs < 1). This
refined analysis has been added to section 4.1.2 and in
Appendix (Table F-2).
SACC
SACC COMMENTS:
Recommendation: Specifically note the criteria used to
assess data relevance for risk assessment in addition to
determining data quality and consider using other data
(including those not considered high quality) in a
corroborative sense to support high-quality studies used to
develop a COC. A simple flow chart on this process may
help clarify these issues in the risk evaluation.
In the methodology presented in Section 3.1.1, it is not
clear how the ECOTOX database was used.
There is a lot of information presented in Appendix G
that would have been clearer if included in the body of
the draft risk evaluation. In Table G-l, many studies
conducted in fish evaluated enzyme induction (which
is not in itself an adverse effect) and some are
Relevance was iteratively assessed throughout the systematic
review process, from data search to data integration. In all
evaluation strategies, professional judgment is employed to
determine the adequacy or appropriateness of the qualitative
rating assigned by the numerical scoring system.
As discussed in the Application of Systematic Review in
TSCA Risk Evaluations, OPPT leveraged EPA's
ECOTOXicology knowledgebase (ECOTOX) as a source of
single chemical toxicity data for aquatic and terrestrial
organisms. Using a modified ECOTOX literature search and
screening protocol, OPPT performed a wide search based on
chemical-specific search terms to gather ecological toxicity
information. Title/abstract and full-text screening decisions
were based on the modified ECOTOX minimum applicability
Page 49 of 210
-------
intraperitoneal exposures that were judged to be high in
data quality. The draft risk evaluation (p. 96, lines
3072-3073) states that 61 of the 73 studies were of
unacceptable data quality (suggesting that they were
excluded); however, Table G-l has more than 12
studies that are rated high in quality. Therefore, it
seems that EPA used other criteria to determine
acceptability in addition to data quality. This was also
inferred in lines 6992-7003. These criteria could be
added to the table as an additional column. Consider
highlighting what studies were selected and used.
criteria that parsed citations into "on-topic" and "off-topic"
bins. The "on-topic" references were further subjected to a
full-text screening step to confirm relevancy. Only citations
that fulfilled the full-text screening criteria moved to the data
evaluation step.
The data quality extraction results for carbon tetrachloride
environmental hazard are presented in Appendix Table F-l.
This table contains citations considered as on-topic according
to the ECOTOX criteria but some of these citations were
excluded from further consideration due to their unacceptable
data quality based on pre-defined data quality evaluation
criteria in the Application of Systematic Review in TSCA Risk
Evaluations and/or were deemed out of scope.
For example, certain environmental studies on carbon
tetrachloride were of high quality but were not biologically
relevant for purposes of environmental hazard assessment due
to the reported endpoints (e.g., glutamic pyruvic transaminase
activity, serum total protein, catalase activity, sodium
concentration in blood, whole body residue). These studies
(Chen et at.. 2004); (de Vera and Pocsidio. 1998); (Barrows
et at.. 1980); (Liu et at.. 2.015); (Jia et at.. 2013); (Kotsanis
and Metcalfe. 1988); (Weber et at ); (Koskinen et at..
2004); (Bander et at.. 2005); (Martins et at.. 2007)) are
contained within the on-topic data evaluation section of
Appendix F. 1, but were not used within the risk evaluation
process. During risk evaluation, EPA made refinements to the
conceptual models resulting in the elimination of the
terrestrial exposure pathway and studies that are not
biologically relevant from further analysis. In the final risk
evaluation, exposures to terrestrial organisms from biosolids
and soils were evaluated qualitatively based on physical-
Page 50 of 210
-------
chemical properties.
EPA/OPPT's quality evaluation method was developed
following identification and review of various published
qualitative and quantitative scoring systems to inform
EPA/OPPT's fit-for-purpose tool. The development process
involved reviewing various evaluation tools/frameworks (e.g.,
OHAT Risk of Bias tool, CRED, etc.; see Table 1 and
Appendix A of the Application of Systematic Review in TSCA
Risk Evaluations document and references therein), as well as
soliciting input from scientists based on their expert
knowledge about evaluating various data/information sources
for risk assessment purposes.
In order to ascertain the quality of the available data, EPA is
using a numerical scoring system to assign a qualitative
rating. This approach adds consistency and transparency to
the evaluation process. Scores will be used for the purpose of
assigning the confidence level rating of High, Medium, Low,
or Unacceptable, and inform the characterization of
data/information sources during the data integration phase.
Of the 75 on-topic environmental hazard sources identified by
the ECOTOX process, 60 citations were considered out of
scope and/or unacceptable in data quality based on the data
quality evaluation metrics and the rating criteria described in
the Application of Systematic Review in TSCA Risk
Evaluations. The data quality evaluation results for the
remaining 15 on-topic studies for carbon tetrachloride
environmental hazard are presented in the document Risk
Evaluation for Carbon Tetrachloride, Systematic Review
Supplemental File: Data Quality Evaluation of Environmental
Hazard Studies (U.S. EPA. 2019V
Page 51 of 210
-------
EPA has incorporated the feedback from SACC for inclusion
of more environmental hazard information in the body of the
risk evaluation document. The summary table of relevant
aquatic toxicity data used to determine the COCs from the
Appendix F-2 Hazard Identification - Aquatic Section into
the environmental hazard section in the main document
(Table 3-1).
Refinements to the evaluation strategies are likely to occur.
EPA already made changes to the physical chemical
properties, environmental hazard, and epidemiological
criteria since the Application of Systematic Review in TSCA
Risk Evaluations was published. These changes were due to
validation and improvement efforts to ensure that the most
relevant studies were included in the TSCA risk evaluations,
and the most up-to-date data quality evaluation criteria will
be available for review in the upcoming the Systematic
Review Protocol Supporting the TSCA Risk Evaluations
document (under development).
The TSCA evaluation strategies consider methodological
design and implementation and reporting within the existing
domains and metrics. Since it is difficult to have high
confidence in data where the underlying methods that are
unreported or poorly reported, EPA assesses reporting and
methodological quality simultaneously. However, EPA
recognizes the challenge of discerning between a deficit in
reporting and a problem in the underlying methodological
quality of the data/information source. Developing a reporting
checklist, guidance document or a separate reporting quality
domain may be possible in the future as EPA uses and
optimizes the evaluation strategies.
Page 52 of 210
-------
SACC SACC COMMENTS:
Recommendation: Describe why more robust methods
(e.g., species sensitivity distributions) could not be used
for the identification of environmental hazards.
EPA explored the use of robust statistical methodologies
including species sensitivity distributions (SSDs) and BMD
modeling as additional lines of evidence for how carbon
tetrachloride exposure could affect the most sensitive
taxonomic group, amphibians. This information has been
added to the Appendix F of the final risk evaluation.
EPA generated SSDs using the SSD Toolbox, a resource
created by EPA's Office of Research and Development
(ORD) (Etterson, 2019). There was insufficient data (n = 4
species) to examine the LC10 data for amphibians using an
SSD. There was enough data (n = 7 species) to preliminarily
examine LCso data from the 4-days post-hatch exposure.
Using the three best-fitting distributions, the model averaged
HC5 (the hazardous concentration intended to be protective
of 95% of amphibians) was = 0.42 mg/L (+/- 0.36 SE). This
value is within range of EPA's COC of 0.09 mg/L (the most
sensitive amphibian LC50 0.9 mg/L, divided by an
Assessment Factor of 10). Although 7 species are not enough
to represent the total variation in sensitivity across the
amphibian taxa, the SSD did reveal that the model frog
Xenopus laevis appeared to be less sensitive than other
species (Figure 1). The American bullfrog (Rana catesbiana)
and the European common frog (Rana temporaria) were the
most sensitive species in the dataset (Figure 1). The SSD
provided a useful line of evidence that EPA used to visually
assess the distribution of the available amphibian toxicity
data. However, due to the collective uncertainties including
unknown total variation in amphibian sensitivity, a small
sample size (n = 7 species, from two studies), and possible
differences across amphibian life stages, EPA used the more
Page 53 of 210
-------
conservative COC generated by dividing the lowest hazard
value by an AF of 10 to assess risk due to acute exposure.
SACC
SACC COMMENTS:
Recommendation: RQs should be made on conservative
data from exposures as well as effects.
The aqueous exposure estimates in Tables E-2 and E-3
of the draft risk evaluation are not conservative
because mean and median values were used. A more
conservative value would be the maximum value or at
least a 90th percentile value, if available from the
model. Several Committee members wondered if the
later data from NPDES permits agreed or disagreed
with the TRI data.
The acute and chronic stream concentrations reported
in Table E-2 were computed in E-FAST using 5-year
mean releases from Table E-l. The mean average
reported in this analysis should include a value other
than zero for early years when the facility was not
manufacturing or using CC14. It should be clearly
indicated that the facility was up and running and
using/producing CC14 for each of these years.
Seven sites show releases only for 2018, the last year
of data. For these sites, the best estimate of average
release is 5X the value presented in Table E-l.
Increases in releases were apparent for three sites. A
more reasonable estimate of mean releases for
increasing sites might be to extrapolate releases for the
next 6 years (timeframe for SDWA review) and use the
average of these values in E-FAST. The existence of an
upward trend in releases should be examined for all 49
sites reporting releases.
The other (default) input values to the E-FAST model
for each of the top release sites are not reported.
The aqueous exposure estimates presented in the final risk
evaluation were based on E-FAST modeling of surface water
carbon tetrachloride concentrations. The E-FAST model used
the conservative, hydrologically-based 7Q10 design flow
statistic (average 7 consecutive day of lowest flow occuring
once every 10 years). The 7Q10 is used by EPA and states for
water quality standards, to estimate toxic wasteload
allocation, and permitted discharge limits in NPDES permits.
For the final risk evaluation, EPA also analyzed carbon
tetrachloride discharges from 5 facilities during biologically
sensitive times of year (e.g., spring and summer) and found
that 90th percentile discharges do not occur during any given
month from these facilities.
Given the variability in carbon tetrachloride discharges for
any given year, EPA averaged facilities' discharges over a
five-year period (2014-2018). EPA added a footnote to clarify
averaging to include zero discharges.
Though it appears that some facilities' releases are increasing,
EPA did not extrapolate releases, instead based the surface
water concentration estimates on the 5-year releases since this
characterizes the variability in discharges over time. A review
of 2019 carbon tetrachloride releases confirms this
assumption, as compared to 2018 levels, releases in 2019
decreased to levels similar to those in 2017.
EPA has added text in Section 2.3 to list all inputs used in E-
FAST modeling.
Page 54 of 210
-------
The analysis in Appendix E tends to focus on the top
21 release sites, but there are only 49 TRI reporting
facilities. The Committee indicated that conducting the
analysis on all 49 sites is not much greater than the
effort for the 21 sites; hence, all 49 sites should be
reported and evaluated.
TRI and DMR data are both facility reported EPA data but
since each has different reporting requirements, comparison
between the two is not always applicable. EPA has added a
chart in Appendix E to present the carbon tetrachloride
release trends from all discharging facilities for each of the
five years (2014-2018) as listed in EPA's DMR database.
SACC
SACC COMMENTS:
Recommendation: Include CC14 transformation products
in the risk evaluation.
The toxicity for major CC14 transformation products,
such as CHCb, should be considered. This is essential
given EPA's reliance on degradation to remove CC14
from water and sediment.
Reasonably available toxicity information was used to assess
the toxicity of carbon tetrachloride to aquatic and sediment
organisms. Information on carbon tetrachloride's fate is used
within other EPA administered regulations {i.e., CWA, CAA)
to determine the safety of measures for carbon tetrachloride
and its transformation products in environmental media.
Section 1.4.3 of the final risk evaluation provides information
on exposure pathways and risks addressed by other EPA
administered statutes.
SACC
SACC COMMENTS:
Recommendation: Discuss the possible impact on
endangered species.
The E-FAST model demonstrated that there is a feature
that allows "searching for endangered species in the
vicinity of specific facilities," which may be useful to
production and use decisions where they are present.
The TSCA risk evaluation focuses on exposures to particular
species and environmental receptors, and appropriately
considered impacts to affected species.
SACC
SACC COMMENTS:
Recommendation: EPA should be very specific in
language describing risks based on what was and was not
assessed. Broad statements of "no risk" are misleading
given that all risks are relative and no condition where
exposure is present is without some level of risk. The
environmental risk characterization should be qualified to
the organisms actually evaluated and the conclusion of no
unreasonable risk based on environmental concentrations
above hazard thresholds be reconsidered.
EPA has added clarifying language in the risk
characterization section 4.1.
While some site-specific RQs, calculated from modeled
release data from particular facilities, are greater than or equal
to 1, indicating risk, uncertainties related to these particular
estimates (discussed specifically in section 4.1) of the risk
evaluation support a determination of no unreasonable risk
for the environment (section 5.2.2).
Page 55 of 210
-------
There may be risks to environmental receptors that are
not assessed in this draft risk evaluation. Only aquatic
receptors were evaluated and there is a reasonable
probability that their exposures are underestimated.
The text points out uncertainties that may overestimate
risk, but it is also possible that these uncertainties could
lead to underestimation.
The language used to describe the scope of this
assessment is insufficient. The limitation to only the
aquatic species and confinement to releases directly to
water must be explicitly stated. The condition of use
language obfuscates the severe limitation of this
assessment.
The Committee concluded that EPA cannot state that
there is no unreasonable risk to environmental
organisms exposed via surface water. The
environmental concentrations are above the hazard
thresholds and the conclusion of no unreasonable risk
is not fully justified.
43
PUBLIC COMMENTS:
EPA ignores unreasonable risks to algae species (four
acute RQs between 6.4 and 18 and two chronic RQs above
1.0), asserting that "[d]ue to the quick regeneration time of
many algae species, impacts to algae populations would be
most likely over long-term consecutive days of release
(i.e., > 20) versus an interval or pulse exposure."
EPA provides no data on the algal regeneration times
and does not consider how severe acute impacts lasting
<20 days may affect that regeneration.
EPA does not justify the assumption that survival of
the species is the only relevant endpoint and acute risks
to algae from releases lasting <20 consecutive days are
reasonable.
The risk determination for algae is based on an RQ > 1 and
>20 days exceedance. The 20-day criterion is derived from
partial life cycle tests (e.g., daphnid chronic and fish early life
stage tests) that typically range from 21 to 28 days in
duration. It is also important to note that the PDM estimates
the total number of days out of 1 year that the COC is
exceeded, and the days are not necessarily consecutive. Thus,
the day criterion is considered likely to protect algae.
EPA considered algal endpoints separately from the other
taxa, because durations normally considered acute for other
species (e.g., 48, 72, or 96 hours) can encompass several
generations of algae. EPA also used a more sensitive hazard
Page 56 of 210
-------
EPA's approach eliminates the possibility of
unreasonable acute risks to algae: even if a single
exposure was sufficient to decimate a local algae
species, EPA would not consider risk to be
unreasonable unless the release were repeated for 20
consecutive days.
EPA has always considered acute algal toxicity as a
relevant endpoint, and the algae COC for CC14 was
calculated based on an acute (72-hour) test.
The final EPA evaluation should determine that CC14
presents an unreasonable risk to the environment.
endpoint (ECio) instead of an acute endpoint (ECso) to
generate a COC relevant to algae. As such, EPA's approach is
protective of acute exposures to algae and is relevant to point
effects beyond mortality (e.g., reductions in growth, yield,
etc.) that are observed to effect at most % 10 of an algae
population.
EPA determined that effects of carbon tetrachloride on the
algae population would likely occur over long-term
consecutive days of release versus an interval or pulse
exposure due to its volatile properties. Therefore, EPA
concludes that there is no unreasonable risk to algae from
carbon tetrachloride under the conditions of use.
23, 30,
43
PUBLIC COMMENTS:
For aquatic species, EPA calculated an acute RQ above 1.0
resulting from CC14 releases from the Dover Chemical site
in Ohio. However, EPA ignores the resulting risks because
"noncompliance and spills are not in the scope of this risk
evaluation." EPA does not even provide data on CC14
releases from a Sea World facility in California because
"the reported level is above permit discharge limits." In
other words, EPA knows of unsafe releases of CC14 to the
environment, but it fails to consider them in the risk
evaluation because it attributes them to spills or releases.
This exclusion violates TSCA, which requires EPA to
consider all exposures from CC14's "intended, known or
reasonably foreseen" conditions of use, including "spills,
leaks, and other uncontrolled discharge[s]." The Ninth
Circuit has also held that "spills, leaks, and other
uncontrolled discharges ... would thus qualify as
'disposals' (and therefore conditions of use)."
Spills and leaks generally are not included within the scope of
TSCA risk evaluations because in general they are not
considered to be circumstances under which a chemical
substance is intended, known or reasonably foreseen to be
manufactured, processed, distributed, used, or disposed of. To
the extent there may be potential exposure from spills and
leaks, EPA is also declining to evaluate environmental
exposure pathways addressed by other EPA-administered
statutes and associated regulatory programs. However, EPA
confirmed that there were regulatory actions outside TSCA
associated with these accidental or noncompliance spills.
First, EPA does not identify carbon tetrachloride spills or
leaks as "conditions of use." EPA does not consider carbon
tetrachloride spills or leaks to constitute circumstances under
which carbon tetrachloride is manufactured, processed,
distributed, used, or disposed of, within TSCA's definition of
"conditions of use." Congress specifically listed discrete,
routine chemical lifecycle stages within the statutory
definition of "conditions of use" and EPA does not believe it
Page 57 of 210
-------
is reasonable to interpret "circumstances" under which carbon
tetrachloride is manufactured, processed, distributed, used, or
disposed of to include uncommon and unconfined spills or
leaks for purposes of the statutory definition. Further, EPA
does not generally consider spills and leaks to constitute
"disposal" of a chemical for purposes of identifying a
condition of use in the conduct of a risk evaluation.
In addition, even if spills or leaks of carbon tetrachloride
could be considered part of the listed lifecycle stages of
carbon tetrachloride, EPA has "determined" that spills and
leaks are not circumstances under which carbon tetrachloride
is intended, known or reasonably foreseen to be
manufactured, processed, distributed, used, or disposed of, as
provided by TSCA's definition of "conditions of use," and
EPA is therefore exercising its discretionary authority under
TSCA section 3(4) to exclude carbon tetrachloride spills and
leaks from the scope of the carbon tetrachloride risk
evaluation. The exercise of that authority is informed by
EPA's experience in developing scoping documents and risk
evaluations, and on various TSCA provisions indicating the
intent for EPA to have some discretion on how best to
address the demands associated with implementation of the
full TSCA risk evaluation process. Specifically, since the
publication of the Risk Evaluation Rule, EPA has gained
experience by conducting ten risk evaluations and designating
forty chemical substances as low- and high-priority
substances. These processes have required EPA to determine
whether the case-specific facts and the reasonably available
information justify identifying a particular activity as a
"condition of use."
With the experience EPA has gained, it is better situated to
Page 58 of 210
-------
discern circumstances that are appropriately considered to be
outside the bounds of "circumstances... under which a
chemical substance is intended, known, or reasonably
foreseen to be manufactured, processed, distributed in
commerce, used, or disposed of' and to thereby meaningfully
limit circumstances subject to evaluation. Because of the
expansive and potentially boundless impacts that could result
from including spills and leaks as part of the risk evaluation
(e.g., due to the unpredictable and irregular scenarios that
would need to be accounted for, including variability in
volume, frequency, and geographic location of spills and
leaks; potential application across multiple exposure routes
and pathways affecting myriad ecological and human
receptors; and far-reaching analyses that would be needed to
support assessments that account for uncertainties but are
based on best available science), which could make the
conduct of the risk evaluation untenable within the applicable
deadlines, spills and leaks are determined not to be
circumstances under which carbon tetrachloride is intended,
known or reasonably foreseen to be manufactured, processed,
distributed, used, or disposed of, as provided by TSCA's
definition of "conditions of use."
Exercising the discretion to not identify spills and leaks of
carbon tetrachloride as a COU is consistent with the
discretion Congress provided in a variety of provisions to
manage the challenges presented in implementing TSCA risk
evaluation. See e.g., TSCA Sections 3(4), 3(12), 6(b)(4)(D),
6(b)(4)(F). In particular, TSCA Section 6(b)(4)(F)(iv)
instructs EPA to factor into TSCA risk evaluations "the likely
duration, intensity, frequency, and number of exposures under
the conditions of use...," suggesting that activities for which
duration, intensity, frequency, and number of exposures
Page 59 of 210
-------
cannot be accurately predicted or calculated based on
reasonably available information, including spills and leaks,
were not intended to be the focus of TSCA risk evaluations.
And, as noted in the preamble to the Risk Evaluation Rule,
EPA believes that Congress intended there to be some
reasonable limitation on TSCA risk evaluations, expressly
indicated by the direction in TSCA Section 2(c) to "carry out
[TSCA] in a reasonable and prudent manner." For these
reasons, EPA is exercising this discretion to not consider
spills and leaks of carbon tetrachloride to be COUs.
Second, even if carbon tetrachloride spills or leaks could be
identified as exposures from a COU in some cases, these are
not forms of exposure that EPA expects to consider in the
carbon tetrachloride risk evaluation. TSCA Section
6(b)(4)(D) requires EPA, in developing the scope of a risk
evaluation, to identify the hazards, exposures, conditions of
use, and potentially exposed or susceptible subpopulations the
Agency "expects to consider" in a risk evaluation. This
language suggests that EPA is not required to consider all
conditions of use, hazards, or exposure pathways in risk
evaluations. EPA has chosen to tailor the scope of the risk
evaluation to exclude spills and leaks in order to focus
analytical efforts on those exposures that present the greatest
potential for risk.
In the problem formulation documents for many of the first
10 chemicals undergoing risk evaluation, EPA applied the
same authority and rationale to certain exposure pathways,
explaining that "EPA is planning to exercise its discretion
under TSCA 6(b)(4)(D) to focus its analytical efforts on
exposures that are likely to present the greatest concern and
consequently merit a risk evaluation under TSCA..." This
Page 60 of 210
-------
approach is informed by the legislative history of the
amended TSCA, which supports the Agency's exercise of
discretion to focus the risk evaluation on areas that raise the
greatest potential for risk. See June 7, 2016 Cong. Rec.,
S3519-S3520.
In addition to TSCA Section 6(b)(4)(D), the Agency also has
discretionary authority under the first sentence of TSCA
Section 9(b)(1) to "coordinate actions taken under [TSCA]
with actions taken under other Federal laws administered in
whole or in part by the Administrator." TSCA Section 9(b)(1)
provides EPA authority to coordinate actions with other EPA
offices, including coordination on tailoring the scope of
TSCA risk evaluations to focus on areas of greatest concern
rather than exposure pathways addressed by other EPA-
administered statutes and regulatory programs, which does
not involve a risk determination or public interest finding
under TSCA Section 9(b)(2). EPA has already tailored the
scope of this risk evaluation using such discretionary
authorities with respect to exposure pathways covered under
the jurisdiction of other EPA-administered statutes and
associated regulatory programs (see section 1.4.3).
Following coordination with EPA's Office of Land and
Emergency Management (OLEM), EPA has found that
exposures of carbon tetrachloride from spills and leaks fall
under the jurisdiction of RCRA. See 40 CFR 261.33(d)
(defining in part a hazardous waste as "any residue or
contaminated soil, water or other debris resulting from the
cleanup of a spill into or on any land or water of any
commercial chemical product or manufacturing chemical
intermediate having the generic name listed [40 CFR
261.33(e) or (f)"|, or any residue or contaminated soil, water
Page 61 of 210
-------
or other debris resulting from the cleanup of a spill, into or on
any land or water, of any off-specification chemical product
and manufacturing chemical intermediate which, if it met
specifications, would have the generic name listed in [40 CFR
261.33(e) or (f)]"); 40 CFR 261.33(f) (listing carbon
tetrachloride as hazardous waste no. U211). As a result, EPA
believes it is both reasonable and prudent to tailor the TSCA
risk evaluation for carbon tetrachloride by declining to
evaluate potential exposures from spills and leaks, rather than
attempt to evaluate and regulate potential exposures from
spills and leaks under TSCA.
Finally, EPA notes that the Ninth Circuit in SCHF v. EPA
presented examples of circumstances that may qualify as
disposal but did not establish a "precise meaning of
'disposal.'" 943 F.3d 397, 426 (9th Cir. 2019). The Court also
did not opine on EPA's authority to determine the
circumstances under which a chemical substance is intended,
known, or reasonably foreseen to be manufactured,
processed, distributed in commerce, used or disposed of.]
EPA has clarified the assessment of the Dover facility (see
Section 4.1 in the Carbon Tetrachloride Risk Evaluation). In
brief, EPA identified an elevated environmental release of
carbon tetrachloride in 2014 at Dover Chemical in Ohio
(NPDES ID OH0007269) due to an unexpected chemical
spill. Because spills and leaks are not included within the
scope of TSCA risk evaluations, the 2014 release was not
included in the analysis. Other releases from the Dover
facility, not due to the chemical spill, were evaluated.
EPA has clarified Sea World carbon tetrachloride discharges.
Specifically, EPA has estimated the surface water
Page 62 of 210
-------
concentration from Sea World-specific carbon tetrachloride
annual loading/releases to Mission Bay using a proxy facility
in San Diego since Sea World permit data was not available
in E-FAST 2014. For discharges into oceans and bays, E-
FAST estimates a dilution factor of 1. EPA has revised the
Carbon Tetrachloride Risk Evaluation to include a greater
explanation of the E-FAST 2014 modeling approach, model
calculations, inputs and results for the one year, 2014, of
carbon tetrachloride releases from Sea World and the
resultant surface water concentration and aquatic exposure
estimates.
26
PUBLIC COMMENTS:
For environmental risk, EPA's own analyses showed that
CC14 presents an unreasonable risk to aquatic organisms
(p. 144), but EPA dismisses this unreasonable risk with
little explanation: "Although the chronic COC was
exceeded by four facilities ranging from 1.2 to 3.4 {i.e.,
worst-case scenario; RQ = 3.4) and the algae COC was
exceeded by four facilities ranging from 6.4 to 18 based on
the 20-day stream concentration and by two facilities
ranging from 1.4 to 1.5 based on the 250-day stream
concentration, these CC14 releases are not continuously
released overtime {i.e., chronic exposure) (p. 144)." Yet
for at least one of these facilities, the chronic COC was
exceeded for 15 days (p. 142). It is clearly reasonably
foreseeable that longer exposures may occur. Based on
EPA's own analyses, EPA found risks to aquatic
organisms from multiple facilities, but EPA dismissed this
risk. This approach is arbitrary and capricious because
EPA refuses to accept the outcomes of its own analyses,
and EPA's conclusions run contrary to the evidence before
the Agency. EPA should find an unreasonable risk to the
environment presented by certain disposal and recycling
EPA has added clarifying language in the risk
characterization section 4.1. All facilities assessed in this risk
evaluation and associated RQs are presented in Table 4 2.
While some site-specific RQs, calculated from modeled
release data from particular facilities, are greater than or equal
to 1, indicating risk, uncertainties related to these particular
estimates (discussed specifically in section 4.1 and 5.2.2) of
the risk evaluation support a determination of no
unreasonable risk for the environment.
Page 63 of 210
-------
conditions of use. The SACC should address EPA's
unwarranted dismissal of these environmental risks.
Occupational Kxposure iintl Releases
Charge Question 3.1: Please comment on the characterization of occupational exposure for workers and ONL s. Is the occupational
exposure characterization supported by the information presented in Section 2.4 of the Draft Risk Evaluation? What other additional
information, or approaches if any, should be considered?
Charge Question 3.2: Please comment on the scientific validity and transparency of EPA's approach and the assumptions EPA used
to characterize exposure for ONUs. Please also comment on the uncertainties related to the assumptions used to characterize
exposures for ONUs.
Charge Question 3.3: Please comment on the approaches and assumptions used and provide any specific suggestions or
recommendations for alternative approaches, models or information that should be considered by the Agency for improving the
workplace exposure assessment. More specifically, if other sources of monitoring data are available to estimate air concentrations for
worker exposures, please provide specific citations.
Charge Question 3.4: Please comment on assumptions used in the absence of specific exposure information (e.g., dermal surface
area assumptions: high-end values, which represents two full hands in contact with a liquid: 890 cm2 (mean for females), 1070 cm2
(mean for males); central tendency values, which is half of two full hands (equivalent to one full hand) in contact with a liquid and
represents only the palm-side of both hands exposed to a liquid: 445 cm2 (females), 535 cm2 (males)).
Charge Question 3.5: Please comment on EPA's approach to characterizing the strengths, limitations and overall confidence for
each OES presented in Section 2.4.1. Please comment on the appropriateness of these confidence ratings for each scenario. Please
#
Summary ol* Comments lor Specific Issues Related to
Charge Question 3
KIW/OPPT Response
Conditions of use considered
SACC,
23, 30,
32, 43
SACC COMMENTS:
The assertion of de minimis exposures is not
adequately supported by citations or data (p. 30, lines
1062-1103). The assertion of "no use" is weakened by
the admission that "CC14 may be present in a limited
number of industrial products with chlorinated
ingredients at a concentration of less than 0.003% by
weight," that is, 30 pounds per million pounds of
The Consumer Product Safety Commission (CPSC) banned
the use of carbon tetrachloride in consumer products
(excluding unavoidable residues not exceeding 10 ppm
atmospheric concentration) in 1970. As a result of CPSC's
ban, EPA does not consider the use of carbon tetrachloride-
containing consumer products produced before 1970 to be
known, intended, or reasonably foreseen. In accordance with
the CPSC ban, carbon tetrachloride is not identified in the
Page 64 of 210
-------
product, but there are millions of pounds of product
created each year.
Recommendation: Add discussion, citations, or data to
better support the assertion of de minimis exposures.
PUBLIC COMMENTS:
In a 2017 preliminary survey of CC14's conditions of use,
EPA identified CC14-containing products available to
consumers; yet there is no discussion of manufacturing,
processing, distribution, use, or disposal of these products
in the draft risk evaluation. This flaw should be remedied
in the final evaluation. There is no discussion of these
products in the draft risk evaluation and no explanation of
why the CC14 levels they contain would be too low to pose
any health concern.
CC14 is known to be released from consumer products and
several products known to contain CC14 remain in use.
Sodium hypochlorite (NaOCl) and many organic
chemicals contained in household cleaning products may
react during use to generate CC14. The SPARC report lists
use of hypochlorite as bleach in domestic applications as a
CC14 emissions source. A 2008 study (Odabaงi, 2008)
measured CC14 concentrations of 0.25-459 [j,g/m3 in
emissions from eight different chlorine bleach-containing
household products. In a search of retail websites, SCHF
identified five consumer sealant products with SDSs
indicating the presence of CC14 at levels of up to 1 percent
by weight. Given the low ambient concentrations of CC14
linked to cancer and other adverse effects, there is no basis
to assume that CC14 releases from these products would be
without concern, particularly when combined with outdoor
California Air Resources Board consumer product database
nor the Washington State Product Testing Data list or the
State of Vermont list of Chemicals in Children's Products and
no consumer uses are listed in the CDR.
As stated in the Problem Formulation, EPA determined after
additional public outreach, literature searches and other
reasonably available information, the consumer uses of
carbon tetrachloride that were initially identified in the Scope
document {i.e., commercially available aerosol and non-
aerosol adhesives/sealants, paints/coatings, and
cleaning/degreasing solvent products) only have the potential
for negligible exposure. Carbon tetrachloride is not a direct
reactant or additive in the formulation of solvents for
consumer use in cleaning and degreasing, adhesives and
sealants or paints and coatings. Trace levels of carbon
tetrachloride in the chlorinated substances used to
manufacture the products are expected to volatilize during the
product manufacturing process.
No additional information was received by EPA following the
publication of the problem formulation that would update the
problem formulation conclusion that carbon tetrachloride is
expected to be present in consumer products at trace levels
resulting in de minimis exposures or otherwise insignificant
risks and therefore that consumer uses do not warrant
inclusion in the risk evaluation.
EPA obtained information indicating that SDSs for industrial
and commercial products manufactured with chlorinated
compounds made with carbon tetrachloride as a process agent
report overestimated range concentrations, and that those
Page 65 of 210
-------
air and drinking water exposures by consumers who also
use the products.
EPA indicates that "direct use of CC14 as a reactant or
additive in the formulation" of consumer products is
prohibited under the MP and CPSC regulations. CPSC
regulations allow "manufacturing residues of CC14 that...
do not result in an atmospheric concentration of CC14
greater than 10 parts per million." EPA maintains that this
residual CC14 is only "present in consumer products at
trace levels resulting in de minimis exposures or otherwise
insignificant risks and therefore consumer uses do not
warrant inclusion in the risk evaluation." TSCA does not
permit exclusion of conditions of use based on the theory
that they lead to de minimis exposure. Further, there is no
way to know if a route of exposure is de minimis unless it
is subject to risk evaluation. The Agency has neither
provided its definition or interpretation of "de minimis" or
"insignificant risk" nor presented any criteria by which one
can determine if a condition of use represents de minimis
or insignificant risk.
EPA's decision not to evaluate these exposure scenarios
was thus arbitrary and unwarranted and results in a
significant understatement of CC14's human health
impacts.
estimates are not based on analytical measured concentrations
or on manufacturing process information.
In exercising its discretion under section 6(b)(4)(D) to
identify the conditions of use that EPA expects to consider in
a risk evaluation, EPA believes it is important for the Agency
to have the discretion to make reasonable, technically sound
scoping decisions. EPA anticipates that any risks presented by
the presence of carbon tetrachloride as a byproduct formed
during the manufacturing, processing or use of the parent
compound will be considered in the scope of the risk
evaluation of the parent compound (see the executive
summary of the Final Scope of the Risk Evaluation for 1,2-
dichlorethane as an example:
https://www.epa.gov/sites/production/files/2020-
09/documents/casrn_107-06-2_12-
di chl oroethanefinalscope. pdf).
Therefore, EPA did not evaluate hazards or exposures to
consumers or bystanders to consumer use in this risk
evaluation in the exercise of the Agency's discretionary
scoping authority under TSCA sec. 6(b)(4)(D). See section
1.4.2.2 of the risk evaluation for more information.
Risks from background concentrations to carbon tetrachloride
are assessed under the EPA NATA. The 2014 NATA reports
a national ambient carbon tetrachloride concentration of 0.53
|ig/m3 and 3 in a million cancer risk.
https://www.epa. gov/national-air-toxics-assessment/2014-
nata-assessment-results#pollutant
Page 66 of 210
-------
SACC
SACC COMMENTS:
Recommendation: Exclusions of conditions of use during
problem formulation should be made more explicit in the
risk evaluation rather than referencing the Scope of Work.
For example, present them in a summary table with the
reasons for exclusion.
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation.
43
PUBLIC COMMENTS:
The final risk evaluation must address both acute and
chronic consumer exposure to CC14. While the consumer
products listed above result in short-term exposure to
CC14, these products (particularly household) may be used
repeatedly over time and consumers are exposed to CC14
in indoor and outdoor air on a continuing basis. Thus,
cancer and noncancer risks to consumers could be
significant and should be assessed and included in a risk
evaluation that encompasses all intended, known, and
reasonably foreseeable conditions of use.
Chronic exposure scenarios resulting from long-term use of
household consumer products are likely to be relatively
infrequent with short durations of use. In addition, the short
half-life of the chemicals in the body does not result in
significant accumulation between uses on different days.
Therefore, even if levels of carbon tetrachloride in consumer
products were measurable, use frequencies would be
considered to be too low to create chronic risk concerns.
26
PUBLIC COMMENTS:
EPA claims "there [is] no data supporting its use in the
[aerospace] industry" (p. 29). EPA's source for this
assumption is a personal communication with the
Aerospace Industries Association (AIA), which EPA has
not corroborated, and the substance of which EPA has not
made available. The email communication directly
contradicts an earlier comment submitted by AIA, which
states: "The aerospace industry uses products/formulations
containing CC14 in the manufacture, operations and
maintenance of aerospace products. CC14 has been
identified in limited use in specific adhesives (including
methacrylate), and for specific cleaning operations."
All reasonably available information, including AIA's
response stating that the previously identified products in
their comment have been discontinued and the lack of data
supporting the use of carbon tetrachloride in the industry,
indicate that there is no known, intended, or reasonably
foreseen use of carbon tetrachloride in the aerospace industry.
In addition, the Montreal Protocol and CAA Title VI prohibit
the direct use of carbon tetrachloride in the formulation of
commercially available products for
industrial/commercial/consumer uses, including aerosol and
non-aerosol adhesives and cleaning/degreasing solvent
products, except as a laboratory chemical.
26
PUBLIC COMMENTS:
EPA states that it "found no evidence to suggest that the
manufacturing of ibuprofen, or any other pharmaceuticals,
While use of carbon tetrachloride as a process solvent in the
manufacture of pharmaceuticals was included in the problem
formulation, upon further analysis, EPA has determined that
Page 67 of 210
-------
still utilizes CC14 or that such use is reasonably foreseen to
resume." However, a cursory Google search suggests that
CC14 is still used in the manufacturing of pharmaceuticals:
Parchem, American Elements, and Olin Chlorinated
Organics advertise uses of CC14 related to
pharmaceutical manufacturing.
A 2019 Market Watch report listed pharmaceutical as
the first in a list of "markets by application" for CC14.
EPA has failed to rely on all reasonably available
information. EPA has broad authority under TSCA to
mandate submissions from industry that would reveal
whether or not this chemical's use as process agent in the
manufacturing of pharmaceuticals is a condition of use.
The SACC should recommend EPA exercise this authority
to obtain information that could be used to confirm or
negate its assumptions.
this use falls outside TSCA's definition of "chemical
substance." Under TSCA ง 3(2)(B)(vi), the definition of
"chemical substance" does not include any food, food
additive, drug, cosmetic, or device (as such terms are defined
in section 201 of the Federal Food, Drug, and Cosmetic Act)
when manufactured, processed, or distributed in commerce
for use as a food, food additive, drug, cosmetic, or device.
EPA has concluded that carbon tetrachloride use as a process
solvent during pharmaceutical manufacturing falls within the
aforementioned definitional exclusion and is not a "chemical
substance" under TSCA. Further, as stated in the draft risk
evaluation, EPA does not have any evidence that carbon
tetrachloride is still being used in the manufacture of
ibuprofen or any other pharmaceuticals. The fact that
distributors of carbon tetrachloride cite pharmaceutical
manufacturing as one of the uses of the chemical substance
does not by itself indicate that it is being used for this
purpose.
31
PUBLIC COMMENTS:
EPA used qualitative assumptions to indicate that exposure
potential was low for reactive ion etching and laboratory
use. The SACC should consider the appropriateness of
these assumptions and provide recommendations regarding
the use of qualitative approaches to support assumptions of
minimal exposure.
EPA requested information on all aspects of risk evaluations
of carbon tetrachloride throughout the risk evaluation process,
including opening public dockets for receipt of such
information, conducting outreach to manufacturers,
processors, users and other stakeholders. The information
received have been incorporated into the risk evaluation.
The TSCA risk evaluation strategies refer to study guidelines
along with professional judgment as helpful guidance in
determining the adequacy or appropriateness of certain study
designs or analytical methods. EPA considered reasonably
available, relevant data and information that conform to the
TSCA science standards when developing the risk evaluation
of carbon tetrachloride.
Page 68 of 210
-------
Due to the performance requirements of products typically
produced via Reactive-ion etching (RIE), carbon tetrachloride
is generally applied in small amounts in a highly controlled
work area (e.g., under a fume hood as per good laboratory
practice), thus eliminating or reducing the potential for
exposures.
40
PUBLIC COMMENTS:
The final risk evaluation should clarify that conclusions of
unreasonable risk do not extent to substances that are not
"chemical substances" as defined in TSCA ง 3(a) and that
the findings are described only to form a basis for
evaluating risk from conditions of use that are governed by
TSCA. Pesticides, tobacco, certain nuclear material,
firearms, shells and cartridges, food, food additives, drugs,
cosmetics, and medical devices are excluded from the
TSCA definition of "chemical substance."
Section 1.4.2 of the risk evaluation specifies that the term
"chemical substance" as defined in TSCA ง 3(2) does not
include any mixture; any pesticide when manufactured,
processed, or distributed in commerce for use as a pesticide;
tobacco or tobacco product; source material, special nuclear
material, or byproduct material; any article the sale of which
is subject to the tax imposed by section 4181 of the Internal
Revenue Code of 1986; and any component of such an
article, or any food, food additive, drug, cosmetic, or device
when manufactured, processed, or distributed in commerce
for use as a food, food additive, drug, cosmetic, or device. For
this reason, any conclusions of unreasonable risk do not
extend to substances that are not defined as chemical
substances under TSCA ง 3(2).
SACC
SACC COMMENTS:
Recommendation: The justification for regrouping
conditions of use should be described in more detail
wherever it was conducted. Surrogate groups should be
named more specifically to distinguish different types. For
example, a chemical surrogate is different from a worker
activity surrogate, although the draft risk evaluation seems
to conflate the two.
The term "surrogate data" can mean different things, with
different levels of uncertainty.
One example is applying monitoring data for the
target chemical to a different condition of use, as in
The regrouping of conditions of use is described in section
2.4.1.6 of the risk evaluation. In addition, EPA expanded
discussions in the final risk evaluation regarding the type of
surrogate data utilized and the associated assumptions in
Section 4.4.1 of the risk evaluation.
Page 69 of 210
-------
the case of the manufacturing and processing
condition of use. The assumptions in this case are
multiple (similar source types, similar processes,
similar worker activities, etc.).
Another example is using monitoring data for a
chemical other than the target (which was not
measured) on the basis that the surrogate chemical
has similar physicochemical properties (or
differences in exposure concentrations could be
estimated from the properties). This type of
surrogate requires fewer assumptions and is likely
to introduce lower uncertainty in exposure
estimates than in the first example.
EPA's hierarchy of exposure estimation approaches
does not distinguish between these two, although
they would be expected to have different levels of
uncertainty, and both seem to co-mingle in the draft
risk evaluation. It can be argued also that EPA uses
workers' exposures as the surrogate for estimating
exposures to ONUs, which is yet another
application of the term "surrogate."
Recommendations: (1) Be specific when using the term
"surrogate" when applying data from one condition of use
to another; (2) ensure that the condition of use and its
surrogate do not have hugely different associated levels of
uncertainty; and (3) better describe the engineering and
worker activities associated with a condition of use and
compare these to their surrogate condition of use to ensure
that they are not significantly different.
26
PUBLIC COMMENTS:
The frequency and magnitude of take-home exposure is
dependent on several factors, including personal hygiene and
Page 70 of 210
-------
EPA excluded a number of reasonably foreseen conditions
of use in the workplace that should have been evaluated,
including: exposures from spills in the workplace; "take-
home exposures;" exposures of maintenance staff,
especially those cleaning up spills and leaks; and
exposures of workers at small or medium facilities where
assumptions of routine PPE use or other protections are
less likely to be valid. Each of these is a "reasonably
foreseen" aspect of the circumstances under which CC14 is
manufactured, processed, distributed, used, or disposed of.
visibility of the chemical on skin or clothing. EPA does not
have methods to reliably predict take-home exposure.
Spills and leaks generally are not included within the scope of
TSCA risk evaluations because in general they are not
considered to be circumstances under which a chemical
substance is intended, known or reasonably foreseen to be
manufactured, processed, distributed, used, or disposed of. To
the extent there may be potential exposure from spills and
leaks, EPA is also declining to evaluate environmental
exposure pathways addressed by other EPA-administered
statutes and associated regulatory programs. See response for
question about spills under Charge Question 1 for additional
detail.
General
population exposures
SACC
SACC COMMENTS:
In accordance with the CPSC ban, carbon tetrachloride is not
identified in the California Air Resources Board consumer
product database and no consumer uses are listed in the CDR.
Consumer products and/or commercial products containing
chlorinated compounds made with carbon tetrachloride as a
process agent are available for public purchase at common
retailers [EPA-HO-OPPT- 733-0003. sections 3 and 4.
( )] However, these products are not expected
to contain measurable amounts of carbon tetrachloride
because carbon tetrachloride is not used in the manufacturing
of the actual products. Trace levels of carbon tetrachloride in
the chlorinated substances used to manufacture the products
are expected to volatilize during the product manufacturing
process.
Concentrations of carbon tetrachloride along with other 37
gas-phase organic air toxics were measured by Logue et al.
Recommendation: Include a summary of residential indoor
and outdoor air concentrations of CC14 as well as personal
air concentrations of the residents.
This information would provide more context for
EPA's decision to not evaluate consumer exposures, as
it did in the evaluation for methylene chloride. The
sources for these data are the same as those cited in
that evaluation.
Page 71 of 210
-------
(2010) over a 2-year period at four different sites in and
around Pittsburgh, PA: a downtown site with substantial
mobile source emissions; two residential sites adjacent to one
of the most heavily industrialized zones in Pittsburgh; and a
regional background site. Concentrations of carbon
tetrachloride exhibited little temporal or spatial variability
with study average concentrations of carbon tetrachloride
varied by less than 25% across the four sites. In a separate
study, carbon tetrachloride was measured and interpreted by
de Bias et al. (2016) with high-time resolution in two sites
(urban and rural) in Northern Spain. One site is an urban area
influenced by the surrounding industry, where measurements
were performed for a one-year period (2007-2008) and the
second site is a rural background area where measurements
were carried out for a non-successive five-year period (2003-
2005, 2010-2011, and 2014-2015 years). Median yearly
carbon tetrachloride mixing ratios (a dimensionless parameter
indicates the abundance of one component of a mixture
relative to that of all other components) were higher in the
urban area (120 parts per trillion by volume) than in
Valderejo Natural Park (80-100 parts per trillion by volume).
The carbon tetrachloride was reported by de Bias et al. (20.1.6)
to be well mixed in the atmosphere and no sources were
reported to impact the rural site. In the urban areas chlorine-
bleach products that are used as indoor cleaning agents could
result a potential source of carbon tetrachloride due to
reactions with organics, soap or surfactants.
Furthermore, background concentrations to carbon
tetrachloride are assessed under the EPA NATA.
SACC
23. 26
SACC COMMENTS:
The current exclusion of exposure pathways to the
general population through releases to ambient air,
As part of the Problem Formulation for carbon tetrachloride
( 2018b). EPA found that exposures to the general
population may occur from the conditions of use due to
Page 72 of 210
-------
32, 38,
41,43
drinking water, ambient water, biosolids, and disposal
pathways could lead to underrepresentation of the
risks.
Recommendation: Consider performing a wider
assessment accounting for these excluded pathways, which
will provide a more reliable measure of the risk.
PUBLIC COMMENTS:
There is extensive evidence of pervasive general
population exposure to CC14 from releases to air, water,
and soil and at levels in ambient air and drinking water that
present significant cancer risks.
Large air emissions of CC14 raise health concerns for the
general population and subpopulations living near
emission sources.
Recent TRI and NEI reporting and scientific studies
indicate substantial ongoing emissions of CC14. The
NTP Report on Carcinogens states that "8 million
people living within 12.5 miles of manufacturing sites
were possibly exposed to CC14 at an average
concentration of 0.5 [j,g/m3 and a peak concentration of
1,580 ng/m3."
ATSDR reports that: "Based on analysis of 4,913
ambient air samples reported in the National Ambient
VOCs Database, the average concentration of CC14
was 0.168 ppb (1.1 ^ig/m3)." It estimates that daily
intake from air ranges from 12 to 511 (J,g/m3, based on
average ambient concentrations of 0.1-4 ppb (0.64-25.6
^g/m3).
A review of EPA's air toxics data reveals that every
census tract in the U.S. has excess cancer risk of about
3.5 in a million due to CC14 in the air.
EPA and USGS report that CC14 has been consistently
releases to air, water or land. The exposures to the general
population via surface water, drinking water, ambient air and
sediment pathways fall under the jurisdiction of other
environmental statutes administered by EPA, i.e., CAA,
SDWA, CWA, and RCRA. As explained in more detail in
section 1.4.3 of the final risk evaluation, EPA believes it is
both reasonable and prudent to tailor TSCA risk evaluations
when other EPA offices have expertise and experience to
address specific environmental media, rather than attempt to
evaluate and regulate potential exposures and risks from those
media under TSCA. EPA believes that coordinated action on
exposure pathways and risks addressed by other EPA-
administered statutes and regulatory programs is consistent
with the statutory text and legislative history, particularly as
they pertain to TSCA's function as a "gap-filling" statute, and
also furthers EPA aims to efficiently use Agency resources,
avoid duplicating efforts taken pursuant to other Agency
programs, and meet the statutory deadlines for completing
risk evaluations. EPA has therefore tailored the scope of the
risk evaluations for carbon tetrachloride using authorities in
TSCA sections 6(b) and 9(b)(1). See section 1.4.3 of the Risk
Evaluation.
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation.
Page 73 of 210
-------
detected in drinking water supplies. ATSDR concludes
that "[iIngestion via contaminated drinking water is an
important route of exposure for the general population
not living in areas where CC14 is extensively used" and
that the general population may also inhale CC14 "from
volatilization of contaminated water during showering
or bathing." There are more than 160 drinking water
systems, serving more than one million people, with
CC14 levels exceeding health-protective standards.
The EPA problem formulation references a study from
New Jersey Department of Environmental Protection
that finds that the "acceptable shower water criteria for
CC14 is 0.15 [j,g/L and the associated shower air
concentration of CC14 would be acceptable at 1.5 x 10-
5 [j,g/m3." The risk evaluation makes no effort to assess
whether these "acceptable" concentrations are being
exceeded.
Hundreds of federal Superfund sites with CC14 in the
soil or groundwater pose a potential threat of vapor
intrusion. Vapor intrusion may provide a partial
explanation for the widespread detection of CC14 in
indoor air. As ATSDR notes "Typical concentrations in
homes in several U.S. cities were about 1 [j,g/m3 (0.16
ppb), with some values up to 9 [j,g/m3 (1.4 ppb)"
These risks are not being effectively reduced under other
environmental laws.
EPA's exclusion of environmental exposure pathways
from risk evaluations will defeat the central TSCA goal
of comprehensively evaluating a chemical's risks to
humans and the environment, and the law's
requirement for EPA to consider all conditions of use,
including those affecting PESS.
Page 74 of 210
-------
EPA has not explained why, in direct contradiction to
how EPA treated background exposures from
hexabromocyclododecane (HBCD) to the general
population, it chose to entirely ignore background
exposures to CC14.
Under the National Air Toxics Assessment (NATA),
EPA calculates the long-term health risks of CC14 by
considering background exposures to the chemical
because it "has a very long residence time, which
makes predictions based on current emissions moot."
The SACC should comment on the human health impacts
of EPA's failure to consider background exposures to
CC14.
SACC,
23, 30,
32, 38,
41, 42,
43
SACC COMMENTS:
Recommendation: Workplace exposure estimates should
be aggregated.
Multiple SACC members favor aggregating
contemporaneous exposures.
PUBLIC COMMENTS:
The Agency has not assessed aggregate exposures for
CC14 or made their unreasonable risk findings based upon
combined exposures, either for a specific condition of use
or with consideration of exposures from non-TSCA-related
scenarios. TSCA provides protections to workers not just
from chemical exposure in the workplace but from air
emissions and other environmental releases as well as
exposures to consumer products.
CC14 levels are likely to be ever greater surrounding the
facilities where CC14 is manufactured and released, which
are the same communities where many of the workers
employed in those facilities live. In tribal communities, a
TSCA section 6(b)(4)(F)(ii) directs EPA to "describe whether
aggregate or sentinel exposures to a chemical substance under
the conditions of use were considered, and the basis for that
consideration" in risk evaluations. EPA defines aggregate
exposures as the combined exposures to an individual from a
single chemical substance across multiple routes {i.e., dermal,
inhalation, or oral) and across multiple pathways {i.e.,
exposure from different sources). 40 CFR 702.33. EPA
defines sentinel exposures as the exposure from a single
chemical substance that represents the plausible upper bound
of exposure relative to all other exposures within a broad
category of similar or related exposures. 40 CFR 702.33. EPA
considered the reasonably available information and used the
best available science to determine whether to consider
aggregate or sentinel exposures for a particular chemical.
EPA has determined that using the high-end risk estimate for
inhalation and dermal risks separately as the basis for the
unreasonable risk determination is a best available science
approach. There is low confidence in the result of aggregating
Page 75 of 210
-------
substantial number of residents have multiple jobs and live
near their community facilities, including disposal
facilities. A single person may be a landfill worker, an
occupational bystander, and a near-facility general
population, as well as a consumer.
The SACC has repeatedly raised concerns about EPA's
failure to consider environmental pathways of human
exposure. Environmental pathways play a major role in
contributing to aggregate exposures and EPA's exclusion
of them means that the Agency is not able to accurately
assess risks, including to PESS.
Congress directed EPA to make an unreasonable risk
determination for the chemical substance as a whole,
taking into account all of its uses. EPA violates that
requirement in this risk evaluation, by proposing use-by-
use determinations of unreasonable risk that fail to
consider the risks to workers who are exposed from
multiple conditions of use, despite noting that "it is not
uncommon for employees at a facility to perform multiple
types of tasks throughout the work day." EPA should
prepare an exposure assessment that examines aggregate
exposure, combining exposures from the inhalation and
dermal pathways, including baseline exposures, under all
conditions of use.
the dermal and inhalation risks for this chemical if EPA uses
an additive approach, due to the uncertainty in the data. EPA
does not have data that could be reliably modeled for the
aggregate exposure, which would be a more accurate
approach than adding, such as through a PBPK model. Using
an additive approach to aggregate risk in this case could result
in an overestimation of risk. Given all the limitations that
exist with the data, EPA's approach is the best available
science. EPA has added language to the Key Assumptions
and Uncertainties section describing these assumptions and
uncertainties. Clarifying language on exposure pathways and
risks under the jurisdiction of other EPA-administered
statutes have been added to section 1.4.3 of the final risk
evaluation document.
EPA did not consider carbon tetrachloride background
exposure that workers might be exposed to in addition to
exposures from TSCA conditions of use. This may result in
an underestimation of risk, and additional discussion of this
underestimation has been added to the document in the Key
Assumptions and Uncertainties section.
Worker exposure estimation: methods, models, and data
SACC
SACC COMMENTS:
Recommendation: EPA should develop a decision tree for
using monitoring data or modeling not just on the basis of
the quality of monitoring data, but also on the quantity of
data.
EPA considered both quality and quantity of monitoring data
when deciding to pursue modeling. For many conditions of
use, there were limited or no reasonably available data to
develop and/or validate simulation models.
Page 76 of 210
-------
SACC,
38, 43
SACC COMMENTS
SACC identified the following data gaps and uncertainties:
lack of exposure data for the scenarios of ONU inhalation,
reactive ion etching, processing agent/aid, additive use,
laboratory use, waste handling, and dermal exposure; and
limited data sets for specialty use. To increase confidence
in risk conclusions, EPA will need to use its statutory
authority to request limited studies to obtain the quality
data to better support these parts of the assessment. This is
particularly important for the determination of
unreasonable risk for ONUs.
One member noted that the availability of workplace
measurements is low and that dependence upon modeling
is, therefore, high in this draft risk evaluation. The
Agency's hesitance to use its authority to request industry
data was again noted by the Committee.
PUBLIC COMMENTS:
If employers did not voluntarily provide monitoring data,
EPA has the authority to compel its production under
TSCA section 8 or to issue subpoenas for "the production
of... documents ... that the administrator deems
necessary" under section 11. In the event that no
monitoring data exist for a condition of use, EPA can order
the generation of such data under TSCA section 4. TSCA
requires EPA to conduct risk evaluations based on
"reasonably available" information, including information
that EPA "can reasonably generate, obtain, and synthesize
for use in risk evaluations." EPA must acquire and
consider that available data, using its TSCA information-
gathering authority to the extent needed
EPA did not find additional reasonably available information
for these sources. EPA requested information on all aspects of
risk evaluations throughout the risk evaluation process,
including opening public dockets for receipt of such
information, conducting outreach to manufacturers,
processors, users and other stakeholders. The data received
have undergone review and interpretations in the risk
evaluation document. In addition, data available from the
peer-reviewed literature are also included in the risk
evaluation document.
EPA had sufficient information to complete the carbon
tetrachloride risk evaluation using a weight of scientific
evidence approach. EPA selected the first 10 chemicals for
risk evaluation based in part on its assessment that these
chemicals could be assessed without the need for regulatory
information collection or development. When preparing this
risk evaluation, EPA obtained and considered reasonably
available information, defined as information that EPA
possesses, or can reasonably obtain and synthesize for use in
risk evaluations, considering the deadlines for completing the
evaluation. In some cases, when information available to EPA
was limited, the Agency relied on models; the use of modeled
data is in line with EPA's final Risk Evaluation Rule and
EPA's risk assessment guidelines.
Page 77 of 210
-------
EPA should conduct dermal exposure monitoring in
representative workplaces.
SACC,
26, 38,
45
SACC COMMENTS:
One Committee member suggested that EPA partner with
the OSHA or the NIOSH to get more sensitive sampling
and analytical methods in use in order to allow estimation
of exposure concentrations closer to 40 ppb or lower, to
allow for actual detected levels. Another Committee
member suggested that the amended TSCA law is a
mechanism for starting a new discussion on occupational
exposure measurement and that the PEL framework is no
longer appropriate.
PUBLIC COMMENTS:
EPA does not discuss its consultation or coordination with
the OSHA on risks to ONUs. TSCA Section 9(a)
contemplates consultation between EPA and OSHA and
authorizes OSHA to decide whether it agrees with EPA's
risk determination concerning worker health. EPA must be
more transparent in its risk evaluations about its
consultations with OSHA.
The CC14 PEL is 50 years old and universally
acknowledged to be unprotective. OSHA promulgated the
PEL for CC14 in 1971 based on research performed during
the 1950s and 1960s, and largely based on acute health
effects. In 1989, OSHA finalized an updated PEL for
CC14: a 2-ppm 8-hour TWA limit. For the original OSHA
limit of 10 ppm, the cancer risk estimate for CC14 was 17.9
excess deaths per 1,000 exposed workers. Even at the limit
of 2 ppm, the predicted risk is 3.7 excess deaths per 1,000
workers. The rule was subsequently vacated by the
Eleventh Circuit. As a result of this decision, the OSHA
EPA engages with all its federal partners as it works to
conduct and refine its risk evaluations. In the 2017
Procedures for Chemical Risk Evaluation Under the
Amended TSCA (82 FR 33726, July 20, 2017), EPA
committed to, by codifying, interagency collaboration to give
the public confidence that EPA will work with other agencies
to gain appropriate information on chemical substances. This
is an ongoing deliberative process and EPA is not obligated to
provide descriptions of predecisional and deliberative
discussions or consultations with other federal agencies. In
the interest of continuing to have open and candid discussions
with our interagency partners, EPA is not intending to include
the content of those discussions in the risk evaluation.
Comparison to the PEL is illustrative only for the purposes of
discussing engineering and administrative controls. OSHA's
Respiratory Protection Standard (29 CFR ง 1910.134)
requires employers in certain industries to address workplace
hazards by implementing engineering control measures and,
if these are not feasible, provide respirators that are applicable
and suitable for the purpose intended. Engineering and
administrative controls must be implemented whenever
employees are exposed above the PEL. If engineering and
administrative controls do not reduce exposures to below the
PEL, respirators must be worn. Respirator selection
provisions are provided in ง 1910.134(d) and require that
appropriate respirators are selected based on the respiratory
hazard(s) to which the worker will be exposed and workplace
and user factors that affect respirator performance and
reliability.
Page 78 of 210
-------
PEL for CC14 remains at 10 ppm, the level adopted in
1971.
SACC,
26, 38
SACC COMMENTS:
Recommendation: Use measured OSHA data in the risk
evaluation to inform "high end" exposures.
Monitoring data from OSHA and/or NIOSH
inspections could be useful for informing exposure
levels. One Committee member commented that the
OSHA inspection data available online reports
measured workplace levels up to 39.5 ppm, which is
much higher than the "high end" exposure level
reported in the draft risk evaluation.
PUBLIC COMMENTS:
For most other conditions of use, EPA did not seek or
receive any monitoring data; however, this does not mean
that such data do not exist.
OSHA requires employers to preserve and maintain
employee exposure records for thirty years. A quick
search of OSHA Chemical Exposure Health Data tool
yielded 321 air samples for CC14 collected as recently
as March 2017.
OSHA's respirator standard also requires that
employers "evaluate the respiratory hazards at their
workplaces," including a quantitative determination of
potential exposures. If respirators were as widely used
as EPA assumes, employers would have significant
amounts of workplace exposure data that would be
reasonably available to EPA. If no such data exist, then
EPA's assumptions of widespread and health-
protective respirator use are wrong.
EPA must acquire all of the relevant OSHA data in order to
comply with the TSCA Section 26 requirement.
EPA is aware of the OSHA data and has reviewed over 300
data points for carbon tetrachloride in the OSHA CEHD. The
reasons for not using these data are the lack of clarity and data
quality on the conditions of use, the date of sampling, and/or
inconsistencies in the sample durations and results. Examples
included:
The samples reported as non-detects (ND) could be due to
the absence of carbon tetrachloride at the site making the
dataset not relevant to carbon tetrachloride;
All samples are short-term samples and not representative of
full-shift exposures;
Samples were collected prior to the Montreal Protocol and
CAA Title VI ban and could include exposures from
phased-out uses;
The condition of use could be a non-TSCA use; and
Sample results did not include sample times such that the
representativeness of operation and exposures are unknown.
The reported respiratory protection and other PPE usages in
workplace are included in the risk evaluation document with
relevant citations. EPA reviewed all relevant and reasonably
available OSHA data.
OSHA data are collected as part of compliance inspections at
various types of facilities. Certain industries are typically
targeted based on national and regional emphasis programs.
Other inspections may be prompted based on complaints or
referrals. As a result, OSHA data may underrepresent PPE
usage throughout the affected industry. Additionally, because
EPA uses the high-end exposure values to account for
Page 79 of 210
-------
uncertainties and variabilities in PPE usage, this is accounted
for in its unreasonable risk determinations.
EPA's approach for developing exposure assessments for
workers is to use reasonably available information and expert
judgment. When appropriate, in the risk evaluation, EPA will
use exposure scenarios both with and without engineering
controls and/or PPE that may be applicable to particular
worker tasks on a case-specific basis for a given chemical.
While EPA has evaluated worker risk with and without PPE,
as a matter of policy, EPA does not believe it should assume
that workers are unprotected by PPE where such PPE might
be necessary to meet federal regulations, unless it has
evidence that workers are unprotected. For the purposes of
determining whether or not a condition of use presents
unreasonable risks, EPA incorporates assumptions regarding
PPE use based on reasonably available information and
professional judgment underlying the exposure scenarios.
These assumptions are described in the unreasonable risk
determination for each condition of use, in section 5.2.
Additionally, in consideration of the uncertainties and
variabilities in PPE usage (e.g., the burden associated with the
use of supplied-air respirators, including the expense of the
equipment and the necessity of fit-testing and training for
proper use), EPA uses the high-end exposure value when
making its unreasonable risk determination in order to address
those uncertainties. EPA has also outlined its PPE
assumptions in section 5.1. Further, in the final risk evaluation
for carbon tetrachloride, EPA has determined that most
conditions of use pose an unreasonable risk to workers even
with the assumed PPE.
Page 80 of 210
-------
SACC
SACC COMMENTS:
A Committee member found a potentially useful
biomonitoring study conducted in Italy (Ghittori et al.,
1994) that is not cited in the draft risk evaluation. That
study collected both environmental and biomarker data for
55 workers exposed to CC14 and potentially could provide
a check on exposure estimates in the risk evaluation.
EPA reviewed the submitted study and incorporated the
exposure monitoring data for the use of carbon tetrachloride
in the revised risk evaluation document. Appropriate citation
and interpretations also included in the risk evaluation
document.
26, 38
PUBLIC COMMENTS:
EPA determined that CC14 presents no unreasonable risk to
workers despite having no exposure data for many
conditions of use and inadequate data for the others. EPA
violated its statutory obligation to consider "reasonably
available information" when evaluating chemical risks.
For CC14 manufacturing, EPA relied exclusively on
exposure data voluntarily submitted by the HSIA.
HSIA's data cover only two manufacturing facilities, a
small fraction of the facilities that manufacture or
process CC14. HSIA did not provide information about
the conditions under which these samples were taken or
the sampling protocols and methodology.
EPA also used this HSIA manufacturing data as a
surrogate to estimate occupational exposures from the
processing of CC14 as a reactant, despite
acknowledging that manufacturing data "are not
directly applicable to processing of CC14 as a reactant."
EPA relied on the HSIA data without questioning its
reliability or representativeness. EPA provides no
justification for its exclusive reliance upon this
potentially biased data without independent validation
and quality assurance reporting.
The data gathering effort to support the risk evaluation was
performed by literature searches and leveraging existing
industry-specific information. HSIA data were provided as
part of continuous industrial hygiene monitoring programs
and were evaluated using the same criteria as other data sets.
The reasonably available data readily attributable to
manufacturing and processing of carbon tetrachloride were
limited and contained their own deficiencies (such as the age
of the studies, lack of discrete data points, and no metadata
information) resulting in low quality ratings. Additionally,
limited exposure data exists due to manufacturing, processing,
and use restrictions enforced under the Montreal Protocol,
CAA Title VI, and the Consumer Product Safety Commission
ban.
29
PUBLIC COMMENTS:
Each facility that manufactures CC14 performs a
documented Qualitative Exposure Assessment in which
EPA does not assess worker exposure through Similar
Exposure Groups (SEGs) because EPA does not have
information available to determine these groups based on the
Page 81 of 210
-------
tasks are assessed and characterized. Components of the
Qualitative Exposure Assessment include full-shift
exposure description, a description of each task that may
contribute to the overall full-shift exposure, and the
frequency/duration/PPE/controls for each task.
Each facility divides employees into Similar Exposure
Groups (SEGs), groups of workers having the same
general exposure profile because of the similarity and
frequency of the tasks performed, materials used,
processes, and controls.
Monitoring data collected for each exposure group is
analyzed to determine the overall exposure potential. If
the 95th percentile analysis results are below the
applicable Occupational Exposure Limits (OELs), then
the exposures are considered acceptable and periodic
monitoring/reassessments are performed to confirm/
validate. Any individual sample results exceeding
applicable OELs is investigated to determine cause(s)
and mitigated.
provided worker activity descriptions. Facility personnel
conducting the monitoring intimately know the facility and
can interview workers to determine SEGs. Additionally,
worker activities and job titles are determined differently at
each facility making an equal comparison very difficult;
therefore, EPA has relied only on designations between
workers and ONUs.
45
PUBLIC COMMENTS:
EPA should delineate a tiered human or environmental
exposure modeling approach for TSCA draft risk
evaluations. This approach will allow EPA to identify and
focus on uses that are high exposure and devote more
resources to determining potential risk presented by those
uses. We propose the Tank Truck and Railcar Loading and
Unloading Release and Inhalation Exposure Model as a
screening level exposure assessment for all occupational
conditions of use that use closed systems. Any conditions
of use that indicate high risk would move to further
analyses and data to confirm high risk levels.
The use of the Tank Truck and Railcar Loading and
Unloading Release and Inhalation Exposure Model as a
screening level tool to determine if further analyses are
required, as suggested in the comment, is inappropriate. This
model only accounts for exposures during the
loading/unloading of bulk containers which is likely only a
portion of the workday and may underestimate total
exposures as described in the uncertainties section of the risk
evaluation document. This estimate could be appropriate for
certain conditions of use where the chemical is primarily used
in closed systems such that the unloading activity is expected
to be the primary exposure activity. However, there could be
other conditions of use where the chemical is used in open
systems that results in significantly higher levels of exposure
Page 82 of 210
-------
than estimated by the Tank Truck and Railcar Loading and
Unloading Release and Inhalation Exposure Model.
31
PUBLIC COMMENTS:
The SACC should consider whether it is appropriate for
EPA to estimate inhalation exposure with a modeling
approach using the Tank Truck and Railcar Loading and
Unloading Release and Inhalation Exposure Model when
monitoring data from tank truck and railcar loading and
unloading are available.
EPA updated the risk evaluation of carbon tetrachloride to
assess the worker exposure during import and/or repackaging
of carbon tetrachloride from the tank truck and railcar loading
and unloading data identified from the monitoring data
submitted by HSIA. Fifteen of the 356 submitted data listed
worker activities for the unloading and/or loading of carbon
tetrachloride into tank trucks or railcars. For this assessment,
EPA only considered the 8-hr TWA data as information to
substantiate 12-hr shifts at repackaging sites were not
identified. Additionally, EPA only used data points if the
worker activities were specifically for carbon tetrachloride
loading/unloading.
ONU exposure estimation: methods, models, and data
SACC
SACC COMMENTS:
Recommendations: Attempt estimation of ONU exposures
where data permit as a check on default assumption of
mean worker exposure.
Consider a hierarchy of ONU exposures to distinguish
extremes within that classification.
EPA's description of the approach and assumptions for
deriving ONU's exposure estimates are adequately
transparent; however, scientific validity is questionable
because the uncertainties, while well described, are
considerable (due in part to data scarcity).
The Agency could use the job categories classified as
ONUs and additional considerations to derive ONU
exposure estimates. In addition, it should be possible to
use exposure or area modeling for at least some of the
conditions of use (for which EPA has, or can request,
data), as a comparison check for exposure estimates.
EPA used a subset of worker data to assess ONU exposure
where appropriate.
EPA has included appropriate modeling considering the
available data.
In the 'Uncertainties' Section 4.3.2.1, the revised document
included: "ONUs are likely a heterogeneous population of
workers, and some could be exposed more than just
occasionally to high concentrations."
Page 83 of 210
-------
One member suggested that the number of sites
actively using CC14 was not so large that EPA could
not request or attempt data collection in a meaningful
sample.
SACC,
22, 29,
39
SACC COMMENTS:
One SACC member requested clarification as to
whether HSIA data that appear to be pertinent to ONU
exposures had been received by EPA and used in the
draft risk evaluation. Those data involve 7 full shift
samples collected from administrative/supervisory
personnel, so it is likely that these are ONU samples.
Concentrations were <0.063-0.066 ppm (under the
LOD of the method).
One Committee member noted that public commenters
indicated that there are monitoring data relevant for
ONUs in the manufacturing and processing sectors that
may be useful to the Agency to consider.
Recommendation: Explain why it was decided not to use
the HSIA administrative/supervisory personnel data, even
if only to compare them to the exposure estimates for
ONUs. There is concern that this may be construed as data
selection bias.
PUBLIC COMMENTS:
Monitoring data on workers at CC14 production facilities
were submitted to EPA as part of HSIA's comments on the
CC14 problem formulation document. The data included
personal breathing zone measurements from both workers
and ONUs.
EPA did not note that certain exposure groups (z. e.,
process supervisors, electricians, utilities control board
technicians) were ONUs and wrongly concluded that
exposure data for ONUs were unavailable.
EPA received the additional information from HSIA to
denote ONU exposure data (EPA-HQ-OPPT-2019-0499-
0022) and has incorporated the ONU data into the risk
evaluation for carbon tetrachloride. These data were used in
the draft risk evaluation but were previously grouped with
worker exposure data. As recommended by the SACC
comment, EPA has revised the assessment to separate these
data from the worker exposure estimate and used these data to
assess ONUs. A total of 17 datapoints were included as ONU
exposure data according to the additional comment provided
by HSIA, EPA-HQ-OPPT-2019-0499-0022. These data
include the 7 full shift samples mentioned in the range
<0.063-0.066 ppm. The HSIA data denoting exposure for
administrative/supervisory personnel data were included in
the ONU exposure assessment for manufacturing.
The ONU exposure data identified by the commenter are not
all non-detect values. However, approximately 60% of the
identified data is below the level of detection. To estimate
exposures from these data, EPA used the Guidelines for
Statistical Analysis of Occupational Exposure Data, which is
summarized in Section 1.4.4.2 of the Supplemental
Information on Releases and Occupational Exposure
Assessment.
For scenarios where ONU data is unavailable, EPA assessed
ONU exposures at the worker central tendency. The
uncertainties of this approach are described in Section 4.4.1
Page 84 of 210
-------
In response to this oversight, HSIA submitted to EPA
ONU monitoring data of 17 breathing-zone full shift
samples showing that exposures are below the
detection limit (<0.063 to <0.21 ppm). The detection
limit provided is likely still much higher than actual
exposures since the evidence is based entirely on non-
detects.
These data demonstrate that using the Workers'
Central Tendency exposure concentration as a
surrogate for ONUs is overly conservative.
EPA should use the ONU monitoring data that were
provided in the docket and consider using the central
tendency estimate of the ONU data. This approach would
still be considered a conservative estimate since it is based
entirely on non-detect concentrations.
of the risk evaluation and such estimates are categorized as
"low confidence."
38
PUBLIC COMMENTS:
The broad range of workers are EPA defines as ONUs is
too large to support any single classification. Under EPA's
definition, ONUs may include cleaning workers, skilled
trade workers, supervisors, and managers. But supervisors
have very different exposure patterns than skilled trade
workers and cleaning workers, and thus face very different
risks from CC14.
EPA uses the central tendency (50th percentile) of worker
inhalation exposures to calculate ONU risks, as opposed to
collecting ONU-specific data or using the higher end
exposure estimates that EPA uses for other workers.
Particularly over a short period (e.g., response to a spill or
equipment maintenance), ONU exposures may be as great
as or greater than other workers, and ONUs are even less
likely to be provided PPE. EPA's failure to collect ONU-
EPA included ONUs who are defined in section 2.4.1 as
"working in the general vicinity of workers but do not handle
chemical substances and do not have direct dermal contact
with chemicals being handled by the workers." Maintenance
staff, cleaning workers, and skilled trade workers are a subset
of ONUs and as such are not excluded from the risk
evaluation.
EPA considers ONUs to be a subset of workers for whom the
potential inhalation exposures may differ based on proximity
to the exposure source. For the majority of carbon
tetrachloride conditions of use, the difference between ONU
exposures and workers directly handling the chemical cannot
be quantified. EPA assumed an absence of PPE for ONUs,
since ONUs do not directly handle the chemical and are
instead doing other tasks in the vicinity of carbon
tetrachloride use. EPA assumed that, in most cases, ONU
Page 85 of 210
-------
specific data and its reliance on central tendency exposure
estimates understates the risks to ONUs.
EPA assumes that ONUs will have no dermal exposures,
an assumption that is unfounded for cleaning workers and
skilled trade workers.
inhalation exposures are assumed to be lower than inhalation
exposures for workers directly handling the chemical
substance. For dermal exposures, EPA assumed that ONUs
do not have direct contact with carbon tetrachloride;
therefore, non-cancer effects and cancer from dermal
exposures from carbon tetrachloride generally were not
assessed.
22,31,
39
PUBLIC COMMENTS:
The SACC should review EPA's assumption that ONUs
are exposed at the central tendency exposure concentration
(50th percentile) of workers for manufacturing and
processing uses. This assumption is overly conservative
and not supported by the ONU personal exposure
monitoring data submitted to EPA for manufacturing and
processing uses.
Using the central tendency value implies that ONU
exposures are 4-fold lower than those of workers in the
near-field. This implication does not align with the data
provided to EPA.
To the extent there are residual data needs for ONUs, a
more appropriate approach to estimate ONU exposures
is the use of ONU-specific exposure models. A cursory
evaluation of near and mid-field plume model shows a
large drop off in concentration with distance. Use of
the same generation rate and air speed calculated for
the Tank Truck and Railcar Loading and Unloading
Release and Inhalation Exposure Model in the near-
field plume model results in a nearly 50-fold reduction
in concentration at a distance of 0.1-1 meter from the
source.
Improvement of the assumptions regarding the mid-
and far-field exposures would have a major impact on
the risk characterization for cancer inhalation.
Where EPA had monitoring or modeled data specific to
ONUs, unreasonable risk determinations were made based on
high-end exposures. For conditions of use where the data did
not distinguish between worker and ONU inhalation
exposures, there was uncertainty regarding ONU exposure.
ONU personal exposures are assumed to be lower than
personal exposures for workers directly handling the
chemical substance. To account for this uncertainty, EPA
considered the workers' central tendency risk estimates from
inhalation exposures when determining ONUs' unreasonable
risk (rather than the high-end inhalation exposures), when
data specific to ONUs was not available.
ONU distance from users are accounted in the uses with
Near-Field/ Far-Field modeling, which is superior to a
method that would use the inverse square law. EPA does not
have a method to account for air exchange rates for potential
use of the inverse square law nor reasonably available data or
information to estimate distance of ONUs from users in the
other assessed uses.
Page 86 of 210
-------
The SACC should consider whether the findings of
unreasonable risks for ONUs are appropriate given that
they are based on the application of worker inhalation
monitoring data to ONUs.
26
PUBLIC COMMENTS:
EPA underestimated exposure to ONUs by assuming
ONUs experience the central tendency exposures
calculated for workers in the absence of PPE because EPA
does not have any monitoring data or modeling specific to
ONUs.
EPA considers ONUs to be a subset of workers for whom the
potential inhalation exposures may differ based on proximity
to the exposure source. For the majority of carbon
tetrachloride conditions of use, the difference between ONU
exposures and workers directly handling the chemical cannot
be quantified. EPA assumed an absence of PPE for ONUs,
since ONUs do not directly handle the chemical and are
instead doing other tasks in the vicinity of carbon
tetrachloride use. EPA also assumed that, in most cases, ONU
inhalation exposures are assumed to be lower than inhalation
exposures for workers directly handling the chemical
substance. For dermal exposures, EPA assumed that ONUs
do not have direct contact with carbon tetrachloride;
therefore, non-cancer effects and cancer from dermal
exposures from carbon tetrachloride generally were not
assessed. To account for those instances where, based on
EPA's analysis, the monitoring data or modeling data for
worker and ONU inhalation exposure could not be
distinguished, EPA considered the central tendency risk
estimate when determining ONU risk.
22, 29,
39
PUBLIC COMMENTS:
ONUs are protected from exposure by engineering and
administrative controls.
If an ONU is in the immediate work environment when
a worker is required to use specific PPE for a task, the
ONU is required to use the same PPE as the worker.
Employees, contractors, and visitors are not allowed in
manufacturing areas without appropriate PPE and
safety training. Locker rooms and lunchrooms are
EPA considers ONUs to be a subset of workers for whom the
potential inhalation exposures may differ based on proximity
to the exposure source. For the majority of carbon
tetrachloride conditions of use, the difference between ONU
exposures and workers directly handling the chemical cannot
be quantified. EPA assumed an absence of PPE for ONUs,
since ONUs do not directly handle the chemical and are
instead doing other tasks in the vicinity of carbon
tetrachloride use. EPA also assumed that, in most cases, ONU
Page 87 of 210
-------
located outside the manufacturing areas. No food or
drinks are allowed in manufacturing areas.
The CC14 production process is a closed system
located in an outdoor area. The only production tasks
that are not closed system involve pulling samples and
collecting waste from the process. These are short,
intermittent tasks (15-30 minutes) performed in the
production area by trained employees, wearing
appropriate PPE.
For maintenance employees performing tasks outside
the production area, a perimeter is established with a
barricade providing a buffer around the area. Real-time
monitoring is done to ensure the buffer prevents
exposure for employees working around the production
area. Anyone working inside the barricaded area must
wear appropriate PPE.
The assumption of significant exposures in the absence
of respiratory protection is not consistent with current
industrial practice.
Even if an ONU is in the general work area, it is
unlikely that an ONU would be there for a full shift.
inhalation exposures are assumed to be lower than inhalation
exposures for workers directly handling the chemical
substance. For dermal exposures, EPA assumed that ONUs
do not have direct contact with carbon tetrachloride;
therefore, non-cancer effects and cancer from dermal
exposures from carbon tetrachloride generally were not
assessed. To account for those instances where, based on
EPA's analysis, the monitoring data or modeling data for
worker and ONU inhalation exposure could not be
distinguished, EPA considered the central tendency risk
estimate when determining ONU risk.
26
PUBLIC COMMENTS:
EPA underestimated exposure to ONU by assuming ONUs
are only present in the "far field zone." ONUs may not stay
within the "far field zone" when they are responding to
spills, maintaining equipment, and otherwise performing
work activities that take them within the "near-field"
workers zone. ONUs may regularly pass into each other's
space to communicate or otherwise interact.
The evaluation of carbon tetrachloride exposure to ONUs
does not use any near-field/far-field models in the evaluation.
Dermal exposure assumptions
SACC
SACC COMMENTS:
The SACC report includes a list of dermal parameters that
are recommended for inclusion in the table of physical-
The Section 2.4.1.8 (Dermal Exposure Assessment) has been
updated with inclusion of a conceptual diagram (Figure 2-4),
and several dermal exposure scenarios of carbon tetrachloride
Page 88 of 210
-------
chemical properties or elsewhere in the risk evaluation.
These include aqueous permeability coefficients, relative
permeability of the stratum corneum to that of the viable
epidermis, theoretical maximum steady-state flux,
octanol/air partition coefficient, stratum corneum/gas
partition coefficient, dermal vapor to inhalation dose ratio
(measured and modeled), and observed absorption flux.
Descriptions, rationales, and references for each parameter
are provided in the SACC report.
A Committee member reported that EPA was again
using a percent absorbed approach based on the Frasch
(2012) interpretation of the work by Kasting and Miller
(2006). The Agency had previously switched to the
Frasch and Bunge (2015) paper, which deals with
absorption of the "skin depot" (post exposure) rather
than the initial load. The Committee did not verify that
the numerical results are correctly computed, but the
change in approach is appropriate.
One member noted that for VOCs, in the absence of
PPE, inhalation would be expected to dominate dermal
vapor exposure. However, if respiratory protection, but
not whole-body vapor protection is provided, dermal
vapor exposure can exceed (PPE-reduced) inhalation
exposure. For instance, if the ratio of inhalation dose to
dermal vapor dose is 10 and an APF of 25 is assumed,
the dermal vapor dose becomes the dominant exposure
pathway.
from the IHSkinPermฎ (developed by American Industrial
Hygiene Association) output using the physical-chemical
properties are summarized in Table 2-23. Description of the
conceptual diagram, synopsis of existing tools/models,
interpretations, and citations of references are also included in
the risk evaluation document.
SACC
SACC COMMENTS:
Experience in the occupational agriculture sector does
suggest that hands are disproportionately exposed. The
hand area data were obtained from the Exposures Factors
The Section 2.4.1.4 of risk evaluation document already
clarified the basis for contact surface area of 1,070 cm2 as an
input parameter for estimating high-end dermal exposure to
Page 89 of 210
-------
Handbook, which in turn derived the estimates from the
Center for Disease Control and Prevention (CDC) National
Health and Nutrition Examination Survey (NHANES)
data. The Agency could use distributed values for hand
surface area as those are available. Section 2.4.1.8 presents
a good discussion. However, EPA does not have
comparable data for many other assumptions.
liquids. This clarification also included that the above value is
equivalent to the 50th percentile surface area of two-hands for
males, the highest exposed population. EPA has no
reasonably available information on actual surface area of
contact with liquid and that the value is assumed to represent
an adequate proxy for a high-end surface area of contact with
liquid that may sometimes include exposures to much of the
hands and also beyond the hands, such as wrists, forearms,
neck, or other parts of the body, for some scenarios. The
above statement also has been included in the Section 2.4.1.4
of risk evaluation document.
43
PUBLIC COMMENTS:
EPA should model a broader range of dermal contact
scenarios based on its own analysis of variations in dermal
exposure conditions.
Based on the variety of number of potential worker exposure
scenarios, EPA considered a general dermal exposure
scenario and used parameters that provide a conservative
estimate.
32, 43
PUBLIC COMMENTS:
The basis for the dermal assessment was highly uncertain
because of the limited data available.
Without test data on dermal absorption rates, EPA
assumed that "the calculated retained dose is low for
all dermal exposure scenarios as CC14 evaporates
quickly after exposure." EPA estimated that
"approximately four percent of the applied dose is
absorbed through the skin" where no gloves are worn
and considerably less in instances of glove use.
EPA assumed no dermal exposure by ONUs.
As EPA acknowledged, rapid volatilization after skin
contact would not occur in all situations; repeated skin
contact with chemicals could have even higher than
expected exposure if evaporation of the chemical
occurs and the concentration of chemical in contact
with the skin increases; wearing of gloves could have
important consequences for dermal uptake; and without
These assumptions were primarily based on the EPA 2-Hand
Dermal Contact with Liquid Model, which is generally
consistent across all risk evaluations. EPA did not find
reasonably available empirical data to develop alternate
estimates of dermal exposure.
Page 90 of 210
-------
any gloves, a splash of the liquid or immersion of the
hand may overwhelm the skin contamination layer .. .if
it is undiluted, then uptake could proceed rapidly.
EPA did not develop alternate estimates of dermal
exposure showing higher levels of absorption in these
scenarios.
26, 32,
38, 43
PUBLIC COMMENTS:
EPA provides little justification for the assumption of a
single dermal exposure event per day. It seems likely that
workers would regularly engage in activities that could
result in multiple exposure events per day. EPA
acknowledges that this assumption "likely underestimates
exposure" but did not to consider those risks or provide
any sort of uncertainty analysis. This is an admitted
violation of TSCA EPA should base dermal exposure
scenarios in the final CC14 evaluation on an assumption of
multiple exposure events per day.
EPA has described events per day (FT) as one of the
uncertainties for dermal modeling in the discussion of
occupational dermal uncertainties (Section 4.4.1). This
discussion also included that the assumption on the number of
events likely underestimates exposure as workers could have
repeated contacts with carbon tetrachloride throughout their
workday.
38
PUBLIC COMMENTS:
EPA improperly assumes worker exposures to CC14
terminate "at the end of the task, shift, or work day." EPA
offers no evidence that all workers clean hands and other
exposed body parts following each shift. In the absence of
cleaning, dermal exposure durations - and associated risks
- may be greater than those estimated by EPA. Clothing
can absorb CC14, and many workers return home in the
same clothes they wear at work. This absorption creates
that potential for additional "take home" exposures that
EPA has not addressed in its draft risk evaluation.
The frequency and magnitude of take-home exposure is
dependent on several factors, including personal hygiene,
good laboratory/industrial practices, and extent and visibility
of the chemical on skin or clothing. EPA does not have
methods to reliably predict take-home exposure.
39
PUBLIC COMMENTS:
Considering the conservatism in the dermal exposure
assumptions, the likely actual estimates for dermal cancer
risk would be below the lxlO"4 benchmark. For most tasks
that involve dermal exposure in chemical manufacturing
EPA has included an explanation of the dermal exposure
assessment parameter assumptions in the Section 2.4.1.4.
EPA stated that the value for the contact surface area is
equivalent to the 50th percentile surface area of two-hands for
males, the highest exposed population. EPA has no
Page 91 of 210
-------
(e.g., sampling a process line or hooking up a transfer
line), there is no likely routine skin contact and certainly
not hours each day. In most routine tasks with any liquid
present, chemical-protective gloves would be used. Any
liquid spills will land on the outside of a glove and largely
evaporate. The full hand surface (or two full hands) would
never be covered with liquid under any normal routine
scenario.
reasonably available dermal exposure data, including
information on actual surface area of contact with liquid.
Exposure uncertainty discussion/confidence ratings
SACC
SACC COMMENTS:
Recommendation: Levels of confidence should be
provided for each route of occupational exposure, in
addition to the overall result.
One member suggested that levels of confidence
should be provided for each route of exposure, in
addition to the overall confidence. In addition, that
member thought that the text should include a more
extensive summary discussion of confidence.
The Committee was of mixed opinion on the merits of
the graphical depiction of the confidence ratings in
Table 2-19. One Committee member commented that
the scale and use of color implies a quantification that
the Agency does not have.
One member noted that Section 4.4.1 does not describe
uncertainties of exposure estimates derived with the
Tank Truck and Railcar Loading and Unloading
Release and Inhalation Exposure Model.
EPA included confidence ratings for both dermal and
inhalation exposure routes.
The graphical depictions of confidence levels are appropriate
as these are qualitative (high, medium, and low are
inappropriate). One SACC member supported the usage of
"higher" and "lower" bands with qualitative markings instead
of arbitrary high, medium, or low assignments.
EPA added a discussion of uncertainties for Modeling
Inhalation Exposures with the Tank Truck and Railcar
Loading and Unloading Release and Inhalation Exposure
Model to Section 4.4.1 of the revised risk evaluation
document.
SACC
SACC COMMENTS:
Occupational exposure data were often not of adequate
quality to support the draft risk evaluation.
Measurements are usually reported as non-detectable
because monitoring methods are typically keyed to the
PEL. Most of the occupational exposure measurements
EPA reviewed information on all aspects of risk evaluations
throughout the risk evaluation process, including NIOSH and
OSHA data. Multiple NIOSH studies are described in the
Supplemental Information on Releases and Occupational
Exposure Assessment, but were not included in any risk
evaluations due to lack of information about the study
Page 92 of 210
-------
used in the draft risk evaluation occurred in scenarios
that are considered well-controlled and for which more
data are available. NIOSH and OSHA measurements
have been useful for evaluating effects on worker
health.
One Committee member suggested that non-detectable
values (below detection limit values) are not so much
inadequate as insufficiently informative for the task of
estimating exposures.
The open burning/open detonation data, which are
below the LOD raise the issue of whether large data
sets with low results are necessarily superior to smaller
data sets with values of the LOD. EPA should clarify
how it assesses the relative merits of data set size and
quality.
Recommendation: Discuss limitations and inadequacies of
occupational monitoring data collected to meet PEL
standards rather than assess relevant health effects.
(number of samples) and lack of temporal representation
(data were collected before the Montreal Protocol and could
misrepresent current worker conditions).
EPA is aware of the OSHA data and has reviewed over 300
data points for carbon tetrachloride in the OSHA CEHD. The
reasons for not using these data are the lack of clarity and
data quality on the conditions of use, the date of sampling,
and/or inconsistencies in the sample durations and results.
Examples included:
The samples reported as non-detects (ND) could be due to
the absence of carbon tetrachloride at the site making the
dataset not relevant to carbon tetrachloride;
All samples are short-term samples and not representative
of full-shift exposures;
Samples were collected prior to the Montreal Protocol and
CAA Title VI ban and could include exposures from
phased-out uses;
The condition of use could be a non-TSCA use; and
Sample results did not include sample times such that the
representativeness of operation and exposures are
unknown.
Regarding non-detectable values and OB/OD, EPA used
established protocols to evaluate occupational exposure data
that were reported as below the LOD. This approach has been
used consistently across the Risk Evaluations and is
summarized in Section 1.4.4.2 of the Supplemental
Information on Releases and Occupational Exposure
Assessment. For datasets including exposure data that were
reported as below the LOD, EPA estimated the exposure
Page 93 of 210
-------
concentrations for these data, following EPA's Guidelines for
Statistical Analysis of Occupational Exposure Data (1994)
which recommends using the LOD/20 5 if the geometric
standard deviation of the data is less than 3.0 and LOD / 2 if
the geometric standard deviation is 3.0 or greater (U.S. EPA,
1994V
SACC
SACC COMMENTS:
The current PEL for CC14 was not set based on the health
outcomes considered in the draft risk evaluation, but was
established many years ago. Lacking an adequate
biological basis for past exposure measures, it is important
that the risk evaluation emphasize the dependency of the
final risk determination on exposure estimates derived
from air concentrations measured as below detection
limits. Since exposures to ONUs are predicted using
worker exposure levels, the same qualifier applies to ONU
risk determinations. Because of this, the Committee
suggested that the risk evaluation should indicate that all
exposure estimates for ONUs are preliminary.
See above response on how exposure data under the LOD
were evaluated. In the Section 4.4.1 EPA discussed the
dependency of ONU exposure estimates on worker exposure
estimates and stated that there is high uncertainty in the
exposure estimations.
SACC
SACC COMMENTS:
A Committee member indicated that the modeling
estimates for exposure require additional analysis and
discussion in terms of uncertainty.
EPA expanded the discussion of the exposure model
uncertainties, as recommended, in Section 4.4.1 of the final
risk evaluation document.
SACC
SACC COMMENTS:
The risk evaluations continue to focus on the intrinsic
quality of the data (i.e., whether exposure sampling and
analytical methods were appropriate and fully reported)
as the single criteria for acceptability. The risk
evaluations have not focused on whether the amount of
data available is adequate to derive reliable estimates
of exposures for occupational scenarios in a condition
of use. The issue of deciding whether there are
EPA indeed considered the number and extent (amount) of
exposure data when making risk evaluations and how they
affect the data quality. When there is limited information
available, EPA acknowledged that there is a greater degree of
uncertainty in the exposures and noted that the data may not
be widely representative of an industry.
Page 94 of 210
-------
adequate samples for supporting exposure estimates
remains essentially unaddressed.
Recommendation: Address directly the issue of how many
and what kinds of samples are adequate to quantify
exposures for condition of use scenarios.
26
PUBLIC COMMENTS:
EPA invokes uncertainty as a basis for excluding
exposures, when the scientifically sound and health-
protective approach would be to include the exposures and
estimate the uncertainty.
EPA did not exclude any occupational exposures due to
uncertainties. Rather, decisions to exclude certain workplaces
were based on information provided by stakeholders and
regulatory bans under the CAA and CPSC.
11 iiin
-------
activities of various compounds have shown that
inflammation drives much of CC14-induced liver
injury. It is unclear why this aspect of CC14-induced
immunotoxicity was not included in the draft risk
evaluation.
It was noted that in the supplemental document
entitled, Inclusion/Exclusion Criteria for Human
Health Hazard Literature, it states that these types of
studies are to be included in the draft risk evaluation.
This does not seem to have happened for this
evaluation.
Recommendations: (1) Include discussion on CC14 effects
on inflammatory and immune effects; and (2) explain why
studies that evaluated the immune responses induced when
CC14 was used as a positive control for inducing liver
inflammation/fibrosis in animals were excluded from data
integration during systematic review.
cirrhosis, fibrosis, organ damage: liver, kidney, and others)
rather than evaluating adverse effects in animals from carbon
tetrachloride exposure . The former type of animal studies
was considered off-topic because it provides limited
applicability for dose-response in the risk evaluation. Also,
sufficient high quality on-topic human health references were
identified for carbon tetrachloride. Appendix B in the
Problem Formulation and section 1.5 of the final risk
evaluation describe the process used to re-screen human
health references for prioritizing the literature for
applicability in the risk evaluation.
30
PUBLIC COMMENTS:
Dermal irritation and sensitization should also be listed as
likely endpoints of concern. Since there are no studies that
evaluate the potential for reproductive effects, this
endpoint should NOT be cited on EPA's list.
The Human Health Hazard section and appendix G identify
irritation and sensitization as hazards associated with carbon
tetrachloride.
Although there are no reproductive toxicity studies for carbon
tetrachloride, observations of reproductive organ tissues in
repeated-dose studies provided some information on the
potential reproductive effects of carbon tetrachloride.
30
PUBLIC COMMENTS:
There is a body of literature on human exposure, both
controlled exposure and epidemiologic studies, that
provide credible information from which to derive acute
PODs and reference values.
Reasonably available information to inform PODs were
considered in the systematic review process for this risk
evaluation.
32, 43
PUBLIC COMMENTS:
To determine PODs for estimating risks, EPA relied on a
single flawed acute toxicity study (classified unacceptable
Due to the lack of reasonably available dermal studies
evaluating non-local or nonlethal effects from exposure to
carbon tetrachloride, the RE presents the alternative approach
Page 96 of 210
-------
in EPA's systematic review) for acute liver effects and
extrapolated a human equivalent dose (HED) for chronic
effects and carcinogenicity from inhalation studies since
no dermal data for these endpoints was available for CC14.
of extrapolating the acute dermal POD from the estimated
chronic dermal POD.
30
PUBLIC COMMENTS:
The chemical is clearly neurotoxic; this endpoint serves as
the basis for the derivation of the acute inhalation exposure
POD and benchmark MOE.
The risk evaluation states that the extrapolation of the acute
dermal POD from acute inhalation POD was not performed
because the critical acute inhalation effects of neurotoxicity
are influenced by the accessibility to brain tissue by inhaled
carbon tetrachloride.
Data used in the chronic noncancer assessment
SACC
SACC COMMENTS:
Adverse effects of CC14 on sperm function and
morphology have not been addressed in the human
health hazards section.
Some reproductive effects have been induced in rodent
studies (Smyth et al., 1936; Adams et al., 1952).
The Committee also referenced studies by El-Faras et
al. (2016) and Turk et al. (2016).
Recommendation: The reproductive toxicity of CC14
should be addressed and incorporated into the document.
The following statement was added to the RE: "As liver
toxicity is identified as the most sensitive effect from repeated
inhalation exposures to carbon tetrachloride, OPPT assumes,
that similarly to developmental toxicity, potential
reproductive effects from carbon tetrachloride exposure are,
at worst, secondary to liver toxicity. For instance, effects on
the reproductive organs (testes, uterus, etc.) have not been
observed in subchronic and chronic animal studies, which
suggest that carbon tetrachloride is not likely to be a
reproductive toxicant, and that any potential reproductive
effects could be only induced, at much higher dose
concentrations than liver toxicity."
El-Faras et al.., (2016) and Turk et al... (2016) studies were not
used to reach a conclusion on the developmental toxicity of
carbon tetrachloride because as explained above, studies in
which carbon tetrachloride was used as a positive control to
induce a disease state in an animal (e.g., cirrhosis, fibrosis,
organ damage: liver, kidney, and others) rather than
evaluating adverse effects in animals from carbon
tetrachloride exposure in animals were considered off-topic
because they provide limited applicability for dose-response
in the risk evaluation. Sufficient high quality on-topic human
Page 97 of 210
-------
health references were identified for carbon tetrachloride.
Appendix B in the Problem Formulation describes the process
used to re-screen human health references for prioritizing the
literature for applicability in the risk evaluation.
SACC
SACC COMMENTS:
Recommendation: Effects of CC14 on the CNS, in rodent
studies, should be addressed.
Several P450s are found in a highly regionalized and
cell-specific fashion in the brain (Navarro-Mabarak et
al., 2018). Furthermore, trans-sulfuration pathways
exist there (Vitvitsky et al., 2006).
There are several studies documenting effects of high-
level CC14 exposure on oxidative stress/lipid
peroxidation markers in the brain of rodents (Ritesh et
al., 2015; Naseem et al., 2014; Al-Olayan et al., 2016).
Acute toxicity studies in humans and animals reported
neurotoxic effects of carbon tetrachloride. The systematic
review process identified on-topic human health references
with human data containing qualitative and quantitative
information on the neurotoxicity effects (CNS depression) in
humans following acute exposures. Further consideration of
the reasonably available animal data was not necessary for the
risk evaluation of this endpoint of concern.
The metabolic activation of carbon tetrachloride by various
P450s found in a highly regionalized and cell-specific fashion
in the brain is a consideration in the discussion of the cancer
MOA presented in the final risk evaluation.
SACC
SACC COMMENTS:
One Committee member suggested inclusion of a more
comprehensive discussion of possible endocrine effects in
the CC14 risk evaluation. A brief summary of data and
references that provide support for an endocrine-related
MOA are provided in the committee report including:
JBRC (1998), Nagano et al. (2007), Colby (1981), and
Narotsky (1997).
Reasonably available information on the endocrine effects of
carbon tetrachloride were considered for hazard
identification. EPA used the approach described in section
Error! Reference source not found, of the final risk
evaluation to evaluate, extract and integrate carbon
tetrachloride's human health hazard and dose-response
information
SACC
SACC COMMENTS:
No justification is provided for why noncancer
endpoints, such as liver fibrosis, are not considered.
The risk evaluation should clearly state why the
noncancer endpoints, identified and discussed in
epidemiological studies, may be less relevant at the
low exposures being considered.
Recommendation: Include a discussion of noncancer
The identified sensitive endpoint of concern (i.e., fatty
changes in the liver, a precursor for liver fibrosis) is based on
the principal study for the derivation of the IRIS RfC is
(Nagano et al.. 2007), which consist of a chronic studv using
two species and preceded by a 13-week subchronic study.
This chronic study is rated of high quality in the systematic
review. Other key subchronic inhalation studies of acceptable
data quality supporting the identified endpoint of concern are
Page 98 of 210
-------
health endpoints from epidemiologic studies.
discussed in the RE.
The limited number of recent epidemiological studies
assessing non-cancer (i.e., Parkinson's disease, autism)
endpoints and with acceptable data quality do not show
association between exposure and non-cancer hazard effects
(see Table 3-1 in RE).
30
PUBLIC COMMENTS:
There are no studies, human or animal, that focus on
characterizing the potential for adverse effects on
reproduction or neurodevelopment. For both acute and
chronic exposures, at least one developmental toxicity
study is needed. For both short-term and chronic
exposures, a one- or two-generation reproductive toxicity
study is needed.
A more systematic evaluation of neurotoxicity and
developmental neurotoxicity is needed, since the worker
population includes women of childbearing age. Once the
risk evaluation is updated to include analyses of any
remaining legacy consumer conditions of use, infants and
young children become a subpopulation of concern.
The RE indicates that liver toxicity is identified as the most
sensitive effect from repeated inhalation exposures to carbon
tetrachloride. Based on the available developmental toxicity
data, developmental toxicity was not identified as the most
sensitive endpoint for inhalation or dermal exposures. OPPT
has concluded that potential reproductive effects from carbon
tetrachloride exposure are, at worst, secondary to liver
toxicity. For instance, effects on the reproductive organs
(testes, uterus, etc.) have not been observed in subchronic and
chronic animal studies, which suggest that carbon
tetrachloride is not likely to be a reproductive toxicant, and
that any potential reproductive effects could be only induced,
at much higher dose concentrations than liver toxicity.
Data used in the cancer assessment, animal and in vitro studies
SACC,
39
SACC COMMENTS:
The observation of significant increases in brain
tumors in multiple studies suggests that additional
examination of this as a potential target organ is
warranted.
However, members of the Committee noted that brain
tumors have not been reported in laboratory animal
bioassays of CC14; that a reported association between
ambient air concentrations and prevalence of
neuroblastomas in one study is not pertinent to this
EPA has conducted a critical and comprehensive evaluation
of the epidemiologic studies and a causal analysis with a
conclusion. EPA has added that evaluation in Section
3.2.4.2.2.
Regarding brain tumors in (Nagano et al. 2007). referred to
in the public comment as JBRC, was a study on F334 rats.
According to a review of 19 studies of the spontaneous
occurrence of astrocytoma in F334/DuCij rats (Nagatani et
al., 2013), the incidence was 0.6% in males and 0.2% in
Page 99 of 210
-------
issue (since neuroblastomas are not tumors of the
brain, contrary to what is stated in the draft risk
evaluation); and that the epidemiological studies
overall are too few and weak to be conclusive.
PUBLIC COMMENTS:
Neither brain toxicity nor brain tumors have been reported
in repeated-dose toxicity studies on CC14. The rat and
mouse 13-week and 2-year inhalation studies by the JBRC
did not find any treatment-related effects (cancer or
noncancer) associated with the brain or nervous system
tissue. Given that EPA considers the JBRC studies to be
of high quality and the basis for its cancer risk assessment,
it can be concluded that adequate data exist from animal
studies to evaluate whether CC14 exposure is associated
with an increased incidence of brain tumors.
females. The review also cites Haseman et al. (1990; 1998)
on the occurrence of astrocytoma in F334 rats as less than
1% in both sexes. At a background incidence of 0.4%, 250
rats would need to be in the control group to have an
expectation of a single brain tumor. To detect an increased
risk of brain tumors would require far more than the standard
group size of 50 per dose group. Given the rarity of
astrocytoma in F334 rats, it is unclear that the lack of
reported effects in (Nagano et al.. 2007) conflicts with the
epidemiologic evidence.
SACC
SACC COMMENTS:
One Committee member noted that the draft risk
evaluation did not appear to use in vitro studies. Another
member noted that caution should be used when
evaluating in vitro CC14 data given its volatility and that
common diluents used for in vitro studies (methanol,
ethanol, dimethyl sulfoxide [DMSO]) were competitive
inhibitors of CYP2E1 and may interfere with CC14
bioactivation in those systems.
In vitro studies were evaluated by EPA when synthesizing
and integrating evidence for human health hazards. EPA
considered quality, consistency, relevancy, coherence and
biological plausibility as specified in Application of
Systematic Review in TSCA Risk Evaluations.
39
PUBLIC COMMENTS:
The animal toxicity data on CC14 do not support brain
tumors being a health concern. The Ritash et al. (2015)
study used only a single oral dose, so information on
dose-response is lacking, including whether the effects in
the brain can occur at lower doses than in the liver.
Nevertheless, the acute oral dose is orders of magnitude
higher than doses that are expected to occur from realistic
Site concordance of tumors can be important evidence,
however site concordance is not always assumed. Brain
tumors are rare in both people and in rats. In F334 rats the
incidence is 0.4-0.5% (additional detail in response to SACC
comment #39). The lack of reported effects in animal studies
may not support the association reported in multiple
epidemiologic studies, but they do not refute those
observations.
Page 100 of 210
-------
human exposure to CC14, and therefore, the study is of
questionable relevance to EPA's risk evaluation.
39
PUBLIC COMMENTS:
The JBRC rodent inhalation bioassays on which the IUR
for CC14 is based were not adequately evaluated by EPA in
the risk evaluation, nor were new scientific data included
in the risk evaluation that provide important evidence for a
cytotoxic-proliferative MOA of CC14 at low doses.
The JBRC rodent inhalation bioassays are described in
(Nagano et aL 2007), which was found to have high data
quality in the systematic review for this risk evaluation. The
findings from the JBRC bioassays are used for cancer MOA
and cancer dose-response in both the IRIS assessment and
this risk evaluation.
39
PUBLIC COMMENTS:
Historical control data in Crj:BDFl mice from 20 studies
at JBRC suggests that the incidence of liver adenomas
was unusually low in control mice in the CC14 study, thus
exaggerating the statistical difference between the 5 ppm
females and the controls. The wide range in the historical
control range for liver adenomas may also indicate that
background rates for these tumors are highly variable.
The incidence of benign adrenal pheochromocytomas was
increased in males at 25 or 125 ppm and females at 125 ppm.
The incidences of hepatocellular adenomas and carcinomas
were elevated in both sexes at >25 ppm. At 5 ppm, the
incidence of liver adenomas in female mice (8/49 or 16%)
was statistically significantly elevated compared to the
concurrent control group and exceeded the historical control
range (2-10%).
The possibility that the increased incidence of liver
adenomas in the 5 ppm female mice is an experimental
artifact from an unusually low incidence of liver adenomas
in the control mice was explored by comparing the incidence
of liver adenomas in the study controls to the historical
laboratory control data. The incidence of liver tumors in
control mice (18% in males and 4% in females for
hepatocellular adenoma and 34% in males and 4% in females
for hepatocellular carcinoma) were similar to historical
control data for liver tumors in Cij :BDF1 mice in 20 studies
at the JBRC (Katagiri et aL 1998). Thus, the historical
control data from the laboratory seems to strengthen the
conclusion that the low dose female adenoma result is likely
compound related.
45
PUBLIC COMMENTS:
There is a considerable degree of variability in the rate of
The incidence of benign adrenal pheochromocytomas was
increased in males at 25 or 125 ppm and females at 125 ppm.
Page 101 of 210
-------
liver adenomas in control BDF1 mice. A separate analysis
of spontaneous liver tumors conducted by JRBC (1998)
reported a relatively high incidence of hepatocellular
adenomas in both sexes of BDF1 mice: specifically, up to
8% of females and 30% of males.
Similarly, Yamate et al. (1990) investigated the rate of
tumorigenesis in BDF1 mice allowed to live out their
lifespan and found that spontaneous hepatocellular
adenomas were common in both male and female mice of
this strain (7/50 males [14%] and 6/50 females [12%]).
EPA should acknowledge both the high rate and the
variability in the rate of spontaneous liver adenomas (and
carcinomas) in this strain of mice in its discussions of
Nagano et al. (2007) and of the plausible MO As for the
carcinogenicity of CC14.
The incidences of hepatocellular adenomas and carcinomas
were elevated in both sexes at >25 ppm. At 5 ppm, the
incidence of liver adenomas in female mice (8/49 or 16%)
was statistically significantly elevated compared to the
concurrent control group and exceeded the historical control
range (210%).
The possibility that the increased incidence of liver adenomas
in the 5 ppm female mice is an experimental artifact from an
unusually low incidence of liver adenomas in the control mice
was explored by comparing the incidence of liver adenomas
in the study controls to the historical laboratory control data.
The incidence of liver tumors in control mice (18% in males
and 4% in females for hepatocellular adenoma and 34% in
males and 4% in females for hepatocellular carcinoma) were
similar to historical control data for liver tumors in Cij :BDF1
mice in 20 studies at the Japan Bioassay Research Center
JBRC flCatagiri et al.. 1998). Thus, the historical control data
from the laboratory seems to strengthen the conclusion that
the low dose female adenoma result is likely compound
related.
39
PUBLIC COMMENTS:
While liver adenomas were increased in the 5 ppm-
exposed female mice in the absence of liver toxicity, this
increase was not statistically significant using the level of
significance (p<0.01) used by NTP and others; this
increase may been artifactual due to an unusually low
incidence of liver adenomas in the control mice.
Therefore, 5 ppm should be considered a no-observed-
effect-concentration NOEC for both liver toxicity and
liver cancer.
The significance of the 8/49 adenomas in the 5ppm dose
female group as compared with 2/50 in the matched controls
is P = 0.05, which is statistically significant in the IRIS
assessments and TSCA risk evaluations.
The study authors published a report on the historical control
incidence in these mice in their lab. CKatagiri et al.. 1998)
reports on spontaneous lesions in the BDF1 mice in 10
bioassays they had conducted. The number of female mouse
adenomas ranged from 1/50 to 4/50, with an overall
incidence of 4.4% as compared with the 8/49=16%) observed
in the low dose carbon tetrachloride females. Thus, the
Page 102 of 210
-------
historical control data from the lab seems to strengthen the
conclusion that the low dose female adenoma result is likely
compound related.
30
PUBLIC COMMENTS:
The Agency should use its enhanced testing authority in
the "new" TSCA to require submission of the studies of
reproduction, genotoxicity, developmental neurotoxicity,
and others relevant to MOA/AOP characterization.
For chronic exposures, studies that would adequately test
for carcinogenic potential by the relevant route(s) of
exposure or that could be extrapolated to those routes of
exposure are needed.
EPA had sufficient information to complete the carbon
tetrachloride risk evaluation using a weight of scientific
evidence approach. EPA selected the first 10 chemicals for
risk evaluation based in part on its assessment that these
chemicals could be assessed without the need for regulatory
information collection or development. When preparing this
risk evaluation, EPA obtained and considered reasonably
available information, defined as information that EPA
possesses, or can reasonably obtain and synthesize for use in
risk evaluations, considering the deadlines for completing the
evaluation.
The JBRC rodent inhalation bioassavs described in (Nagano
et al... 20071 were found to be high quality inhalation
bioassays in the systematic review for this risk evaluation.
The lack of chronic dermal studies is acknowledged in the
risk evaluation as a key source of uncertainty.
Data used in the cancer assessment, epidemiological studies
SACC,
23, 32,
33, 39,
43,45
SACC COMMENTS:
The risk evaluation should include the Heineman et al.
(1994) study as well as other epidemiological studies
on CC14 that may have investigated the occurrence of
gliomas, brain tumors, and other types of cancer.
Standard epidemiological approaches for extracting the
data should be employed with the full range of risk
estimates presented.
The Committee noted that Heineman et al. (1994)
reported that after adjusting for co-exposure to other
chlorinated aliphatic hydrocarbons, the association of
CC14 to brain cancer was no longer statistically
EPA has added a critical and comprehensive evaluation of the
epidemiologic studies of carbon tetrachloride and brain
cancer (including (Heineman et al... 1994)) and a synthesis of
the available evidence for carcinogenicity that takes into
account the considerable research in animals showing that
carbon tetrachloride can pass through the blood-brain barrier,
is rapidly absorbed by the brain and liver, causes oxidative
stress in the brain, and is metabolized in the brain. Evaluation
of the epidemiologic studies of carbon tetrachloride and brain
cancer included application of the Bradford-Hill
considerations as well as discussion of any potential biases
and the evidence integration weighed that evidence across the
Page 103 of 210
-------
significant and the odds ratio (OR) at the highest level
of exposure was actually decreased from the medium
exposure level.
One Committee member recommended that citation
and discussion of the older epidemiologic studies be
added to Tables 3-7 and 3-8 of the draft risk
evaluation. Though the Committee member
understands that the studies were part of the previous
evaluation, they appear to add weight of evidence for
the overall evaluation of the chemical.
Recommendations: (1) A critical and more comprehensive
evaluation of the reported associations between CC14 and
brain cancer is needed; and (2) expand the discussion of
the Heineman et al. (1994) study.
Recommendation: Revise the table listing epidemiologic
studies per the example given in the SACC report Table 2,
and apply Bradford-Hill criteria in assessing study
strengths. Endpoints to consider should be chosen a priori,
and then reported uniformly across studies.
PUBLIC COMMENT PART 1: Conclusions from
brain cancer studies are not reliable
Considering the risk of bias, lack of consistency, and high
contribution of chance and confounding, it was concluded
that the five studies by Nelson et al. (2012), Neta et al.
(2012), Heck et al. (2013), Ruder et al. (2013), and
Heineman et al. (1994) do not show an increased risk of
brain and nervous system tumors due to CC14 exposure.
While EPA reviewed each study across six domains with
respect to quality and risk of bias, there was no discussion
regarding causal inference or WOE across studies. It is
body of the literature regarding causal inference. EPA has
added that evaluation in Section 3.2.4.2.2.
The four epidemiologic studies of brain cancer are reviewed
and discussed in section 3.3.4.2 of the final risk evaluation .
Findings from the newer epidemiologic data on
carcinogenicity have been included, qualitatively, in the
cancer MOA and dose-response conclusions.
Page 104 of 210
-------
important to note in these small epidemiology studies of
rare diseases and uncommon exposures that artificially
high risk estimates can occur from random variability,
resulting is a phenomenon of effect size magnification.
The results may be statistically significant but with very
wide CIs that indicate imprecision. The lack of precision in
low powered studies may be quantified by calculating the
ratio of the upper and lower confidence limits (Poole,
2001). The NTP OHAT guidelines deem the risk of bias to
be "very serious" if the CI ratio is >10. Other reviewers
have considered the measures to be precise if the CI ratio
is below 4 (Schinasi and Leon, 2014). This imprecision is
seen in all five of the epidemiology studies, with the
exception of the case-control study by Ruder et al. (2013),
which showed no association between brain tumors and
CC14 exposures.
PUBLIC COMMENT PART 2: Conclusions from
brain cancer studies are reliable
Section 3.2.3.3.2 (Carcinogenicity) provides very little
discussion of the body of epidemiological studies and
provides no discussion of the implications of recent studies
of nervous system cancers (Heck et al., 2013; Nelson et al.,
2012; Neta et al., 2012; Ruder et al., 2013).
Given their high quality, significant results, and
consistency with each other, the three positive brain cancer
studies (Nelson et al., 2012, Neta et al., 2012, Heck et al.,
2013) should be used in assessing CC14 cancer risks (the
one study by Ruder et al. that failed to identify a cancer
risk should not be relied upon, as it lacked detailed
information on exposures, and instead assumed that
workplace levels were within the ranges reported in the
Page 105 of 210
-------
literature, making it too limited to support a no-risk
finding).
Although describing these studies, the draft evaluation
does not include them in its analysis of the weight of the
scientific evidence for carcinogenicity, its determination of
a cancer inhalation unit risk or its risk estimations for
cancer effects. Based on these studies, EPA should classify
CC14 as "Carcinogenic to Humans" under its cancer risk
assessment guidelines because "there is convincing
epidemiologic evidence of a causal association between
human exposure and cancer."
PUBLIC COMMENT PART 3: Further comments on
Heck et al. 2013
EPA incorrectly described Heck et al. (2013) as a study of
brain cancer, but it was actually of neuroblastomas, a
childhood cancer arising from cells that form the
sympathetic nervous system, which is not the brain. This
should not be considered as a brain cancer with the other
studies of adult occupational exposure to CC14.
Evidence is inconclusive due to small number of exposed
cases, poor precision in risk estimates, and low-quality
exposure assessment. Limitations include: (1) the exposure
assessment was low quality and was inferred based only
upon residence at birth; (2) methods for calculating the
mean concentrations were vague; (3) no information was
provided for the actual concentrations of CC14 (and other
pollutants) over time or by location; (4) sensitivity analysis
would have added confidence to the results; and (5)
analytic techniques are available to model the impact of
greater or lesser mobility upon the exposure-outcome
Page 106 of 210
-------
models. Strengths include: (1) record linkage studies are
not subject to participation rate and recall bias; (2)
exposure metrics were based on actual stationary monitors,
omitting the need for self-reporting of exposure and/or job
history; and (3) with the use of monitors, concentrations
were specific to CC14 (versus chlorinated solvents). WOE:
Because of limitations in exposure assessment, it is likely
that misclassification occurred. It is unknown how the
children born in the 1990-1998 period, for whom only zip
code was available, were included in these analyses.
Heck et al. (2013) is limited by its ecological design in
which exposure was estimated relatively crudely;
specifically, using ambient air pollution monitoring
stations and classified according to distance from these
monitors. Rates of cancers may vary geographically due to
differences in socioeconomic status, underlying prevalence
of other risk factors, and so forth; therefore, the cause(s) of
any differences in cancer rates cannot be elucidated.
PUBLIC COMMENT PART 4: Further comments on
Nelson et al. 2012
The evidence from Nelson et al. (2012) is inconclusive due
to small number of exposed cases, poor precision in risk
estimates, and low-quality exposure assessment.
Limitations include that the statistical power was low due
to the small number of glioblastoma multiforme cases (N =
9) and only two cases had probable exposure to CC14.
Strengths include that data were collected prospectively
before the subjects were ill, which reduces the problem of
information bias and low participation rates. WOE:
Because of limitations in exposure assessment, it is likely
that misclassification occurred. In addition to poor
Page 107 of 210
-------
exposure information, this study had very few cases,
resulting in incidence rates and risk measures with a large
magnitude of uncertainty, evidenced by wide CIs.
PUBLIC COMMENTS Part 4: Further comment on
Neta et al. (2012)
Neta et al. (2012): This study of glioma and meningioma
was inconclusive. The risk estimates when comparing high
to low exposed are statistically significant but imprecise.
No association was observed for meningioma and CC14.
Limitations include: (1) differential information bias may
have occurred from the cases being more motivated to
contribute detailed occupational information; and
(2) exposure to CC14 was based upon the job history and
likely affected by recall bias. Strengths include that_cases
were identified and enrolled in the study very quickly; the
study was of incident cases not deaths, reducing the
number of proxy interviews; and the authors conducted
sensitivity analyses to test various hypotheses and reran
different statistical models. WOE: The authors conducted
analyses in two ways: one using unexposed as the referent
and another using low exposed as a referent. Their
rationale was provided that "unexposed persons may be
substantially different from exposed persons in ways that
cannot be adjusted for in our analysis." However, they do
not discuss how or why this may occur.
PUBLIC COMMENTS Part 5: Further comment on
Ruder et al. (2013)
Ruder et al. (2013): No increased risk was observed for
gliomas and exposure to CC14. These results were not
statistically significant, but the CI ratio was <4, indicating
precision. Limitations include: (1) all exposure information
Page 108 of 210
-------
was collected retrospectively, with a high proportion from
proxies; (2) the focus of the study was on agricultural
exposures and the participants may have forgotten relevant
exposed jobs.; (3) the estimates of job-based exposures to
CC14 were based upon models reported in the literature;
and (4) as the authors noted, they were unable to determine
if their study participants' experiences were consistent
with these estimates. Strengths included: (1) the study was
based upon confirmed incident cases of glioma (versus
cases from death certificates); (2) the authors stratified
their results by respondent type {i.e., proxy) so that
information bias, if present, could be quantified; (3) there
were a large number of exposed cases permitting sufficient
statistical power to evaluate solvent exposures; and (4)
genotypes for glutathione-S-transferase were evaluated to
test for genetic susceptibility. WOE: Adequate design,
high outcome ascertainment and a specific exposure
metric. No increase in exposure to CC14 was observed in
any analysis of glioma.
PUBLIC COMMENTS PART 6: further discussion of
Heineman et al. 1994
Heineman et al. (1994): There is no increased risk of
astrocytic brain cancer when limited to subjects with high
probability of exposure (odds ratio = 0.8) and when
controlling for other solvents. Limitations include: (1) all
of the exposure and lifestyle information was based upon
interviews with a proxy, which is likely to be incorrect
recall especially for jobs in the distance past; (2) the data
available on each job lacked specificity for unique solvents
and poor temporal detail; and (3) the overall participation
was poor, which may introduce bias if participation was
influenced by perception of exposure. Strengths include
Page 109 of 210
-------
reporting by probability of use, which permits the reader to
evaluate results for the group with the highest confidence
of exposure. WOE: The sample size is greatly reduced
from "ever" exposed to "high probability" of exposed.
Most analyses show no excess risk and are not statistically
significant.
Methods used in the noncancer assessment: evidence synthesis ant
POD selection
SACC
SACC COMMENTS:
Recommendation: Improve the discussion and include
more details about the selection and derivation of the
PODs (including calculations where possible).
For example, in Section 3.2.5.1.2 on p. 128 of the draft
risk evaluation, why was the 25 parts per million (ppm)
from the rat data in the Nagano study selected to derive
the POD rather than the mouse data from the same
study, or why weren't the eosinophilic granules seen in
the rats at 5 ppm used?
More information is needed on the calculation of the
HECs and the adjustments to convert from continuous
exposure to the 8- and 12-hour occupational exposures.
Which BMDLio calculation was used and how was it
used to derive the 14.3 mg/m3 value?
Most of the BMDLio outcomes from the modeling in
the appendix appear in [j,mol/L units, which is very
different from most EPA assessments. Similarly, it
would help the reader if the actual calculations were
provided in the footnote of Table 3-14 to illustrate how
the occupational exposure levels were calculated.
The basis for all of the critical PODs, especially those
in Table 3-17, should be shown.
The following language was added to Section 3.2.5.1.2:
"Fatty change in the liver of rats was selected as the endpoint
for dose-response analysis because this histopathologic
lesion, which is indicative of cellular damage was a more
sensitive endpoint than other histopathologic changes that
were also observed in rats exposed to 25 ppm from the
(Nagano et al 2007) studv. The onlv histopathological
change observed in the 5 ppm group in the chronic rat study
is an increase in eosinophilic granules in the nasal cavity of
the female rats. This histopathological change is not
considered an adverse effect by itself because it is not
accompanied by other adverse effects in the nasal cavity.
Furthermore, while severe renal and hepatic effects are
observed in the high-exposure group, the nasal lesion is only
of moderate severity in such exposure group."
The dose response analysis included the use of the PBPK
model and BMD modeling methodology used in the IRIS
Toxicological Review (U.S. EPA. 2010) to estimate internal
doses and analyze the relationship between the estimated
internal doses and fatty change (i.e., response). The resulting
BMDL values were converted to estimates of equivalent
HECs by applying a human PBPK model. Estimated values
for HECs corresponding to BMDLIO values for fatty changes
of the liver for alternative values of Vmaxc in the rat and
Page 110 of 210
-------
human are presented in in Tables 5-6 and 5-7 of the IRIS
Toxicological Review (U.S. EPA. 2.010). A human Vmaxo
estimated from in vitro human data can reasonably be
presumed to be more relevant than a human Vmaxc based
entirely on rodent data. Because the MOA for carbon
tetrachloride-induced hepatotoxicity involves metabolism to
reactive metabolites in the liver, HECs based on the mean rate
of metabolism in the liver dose metric is the most proximate
to the critical effect. The resulting BMCLio[hec] based on data
for the male rat is 14.3 mg/m3 for continuous exposures.
Language in Section 3.2.5.2.2 already states that the BMDLio
value for continuous exposures was extrapolated to shorter
exposure durations using the equation Cn x t = k, where an
empirical value of n was determined to be 2.5 on the basis of
rat lethalitv data (Ten Berge et aL 1986)
This language was modified as follows:
"BMDLIO value for continuous exposures was extrapolated
to shorter occupational exposure durations (8-hr/day and 12
hr/day) using the equation Cn x t = k, where an empirical
value of n was determined to be 2.5 on the basis of rat
lethalitv data (Ten Berge et aL 1986). Further information on
this temporal scaling equation can be found in (NRC. 2014)."
The column 'Basis for Selection' in Table 3-17 was also
updated.
SACC
SACC COMMENTS:
Some Committee members noted that the HEC computed
in the draft risk evaluation is below doses observed in the
original animal study and similarly below the current
PEL. The basis for the chronic inhalation POD was set
The chronic POD for inhalation exposures is based on a
study observing increased fatty changes in rodent livers
(Nagano et aL 2007). The lowest exposure concentration (5
ppm) in the 104-weeks inhalation study with F344/DuCij
rats (Nagano et aL. 2007) was considered a NOAEC based
Page 111 of 210
-------
using the NOAEC for liver cancer of 5 ppm based on the
Nagano et al. (2007b) rodent study (p. 130, lines 4176-
4178). More data are needed to validate use of such a low
HEC value. Such data might be obtained by NTP via a 13-
week inhalation study using 4-5 concentrations between
50 ppb and 5 ppm. Barring this study, the risk
characterization in the risk evaluation should be labeled as
preliminary, primarily due to this low-dose extrapolation.
on liver and kidney toxicity at > 25 ppm. Interpretation of the
observed proteinuria and the renal lesions in the F344 rat is
difficult because this strain has a high spontaneous incidence
of renal lesions. Increases in the incidence and severity of
nonneoplastic liver lesions (fatty change, fibrosis, cirrhosis)
were seen at 25 and 125 ppm in both males and females.
The HEC (in mg/m3) consisting of BMDLio for fatty changes
of the liver of 14.3 mg/m3 for continuous exposures was
estimated using a PBPK model in the peer-reviewed IRIS
Toxicological Review for Carbon Tetrachloride.
SACC
SACC COMMENTS:
A Committee member commented that they would
like to see more discussion as to why a NOAEL of 5
ppm is used when there were effects seen in JBRC
(1998) (e.g., spleen, urine analysis, white blood cell
count) at 5 ppm that do not seem to be reported in
Nagano et al. (2007a). This Committee member
suggested that these observed effects could inform
levels impacting developmental neurotoxicity and
immune effects.
Despite the many changes observed in the JBRC
studies at the lowest doses, the draft risk evaluation
reports lowest and mid doses as NOAELs for key
endpoints in Appendix H and line 4175 that are
higher.
The systematic review for this risk evaluation identified
(Nagano et al... 2007) as a high qualitv studv. (Nagano et al..
2007) is based in JBRC studv described in the JBRC, 1998
reference.
The JBRC 1998 reference was specifically evaluated in the
peer-reviewed IRIS Toxicological review for Carbon
Tetrachloride. IRIS evaluation of JBRC 1998 did not identify
adverse immune effects at non-hepatotoxic doses.
SACC
SACC COMMENTS:
Recommendation: The toxicokinetics discussion should
be updated and expanded, particularly on the influence of
exposure route on systemic disposition and effects, as well
as inter- and intra-species differences in metabolic
activation and susceptibility.
It was noted on lines 4085-4091 that the utility of the
The final risk evaluation contains the following statement
explaining the limited utility of the oral study for risk
characterization: "oral exposures to carbon tetrachloride
undergo first-pass metabolism in the liver, the organ with the
highest concentration of CYP2E1 enzymes involved in the
generation of carbon tetrachloride's toxic metabolites. This
major difference in the metabolism of carbon tetrachloride
Page 112 of 210
-------
oral developmental study of Narotsky et al. (1997)
was limited; however, the reason was not clearly
described, namely that first-pass hepatic metabolism
following ingestion reduced the amount of CC14
reaching the systemic (arterial) circulation and extra
hepatic organs. With low dose oral exposures, the
liver and lungs, acting in concert, eliminated/removed
virtually all CC14 and other VOCs before they enter
the systemic circulation. High oral doses, however,
can exceed the uptake and metabolic capacity of the
liver systemic circulation.
The findings of Sanzgiri et al. (1997) are applicable
here. They characterized the influence of route and
rate of administration of CC14 on blood and tissue
levels of CC14 in rats. Presystemic elimination of
CC14 can be protective of extrahepatic organs, but the
liver often "bears the brunt" of adverse effects
(Sanzgiri et al., 1995).
There are no descriptions in Section 3.2.2 on
toxicokinetics of the time-course of CC14 or its key
metabolites, for use in understanding the chronicity of
adverse effects of single and multiple exposures. The
Committee suggests using data from Kim et al. (1990)
and Rao and Recknagel (1968, 1969).
The Committee agreed that it would be worthwhile to
expand the description of CC14 metabolism, and to
link the chemical's bioactivation to its MOA. The
Committee noted that there was an excellent
publication by Slater (1987) describing biochemical
reactions and effects of the chloromethyl peroxyl
radical and subsequent products of lipid peroxidation.
The experimental protocol of an unpublished study by
Benson and Springer (1999) is described on pp. 107
between oral and inhalation routes of exposure limits the
usefulness of extrapolating a developmental inhalation POD
from the oral developmental study, given that different
developmental toxicity processes may be involved between
the two routes of exposure."
The Toxicokinetics section was expanded to include the
following language:
The toxicokinetics of carbon tetrachloride have been
comprehensively described in previous toxicological
assessments (see Error! Reference source not found.). In
summary, the IRIS assessment describes that carbon
tetrachloride is "rapidly absorbed by any route of exposure."
However, it is noted that dermally absorbed fraction would
be "negligible for exposures to carbon tetrachloride vapor
(Mccollister et al.. 1951)
Once absorbed, carbon tetrachloride is widely distributed
among tissues, especially those with high lipid content,
reaching peak concentrations in <1-6 hours, depending on
exposure concentration or dose. Animal studies show that
volatile metabolites are released in exhaled air, whereas
nonvolatile metabolites are excreted in feces and to a lesser
degree, in urine.
Findings from (Sanzgiri ar kner. 1997). in which tissue
distribution of inhaled carbon tetrachloride was compared to
the equivalent oral dose show that maximal levels in fat were
considerably in excess of the maximal levels in other tissues,
regardless of route of exposure. Among tissues other than
fat, distribution kinetics were generally similar for the
tissues, except that maximal levels were higher and attained
Page 113 of 210
-------
more quickly in the liver than in other tissues following
bolus oral administration.
The metabolism of carbon tetrachloride has been extensively
studied in in vivo and in vitro mammalian systems. Carbon
tetrachloride is metabolized in the body, primarily by the
liver, but also in the kidney, lung, and other tissues
containing CYP450. Based on reasonably available
information, the initial step in biotransformation of carbon
tetrachloride is reductive dehalogenation: reductive cleavage
of one carbon-chlorine bond to yield chloride ion and the
trichloromethyl radical. Biotransformation of carbon
tetrachloride to reactive metabolites, including the
trichloromethyl radical, is hypothesized to be a key event in
the toxicity of carbon tetrachloride. The fate of the
trichloromethyl radical depends on the availability of oxygen
and includes several alternative pathways for anaerobic or
aerobic conditions. Anaerobic dimerization forms
hexachloroethane, while aerobic trapping by oxygen forms a
trichloromethyl peroxy radical. The trichloromethyl peroxy
radical is the primary initiator of lipid peroxidation that
occurs from exposure to carbon tetrachloride (Rao and
Reckmaeet. 1969)
Cytochromes CYP2E1 and CYP2B, the primary enzymes
responsible for biotransformation of carbon tetrachloride in
rodents, were measured in all exposed and control animals in
the metabolic studies by (Benson and Springer. 1999). In all
species, microsomal measurement of these enzymes
indicated that while enzyme induction increased several fold
as dose increased, catalytic activity was not significantly
altered. In addition, the rate of carbon tetrachloride
metabolism was measured in rat, mouse and hamster species.
Page 114 of 210
and 108 of the draft risk evaluation, but relatively few
of their findings or conclusions are mentioned.
Derivation of human metabolic rate constants was
mentioned, but no results were provided. Was any
information obtained to assess the existence of
genotoxic versus non-genotoxic mechanisms of liver
tumors? Thrall et al. (2000) did report the following
rank order of CC14 metabolism: hamster > mouse >
rat > human.
-------
The metabolic rate of carbon tetrachloride did not vary more
than 2-fold between the three speciesf Benson and Springer.
?).
(Thrall et al 2000) and (Benson and Springer. 1999) used in
vitro data on metabolism of carbon tetrachloride by human
liver microsomes and in vitro and in vivo rodent data, to
estimate the in vivo human metabolic rate constants. Those
rate constants were used by the IRIS Program for
interspecies extrapolation (i.e., rat-to-human, mouse-to-
human) and route-to-route extrapolation of carbon
tetrachloride inhalation dosimetry using a human PBPK
model, which has been described in (Paustenbach et aL
3)), (Thrall et aL 2.000) and (Benson and Springer.
?).
Methods used in the cancer assessment: selection of tumor type, IV
OA, POD, and I.UR calculation
SACC,
39
SACC COMMENTS:
In the JBRC (1998) inhalation cancer bioassay, there
were increased adrenal, endometrium, ovary, and
thyroid (as well as pancreas, spleen, and subcutis)
tumors reported at low and mid doses in female rats.
While many of these did not reach statistical
significance, taken together, they are notable, and it
would be a more complete presentation to include a
summary of these findings in the risk evaluation.
These endocrine tumors are consistent with evidence
of an endocrine MOA for some noncancer and cancer
endpoints observed with CC14 and further discussion
on this point would contribute to discussion of cancer
MOA.
Recommendation: Include a summary table of tumors
observed in endocrine-associated tissues in the JBRC
(1998) inhalation study, particularly for female rats, and
EPA relies on current agency guidance and risk assessment
practice for developing cancer assessments in TSCA risk
evaluations. Adding up different type of tumors to reach
statistical significance or use of the Haseman Rule are not in
agreement with current Agency guidelines for cancer
assessment.
One of the general considerations for MOA analysis in the
Guidelines for Carcinogen Risk Assessment (U.S. EPA.
2005) for analyzing an agent's influence in the development
of tumors is the consideration of an agent working by more
than one MOA at different sites and at the same tumor site.
Therefore, the cancer MOA cannot be generalized to other
tissues or cell types without additional analyses.
Page 115 of 210
-------
include a discussion of their significance.
PUBLIC COMMENTS:
The NTP uses what is known as the "Haseman Rule,"
which tests the significant differences in tumor incidence
between the control and dose groups at 0.05 for rare
tumors and at 0.01 for common tumors. Based on the
"Haseman Rule," the increased incidence of liver
adenomas in the 5 ppm female mice is not statistically
significant at p<0.01 and should therefore not be
considered treatment-related.
SACC
SACC COMMENTS:
Recommendation: Major points about genotoxicity of
CC14 should be brought forward from Appendix I,
including overall conclusions reached about strengths,
weaknesses, and limitations of existing studies, WOE, and
data needs.
The SACC report offers a detailed discussion of a
proposed MO A, including noting that many studies
have demonstrated that CC14 impairs the immune
system, and that immune suppression promotes tumor
growth.
One Committee member recommended that EPA
better explain why genomics, proteomics,
genotoxicity, indirect genotoxicity, changes in gene
expression, or messenger ribonucleic acid (mRNA)
levels were excluded while evaluating CC14 MOA
studies of in vitro models.
Major points about genotoxicity have been brought forward
from Appendix I.
EPA considered the genotoxicity, indirect genotoxicity,
changes in gene expression studies while evaluating carbon
tetrachloride MOA studies of in vitro models. Those studies
were used to identify the key events in the MOA. Other
studies {i.e., proteomics and genomics) that provided more
detailed mechanistic information within each key event were
considered off topic.
Information on the criteria for determination of on topic and
off topic studies can be found in section 1.5.1 of the final risk
evaluation.
SACC,
39
SACC COMMENTS:
Specific molecular and cellular mechanisms through
which CC14 exerts its toxicity have been thoroughly
investigated and this deep knowledge of the
mechanisms of action of CC14 should be carried
The discussion on the carcinogenicity MOA has been
expanded in the final risk evaluation. However, Khan and
Younus (2011) is an in vitro study in which carbon
tetrachloride was used as a positive control to induce a
disease state. This type of study was considered off-topic in
Page 116 of 210
-------
through to the human health risk assessment.
Key steps occurring in both liver and adrenal gland
tumor formation should be compared. A diligent effort
should be made to update the literature review in this
area as there are likely to be relevant studies that have
been published in the 10-15 years since the IRIS
document was written.
For example, a study by Khan and Younis (2011)
describes oxidative damage occurring in the adrenal
gland following CC14 administration. Also, the U.S.
EPA (2010) evaluation missed key studies, such as
Slater (1987).
Recommendation: Expand the discussion of CC14's MOA
for carcinogenicity in both the liver and adrenal gland.
PUBLIC COMMENTS:
Below are several key points suggesting similar low-dose
threshold MO As for both liver and adrenal medulla
tumors:
Adrenal medulla cells have the same basic cell
structure as liver cells.
CC14 is expected to be metabolized to trichloromethyl
and trichloromethyl peroxy radical metabolites in the
endoplasmic reticulum.
Reactive CC14 radical mechanisms in adrenal medulla
cells are expected to be similar to liver cells.
Antioxidant defense mechanisms in adrenal medulla
cells are expected to be similar to liver cells.
Mutagenic MOA for tumors is not supported by
genotoxicity data.
the systematic review because it provides limited
applicability for dose-response in the risk evaluation. In
addition, sufficient high quality on-topic human health
references were identified for carbon tetrachloride. Appendix
B in the Problem Formulation describes the process used to
re-screen human health references for prioritizing the
literature for applicability in the risk evaluation.
Furthermore, the findings from Khan and Younus ( )
show that carbon tetrachloride does induce oxidative stress.
This conclusion has been reached in the IRIS assessment and
this risk evaluation without the need of that study. The final
risk evaluation indicates that metabolism of carbon
tetrachloride leads to the production of free peroxy radicals
which induce oxidative stress with, which can damage
proteins, DNA and lipids. The IRIS assessment indicates that
in vitro studies bv (Colby et al 1994) showed that
preincubation of adrenal microsomes with 1-
aminobenzotriazole, a CYP450 suicide inhibitor, prevented
the effects of carbon tetrachloride on lipid peroxidation and
covalent binding. Nevertheless, there is not sufficient
information to elucidate the key events for cancer induction
in the adrenal gland and astrocytic brain tissues.
Slater 1987 consist of a lecture transcript. The studies on
carbon tetrachloride cited in the lecture were evaluated in the
IRIS assessment, which is one of the assessments considered
in the systematic review for this risk evaluation.
SACC
SACC COMMENTS:
Recommendation: The contribution of inhibition of
immune function to an indirect carcinogenic MOA should
Based on the review of the on topic human health references
in the systematic review, EPA has concluded that carbon
tetrachloride immunological effects were, at least in part,
Page 117 of 210
-------
be discussed.
secondary to hepatotoxicity and the process of hepatic repair,
which produces adverse effects on T-cell-dependent
immunity at doses that are hepatotoxic. However, elucidation
of the exact mechanism by which carbon tetrachloride
induces tumors is outside the scope of this risk evaluation.
SACC,
30,31,
39, 43,
45
SACC COMMENTS: points against low-dose linear
mechanism of action
Although the draft risk evaluation claims to have
"Evaluated the weight of the scientific evidence based
on the available human health hazard data for carbon
tetrachloride," the Committee noted that convincing
support for this claim is lacking.
In particular, the draft risk evaluation refers repeatedly
to a concern that low-level exposures to CC14 may
somehow act through genotoxic mechanisms
(evidence for this notwithstanding); indeed, this
concern is its underlying justification for using the
"default" approach of applying a linearized model to
the tumor mouse bioassay data in order to predict low-
dose cancer risk. But the WOE clearly indicates that
any genotoxicity caused by CC14 can occur only at
exceedingly high levels of exposure, and is caused not
by CC14 directly, but only indirectly after high levels
of lipid peroxide byproducts (such as reactive
aldehydes) have accumulated intracellularly (see, for
example, Slater, 1987; MAK, 2000; Weber et al.,
2003; Eastmond, 2008; Hernandez et al., 2009;
Borgert et al., 2015).
No support is provided for EPA's designation of an
"alternate MO A" that combines cytotoxic mechanisms at
relatively high CC14 doses with "alternate, non-cytotoxic
mechanisms" at lower doses.
The evidence on cancer MOA has been revisited and
expanded for liver, adrenal and brain tumors. In addition, the
key events for the liver tumors MOA and uncertainties of
alternate MO As are presented in appendices of the final risk
evaluation.
The cancer assessment relies in the 2010 IRIS Toxicological
Review of Carbon Tetrachlorided '.S. EPA. 2010) findings,
newer epidemiological studies presenting additional
evidence of an association between carbon tetrachloride
exposure and neuroblastomas (adrenal gland tumors in
infants) and brain cancers and alternate MOA information.
The final risk evaluation includes evaluation of the available
carcinogenicity studies and MOA information in support of
evaluating the potential cancer risk for carbon tetrachloride.
MOA information on carbon tetrachloride has been
evaluated in the context of EPA's "MOA framework" as
presented in EPA's 2005 Guidelines for Carcinogen Risk
Assessment (U.S. EPA. 2005). (see Chapter 2.4 of EPA's
2005 cancer guidelines). The new epidemiological
information provides evidence on carbon tetrachloride
carcinogenicity in humans when considered with the site
concordance with pheochromocytomas (adrenal gland
tumors) in mice and other evidence of hepatic tumors in
multiple species.
The key events underlying the MOA for induction of liver
Page 118 of 210
-------
What is meant by an "alternate non-cytotoxic
mechanism" (p. 124, line 4005)? This appears to be
speculation that something must be occurring to
produce an increased incidence in liver adenomas in
the female mice dosed at 5 ppm.
Consideration should be given to the possibility that
this was a chance occurrence in a single study. The
historical incidence of this benign tumor in control
Crj:BDFl mice is as high as 10%.
Had 3 of 50 control females exhibited liver adenoma
in this particular experiment, the difference between
them and the 5 ppm dose group would not have been
statistically significant. There was no increase in liver
carcinoma incidence in the females dosed at 5 ppm
and no significant increase over controls in combined
benign and malignant liver tumors.
It should also be noted that there was no increase in
hepatocellular adenoma or carcinoma in the male mice
dosed at 5 ppm. Male mice metabolically activate more
CC14 and experience a higher incidence of liver cancer
than females.
PUBLIC COMMENTS: points against low-dose linear
mechanism
EPA should prepare an independent, clear, and robust
MOA analysis for both alternatives. EPA is obligated,
under the statute, regulation, and Agency-wide guidance,
to calculate potential risks from the alternative MOA, and
the default option, and to characterize each fully, both
narratively and quantitatively, for the risk manager.
EPA should utilize an established framework to organize
evidence for MOA based on side-by-side WOE
tumors by carbon tetrachloride have been extensively
investigated. Metabolism is identified as the first key event
for the induction of liver, adrenal tumors and brain tumors by
carbon tetrachloride. The other key events by which carbon
tetrachloride induces pheochromocytomas in mice and
neuroblastomas and brain tumors in humans are currently
unknown due to lack of mechanistic information on these
tumor types.
Biological support exists for a hypothetical MOA involving
metabolism of carbon tetrachloride by CYP2E1, sustained
cytotoxicity, and regenerative cell proliferation as key events
driving the steep nonlinear increase in liver tumor dose-
response at relatively high carbon tetrachloride exposures.
However, several pieces of evidence suggest that carbon
tetrachloride carcinogenicity is not explained by a cytotoxic-
proliferative MOA in tumor types other than liver.
At lower exposure levels, the correspondence between
hepatocellular cytotoxicity and regenerative hyperplasia and
the induction of liver tumors is inconsistent. In particular,
liver findings from the JBRC bioassay (Nagano et at.. 2007)
suggest that mouse hepatocarcinogenicity cannot be
explained in terms of the cytotoxic-proliferative MOA. An
increased incidence of hepatocellular adenomas occurred in
the low-exposure (0.9-ppm adjusted) female mouse in the
absence of nonneoplastic liver toxicity, raising the possibility
of another MOA operating in addition to or in conjunction
with the cytotoxic-proliferative MOA. Other considerations
suggest that the carbon tetrachloride database is insufficient
to rule out other MO As at low exposure levels, in particular
considerations related to the compound's genotoxicity and
general reactivity.
Page 119 of 210
-------
comparison of alternative plausible MO As (e.g., AOPs,
IPCS, Becker et al., 2017). A systematic and explicit
approach must be uniformly implemented to compare
potentially relevant MO As. One method for doing this
involves deriving WOE confidence scores based on the
IPCS framework and Bradford Hill causation criteria.
Significant effort has been directed to characterizing the
MO A/AOPs at these sites, with agreement on this point
not yet realized. Some additional work is needed, which
will also lead to consensus on the appropriate choice(s)
for dose response assessment. At present, the MOA
analysis in the draft risk evaluation summarizes EPA's
prior IRIS analysis with no updates or use of WOE
analysis methods. Further, the IRIS analysis was
published 10 years ago; thus, EPA should examine
whether those conclusions still reflect the current state of
the science.
The SACC should discuss and advise EPA on providing a
more thorough discussion surrounding the uncertainty for
each alternative and on whether EPA should also include
a determination of confidence in the selection of a
particular MOA.
EPA's position in the risk evaluation of a low-dose linear
MOA for liver tumors is untenable in light of the most-up-
to-date scientific studies on CC14 toxicity. Uehara et al.
(2013) showed that there was no secondary DNA damage
associated with CC14 radical-induced lipid peroxidation
and/or cytotoxicity at the time points measured at a
relatively low dose of CC14 that also resulted in liver
tumors in mice. This lack of concordance between DNA
Therefore, EPA considers alternate MO As such as (1)
cytotoxic mechanisms at high doses with alternate, non-
cytotoxic mechanisms as lower doses, and (2) cytotoxicity
and regenerative hyperplasia for liver tumors, in conjunction
with the lack of MOA information on other tumor types
induced by carbon tetrachloride in the animal and human
data.
The alternate MO As and uncertainties in the cancer MOA
for the different tumor types are addressed in the final risk
evaluation by cancer dose response analysis and cancer risk
calculations based on both linear and nonlinear approaches
which encompass all the considered MO As.
Page 120 of 210
-------
adducts and cellular oxidative stress in liver tumor-
bearing mice dosed with CC14 provides critical evidence
supporting a cytotoxic-proliferative (non-linear) MOA for
CC14 carcinogenicity at low doses.
In describing the cytotoxic MOA in Table 3-11, EPA
should consider whether the results in female mice are
consistent with other animal studies and describe other
data that substantiate the counterfactual argument against
this MOA. Given the uncertainties in the current draft
MOA analysis, EPA needs to revisit this entire section
and provide a more comprehensive evaluation using, for
example, the evolved qualitative MOA framework of
WHO/IPCS or the quantitative MOA confidence scoring
method described in Becker et al. (2017).
Based on a considerable number of scientific studies, the
MOA can be explained by the involvement of cytotoxicity
and proliferation from the highly reactive radical
metabolites of CC14. The best available science and the
weight of the scientific evidence indicate that CC14 is
carcinogenic in the liver only via a MOA with a non-
linear (threshold) dose-response.
PUBLIC COMMENTS: points for low-dose linear
mechanism
EPA's final evaluation should continue to conclude that
evidence for a non-linear MOA is inadequate.
SACC,
31,39,
43,45
SACC COMMENTS: low dose risk calculation
The draft risk evaluation appears to present two
approaches to calculating low-dose risk (a low dose linear
approach and a non-linear approach), these two approaches
appeared to be melded into a single risk assessment model
The final risk evaluation presents a low dose linear
extrapolation and threshold risk assessment approaches. The
evidence on cancer MOA has been revisited and expanded for
liver, adrenal and brain tumors. In addition, the key events for
the liver tumors MOA and uncertainties of alternate MO As
Page 121 of 210
-------
(lines 621 and 4325). The low dose linear approach was
used for "low dose exposures of carbon tetrachloride." The
nonlinear approach was used only for doses "exceeding the
POD" of 18 mg/m3.
It was not clear to the Committee how the non-linear
approach is to be implemented. Both the linear and
non-linear approaches are alternative approaches for
quantifying low dose risk.
The Committee does not understand what it meant for
the non-linear approach to be implemented for high
doses only.
Some members noted that this confusion was due to
the draft risk evaluation's reliance on IRIS (U.S. EPA,
2010) for its human health risk evaluations. Given this,
the Committee wondered whether any discussion of
cancer induction mechanism was needed in the risk
evaluation, since, again, none of it seems to be used for
purposes of either qualitative or quantitative human
health risk evaluation.
Some Committee members agreed with EPA's
determination while other disagreed based on the MOA
for this chemical and its free radical metabolites. Some
Committee members would like to see a nonlinear
threshold-type of approach also presented for the
cancer risks based on long-standing, published, peer-
reviewed evidence regarding the peroxyl radical-based
mechanisms by which CC14 induces tumors.
Recommendation: State clearly and justify whether a
low-dose linear risk assessment approach or a non-
linear risk assessment approach is preferred.
The Committee concluded that the weight of a
considerable body of scientific evidence indicates that
the relationship between CC14 dose/exposure and its
are presented in appendices of the final risk evaluation.
(LL
2010) concludes that the key events in the cancer
MOA for liver tumors described in section Error! Reference
source not found, of the final risk evaluation appear to play a
significant role at high exposure doses. Therefore, EPA
considers an alternate MOA that combines cytotoxic
mechanisms at high doses with alternate, non-cytotoxic
mechanisms as lower doses.
Metabolism is identified as the first key event for the
induction of liver, adrenal tumors and brain tumors by carbon
tetrachloride. The other key events by which carbon
tetrachloride induces pheochromocytomas in mice and
neuroblastomas and brain tumors in humans are currently
unknown due to lack of mechanistic information on these
tumor types.
There is general consensus that metabolism of carbon
tetrachloride leads to the production of free peroxy radicals
which induce oxidative stress that can damage proteins, DNA
and lipids. As described in the IRIS assessment, in vitro
studies by (Colby et al. 1994) showed that preincubation of
adrenal microsomes with 1-aminobenzotriazole, a CYP450
suicide inhibitor, prevented the effects of carbon tetrachloride
on lipid peroxidation and covalent binding. Nevertheless,
there is not sufficient information to elucidate the key events
for cancer induction in the adrenal gland and brain tissues.
One of the general considerations for MOA analysis in the
Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2.005)
for analyzing an agent's influence in the development of
tumors is the consideration of an agent working by more than
Page 122 of 210
-------
genotoxic response is nonlinear with a steep dose-
response.
The Committee noted that pheochromocytomas are
tumors of chromaffin cells in the adrenal gland. CC14
is among the small number of chemicals that can cause
adrenal tumors in mice, and also cause liver tumors.
The Committee briefly discusses available data
relevant to CC14 and adrenal tumors, concluding that
genotoxic events in the adrenal appear to be
attributable to the indirect action of free radicals.
If one assumes that the key steps are the same in
adrenal gland and liver tumors, extrapolation using a
non-linear, threshold model would seem appropriate.
This would also be supported by the in vitro and the
systemic in vivo genotoxicity data for CC14, which are
generally negative.
One Committee member suggested that CC14, like
other carcinogens, with multiple interacting MO As will
operate as additive to background. As a result, the
dose-response relationship may look quite linear,
especially in a heterogenous population of humans
(Crump, 2018).
Several points supporting the SACC conclusion are
provided in the report.
The Agency needs to be clear about what the terms "linear
low-dose" or a non-linear or threshold dose-response
means.
Rather than separately defining low-dose sub-linear
and threshold, EPA (2005) defines "low-dose
nonlinear" as a dose-response "whose slope is zero at a
dose of zero." Note that this includes both low-dose
sub-linear and threshold dose-responses as defined by
one MOA at different sites and at the same tumor site.
Therefore, the cancer MOA cannot be generalized to other
tissues or cell types without additional analyses. Based on the
reasonably available information and cancer MOA
considerations in (U.S. EPA. 20051 EPA concludes that all
the key events in the MOA for carbon tetrachloride
carcinogenicity in adrenal gland and brain tissues across all
exposure levels is unknown at this time.
Therefore, EPA considers alternate MOAs such as (1)
cytotoxic mechanisms at high doses with alternate, non-
cytotoxic mechanisms as lower doses, and (2) cytotoxicity
and regenerative hyperplasia for liver tumors, in conjunction
with the lack of MOA information on other tumor types
induced by carbon tetrachloride in the animal and human
data.
The alternate MOAs and uncertainties in the cancer MOA for
the different tumor types are addressed in the final risk
evaluation by cancer dose response analysis and cancer risk
calculations based on both linear and nonlinear approaches
which encompass all the considered MOAs.
Page 123 of 210
-------
Crump (2011) and the EPA cancer guidelines (U.S.
EPA 2005) but does not include supra-linear dose-
responses.
The EPA guidelines do not discuss or define supra-
linearity.
In order to conclude that the low dose-response is non-
linear or threshold, it is not sufficient to conclude that
carcinogenicity is not produced via a mutagenic MOA.
There are mechanisms other than mutagenicity that can
produce a low-dose linear response.
Recommendation: EPA should apply a non-linear model in
estimating cancer risks, in light of the preponderance of
evidence that lipid peroxidation- and endonuclease-derived
mutations, and other cytotoxic effects, are the origins of
tumors of the liver and adrenal gland.
Recommendation: Consider adopting a threshold-type
MOA in estimating carcinogenic risk and consider
applying UFs for database deficiencies due to more limited
mechanistic information about adrenal gland tumors in
mice and reported associations of occupational CC14
exposure and increased incidence of gliomas in workers.
Public commenters suggested, and the Committee agreed,
that when there was conflicting information on the cancer
MOA, EPA should, at a minimum, include a risk
characterization for both linear and non-linear dose-
response models to allow for comparison of the results.
The SACC suggested selecting the most conservative
model for the evaluation of risks.
PUBLIC COMMENTS: low dose risk calculation
Given the strong evidence supporting the hypothesized
Page 124 of 210
-------
alternative approach (threshold cytotoxicity MO A), and
the uncertainties in the MOA that EPA has postulated
invokes the no-threshold, low-dose linearity default, EPA
must quantify risks for both approaches fully. In its TSCA
risk evaluations, EPA should more clearly and
transparently present biologically robust, MOA
assessments where the WOE is integrated fully.
Ultimately, EPA should carry any biologically plausible
alternative MO As and the default MOA option through the
entire assessment and present all risk calculations in the
risk characterization section. To do otherwise is
inconsistent with the TSCA statute, the TSCA Risk
Evaluation Rule, and the Agency's Cancer Guidelines.
In the 2010 CC14 IRIS assessment, EPA concluded that
there is insufficient information on the MOA of CC14 for
mouse liver tumors at low doses and the mouse
pheochromocytomas to support a non-linear dose-response
approach for assessing cancer risk. In spite of that
conclusion, a majority (four out of six) of the EPA Science
Advisory Committee review for the 2010 CC14 IRIS
assessment recommended that the CC14 cancer risk should
be preferably based on a non-linear threshold method.
EPA did not refer to the impact on the risk estimate of the
policy chosen dose-response model, the linearized
multistage model (LMS). Alternative models would give
risk values several orders of magnitude lower than the
LMS model.
EPA should provide added justification for moving
forward with quantification of risk associated with only
one of the MO As. Additionally, the SACC should discuss
Page 125 of 210
-------
and evaluate whether EPA should quantify risks for both
alternatives, or at a minimum, include a sensitivity analysis
to examine whether the MOA analysis influences the risk
conclusions.
45
PUBLIC COMMENTS:
The draft risk evaluation describes the two quantitative
approaches for assessment of carcinogenicity in the IRIS
Toxicological Review (U.S. EPA, 2010), but states in
number 2 on p. 135, "This threshold approach is used in
this risk evaluation for high exposures based on a
benchmark MOE of 30." However, in the risk evaluation,
the threshold approach is not described further and does
not appear to be used in this manner.
The statements on threshold approach were corrected or
eliminated in the final risk evaluation.
SACC
SACC COMMENTS:
There appeared to be no description of the calculation
of the POD of 18 mg/m3 in the document.
Recommendation: Explain the basis and the calculations
in determining the PODs.
The basis and calculations in determining the PODs have
been incorporated in the final risk evaluation.
SACC,
29
SACC COMMENTS: IUR calculation
Recommendation: Key details on the derivation of the
IUR, similar to that provided in the IRIS summary {i.e.,
species, cancer type, extrapolation model, risk levels,
etc.), should also be provided in this risk evaluation.
PUBLIC COMMENTS:
As pheochromocytomas occurred in mice at exposure
concentrations that also resulted in toxicity in liver cells,
estimation of human cancer risk based on liver toxicity
would be adequately protective for both tumor types.
Key details on IUR derivation have been added to section
3.2.5.2.5.
IUR estimates based on the tumor data sets in (Nagano et al.
2007) were calculated using the following equation: IUR =
BMR HEC, where BMR = benchmark response, HEC =
human equivalent concentration. The highest estimated IUR
for carbon tetrachloride via the inhalation pathway is 6 x 10"6
([j,g/m3)"l, which is associated with pheochromocytomas in
the male mouse. The data set on pheochromocytomas in the
male mouse was determined to be applicable, scientifically
sound, and yielded the highest estimate of risk and is
supported by the EPA IRIS Program ("U.S. EPA. 2005).
30
PUBLIC COMMENTS:
Temporal adjustments were performed using experimental
data and/or PBPK modeling described in sections 3.2.5.2.1.
Page 126 of 210
-------
Studies are needed that would illuminate the potential for
general systemic toxicity over exposure duration(s)
commensurate with that/those of the actual exposure
scenario(s) under evaluation or, if long term, that could be
extrapolated from shorter-term exposure studies
accompanied by the application of a UF representing that
extrapolation (e.g., acute short term or subchronic to
chronic).
and 3.5.2.5.2.2, therefore EPA didn't apply UFs when
extrapolating for exposure time duration.
43
PUBLIC COMMENTS:
The 2010 IRIS assessment and the 2014 NAT A show that
the risk to most Americans from ambient air exposure to
CC14 exceeds the 1-in-a-million lifetime risk level. Yet
EPA's risk evaluation ignores this evidence of excess
cancer risk to the general population, as well as to
particularly exposed subpopulations, based on its
exclusion of all air emissions from the evaluation's scope.
EPA also fails to consider the impacts of these
background CC14 concentrations on the workers and
ONUs studied in the risk evaluation who are exposed in
the workplace, and thus understates the risks to this
population from aggregate exposure to CC14.
EPA did not consider background exposure that workers
might be exposed to in addition to exposures from TSCA-
conditions of use. This may result in an underestimation of
risk, and additional discussion of this underestimation has
been added to the document in the Assumptions and Key
Sources of Uncertainty section. EPA relied on NIOSH
guidance in order to establish 10"4 as the cancer risk
benchmark for workers, although acknowledging that other
laws have standards that differ from TSCA's.
In addition to assessing the cancer risk using a linear
extrapolation approach and comparing the results to the
standard cancer benchmark of lxlO"4, EPA also assessed
cancer risk using a threshold approach. Based on the
threshold approach, EPA identified MOEs for cancer risks.
EPA used both the risk estimates derived from the linear
extrapolation approach and the MOEs derived from the
threshold approach for the unreasonable risk determinations
for individuals exposed to carbon tetrachloride.
TSCA section 6(b)(4)(F)(ii) directs EPA to "describe
whether aggregate or sentinel exposures to a chemical
substance under the conditions of use were considered, and
the basis for that consideration" in risk evaluations. EPA
defines aggregate exposures as the combined exposures to an
Page 127 of 210
-------
individual from a single chemical substance across multiple
routes (i.e., dermal, inhalation, or oral) and across multiple
pathways (i.e., exposure from different sources). 40 CFR
702.33. EPA defines sentinel exposures as the exposure from
a single chemical substance that represents the plausible
upper bound of exposure relative to all other exposures
within a broad category of similar or related exposures. 40
CFR 702.33. EPA considered the reasonably available
information and used the best available science to determine
whether to consider aggregate or sentinel exposures for a
particular chemical. EPA has determined that using the high-
end risk estimate for inhalation and dermal risks separately
as the basis for the unreasonable risk determination is a best
available science approach. There is low confidence in the
result of aggregating the dermal and inhalation risks for this
chemical if EPA uses an additive approach, due to the
uncertainty in the data. EPA does not have data that could be
reliably modeled into the aggregate, which would be a more
accurate approach than adding, such as through a PBPK
model. Using an additive approach to aggregate risk in this
case would result in an overestimate of risk. Given all the
limitations that exist with the data, EPA's approach is the
best available science. EPA has added language to the Key
Assumptions and Uncertainties section describing these
assumptions and uncertainties.
39
PUBLIC COMMENTS:
The repetition of the 2010 CC14 IRIS assessment for the
risk evaluation does not fulfill the requirements of the
Lautenberg Act with the use of the best available science
and decisions based on the weight of the scientific
evidence.
EPA has used information consistent with the best available
science, as required by TSCA Section 26(h). EPA
comprehensively reviewed key studies from the 2010 IRIS
assessment in addition to epidemiological and animal studies
as well as invitro information published after publication of
the 2010 IRIS assessment.
Page 128 of 210
-------
39
PUBLIC COMMENTS:
The draft risk evaluation continues to rely on the same
methodology that EPA has followed for 40 years, as
evidenced inter alia by its references to the 2005
Guidelines for Carcinogen Risk Assessment and the 2010
IRIS review of CC14. The methodology incorporated
generic policy choice default assumptions that date from
the 1970s. The criteria for data interpretation and analysis
are policy choices resulting in the regulatory use of an
upper confidence limit value calculated using only a
selected part of the data. This is not in accordance with
TSCA ง 26(h) and (i).
When synthesizing and integrating evidence for each human
health hazard endpoint, EPA considered quality, consistency,
relevancy, coherence and biological plausibility as specified
in Application of Systematic Review in TSCA Risk
Evaluations. Sections 3.2.1 and 3.2.4 describe EPA's process
of weighing and integrating scientific evidence for hazard
endpoints. EPA is developing and implementing more formal
and structured data integration strategies for the next set of
TSCA chemical risk evaluations. In addition, EPA
anticipates feedback from the NASEM TSCA Committee on
its systematic review process and will carefully review and
implement relevant recommendations.
43
PUBLIC COMMENTS:
EPA's risk evaluation should account for acute cancer
risks to workers and consumers. There exists a recognized
methodology for extrapolating from findings of
carcinogenicity in long-term studies to exposures of short
duration (NRC, 2011). Rather than summarily dismissing
acute cancer risks as impossible to estimate, EPA should
have quantified these risks using the framework outlined
by the National Research Council (NRC).
EPA relies on current agency guidance and risk assessment
practice for developing cancer assessments in TSCA risk
evaluations.
Methods used for dermal exposures
SACC,
39, 43,
45
SACC COMMENTS:
The calculation of the dermal slope factor on pp. 134-
135 of the draft risk evaluation is incorrect. To estimate
a slope factor based on absorbed rather than gross dose,
correction for an assumed pulmonary bioavailability of
63% is attempted. However, division by 63, rather than
0.63 results in a 100-fold underestimation of the
dermal slope factor. Since carcinogenic risk is linearly
related to carcinogenic potency, this means that all
worker dermal pathway cancer risk evaluations are also
underestimated by a factor of 100.
Calculations have been corrected in section 3.2.5.2.5 of the
risk evaluation.
Page 129 of 210
-------
The SACC report highlights mistakes in the equation
strings on lines 4334-4342 (pp. 134-135).
A member of the Committee warned that the primary
conclusions of the draft risk evaluation were very
sensitive to the error made in calculation of the dermal
carcinogenic slope factor (DSF). The error (100X) was
so large that no plausible adjustment to the assumed
glove PFs could compensate for it. The final risk
evaluation for CC14 should find unreasonable risk for
all worker conditions of use.
Recommendation: Correct the calculation of the cancer
slope factor for dermal exposure and adjust the risk
calculations accordingly.
PUBLIC COMMENTS:
An error in calculating the cancer slope factor for dermal
exposure resulted in underestimating cancer risk from the
route by two orders of magnitude.
EPA uses Equation 3-2 (p. 134 of the draft risk evaluation)
to calculate a POD for chronic dermal exposures for a
noncancer endpoint. In this equation, the dermal
absorption factor is eliminated because an external
inhalation exposure concentration is extrapolated to a
dermal retained dose. Based on the information provided
in the risk evaluation, the HEDdermai is 31.1 mg/m3 x 1.25
m3/hour x 8 hours/day x 0.63 retained inhaled dose fraction
/ 80 kg = 2.45 mg/kg-day. EPA seems to have used the
percent value (63%) rather than the fraction (0.63),
resulting in the HEDdermai being 100-fold greater or 245
mg/kg-day.
Correcting this error will cause the estimated cancer risk to
Page 130 of 210
-------
increase significantly and impact EPA's determinations of
unreasonable risk for workers and other subpopulations.
The adjustment for inhalation absorption should be 1/0.63,
not 1/63. Thus, the correct dermal cancer slope factor
(CSF) is 8xl0"2 per mg/kg-day.
For the cancer risk estimate, the corrected dermal slope
factor is 8xl0"2 per mg/kg-day as retained dose. Based on
the estimated dermal chronic retained dose for cancer of
0.1 mg/kg-day for central tendency and 0.39 mg/kg-day
high end, the corresponding risk estimates are 8xl0"3 and
3xl0"2, respectively. Thus, as with the chronic noncancer
endpoint, appropriate glove use in a production facility
with a PF of 20 would result in cancer risk close to lxlO"4
for the central tendency and slightly above for the high-end
dermal exposures.
It is unclear why EPA referred to a value of 0.8% to adjust
the IUR for dermal absorption (see p. 149). The dermal
CSF calculated in the risk evaluation is based on the
retained dose from inhalation exposure and is used to
calculate risk from retained dose from dermal exposure
assuming 4% absorption. Use of 0.8% dermal absorption
rather than the 4% value results in an additional 5-fold
reduction in risk.
SACC,
39
SACC COMMENTS:
The Jongeneel study was presented as a Rijksinstituut
voor Volksgezondheid en Milieu (RIVM) letter report
to the National Institute for Public Health, which may
mean that it is considered gray literature rather than
peer reviewed. It was not referenced in PubMed,
making it difficult to determine if it is routinely
referenced as an appropriate approach.
This equation is similar to equations used by other First-10
chemicals {i.e., methylene chloride) risk evaluations.
Nonetheless, the equation was replaced with a peer-reviewed
equation used in previous TSCA risk evaluations.
Page 131 of 210
-------
Recommendation: Provide justification for the use of the
Jongeneel equation to extrapolate chronic inhalation HEC
to chronic dermal HED.
SACC
SACC COMMENTS:
Regarding the POD for occluded conditions:
Using liver toxicity data from a single animal in an
unacceptable study to determine the NOAEL seems
questionable. One cannot assume that induction of liver
toxicity is unlikely for animals dermally exposed for 4
hours to 0.5 ml CC14 just because that toxicity was not
observed at that time point in one animal. A 4-hour
exposure could induce liver toxicity that did not
manifest until a later time point. Thus, this cannot be
used to obtain a NOAEL. Third, the NOAEL seems to
be calculated under the assumption that the animals in
the Kronevi (1979) study were exposed to 0.5 ml of
CC14, when they were in fact exposed to 1.0 ml. This
indicates that a faulty NOAEL of 110 mg/kg/day,
rather than 216 mg/kg/day, was used.
EPA noted that the POD of 2,750 mg/kg/day is similar
to a POD of 2,450 mg/kg/day derived by using the
chronic inhalation values to extrapolate a chronic
dermal value, and then further extrapolating an acute
dermal POD by inexplicably multiplying by a factor of
10. Just because two questionable methods end up with
similar values did not seem to be sufficient justification
for their use.
Recommendations: (1) Explain why a poor-quality study
(Kronevi, 1979) was used to establish the acute dermal
POD when so many other better quality studies were
dismissed; (2) acknowledge that there are insufficient data
to devise an acute dermal NOAEL and POD using the
Kronevi (1979) study; (3) use the LOAEL from the
Non cancer dermal POD is now extrapolated from inhalation
information due to dermal data limitations.
Page 132 of 210
-------
Wahlberg and Boman (1979) study to determine the POD
for acute occluded dermal exposure to CC14; and (4) use
the POD for occluded dermal exposure derived from the
Wahlberg and Boman (1979) LOAEL to calculate a POD
for acute non-occluded dermal exposure.
30
PUBLIC COMMENTS:
EPA explicitly asserts that the inhalation assessment is
protective of heavy alcohol users and is silent on that
point with regard to the dermal assessment, although one
might interpret equivalency.
Information on Intraspecies UF has been updated based on
SACC recommendations, which are applicable to both
inhalation and dermal exposures (See section 3.2.5.2 of the
risk evaluation).
Risk ( h:ii:K
-------
#
Summary of Com incuts lor Specific Issues Related to
Charge Question 5
I PA/OPPI Response
Overall risk approach
SACC,
26, 38,
43
SACC COMMENTS: cancer benchmark
A Committee member found the discussion in Section
5.1.2.2, Determining Cancer Risks (p. 173), to be
unclear and disagreed with the choice of 10"4 as an
acceptable risk.
EPA should also consider the approach described in
Chiu and Crump (2012) in the derivation of unit risk.
PUBLIC COMMENTS: cancer benchmark
EPA cites a NIOSH guidance document that recommends
the use of a 1 in 10,000 cancer threshold when determining
risk management limits (RMLs) for carcinogens. NIOSH,
however, is not required to set RMLs at levels that avoid
unreasonable risk to potentially exposed and susceptible
subpopulations. Moreover, as NIOSH has explained, "[a]n
excess lifetime risk level of 1 in 10,000 is considered to be
a starting point for continually reducing exposures in order
to reduce the remaining risk ... [F]or most carcinogens,
there is no known safe level of exposure ... [and] NIOSH
will continue to recommend that employers reduce worker
exposure to occupational carcinogens as much as possible
through the hierarchy of controls, most importantly
elimination or substitution of other chemicals that are
known to be less hazardous.
EPA's use of a 1 in 10,000 cancer risk level as reasonable
for workers is deeply flawed. EPA's decision is wholly at
odds with its own acknowledgment that other laws have
standards that differ from TSCA's (p. 172, footnote 21).
As noted in the draft risk evaluation, EPA relied on Agency
precedent and NIOSH guidance when choosing the 10"4
cancer risk benchmark to evaluate risks to workers from
carbon tetrachloride exposure. NIOSH's mandate, on pg. iii
of fWhittaker et al.. 2016), is to: "... describe exposure levels
that are safe for various periods of employment, including but
not limited to exposure levels at which no employee will
suffer impaired health or functional capacities or diminished
life expectancy as a result of his work experience." Although
NIOSH guidance, p. 20, states that: "exposures should be
kept below a risk level of 1 in 10,000, if practical [emphasis
added]" EPA adheres to the 1 in 10,000 benchmark during the
risk evaluation stage for TSCA chemicals.
The standard cancer benchmarks used by EPA and other
regulatory agencies range from 1 in 1,000,000 to 1 in 10,000
{i.e., lxlO"6 to lxlO"4) depending on the subpopulation
exposed. EPA has consistently applied a cancer risk
benchmark of lxlO"4 for assessment of occupational scenarios
under TSCA. This is in contrast with cancer risk assessments
for consumers or the general population, for which lxlO"6 is
applied as a benchmark.
EPA, consistent with 2017 NIOSH guidance, used lxlO"4 as
the benchmark for the purposes of unreasonable risk
determinations for individuals exposed to carbon
tetrachloride in industrial and commercial work
environments, including workers and ONUs. lxlO"4 is not a
bright line and EPA has discretion to make unreasonable risk
determinations based on other benchmarks as appropriate.
Page 134 of 210
-------
EPA is required to protect workers, both generally and as a
"potentially exposed or susceptible subpopulation," under
TSCA, not under OSHA. The 2016 amendments to TSCA
strengthened EPA's already-existing mandate to protect
workers. TSCA's new definition of "potentially exposed or
susceptible subpopulation" has no asterisk next to workers,
and there is no basis in TSCA for EPA to provide less
protection to workers than any other such subpopulation,
let alone than the general population. Yet that is exactly
what EPA has done here.
The 2016 amendments to TSCA also explicitly preclude
EPA from considering feasibility or other non-risk factors
when determining whether a chemical presents an
"unreasonable risk," including to workers. EPA cannot
point to any legislative history suggesting that TSCA
adopted OSHA's standard. Moreover, if Congress had
intended to adopt the Benzene standard under TSCA, it
would have required that EPA regulate "significant risks,"
not "unreasonable risks." Indeed, the significant
differences between the language and structure of the two
statutes strongly indicates that Congress meant to adopt a
different standard in TSCA, not the standard articulated by
the Court in the Benzene case. When Congress amended
TSCA to include the unreasonable risk standard, it did so
knowing that agency practice was to regulate cancer risks
at the 10"6 risk level. It should be presumed that Congress
meant to adopt this risk standard when codifying the
unreasonable risk standard.
EPA blurs a critical distinction made when EPA has
invoked the less stringent level of protection from cancer
risks: the level set to reflect the maximum risk faced by
See section 5.1.1.2 of the risk evaluation for additional
information.
In addition to assessing the cancer risk using a linear
extrapolation approach and comparing the results to the
standard cancer benchmark of lxlO"4, EPA also assessed
cancer risk using a threshold approach. Based on the
threshold approach, EPA identified MOEs for cancer risks.
EPA used both the risk estimates derived from the linear
extrapolation approach and the MOEs derived from the
threshold approach for the unreasonable risk determinations
for individuals exposed to carbon tetrachloride.
In consideration of the uncertainties and variabilities in PPE
usage, EPA uses the high-end exposure value when making
its unreasonable risk determination in order to address those
uncertainties.
Page 135 of 210
-------
any individual versus the level set to protect a broader
population. EPA invokes the "two-step approach" used
under the CAA and the two-step, risk-based decision
framework for the National Emission Standard for
Hazardous Air Pollutants (NESHAP). But in this risk
evaluation, EPA has set a risk level for the entire worker
population that is the same as the level that EPA elsewhere
set for the most exposed individual in a population. EPA
then erroneously invokes this level repeatedly to find a
number of conditions of use of CC14 to pose no risk to any
workers, thereby subjecting many tens of thousands of
workers to cancer risks that are as much as two orders of
magnitude higher than warranted. This approach must be
rejected on scientific as well as legal grounds.
EPA should use a benchmark of lxlO"6 to determine
whether cancer risks to workers and consumers are
unreasonable under TSCA. The SACC has previously
stated that EPA has not provided "adequate explanation
and justification" for this reduced threshold and the CC14
draft evaluation also fails to justify EPA's approach.
EPA's recent draft risk evaluations maintains that risks
-------
1,000,000 unreasonable risk threshold, all of those
occupational risks would have been classified as
unreasonable and regulated under TSCA. However,
because EPA used a less protective risk threshold for
workers, no workers who manufacture or directly use CC14
will be protected.
SACC,
23, 26,
30, 32,
38, 42,
43
SACC COMMENTS: aggregate exposure
Recommendation: Consider assessing combined dermal
and inhalation exposure for workers since it is very
unlikely that dermal exposure to CC14 would occur in the
absence of inhalation exposure.
PUBLIC COMMENTS: aggregate exposure
Of greatest concern is EPA's failure to aggregate dermal
and inhalation exposure and derive composite risk
estimates even though the draft risk evaluation indicates
that "inhalation and dermal exposures are assumed to
occur simultaneously for workers." EPA acknowledges
that its "glove protection factors are based on . . . 'what-if
assumptions and are highly uncertain" and that it "does
not know the actual frequency, type, and effectiveness of
glove use in specific workplaces of the occupational
exposure scenarios." Given these admissions, it is hard to
understand how EPA can dismiss aggregate inhalation and
dermal exposure as "highly unlikely." EPA should: (1)
model a broader range of dermal contact scenarios based
on its own analysis of variations in dermal exposure
conditions; and (2) aggregate dermal and inhalation
exposures since these two routes of exposure occur
simultaneously and EPA has no plausible basis to
conclude that use of gloves will prevent dermal contact
with CC14.
TSCA section 6(b)(4)(F)(ii) directs EPA to "describe
whether aggregate or sentinel exposures to a chemical
substance under the conditions of use were considered, and
the basis for that consideration" in risk evaluations. EPA
defines aggregate exposures as the combined exposures to an
individual from a single chemical substance across multiple
routes (i.e., dermal, inhalation, or oral) and across multiple
pathways (i.e., exposure from different sources). 40 CFR
702.33. EPA defines sentinel exposures as the exposure from
a single chemical substance that represents the plausible
upper bound of exposure relative to all other exposures
within a broad category of similar or related exposures. 40
CFR 702.33.
EPA has determined that using the high-end risk estimate for
inhalation and dermal risks separately as the basis for the
unreasonable risk determination is a best available science
approach. There is low confidence in the result of
aggregating the dermal and inhalation risks for this chemical
if EPA uses an additive approach, due to the uncertainty in
the data. EPA does not have data that could be reliably
modeled for the aggregate exposure, which would be a more
accurate approach than adding, such as through a PBPK
model. Using an additive approach to aggregate risk in this
case could result in an overestimate of risk. Given all the
limitations that exist with the data, EPA's approach is the
best available science. EPA has added language to the Key
Page 137 of 210
-------
EPA claims that it "chose not to employ additivity of
exposure pathways ... because of the uncertainties present
in the current exposure estimation procedures that may
lead to an underestimate of aggregate exposure." Even if
combining exposure routes "may lead to an underestimate
of aggregate exposure," the failure to combine routes is
known to lead to an even greater underestimate, since it
unrealistically assumes that no worker will have both
dermal and inhalation exposures. There is no basis for
EPA to rely on false exposure assumptions or to ignore
known combinations of inhalation and dermal exposures
just because the calculation of more accurate, combined
exposures presents "uncertainties."
The lack of aggregation leads to an underestimate of
exposure and risk and, potentially, an incorrect declaration
of "no unreasonable risk" when one actually exists,
assessments. Aggregation can be done relatively easily for
the chronic exposure scenarios. The same study and set of
endpoints are used for both the inhalation and dermal
assessments, as the latter is extrapolated from the same
data used for the inhalation assessment. This is true for
both the noncancer (endpoint = fatty liver) and cancer
(endpoint = increased tumor incidences [liver and
pheochromocytoma])
Some extra effort would be required to do an aggregate
assessment in the case of the acute exposure scenarios,
given that different studies and different endpoints (one
study in humans - neurotoxicity, the other in guinea pigs
- liver) were used to derive PODs for each acute route of
exposure. In addition to doing the necessary math to
convert the administered or internal dose for each route to
Assumptions and Uncertainties section describing these
assumptions and uncertainties.
EPA did not consider background exposure that workers
using products containing carbon tetrachloride might be
exposed to in addition to exposures from TSCA conditions
of use. Risks from background concentrations to carbon
tetrachloride are assessed under the EPA NATA. The 2014
NATA reports a national ambient carbon tetrachloride
concentration of 0.53 |ig/m3 and 3 in a million cancer risk.
https://www.epa. gov/national-air-toxics-assessment/2014-
nata-assessment-results#pollutant. This may result in an
underestimation of risk, and additional discussion of this
underestimation has been added to the document in the Key
Assumptions and Uncertainties section. Clarifying language
on exposure pathways and risks under the jurisdiction of
other EPA-administered statutes have been added to section
1.4.3 of the final risk evaluation document.
The products available for purchase by consumers are not
expected to contain measurable amounts of carbon
tetrachloride because carbon tetrachloride is not used in the
manufacturing of the actual products. Trace levels of carbon
tetrachloride in the chlorinated substances used to
manufacture the products are expected to volatilize during
the product manufacturing process. Furthermore, background
concentrations to carbon tetrachloride are assessed under the
EPA NATA. Therefore, consumer conditions of use were
removed from the risk evaluation in the exercise of EPA's
discretionary scoping authority under TSCA sec. 6(b)(4)(D)
and EPA did not evaluate hazards or exposures to consumers
or bystanders to consumer use in this risk evaluation.
Page 138 of 210
-------
the same metric, a decision would have to be made as to
what the appropriate benchmark MOE would be.
EPA fails to consider the impacts of background CC14
concentrations on the workers and ONUs studied in the
risk evaluation, and thus understates the risks to this
population from aggregate exposure to CC14. EPA lacks
adequate occupational exposure data to support its
findings of no unreasonable risk, and it fails to account for
the background CC14 concentrations that workers are
exposed to outside the workplace. These errors and
omissions understate CC14's occupational risks, in
violation of TSCA's express requirement to protect
workers.
EPA disregards environmental pathways of human
exposure that raise serious health concerns and makes the
mistaken assumption that consumers are not exposed to
CC14.
SACC
SACC COMMENTS: working lifetime
Recommendation: Use a 45-year working lifetime instead
of a 40-year lifetime to align with the NIOSH policy. It
would also be useful to calculate risk using ranges of work
lifetimes.
The cancer assessment for carbon tetrachloride is based on
general exposure frequency (i.e., the amount of days per year
for workers or ONUs exposed to carbon tetrachloride) of 250
days per year and the occupational exposure duration was 40
years over a 70-year lifespan.
The risk evaluation states that it is recognized that these
exposure assumptions are likely yielding conservative cancer
risk estimates, but EPA does not have additional reasonably
available information for further refinement.
26, 30,
39,41,
43
PUBLIC COMMENTS: risk evaluation does not fulfill
statutory requirements
EPA appreciates this feedback. Additional discussion of risk
underestimation has been added to the document in the
Assumptions and Key Sources of Uncertainty section.
Page 139 of 210
-------
Regrettably, the draft risk evaluation does not fulfill the
requirements of the Lautenberg Act. Its hazard assessment
is not based on the best available science; its exposure
assessment does not utilize all of the available
occupational exposure information; and it does not reflect
the current industrial hygiene practices in place at
facilities where CC14 is produced. To maintain the
credibility of its regulatory efforts under TSCA, it is
imperative that EPA build upon the available information
to construct a more realistic risk assessment before
proceeding with rulemaking.
EDF's analyses identify and quantify several major ways
in which EPA has underestimated occupational risks,
including through: its unsupported assumptions regarding
worker use of PPE for all conditions of use; its use of a
cancer risk benchmark level for workers that fails to
protect them as a vulnerable subpopulation as required by
TSCA; its failure to consider combined exposures of
workers from multiple sources; its failure to identify
unreasonable risks for the most highly exposed, and hence
most vulnerable, of workers; and its suggestion that it may
dismiss the few unreasonable risk findings it made by
invoking "uncertainty."
EPA finds CC14 presents risks of concern for some
conditions of use, and particularly for ONUs. However,
due to critical scientific flaws in EPA's risk assessment
approaches that lead to underestimation of risk, the actual
risks are of greater magnitude that stated by EPA and
additional conditions of use present unreasonable risks.
Piecemeal determinations that isolated conditions of use
Under TSCA ง 6(b), EPA is required to conduct risk
evaluations to determine whether a chemical substance
presents unreasonable risk of injury to health or the
environment, under the conditions of use, without
consideration of costs or other non-risk factors, including an
unreasonable risk to potentially exposed or susceptible
subpopulations, identified as relevant to the risk evaluation.
Per 40 CFR 702.47 ".. EPA will determine whether the
chemical substance presents an unreasonable risk of injury to
health or the environment under each condition of use within
the scope of the risk evaluation..This approach in the
implementing regulations for TSCA risk evaluations, is
consistent with statutory text in TSCA section 6(b)(4)(D),
which instructs EPA to conduct risk evaluations to determine
whether a chemical substance presents unreasonable risk of
injury to health or the environment "under the conditions of
use."
Per the statute (see TSCA section 6(b)(4)(A)) and the
implementing regulations for risk evaluations (40 CFR part
702, subpart B), EPA must make the unreasonable risk
determination at the time of the risk evaluation. Upon finding
unreasonable risk, EPA will apply risk management actions
to the extent necessary so that the chemical no longer
presents such risk, in accordance with TSCA section 6(a).
As required by TSCA ง (6)(b), EPA established, by rule, a
process to conduct these risk evaluations. TSCA ง 26(h) and
(i) require EPA, when conducting risk evaluations, to use
scientific information, technical procedures, measures,
methods, protocols, methodologies and models consistent
Page 140 of 210
-------
of CC14 pose "no unreasonable risk" violates TSCA's
plain text. EPA must revise its risk evaluation for CC14 to
make a single risk determination for the chemical
substance as a whole. Based on EPA's findings that some
conditions of use present unreasonable risks to health,
EPA must conclude under TSCA section 6(b)(4)(A) that
CC14 presents an unreasonable risk to human health.
EPA should re-evaluate all conditions of use for both the
worker and ONU populations, implementing
modifications to the exposure assessments, PODs, and
benchmark MOEs recommended above. It is expected that
some number of scenarios would flip from a declaration
of "no unreasonable risk" to one of "an indication of
unreasonable risk," increasing the number of scenarios
requiring risk mitigation.
with the best available science and to base its decisions on
the weight of the scientific evidence. While the law does not
specifically define this term "unreasonable risk", during the
risk evaluation process EPA weighs a variety of factors
including the effects of the chemical on humans or the
environment, the population exposed (including any
sensitive subpopulations), the severity of the hazard, and
uncertainties. This approach is outlined in EPA's 2017
Procedures for Chemical Risk Evaluation Under the
Amended TSCA rule (risk evaluation rule) preamble on how
risk evaluations will be conducted. [82 FR 33726, at 33735
(July 20, 2017)]
To meet these TSCA ง 26 science standards, EPA used the
TSCA systematic review process described in the
Application of Systematic Review in TSCA Risk Evaluations
document. EPA believes the risk evaluation for carbon
tetrachloride meets all requirements for risk evaluations
identified under TSCA and its implementing regulations.
In making the risk determinations, EPA considers relevant
risk-related factors, including, but not limited to: the effects
of the chemical substance on health and human exposure to
such substance under the conditions of use (including cancer
and non-cancer risks); the effects of the chemical substance
on the environment and environmental exposure under the
conditions of use; the population exposed (including any
potentially exposed or susceptible subpopulations (PESS));
the severity of hazard (including the nature of the hazard, the
irreversibility of the hazard); and uncertainties.
EPA considers the uncertainties associated with each
condition of use, and how the uncertainties may result in a
Page 141 of 210
-------
risk estimate that overestimates or underestimates the risk.
Based on such analysis, EPA determines whether or not the
identified risks are unreasonable. Such consideration carries
extra importance when the risk estimates are close to the
benchmarks for risks from acute and chronic non-cancer
health effects and cancer.
To determine whether or not a condition of use presents
unreasonable risks, EPA incorporates assumptions based on
information and judgment underlying the exposure scenarios.
These assumptions, which include assumptions regarding
PPE use, are described in the unreasonable risk
determination for each condition of use, in section 5.2. It is
important to note that the benchmarks for cancer and non-
cancer risk estimates are not bright lines, and EPA has
discretion to make unreasonable risk determinations based
on other risk benchmarks or factors as appropriate.
EPA is making its unreasonable risk determinations on the
high-end exposure value for workers and consumers and
either the high-end exposure value or central tendency for
ONUs, depending on the data, and factoring in the
uncertainties due to UF factors. Additionally, EPA makes an
unreasonable risk determination and makes no determination
on reasonable risk.
26
PUBLIC COMMENTS:
EPA inappropriately fails to find unreasonable risk to
workers despite exceedances of its benchmarks for high-
end exposures. Among other concerns, EPA's approach is
at odds with its obligation under TSCA to conduct risk
evaluations that ensure protection of "potentially exposed
or susceptible subpopulations," which TSCA explicitly
defines as including workers. TSCA does not permit EPA
The use of the high-end exposure value when making the
unreasonable risk determination for workers accounts for
potentially exposed or susceptible subpopulations (PESS).
EPA found that there is unreasonable risk to workers for
dermal exposures. For inhalation exposures, based on the
high-end exposure value, EPA found that there is no
unreasonable risk to workers when assuming use of PPE.
Page 142 of 210
-------
to protect against only the "average or typical exposure;"
in fact, when it comes to workers and other "potentially
exposed or susceptible subpopulations," EPA is required
to protect all of them.
TSCA section 3(12) lists examples of "potentially exposed or
susceptible subpopulations" but neither that provision nor
TSCA section 6(b) specifies subpopulations that must be
considered PESS in any given risk evaluation. EPA therefore
has the discretion to identify PESS that are relevant to a risk
evaluation.
26, 38
PUBLIC COMMENTS:
Despite assuming that ONU exposures "are expected to be
lower than ... exposures for workers directly handling the
chemical," EPA concludes the only ONUs, and not direct
occupational users, face unreasonable risk from CC14.
EPA is clearly suggesting that it may deem these four-fold
exceedances of its own too-lax cancer risk benchmark by
central tendency exposures not to constitute unreasonable
risk because of the uncertainty in its estimates. Set aside
that this uncertainty is the result of EPA having made no
effort to obtain any actual exposure data for ONUs.
EPA's own analyses in these cases showed that CC14
presents an unreasonable risk, but EPA indicates that it
may dismiss this unreasonable risk by invoking uncertainty
in the data. This approach is arbitrary and capricious
because EPA refuses to accept the outcomes of its own
analyses, and EPA's conclusions run contrary to the
evidence before the Agency. Based on the analysis
presented in the draft risk evaluation, EPA should affirm
that an unreasonable risk is presented to ONUs by these
conditions of use.
EPA considers ONUs to be a subset of workers for whom the
potential inhalation exposures may differ based on proximity
to the exposure source.
EPA assumed an absence of PPE for ONUs, since ONUs do
not directly handle the chemical and are instead doing other
tasks in the vicinity of carbon tetrachloride use. EPA also
assumed that, in most cases, ONU inhalation exposures are
lower than inhalation exposures for workers directly handling
the chemical substance. For dermal exposures, because ONUs
are not dermally exposed to carbon tetrachloride, dermal risks
to ONUs were not assessed.
Based on comments received on the draft risk evaluation,
EPA was able to evaluate ONU inhalation exposures
separately from workers for several carbon tetrachloride
conditions of use, including domestic manufacturing.
Consistent with the way that unreasonable risk determinations
are made for workers, for these conditions of use with ONU-
specific exposure estimates, EPA uses the high-end exposure
value when making its unreasonable risk determinations in
order to capture exposures for PESS.
For the rest of the conditions of use, the difference between
ONU exposures and workers directly handling the chemical
cannot be quantified. EPA assumed that, in most cases, ONU
inhalation exposures are lower than inhalation exposures for
Page 143 of 210
-------
workers directly handling the chemical substance. For
inhalation exposures, to account for those instances where,
based on EPA's analysis, the monitoring data or modeling
data for worker and ONU inhalation exposure could not be
distinguished, EPA considered the central tendency risk
estimate when determining ONU risk.
The final unreasonable risk determinations are based on the
risk estimates in the final risk evaluation, which may differ
from the risk estimates in the draft risk evaluation due to peer
review by the SACC and public comments. In the final risk
evaluation, EPA has determined that most of the conditions of
use present unreasonable risks to ONUs.
33
PUBLIC COMMENTS:
EPA should be encouraged to consider conducting a multi-
site risk estimate that accounts for the risks to multiple
sites. Multi-site additivity is used by EPA if the tumors are
occurring in the same strain, sex, and study in animal
laboratory studies; this should have been done for human
epidemiologic data. The lessons from the IRIS chloroprene
assessment can be applied to this CC14 assessment: linear
extrapolation; use of age-dependent AFs; and evaluation of
multi-site cancer risks.
Human epidemiological data on carbon tetrachloride has been
used in a qualitative manner due to data limitations outlined
in the TSCA risk evaluation.
Characterization of uncertainty and assumptions
SACC,
30, 32,
43
SACC COMMENTS: IntrasDecies UF
To clarify the basis for the UFs, the following language was
added to section 3.2.5, Dose Response Assessment:
"EPA applied a composite UF of 30 for the chronic inhalation
benchmark MOE, based on the following considerations:
1) Interspecies uncertainty/variability factor (UFA) of
3 to account for species differences in animal to
human extrapolation. An interspecies
The two UFs generally applied (UFa= 3 and UFh= 10)
do not account for the 10% risk at the BMDL or the
uncertainty as to whether the NOAEL was actually a
no-effect level. Therefore, another factor is needed to
reduce the risk to an acceptable threshold, if a
threshold was being assumed. It could also be argued
that a factor is needed to account for the seriousness of
the health effect.
Page 144 of 210
-------
EPA should review the UFs used for setting the
maximum workplace concentration or Occupation
Exposure Limits for CC14 in Germany, the MAK
(2000), that is based on its potential to cause toxicity,
including tumors, in humans.
Another Committee member noted that Table 1-3 (pp.
27-28) includes assessments done by other countries,
including the Health Canada guidelines for drinking
water. The German MAK assessment should be added
to Table 1-3. In addition, the risk evaluation should
include more details on how the completed
assessments were used in this risk evaluation. A 12-
fold intra-human variability was found in the quantities
of hepatic microsomal CYP2E1 (Snawder and
Lipscomb, 2000). For this reason, the Committee
member questioned if the UF for intra-human
variability should be greater than 10 and suggested that
a factor of 12 be used.
Another Committee member commented that
sensitivity to CC14 toxicity is directly correlated to the
levels of CYP2E1 present in the individual.
Recommendations: (1) Describe what the two UFs (UFa
and UFh) represent and give some basis for their values;
and (2) review and discuss UFs used by other expert
bodies for CC14 and consider any changes needed for this
risk evaluation; explain how assessments from other
jurisdictions were, or were not, considered for this risk
evaluation.
There is a need to expand and clarify the UF
discussions in the risk evaluation. The draft risk
evaluation lacks a separate section that discusses how
UFs should be applied under TSCA. It is difficult to
uncertainty/variability factor of 3 (UFA) was applied
for toxicodynamic differences between species. This
UF is comprised of two separate areas of uncertainty
to account for differences in the toxicokinetics and
toxicodynamics of animals and humans. In this
assessment, the toxicokinetic uncertainty was
accounted for by the PBPK modeling. As the
toxicokinetic differences are thus accounted for, only
the toxicodynamic uncertainties in extrapolating from
animals to humans remain, and an UFA of 3 is
retained to account for this uncertainty.
2) Intraspecies uncertainty/variability factor (UFH)
of 10 to account for variation in sensitivity within
human populations. An intraspecies
uncertainty/variability factor of 10 (UFH) was applied
for toxicokinetics and toxicodynamic differences in
the human population due to humans of varying
gender, age, health status, or genetic makeup might
vary in response to carbon tetrachloride, including
reasonably available quantitative information on
human variability in CYP2E1 enzyme in adults."
The following footnote was added to Table 1-3 (Assessment
History of Carbon Tetrachloride) "The information in this
table is based on Tablel-1 in the Problem Formulation
document and is not meant to be inclusive for all assessments
from other countries."
The following language was added to the PESS section
"Heterogeneity in the human population distribution of
microsomal enzymes metabolizing carbon tetrachloride has
influence in the susceptibility to this chemical because
Page 145 of 210
-------
determine from reading the many sections that discuss
UFs in the draft risk evaluation the extent to which all
pertinent factors are used to inform UFs, and whether
the UFs applied adequately account for the
uncertainties in the data and the methods used to derive
risk estimates.
One Committee member noted that the draft risk
evaluation is inconsistent in discussing uncertainties
and data limitations associated with methodology
limitations, and in particular how this impacts the
assessment of health risks for PESS.
Transparency would be increased by having separate
paragraphs for each of the PESS categories; this is
recommended for alcohol consumption and variability
in CYP2E1 status.
The Committee suggests summarizing the pertinent
information in terms of what is known about each
specific susceptibility category and indicating how this
information has been included in the risk
characterization.
Recommendation: Expand and clarify the UF discussions,
especially regarding PESS.
PUBLIC COMMENTS: Intraspecies UF
The draft evaluation fails to apply UFs necessary to
account for elevated risks to vulnerable subpopulations and
gaps in the CC14 database.
The Agency should provide substantive documentation
that the 10-fold intra-human UF was, in fact, sufficient to
accommodate for the impact of heavy alcohol use - a not-
metabolism is a hypothesized key event in carbon
tetrachloride toxicity. Reasonably available quantitative
information on the variation in human hepatic levels of the
main metabolic enzyme, CYP2E1, demonstrates considerable
intrahuman variability. For example (Lipscomb et at... 1997)
reported a sevenfold range in activity of CYP2E1 among
hepatic microsomal samples from 23 subjects. Snawder and
Lipscomb (Snawder and Lipscomb. 2000) demonstrated 12-
fold differences in CYP2E1 protein content between the
highest and lowest samples from 40 samples of microsomes
from adult human liver organ donors. Consideration of this
PESS quantitative information is incorporated in the UFs
used in the risk characterization."
Section 3.2.5.4 of the final risk evaluation states that the
variability in the response to carbon tetrachloride in relation
to alcohol consumption is emphasized by the fact that an
estimated exposure at 63 ppm-h was fatal in a heavy drinker
whereas controlled exposures at 190 ppm-h were without
effect for individuals not categorized as heavy drinkers. The
following language was added for clarity: "This exposure
information indicates that a 3-fold exposure reduction to the
NOEC value produces an extreme toxic response in heavy
drinkers, suggesting that an UF of 10 for intraspecies
variability is protective of heavy drinkers."
Page 146 of 210
-------
unexpected lifestyle practice of some among the
populations being assessed in this risk evaluation. Without
such documentation, one might consider it appropriate to
expand the UFh to 12-15 and the benchmark MOE to 12-
15 from 10.
As for the acute exposure scenarios, the Agency must
provide adequate documentation that the 10X intra-human
UF adequately covers the special populations that it
acknowledges. Without such documentation, one might
consider it appropriate to expand the range of UFh to 12-
15. The resulting noncancer chronic benchmark MOE,
which would encompass the uncertainties related to
interspecies toxicodynamic and intra-human variability
and database deficiencies, would increase from 30 to 120
or 150 (UFa x UFh x UFd = benchmark MOE: 3.16 x 12 x
3.16 = 120 or 3.16 x 15x3.16 = 150).
EPA has identified specific subgroups with biological
characteristics that make it likely that they will experience
adverse effects from CC14 at lower concentrations than
healthy adults. To provide protection to these groups, a UF
beyond the default intraspecies 10X factor should be
applied, as EPA has previously done for other susceptible
groups such as infants and children. The SACC should
recommend that EPA apply a UF of 20X.
SACC,
23, 30,
32, 43
SACC COMMENTS: In favor of database UF
A Committee member commented that there appears to be
an important data gap and uncertainty about what exposure
level will protect a developing fetus for a pregnant woman
exposed in the workplace.
There is no universal list of hazard data required when
evaluating chemical risks under TSCA. Furthermore, for
carbon tetrachloride, EPA has sufficient, reasonably available
hazard information to conduct a risk evaluation and support
the use of the chosen hazard endpoints. Therefore, EPA did
not use a database UF in the carbon tetrachloride risk
Page 147 of 210
-------
PUBLIC COMMENTS: In favor of database UF
There are no studies that evaluate the potential for
reproductive effects, a significant deficiency, given that
men and women of active reproductive age are likely to be
members of both the worker and ONU populations. A
database deficiency UF >1X (at least 3X) should be
incorporated when deriving the chronic noncancer
benchmark MOE, raising it from the current agency choice
of 30 to at least 100.
The draft risk evaluation identifies developmental toxicity
as an endpoint with limited data, and there is also no
neurodevelopmental toxicity study on CC14, an area of
potential concern given its serious neurotoxic effects. No
endocrine effects data are available either. Given the extent
of these data gaps, we believe a UF of 10 is warranted. The
paucity of any toxicology data on CC14's effects by the
dermal route of exposure, combined with the lack of
dermal absorption studies, create a high level of
uncertainty in EPA's assessment of dermal risks. EPA
should add a UF of 10 to its current benchmark MOEs for
dermal exposure of 100 (acute) and 30 (chronic).
PUBLIC COMMENTS: Against database UF
EPN sees no need for a database UF to be employed in the
acute exposure assessments.
evaluation.
SACC
SACC COMMENTS: Acute UF
A Committee member commented that the NAS in
their recommendations for operating procedures in the
setting of AEGLs (NRC, 2001) provided more leeway
in the choice of UFs than may be indicated by the
Agency's own guidance.
EPA should consider adapting this type of decision
The following language was added to section 3.2.5, Dose
Response Assessment:
EPA applied a composite UF of 10 for the acute inhalation
benchmark MOE, based on the following considerations:
1) Interspecies uncertainty/variability factor (UFA)
Page 148 of 210
-------
roadmap as described in Sections 2.5 and 2.6 in NRC
(2001) in order to increase clarity and transparency
when adopting UFs.
Another Committee member commented that while
EPA used the UF of 10 for acute CNS depression, the
AEGL committee determined that the value of 3 was
sufficient. Therefore, EPA should clarify that the UF
would protect against liver toxicity for all purposes.
A 12-fold intra-human variability was found in the
quantities of hepatic microsomal CYP2E1 (Snawder
and Lipscomb, 2000). For this reason, the Committee
member questioned if the UF for intrahuman
variability should be greater than 10 and suggested
that a factor of 12 be used.
Recommendation: Consider whether additional UFs are
needed.
of 1 Accounting for differences between animals and
humans is not needed because the POD is based on
data from humans
2) A default intraspecies uncertainty/variability
factor (UFH) of 10 To account for variation in
sensitivity within human populations due to limited
information regarding the degree to which human
variability may impact the disposition of or response
to, carbon tetrachloride including reasonably available
quantitative information on human variability in
CYP2E1 enzyme in adults.
43
PUBLIC COMMENTS: Cancer UF
EPA should consider adding a UF to its cancer risk
estimates to acknowledge that they do not account for the
multiple tumor types associated with CC14.
EPA evaluated cancer risk from carbon tetrachloride and
other chemicals using an approach consistent with the EPA
Guidelines for Carcinogen Risk Assessment, thus and
additional UF was not applied.
SACC
SACC COMMENTS:
EPA stated that the conservative assumptions used to
derive PODs were likely to result in overestimation of
risk. However, some Committee members disagreed
with this statement. It was the opinion of some
members, regarding mortality observed in the
Wahlberg and Boman (1979) study (the only dermal
study with an acceptable rating), that it was important
to refrain from underestimating risk.
The Committee also noted that PODs could be
erroneously calculated for acute occluded and non-
The description of uncertainties in dermal risk and dermal
PODs were revised in the risk evaluation
Page 149 of 210
-------
occluded dermal exposure. The Agency should address
over- or underestimating risks prior to its
determination.
Recommendation: Re-evaluate the description of
uncertainties in dermal risk after addressing the faulty
calculations used in estimating the dermal POD.
26
PUBLIC COMMENTS:
To the extent that there are uncertainties in EPA's analysis,
such uncertainties counsel in favor of a finding of
unreasonable risk - EPA could as easily be
underestimating the risk presented by these conditions of
use as overestimating them. Uncertainty increases the
chances of an unreasonable risk; it does not diminish them.
Uncertainty, standing alone, does not justify a finding of
no unreasonable risk when EPA's own analyses support a
finding of unreasonable risk.
To determine whether or not a condition of use presents
unreasonable risks, EPA incorporates assumptions based on
information and judgment underlying the exposure scenarios.
These assumptions, which include assumptions regarding
PPE use, are described in the unreasonable risk determination
for each condition of use, in section 5.2. It is important to
note that the benchmarks for cancer and non-cancer risk
estimates are not bright lines, and EPA has discretion to make
unreasonable risk determinations based on other risk
benchmarks or factors as appropriate.
EPA uses the high-end exposure value when making its
unreasonable risk determination in order to address
uncertainties around PPE usage as well as to capture
exposures for PESS. EPA is making its unreasonable risk
determinations on the high-end exposure value for workers
and consumers and either the high-end exposure value or
central tendency for ONUs, depending on the data, and
factoring in the uncertainties due to UF factors. Additionally,
EPA makes an unreasonable risk determination and makes no
determination on reasonable risk.
38
PUBLIC COMMENTS:
By assuming extensive use of PPE, without any evidence
that is the case, EPA leaves all workers exposed below the
OSHA PEL subject to the voluntary whims of their
employer, with no mandatory, enforceable duty under
either OSHA or TSCA that workers be provided protection
For the purposes of determining whether or not a condition of
use presents unreasonable risks, EPA incorporates
assumptions regarding PPE use based on information and
judgment underlying the exposure scenarios. These
assumptions are described in the unreasonable risk
determination for each condition of use, in section 5.2 of the
Page 150 of 210
-------
against the risks posed by CC14. Leaving workers in this
void violates TSCA. EPA should revise the draft risk
evaluation to address these issues and promptly take action
to eliminate all of CC14's unreasonable risks.
risk evaluation. Additionally, in consideration of the
uncertainties and variabilities in PPE usage, EPA uses the
high-end exposure value when making its unreasonable risk
determination in order to address those uncertainties. EPA has
also outlined its PPE assumptions in section 5.1 of the risk
evaluation. Further, in the final risk evaluation for carbon
tetrachloride, EPA has determined that most conditions of use
pose an unreasonable risk to workers even with the assumed
PPE.
Validity of confidence summaries
SACC,
45
SACC COMMENTS:
A Committee member commented that the confidence
summaries are appropriate as written, while also
expressing the sentiment that it would be more useful
to have confidence expressed in a more quantified
manner.
Another member commented that, in general, there
appears to be discrepancies between the types and
levels of uncertainties described by the Agency, and
the resulting levels of confidence. For example, high
confidence relating to environmental risk appears
overstated given the uncertainties described as related
to environmental risks. Similarly, the high level of
confidence for surrogate scenarios is not well justified.
Specific data inadequacies/uncertainty and assumption
uncertainties are not carried through to confidence
assessment of risk estimates. A formal process needs to
be established, described, and consistently followed.
Recommendations: (1) Section 4.5 of the risk evaluation
should present a more detailed discussion of the links
between uncertainties in exposure as well as hazard
assessment, and the overall level of confidence assigned to
each risk estimate; little was stated about uncertainties in
EPA considered the key assumptions and uncertainties
described in section 4.4 when determining the overall
confidence for the risk estimates.
EPA updated the confidence rating for environmental
receptors in Section 4.5.1 to "moderate" to reflect
uncertainties associated with risk estimates, which are
described in Section 4.1. In addition, a species sensitivity
distribution was added in Appendix F.4, to explore sensitivity
among the most sensitive taxonomic group, amphibians.
Section 4.5 has been edited to include additional discussion of
uncertainties.
Page 151 of 210
-------
the hazard assessment as compared to the exposure
estimation in this section of the draft risk evaluation; and
(2) confidence statements on risk estimates should
synthesize uncertainties in data and assumptions.
Additional clarity is needed on how the uncertainties
propagate and are translated into the levels of
confidence about risk the estimates, or in decisions
about AFs.
Recommendation: Consider scoring data and assumption
uncertainty to derive a final confidence score.
A Committee member indicated that the confidence rating
of "high" presented in Section 3.1.2 (p. 97, lines 3108-
3112) for risk to environmental receptors is not well
supported when compared to statements on p. 160, lines
5173-5179, with respect to confidence in human health
risk, and considering the complexity that includes
environmental breakdown products and potential for
indirect effects (e.g., lack of invertebrate prey base),
neither of which were evaluated. As a result, the
confidence rating of "high" is not well supported. Raising
UFs and/or AFs should correspond to raising confidence
scores.
PUBLIC COMMENTS:
EPA should further explain what constitutes high
confidence. For example, what were the results of the data
quality evaluation, how many acute and chronic
studies/data points were available, were all taxonomic
groups represented (i.e., fish, invertebrates, algae, etc.),
were data consistent and comparable, what were the most
sensitive species (a species sensitivity distribution would
Page 152 of 210
-------
be informative). At present, this section is lacking in
information for the reader to confirm the conclusions in
this section.
Objectivity of assumptions and data
SACC,
26, 38
SACC COMMENTS:
The Committee understood that monitoring and
regulation of ambient air levels of CC14 (and other
similar volatile chemicals) fall under the purview of the
CAA, but this fact should not excuse not including
ambient CC14 concentration in exposure calculations
for workers, ONUs, and consumers. There are concerns
that ambient CC14 concentration values, in some
locations, exposes workers, ONUs, and consumers
living in these areas to greater risk for CC14 and
subsequent health effects.
Recommendation: Include background exposures in the
assessment for workers and ONUs or alternatively provide
a more detailed justification why background exposures
are not considered.
PUBLIC COMMENTS:
According to EPA, "[m]ost risk from NATA background
concentrations is from carbon tetrachloride." EPA has
failed to explain why it completely dismissed background
exposures to CC14 in the draft risk evaluation when the
Agency has, very recently, calculated ongoing risk to the
general population from background exposures to this
chemical. EPA has not explained why, in direct
contradiction to how EPA treated background exposures
from HBCD to the general population, it chose to entirely
ignore background exposures to CC14.
Risks from background concentrations to carbon tetrachloride
are assessed under the EPA NATA. The 2014 NATA reports
a national ambient carbon tetrachloride concentration of 0.53
|ig/m3 and 3 in a million cancer risk.
httos://www.eoa.eov/national-air-toxics-assessment/2014ป
nata-assessment-results#Dollutant.
EPA did not consider background exposure that workers
using products containing carbon tetrachloride might be
exposed to in addition to exposures from TSCA conditions of
use. This may result in an underestimation of risk, and
additional discussion of this underestimation has been added
to the document in the Key Assumptions and Uncertainties
section.
Justification for not including background concentrations is
presented in the final risk evaluation (see section 1.4.2.2).
26
PUBLIC COMMENTS:
HSIA is the main trade association for manufacturers of
halogenated solvents, such as CC14, and, as such, it has a
The data gathering effort to support the risk evaluation was
performed by literature searches and leveraging existing
industry-specific information. HSIA data were provided as
Page 153 of 210
-------
vested interest in EPA finding that the chemicals do not
present unreasonable risk. This vested interest calls into
question the reliability and completeness of the data
voluntarily submitted by HSIA.
part of continuous industrial hygiene monitoring programs
and were evaluated using the same criteria as other data sets.
The reasonably available data readily attributable to
manufacturing and processing of carbon tetrachloride were
limited and contained their own deficiencies (such as the age
of the studies, lack of discrete data points, and no metadata
information) resulting in low quality ratings. Additionally,
limited exposure data exists due to manufacturing,
processing, and use restrictions enforced under the Montreal
Protocol, CAA Title VI, and the Consumer Product Safety
Commission ban.
26
PUBLIC COMMENTS:
In its systematic review process, EPA rated the data that
HSIA submitted in 2019 as 1.8, or "Medium." In doing so,
EPA made some questionable decisions. First, EPA
assigned the data a score of "1" for Geographic Scope
apparently because the data come from U.S. facilities.
However, it appears that the data represent only two
manufacturing facilities (as EPA identifies them only as
"Company A" and "Company B," p. 69), and it is far from
clear that they are at all representative of the entire country
as they comprise only a minority of facilities making or
importing this chemical. Second, as EPA acknowledges in
its systematic review, HSIA has not provided a standard
description of the methods used to collect the data or to
analyze the samples. EPA assigned the data HSIA
submitted in 2019 a "3" for Methodology with the
comment "not specified." However, because of EPA's
approach to weighting criteria, which is inconsistent with
best practices in systematic reviews, this "Low" score for
Methodology has little impact on its overall score.
The assigned scores to both metrics are consistent with the
approaches outlined in the Application of Systematic Review
in TSCA Risk Evaluations document
(httDs://www.eDa.gov/assessing-and-~rnanaging-chernicals-
un der-tsca/ aDoli cati onปsv stem ati c~revi e w-tsca-ri sk-
evaluations). The geographic scope onlv considers whether
the data were collected from site(s) within the United States
to receive a "1" or "high" rating. Considerations of whether
the data addresses variability between sites are considered
when scoring the "variability and uncertainty" metric. This
criterion received a score of "3" or "low" as the data does not
address this topic. As indicated in the comment, the
methodology scored a "3" or "low" as the sampling and
analytical methods used were not specified. Companies A and
B are the only companies manufacturing carbon tetrachloride.
Risk evaluation of potentially exposed or susceptible subpopulations
SACC,
SACC COMMENTS:
Clarifying language on PESS has been added to section
Page 154 of 210
-------
It was not clear as to why Section 2.5.1 appears to be
abbreviated versus later discussions in Section 4.3,
which describes potential PESS within workers and
ONUs.
The draft risk evaluation lists the variables that define
PESS in both Sections 3.2.5.4 and 4.3. However, each
category is described to different extents, in some
places extensively, in other places briefly, and
occasionally not at all. For example, there are no
descriptive discussions regarding subpopulations with
pre-existing disease, beyond identifying the
subpopulations as a category. The paragraph in lines
4964-4966 does not offer additional explanations in
Section 4.3 due to workers and ONUs as being
identified as PESS earlier in previous paragraphs.
The PESS section does not mention if intraspecies UFs
of 10 were applied; UFs of 10X are generally used to
account for variation among people and not for known
PESS.
The Agency describes how a UF was used to account
for variations of sensitivity, but it is not clear whether
EPA did a separate assessment for these more
susceptible individuals. It would seem that EPA is
considering them as part of the workers and ONU
groups, but this explanation is not clear in the
document.
A Committee member noted that the discussion of
PESS appears disjointed and not very compelling
without a risk evaluation for susceptible populations.
Recommendation: Consolidate explanation of PESS into
one section and develop more protective guidelines for
PESS.
3.2.5.4
Page 155 of 210
-------
PUBLIC COMMENTS:
Under TSCA, EPA must account for and protect not only
exposed workers, but also those subpopulations who are
most susceptible to a chemical's risks. The draft CC14 risk
evaluation fails to do so.
SACC
SACC COMMENTS:
The data on cancer endpoints suggest that there may be
differences with age (adults versus children for brain
cancer, for example), race (Japanese Americans versus
White Americans), and metabolic germline
polymorphisms. None of this is discussed or analyzed
in depth in this document.
There are novel genome-wide association studies
(GWAS) studies that suggest genetic differences that
may modulate acute exposure effects. These should be
identified and discussed.
Recommendation: The discussion on PESS should include
subgroups and conditions identified in epidemiologic
studies and in more recent GWAS research.
The brain cancer studies were all conducted in adult
populations so there were no differences by age. While one of
the studies (Nelson et al.! ) reported increased risks
within a cohort of Japanese American men, other studies
reported increased risks among people from across the U.S.,
which did not suggest differences by race.
SACC
SACC COMMENTS:
Hereditary hemochromatosis is an autosomal recessive
disorder that affects about 1 in 200-500 individuals.
Those who are either homozygous or heterozygous for
this condition should be included among the groups
that would be more sensitive to CC14-induced
oxidative and peroxidative damage.
Another Committee member commented that the
embryo and fetus (pregnant female workers) should be
considered a PESS based on the neuroblastoma risk.
Recommendation: Consider including and discussing
individuals who are sensitive to oxidative damage and the
embryo/fetus of pregnant female workers as PESS.
The following language was added to the PESS section:
Heterogeneity in the human population distribution of
microsomal enzymes metabolizing carbon tetrachloride has
influence in the susceptibility to this chemical because
metabolism is a hypothesized key event in carbon
tetrachloride toxicity. Reasonably available quantitative
information on the variation in human hepatic levels of the
main metabolic enzyme, CYP2E1, demonstrates considerable
intrahuman variabilitv. For example (Lipscomb et al.. 1997)
reported a sevenfold range in activity of CYP2E1 among
hepatic microsomal samples from 23 subjects. Snawder and
Lipscomb (Snawder and Lipscomb. 2000) demonstrated 12-
fold differences in CYP2E1 protein content between the
highest and lowest samples from 40 samples of microsomes
Page 156 of 210
-------
from adult human liver organ donors. Consideration of this
PESS quantitative information is incorporated in the UFs
used in the risk characterization. Qualitative information on
susceptibility could not be incorporated in the UFs.
32, 38,
43
PUBLIC COMMENTS:
There are two significant ways in which the draft risk
evaluation uses insufficiently protective UFs and
understates risks as a result. "[CJases of acute toxicity from
occupational exposures indicate that heavy drinkers are
more susceptible to carbon tetrachloride and this
observation has been verified in numerous animal studies."
In addition, "reduced kidney function and increased
CYP3 A activity in the liver (indicated by animal studies)
suggest that older populations could be at greater risk of
carbon tetrachloride-associated kidney damage."
In its draft risk evaluation, EPA identified "human
subpopulations that may have greater susceptibility to
carbon tetrachloride than the general population,"
including moderate to heavy alcohol users, people with
preexisting liver disease, and populations with certain
genetic polymorphisms. However, EPA does not evaluate
the risks facing these specific subpopulations, but instead
relies on a default intraspecies UF to account for all of
them. For instance, EPA does not consider alcohol
consumption rates within the exposed worker population
or separately adjust its risk calculations to account for
these susceptibilities. Under TSCA, EPA must calculate
risks for these PESS, or at a minimum demonstrate that its
chosen UF is sufficient to account for all such populations.
Section 3.2.5.4 of the final risk evaluation states that the
variability in the adverse response to carbon tetrachloride
exposure in relation to alcohol consumption is emphasized by
the fact that an in a heavy drinker whereas controlled
exposures at 190 ppm-h were without effect for individuals
not categorized as heavy drinkers.
The following language was added for clarity: This exposure
information indicates that a 3-fold exposure reduction to the
NOEC value produces an extreme toxic response in heavy
drinkers, suggesting that a UF of 10 for intraspecies
variability is protective of heavy drinkers.
30
PUBLIC COMMENTS:
This draft risk evaluation includes the assessment of risk to
workers and ONUs from acute and chronic inhalation and
EPA did not identify any legacy uses or associated disposals
for carbon tetrachloride. EPA did not assess exposures from
legacy disposals, or disposals that have already occurred,
Page 157 of 210
-------
dermal exposures. However, neither pregnant women nor
male workers considering a family nor the general
population, which also includes infants and young
children, have been specifically addressed. This becomes
particularly important once the risk evaluation is updated
to include the analysis of legacy consumer conditions of
use.
because they are not considered conditions of use. Clarifying
language about what pathways are under the jurisdiction of
other EPA-administered statutes has been added to Section
1.4.3 of the Risk Evaluation.
EPA did not evaluate risks to the general population from any
conditions of use and the unreasonable risk determinations do
not account for exposures to the general population.
Additional details regarding the general population are in
Section 1.4.3.
The products available for purchase by consumers are not
expected to contain measurable amounts of carbon
tetrachloride because carbon tetrachloride is not used in the
manufacturing of the actual products. Trace levels of carbon
tetrachloride in the chlorinated substances used to
manufacture the products are expected to volatilize during the
product manufacturing process. Furthermore, background
concentrations to carbon tetrachloride are assessed under the
EPA NAT A. Therefore, consumer conditions of use were
removed from the risk evaluation in the exercise of EPA's
discretionary scoping authority under TSCA sec. 6(b)(4)(D).
EPA does account for exposures to potentially exposed or
susceptible subpopulations (PESS) by using the high-end
exposure value when making its unreasonable risk
determination for workers.
38
PUBLIC COMMENTS:
The statute specifically defines PESS to include "workers,"
reflecting Congress's intent that EPA evaluate and address
occupational risks under TSCA.
EPA identified the following potentially exposed or
susceptible subpopulations based on their greater exposure to
carbon tetrachloride: workers and ONUs.
TSCA section 3(12) lists examples of "potentially exposed or
susceptible subpopulations" but neither that provision nor
Page 158 of 210
-------
TSCA section 6(b) specifies subpopulations that must be
considered PESS in any given risk evaluation. EPA therefore
has the discretion to identify PESS that are relevant to a risk
evaluation.
26
PUBLIC COMMENTS:
Workers at any facility - whether small, medium, or large
- where use of effective PPE cannot be thoroughly
documented should be considered vulnerable
subpopulations and the risk they face be specifically
assessed. For these subpopulations, EPA must determine
risk based on exposures without assuming any use of PPE.
PPE use expectation is applicable to all facilities (OSHA
regulations cover large and small facilities).
EPA has recognized in the draft and final risk evaluations
OSHA's hierarchy of controls and recognized that there can
be reliability issues associated with PPE. EPA's risk
evaluation characterizes risks with and without PPE
considerations, with considerations of engineering and
administrative controls. Additionally, in consideration of the
uncertainties and variabilities in PPE usage, EPA uses the
high-end exposure value when making its unreasonable risk
determination in order to address those uncertainties. EPA has
also outlined its PPE assumptions in section 5.1.
26,41,
42
PUBLIC COMMENTS:
As part of this analysis, EPA should identify people living
near all disposal sites containing CC14 as PESS. These
groups include (but are not limited to) those living near
Superfund or NPL sites. To be clear, many disposal sites
are associated with activities that reflect ongoing or
prospective manufacturing, processing, distribution, or use,
so EPA must also analyze those disposals and disposal
sites and populations living in proximity to them.
Additionally, EPA should include all communities living
near facilities that report releases of CC14 under TRI.
In order to make an accurate risk characterization of tribal
communities, EPA needs to consider releases of CC14
from landfills. In the case of many tribal and rural
communities, the disposal site may be in close proximity to
Clarifying language on exposure pathways and risks under
the jurisdiction of other EPA-administered statutes have been
added to section 1.4.3 of the final risk evaluation document.
EPA did not identify any legacy uses or associated disposals
for carbon tetrachloride. EPA did not assess exposures from
legacy disposals, or disposals that have already occurred,
because they are not considered to be "conditions of use."
Page 159 of 210
-------
residents, may be unlined, may be open access, and may
include open burning as a management practice, all of
which present multiple exposure pathways and routes for
intake and uptake. It cannot be assumed that all CC14
product disposal would be at Subtitle C landfills. For
example, there is not a single Subtitle C landfill in the
State of Alaska. The multiple exposure scenarios
associated with proximity to unlined disposal site releases
to environmental media must be analyzed for both
individual exposures and the cumulative exposures tribal
members face from their customary and traditional tribal
lifeways (inhalation, dermal, ingestion).
EPA provides no analysis of whether those living in
proximity to the conditions of use are at greater risk due to
greater exposure. EPA should analyze these exposures and
should analyze these potentially exposed subpopulations.
EPA's failure to consider this relevant aspect of the
problem is arbitrary and capricious.
42
PUBLIC COMMENTS:
Tribal lifeways can lead to chronic exposures to toxins in
the environment, due to the much longer duration and
higher frequency of exposures tribal people may
experience, as well as the higher cumulative dose from
multiple exposure pathways {i.e., differences in diet,
housing, worker safety, and water use). Tribes must be
considered as a sensitive subpopulation under TSCA.
NTTC has in previous comment letters informed EPA in
detail about the unique characteristics of disposal sites on
tribal lands and in tribal communities and we are able and
willing to provide extensive photographic and narrative
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation.
EPA did not identify any legacy uses or associated disposals
for carbon tetrachloride. EPA did not assess exposures from
legacy disposals, or disposals that have already occurred,
because they are not considered to be "conditions of use."
EPA did not evaluate risks to the general population from any
conditions of use and the unreasonable risk determinations do
not account for any risks to the general population. Additional
details regarding the general population are in Section 1.4.3.
Because "the term 'potentially exposed or susceptible
Page 160 of 210
-------
evidence that exposure through disposal is very likely for
tribal people.
subpopulation' means a group of individuals within the
general population identified by the Administrator who, due
to either greater susceptibility or greater exposure, may be at
greater risk than the general population of adverse health
effects from exposure to a chemical substance or mixture,
such as infants, children, pregnant women, workers, or the
elderly," EPA believes that the statutory directive to consider
potentially exposed or susceptible subpopulations (PESS) and
the statutory definition of PESS inherently include tribes.
Therefore, the UF applied to account for PESS do cover tribal
exposures.
EPA did not consider background exposure that workers
using products containing carbon tetrachloride might be
exposed to in addition to exposures from TSCA conditions of
use. This may result in an underestimation of risk, and
additional discussion of this underestimation has been added
to the document in the Key Assumptions and Uncertainties
section.
42
PUBLIC COMMENTS:
The SACC has previously stated that EPA must consider
all exposure routes and give "special consideration to
specific populations (e.g., tribal, arctic inhabitants, etc.)
who depend on fish as a major source of food because of
cultural considerations and provide some quantitative
sense of how much extra risk exists for these populations.
In considering special and susceptible population
exposures, more consideration needs to be given to
populations with specific preexisting conditions, such as
metabolic disease and obesity, as well as to tribal, ethnic
and other subpopulations that depend heavily on
potentially contaminated foods, such as Native American
subsistence fishers."
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation.
EPA does account for exposures to potentially exposed or
susceptible subpopulations (PESS) by using the high-end
exposure value when making its unreasonable risk
determination for workers. Because "the term 'potentially
exposed or susceptible subpopulation' means a group of
individuals within the general population identified by the
Administrator who, due to either greater susceptibility or
greater exposure, may be at greater risk than the general
population of adverse health effects from exposure to a
chemical substance or mixture, such as infants, children,
Page 161 of 210
-------
Clearly, tribes experience exposures even where
responsibility rests on other environmental statutes
(RCRA, SDWA, CWA CAA) and NTTC strongly urges
EPA to comply with their statutory obligation to consider
all exposures, particularly for susceptible and highly
exposed populations, such as tribes.
NTTC has expressed concern at the paucity of data on
tribal risks, as well as the observation that tribal people are
absent from or underrepresented in EPA's risk evaluations
and proposed actions. It is well documented in the
scientific literature that Native Americans experience
significant health disparities from the general population
and the practice of leaving them out of any protections will
only contribute to further health disparities.
NTTC urges EPA to include consideration of Tribal data
that may be submitted by the Tribe that produced it. Where
data are not available, modeling should be employed so
that all significant Tribal exposures are captured.
pregnant women, workers, or the elderly," EPA believes that
the statutory directive to consider potentially exposed or
susceptible subpopulations (PESS) and the statutory
definition of PESS inherently include tribes.
In addition, based on its physical-chemical properties, carbon
tetrachloride does not partition to lipid or bioaccumulate in
fish (BCF is estimated at 40, whereas the threshold for
bioaccumulative chemicals is 1,000). Therefore, elevated fish
consumption by individuals {i.e., such as indigenous
populations) is not a factor for carbon tetrachloride
susceptibility.
Residual concentrations of carbon tetrachloride in surface
waters not used for drinking water are also regulated via the
CWA Ambient Water Quality Criteria for human health
consumption of water and organisms (0.4ug/L). CWA(a)(l).
42
PUBLIC COMMENTS:
The SACC also recommended that "the context of the
assessment would be improved by including a graphic
similar to the one presented by the National Tribal Toxics
Council at the public meeting, that illustrates exposure
routes for potentially sensitive or highly exposed
populations" (reference to the conceptual model).
EPA appreciates this suggestion, which will be considered for
future risk evaluations.
42
PUBLIC COMMENTS:
In this draft risk evaluation, EPA limited its analysis to
only considering people who have higher susceptibility to
CC14 due to genetic polymorphism in its metabolizing
enzymes. However, other than the consideration of worker
Section 3.2.5.4 explains how PESS quantitative information
is incorporated in the UFs used in the risk characterization.
Qualitative information on susceptibility could not be
incorporated in the UFs.
Page 162 of 210
-------
and ONU exposures, EPA did not consider whether any
subpopulations might face greater risk due to greater
exposure to CC14. EPA must consider and analyze each of
these types of subpopulations, as mandated by the
Lautenberg Act.
Risk evaluation of workers with PPE
SACC,
23, 26,
30, 32,
38, 42,
43
SACC COMMENTS:
EPA is not adequately considering the hierarchy of
controls in occupational hygiene and is emphasizing the
last step, which is PPE.
Hard empirical evidence for assumed levels of PPE
efficacy linked to the conditions of use being described is
not provided. The Agency relies upon generic tabulated
values. This approach entails substantial uncertainty.
EPA is not adequately considering issues of training,
availability of appropriate materials, and human factors in
compiling tables of PPE efficacy. Discussion on pp. 62-63
of the draft risk evaluation describes results of a NIOSH
survey of U.S. employers regarding the use of respiratory
protective devices between August 2001 and January 2002
that suggested that full adherence to best PPE practice is
likely a minority occurrence. Estimation of central
tendency and high-end exposures with assumption that
high degrees of protection are routinely achieved is
problematic.
Recommendation: Provide a brief description of the
rationale for linking the information on occupational
exposure control to the decision to apply respirator and
glove PFs.
OSHA's hierarchy of controls is a method for eliminating
workplace hazards. EPA will manage unreasonable risks
presented by chemical substances when the Agency
undertakes regulatory action for conditions of use determined
to have unreasonable risk. Utilization of the hierarchy of
controls to recommend or require risk management actions in
the risk evaluation would be premature and inappropriate.
Assumed PPE is reflected by the type of use, whether
industrial, commercial, or consumer, and the anticipated
presence of an industrial hygiene program. EPA does not
assume that the use of SDSs are sufficient to ensure PPE use
and EPA does not make PPE use assumptions based on SDSs.
The OSHA regulations at 29 CFR 1910.132 require
employers to assess a workplace to determine if hazards are
present or likely to be present which necessitate the use of
PPE. If the employer determines hazards are present or likely
to be present, the employer must select the types of PPE that
will protect against the identified hazards, require employees
to use that PPE, communicate the selection decisions to each
affected employee, and select PPE that properly fits each
affected employee. OSHA has established a PEL of 10 ppm
(8-hour TWA) for carbon tetrachloride at 29 CFR 1910.1000.
However, as noted on OSHA's website, "OSHA recognizes
that many of its PELs are outdated and inadequate for
ensuring protection of worker health. Most of OSHA 's PELs
were issued shortly after adoption of the Occupational Safety
Page 163 of 210
-------
PUBLIC COMMENTS:
The CC14 risk evaluation provides a detailed discussion of
the role of PPE in workplace protection strategies (i.e.,
hierarchy of controls), which demonstrates that PPE are
not a substitute for more effective controls on workplace
exposure and that there is considerable uncertainty about
whether PPE is consistently used even where legally
required. Thus, to rely entirely on PPE without first
requiring engineering controls and other protections - as
EPA effectively does in the CC14 risk evaluation - is
contrary to accepted principles of worker protection.
EPA relies on OSHA's Hazard Communication Standard
to support its "expectation]" that workers will be provided
"appropriate PPE consistent with the applicable SDSs in a
manner adequate to protect them." However, the Hazard
Communication Standard merely requires the provision of
SDSs, not PPE, and OSHA has made clear that employers
are under no obligation to follow the recommendations in
an SDS. In the absence of such a requirement, there is no
basis for EPA's assumption that the Hazard
Communication Standard will result in the uniform use of
appropriate PPE.
The information and recommendations included in SDSs
are based on manufacturers' judgment. As a result, they
are often vague and inconsistent. Further, SDS
recommendations are not binding on employers. EPA has
no basis for assuming that specific glove PFs in its draft
risk evaluation.
and Health (OSH) Act in 1970, and have not been updated
since that time." OSHA provides an annotated list of PELs on
its website, including alternate exposure levels. For carbon
tetrachloride, the alternates provided are the California OSHA
PEL of 2 ppm and the ACGIH TLV of 5 ppm.
(https://www.osha.eov/dsg/aimotated-pels/tablez-Lhtml).
EPA agrees that there are challenges associated with use of
PPE; they are described in Section 5.1.1.3. By providing risk
estimates that account for use of PPE, EPA is not
recommending or requiring use of PPE. Rather, these risk
estimates are part of EPA's approach for developing exposure
assessments for workers that relies on the reasonably
available information and expert judgment.
When appropriate, in the risk evaluation, EPA will use
exposure scenarios both with and without engineering
controls and/or PPE that may be applicable to particular
worker tasks on a case-specific basis for a given chemical.
Again, while EPA has evaluated worker risk with and without
PPE, as a matter of policy, EPA does not believe it should
assume that workers are unprotected by PPE where such PPE
might be necessary to meet federal regulations, unless it has
evidence that workers are unprotected.
For the purposes of determining whether or not a condition of
use presents unreasonable risks, EPA incorporates
assumptions regarding PPE use based on reasonably available
information and professional judgment underlying the
exposure scenarios. These assumptions are described in the
unreasonable risk determination for each condition of use, in
section 5.2. In the case of carbon tetrachloride, which is
manufactured, processed, and used in industrial settings,
Page 164 of 210
-------
EPA assumes that workers will not only be provided with
appropriate respirators with an average PF up to 50 and
chemical-resistant gloves with a PF up to 20, but will
receive the training, fit testing, and medical evaluations
required to ensure proper respirator use. Does the draft risk
evaluation provide adequate support for those
assumptions? EPA assumes that workers exposed to CC14
will wear respirators. These assumptions are legally and
factually baseless.
Small facilities are much less likely to require routine and
effective use of PPE or to employ engineering controls,
such as closed systems. Smaller businesses and facilities
are the norm in Indian Country, including Alaska Native
villages, and they are subject to OSHA exemptions to the
Respiratory Protection Standard, as well as to reporting
and inspection requirements. Self-employed workers are
also exempt from many OSHA requirements and self-
employment is common in tribal communities. For
accurate risk characterization of tribal members, NTTC
would like to see a risk determination for workers and
ONUs, both self-employed and in small businesses, that
incorporates OSHA's exemptions and practical exceptions.
In these communities, take-home exposures are also very
likely.
CC14 is produced and used by thousands of workers across
a range of different sectors. Even within a given condition
of use (e.g., disposal and waste handling), there often are a
wide range of employers and workplaces. However, EPA
arbitrarily assumes that all workers will be provided with,
and will use, PPE, without any supporting evidence. EPA
must make risk determinations about CC14 use under the
where there are typically strong industrial hygiene programs
that include training and oversight, EPA believes that it is
reasonable to assume a PF of 20 for dermal protection
(gloves) and APF of 50 for inhalation protection (respirators).
Additionally, in consideration of the uncertainties and
variabilities in PPE usage including the duration of PPE
usage, EPA uses the high-end exposure value when making
its unreasonable risk determination in order to address those
uncertainties.
EPA has also outlined its PPE assumptions in section 5.1.
Further, in the final risk evaluation for carbon tetrachloride,
EPA has determined that most conditions of use pose an
unreasonable risk to workers even with the assumed PPE.
Page 165 of 210
-------
foreseen (and known) circumstances, assuming such
respirators are not worn.
Mendel off et al. (2013) noted that "the pattern of
noncompliance across industries mostly mirrored the
survey findings about the prevalence of requirements for
respirator use." Based on this study EPA concluded "The
likelihood of respirator use may not be widespread or
effective" (p. 63).
EPA identifies no data concerning the use of respirators by
workers exposed to CC14, and it acknowledges "the
likelihood of respirator use may not be widespread or
effective." In the absence of chemical specific data, EPA
relies on a generic 2003 NIOSH survey, which reports that
among the small fraction of employers that require
respirators, most do not conduct the planning, training, and
testing required to ensure that respirators are serving their
intended function. These data show wide gaps in use of
appropriate respirators and measures of effectiveness.
EPA has previously acknowledged that "not all workers
may be able to wear respirators." In particular, EPA
explained that "[iIndividuals with impaired lung function
due to asthma, emphysema, or chronic obstructive
pulmonary disease ... may be physically unable to wear a
respirator." Workers' facial hair, including beards and
sideburns, can also interfere with the seal of a respirator,
rendering it ineffective. Other workers cannot wear
respirators because they "may also present communication
problems, vision problems, worker fatigue, and reduced
work efficiency." OSHA and NIOSH have similarly found
that respirators can cause discomfort, skin irritation, heat
Page 166 of 210
-------
stress, communication difficulties, and vision limitations,
and that they often create other hazards for workers, such
as trips, falls, and "struck by" hazards.
NIOSH has found that respirator programs often provide
inadequate protection even where respirator use is legally
required and there is serious doubt whether respirator use
at many facilities is consistent, reliable, and protective.
There is no basis for EPA to assume that employers will
voluntarily exceed the OSHA standard and provide
additional respiratory protection to eliminate the risks
below the PEL.
EPA proposes to determine that CC14's risks to workers
are not unreasonable where the "expected" use of
respirators and gloves would reduce exposures to levels
that provide "acceptable" MOEs and cancer risk levels as
compared to EPA's benchmarks. However, as the SACC
has repeatedly underscored and EPA's draft evaluations
recognize, this "expectation" of universal PPE use is not
grounded in data, departs from established workplace
protection policy, and is contrary to the realities of worker
exposure to unsafe chemicals. Risk estimates should be
presented without the use of PPE as reasonable worst case.
This will result in a determination that workers are at
unreasonable risk from CC14 (cancer and noncancer risks).
The worker protection measures necessary to protect
workers from this risk should be in risk management
rulemaking under TSCA section 6(a).
SACC
SACC COMMENTS:
Recommendation: EPA should replace the assumed APFs
in Table 4-13 with data-based estimates. If no reliable
estimates can be developed, only risk estimates assuming
EPA considers ONUs to be a subset of workers for whom the
potential inhalation exposures may differ based on proximity
to the exposure source. EPA assumed an absence of PPE for
ONUs, since ONUs do not directly handle the chemical and
Page 167 of 210
-------
no PPE use should be presented, with appropriate caveats
in the discussion.
SACC members noted that ONU exposures were
deemed unreasonable, but worker exposures were
deemed reasonable. Workers are assumed to be
exposed both via inhalation (in higher concentration
environments than ONUs) and by dermal contact to
liquids (while ONUs are not). The counterintuitive
finding that ONUs are at higher risk highlights the
assumption made that workers routinely have access to
appropriate PPE and use it effectively. Several SACC
members expressed doubts regarding this assumption.
(It was noted that the finding of no unreasonable risk to
workers via dermal contact to liquid was an artifact of
an error in calculating the cancer slope factor; see
charge question 4.)
are instead doing other tasks in the vicinity of carbon
tetrachloride use. For dermal exposures, because ONUs are
not dermally exposed to carbon tetrachloride, dermal risks to
ONUs were not assessed.
Based on comments received on the draft risk evaluation,
EPA was able to evaluate ONU inhalation exposures
separately from workers for several carbon tetrachloride
conditions of use, including domestic manufacturing.
Consistent with the way that unreasonable risk determinations
are made for workers, for these conditions of use with ONU-
specific exposure estimates, EPA uses the high-end exposure
value when making its unreasonable risk determinations in
order to capture exposures for PESS.
For the rest of the conditions of use, the difference between
ONU exposures and workers directly handling the chemical
cannot be quantified. EPA assumed that, in most cases, ONU
inhalation exposures are lower than inhalation exposures for
workers directly handling the chemical substance. For
inhalation exposures, to account for those instances where,
based on EPA's analysis, the monitoring data or modeling
data for worker and ONU inhalation exposure could not be
distinguished, EPA considered the central tendency risk
estimate when determining ONU risk.
In the risk evaluation for carbon tetrachloride, EPA used the
high-end exposure value when considering worker risks in
order to address the uncertainties and variability in PPE
usage. For inhalation exposures, EPA, where possible,
estimated ONU exposures and described the risks separately
from workers directly exposed. To account for those instances
where, based on EPA's analysis, the monitoring data or
Page 168 of 210
-------
modeling data for worker and ONU inhalation exposure could
not be distinguished, EPA considered the central tendency
risk estimate when determining ONU risk.
EPA considered the high-end no PPE scenario within each
OES as the sentinel exposures for workers. In presenting the
inhalation results, high intensity use was characterized by the
model iteration that utilized the 95th percentile duration of
use and mass of product used and the maximum weight
fraction derived from product specific SDS, when available.
Dermal exposures for high intensity use were characterized
by the model iteration that utilized the 95th percentile
duration of use and maximum weight fraction.
SACC
SACC COMMENTS:
There are gaps between the description of the exposure
control hierarchy and the application of the PPE PFs that
reduce clarity. The risk evaluation should provide 1-2
paragraphs describing the decision process between
acknowledgement of guidelines for exposure control and
the application of PFs for PPE.
A Committee member indicated that the description of
exposure controls, PPE, and the effectiveness of PPE
should also be briefly summarized in the risk
characterization, even if it is discussed in detail elsewhere
in the document.
See the Executive Summary, updated Risk Characterization
(Section 4), and updated Risk Determination (Section 5) for
more clarification on how these sections support each other
and how new information submitted to or obtained by EPA
following publication of the draft risk evaluation is
incorporated.
SACC
SACC COMMENTS:
The Committee is concerned about the use of respirator
PFs, particularly for exposures for manufacturing and
processing as reactant/intermediate (8- and 12-hour
TWA). Even though EPA estimated high-end chronic
inhalation exposures with noncancer MOEs below the
benchmark and cancer risks greater than the benchmark.
EPA did assess the risk to workers in the absence of PPE and
with PPE; those risk estimates are in Tables 4-7 through 4-13
in Section 4, Risk Characterization.
EPA considers each condition of use and uses exposure
scenarios with and without PPE that may be applicable to
particular worker tasks on a case-specific basis for a given
Page 169 of 210
-------
EPA drew a conclusion that there was no unreasonable
risk in any of the manufacturing scenarios because of the
exposure reductions expected as a result of use of
respirators, which lead to MOEs greater than the
benchmark MOE (Section 4.2.8, p. 158, Table 4-13). In a
similar manner, there were cancer risks above the cancer
risk benchmark for the high-end exposures for the
additive, processing agent/aid, import and repackaging,
specialty uses-Department of Defense (DoD), and
disposal/recycling conditions of use, but EPA also
accounted for the use of respirator PFs to conclude that
the cancer risk was below the benchmark.
chemical. For the purposes of determining whether or not a
condition of use presents unreasonable risks, EPA
incorporates assumptions regarding PPE use based on
information and judgment underlying the exposure scenarios.
SACC
SACC COMMENTS:
EPA does not discuss glove life/replacement when faced
with chemical and physical challenges.
Proper care, maintenance, useful life and disposal of PPE are
recommended by OSHA. Several OSHA citations included
in the risk evaluation document indicate recommended
practices. The glove replacements are generally included in
the industry-specific health and safety plan. These
discussions are not within the scope of risk evaluation.
SACC
SACC COMMENTS:
Recommendation: EPA should provide sufficient detail on
the use of the conceptual model in Cherrie et al. (2004) so
that a reader could reproduce the values reported in Table
2.3.
If glove PFs depend upon flux and time, an
explanation is needed as to why the values reported in
Table 2.3 depend on neither.
The citations of relevant peer-reviewed articles are included
in the risk evaluation document. The Table 2-5 (Table 2-3 in
the previous version of the draft risk evaluation document)
includes the cited reference.
SACC
SACC COMMENTS:
A Committee member commented that any use of glove
PFs listed in various tables or in discussion should clearly
reference OSHA guidelines.
The relevant source of the glove PFs cited as footnote of the
Table 2-5 as suggested.
SACC
SACC COMMENTS:
A Committee member stated as part of its Risk21ฎ effort,
the Health and Environmental Sciences Institute (HESI)
EPA appreciates the information on different tools for
conveying risk with and without PPE. As EPA continually
refines its risk evaluations, it will consider this tool as a
Page 170 of 210
-------
developed a graphic that may prove useful in
demonstrating risks with and without the use of PPE. The
graphic can be generated using a web-based tool available
from the risk21.org web site.
possible option.
39
PUBLIC COMMENTS:
HSIA submitted a description of the industrial hygiene
practices at CC14 production facilities, including details on
tasks by exposure groups and generalized PPE
requirements (EPA-HQ-OPPT-2019-0499-0029).
EPA has reviewed industrial hygiene practice reports on
carbon tetrachloride submitted by the commenter as well as
the details on the tasks by exposure groups and generalized
PPE requirements submitted by the commenter.
26, 32,
38, 39,
43
PUBLIC COMMENTS:
EPA's assessment of dermal exposure likely
underestimates exposure. EPA does not have any data on
glove use and efficacy, which is necessary to accurately
assess dermal exposure. EPA acknowledges that gloves are
likely to provide only limited protection from CC14, given
that the chemical can break through gloves made of certain
materials. EPA recognizes the potential for occlusion,
whereby glove use can increase skin exposure (p. 60).
However, the dermal exposure estimates do not account
for occluded conditions.
EPA's document provides contradictory discussion of
occlusion for CC14. EPA appears to acknowledge the
limitations of gloves and their potential to increase skin
absorption, but then to simply assume that gloves actually
provide preset levels of protection over no gloves -
regardless of the potential for occlusion - without citing
any evidence to support these values. The premise seems
to be that if the most protective gloves potentially available
can be assumed to provide a PF that reduces risk to below
the benchmark, then EPA can conclude that there is no
unreasonable risk. This approach will allow clear risks to
occur whenever a worker uses anything less than the most
The occlusion, no gloves use and gloves use with various PFs
have been discussed in "10 CC14 Supplemental File
Engineering Report" that is attached with the final risk
evaluation document. In addition, the revised risk evaluation
has been updated with real word usage scenarios with
citations of peer-reviewed publications (see Section 2.4.1.5 -
Consideration of Engineering Controls and Personal
Protective Equipment).
Page 171 of 210
-------
protective gloves (or no gloves), or when there is
occlusion; these scenarios are quite likely to occur in the
real world.
The extent to which the preconditions for effective glove
use are in fact followed in workplaces is highly uncertain.
Overall, EPA concedes that it "does not know the actual
frequency, type, and effectiveness of glove use in specific
workplaces of the occupational exposure scenarios." The
Agency assumes fixed PFs of 5, 10, and 20X, which do not
appear to be supported by any empirical data that account
for the complexities of glove use in the real world. EPA
should revise the CC14 risk evaluation so that its
unreasonable risk determinations for workers are based on
workplace exposure levels in the absence of PPE. Where
unreasonable risk is demonstrated, PPE, along with other
worker protection measures, should be considered in
determining how best to eliminate the unreasonable risk.
23, 26,
38
PUBLIC COMMENTS:
The draft risk evaluation states that, because CC14 "is a
skin irritant and sensitizer," workers will be "persuaded on
their own (in addition to the workplace industrial hygiene
program and OSHA regulations) to wear gloves when
handling the chemical." EPA does not explain how
workers, many of whom are exposed to chemicals other
than CC14, will be able to identify the specific source of
their rash or skin irritation, in order to identify the
appropriate PPE. Nor does EPA indicate how workers who
are able to diagnose their own injuries will be assured
access to the proper type of protective equipment, which
their employers may or may not supply.
The language has been replaced with the following:
"carbon tetrachloride is identified and labeled as a skin
irritant and sensitizer, which suggests that workers are less
likely to not be wearing gloves when handling the chemical."
38
PUBLIC COMMENTS:
EPA must evaluate the conditions of use it expects to
consider under TSCA in the risk evaluation and determine
Page 172 of 210
-------
The Benzene decision has no bearing on EPA's duty to
identify and manage unreasonable risks under TSCA.
Consistent with NIOSH recommendations, EPA should
reduce exposure to occupational carcinogens such as CC14
"as much as possible," the extent of which should be
decided during risk management and not during risk
evaluation.
whether the condition of use presents unreasonable risk. If
necessary, any risk management activities will only occur
after EPA has completed the risk evaluation.
The standard cancer benchmarks used by EPA and other
regulatory agencies range from 1 in 1,000,000 to 1 in 10,000
(i.e., lxlO"6 to lxlO"4) depending on the subpopulation
exposed. EPA, consistent with 2017 NIOSH guidance, used
lxlO"4 as the benchmark for the purposes of unreasonable risk
determinations for individuals exposed to carbon tetrachloride
in industrial and commercial work environments, including
workers and ONUs. lxlO"4 is not a bright line and EPA has
discretion to make unreasonable risk determinations based on
other benchmarks as appropriate. See section 5.1.1.2 of the
risk evaluation for additional information.
In addition to assessing the cancer risk using a linear
extrapolation approach and comparing the results to the
standard cancer benchmark of 1 x 10"4, EPA also assessed
cancer risk using a threshold approach. Based on the
threshold approach, EPA identified MOEs for cancer risks.
EPA used both the risk estimates derived from the linear
extrapolation approach and the MOEs derived from the
threshold approach for the unreasonable risk determinations
for individuals exposed to carbon tetrachloride.
26, 38
PUBLIC COMMENTS:
By assuming extensive use of PPE at the risk evaluation
stage, EPA also conflates risk evaluation with risk
management. TSCA requires EPA to complete a risk
evaluation and to make a determination of unreasonable
risk before it considers how such risks may be managed.
Per the statute (see TSCA section 6(b)(4)(A)) and the
implementing regulations for risk evaluations (40 CFR part
702, subpart B), EPA must make the unreasonable risk
determination at the time of the risk evaluation. Upon finding
unreasonable risk, EPA will apply risk management actions to
the extent necessary so that the chemical no longer presents
Page 173 of 210
-------
PPE is a risk management tool, albeit a poor one that may
be used only when preferable options are not available. As
such, PPE may only be considered, if at all, during the risk
management stage when it can be weighed against more
effective means of risk reduction.
such risk, in accordance with TSCA section 6(a).
EPA considers the uncertainties associated with each
condition of use, and how the uncertainties may result in a
risk estimate that overestimates or underestimates the risk.
Based on such analysis, EPA determines whether or not the
identified risks are unreasonable. Such consideration carries
extra importance when the risk estimates are close to the
benchmarks for risks from acute and chronic non-cancer
health effects and cancer.
38
PUBLIC COMMENTS:
EPA notes that "engineering controls" should be "the
primary means to control air contaminants" such as CC14.
However, because EPA assumes extensive respirator use to
avoid unreasonable risk determinations, EPA will never
proceed to the risk management stage where it can
consider whether other, more cost-effective control options
exist. This is particularly true with a chemical such as
CC14, which requires respirators with PFs up to 50. Such
respirators have significant costs, both in the ability of
workers to wear them while doing their jobs safely, and in
the expense to employers of ensuring their comprehensive
respirator program is adequate.
OSHA's hierarchy of controls is a method for eliminating
workplace hazards. EPA will manage unreasonable risks
presented by chemical substances when the Agency
undertakes regulatory action for conditions of use determined
to have unreasonable risk. Utilization of the hierarchy of
controls to recommend or require risk management actions in
the risk evaluation would be premature and inappropriate.
Other as
pects of the human health risk characterization
SACC
SACC COMMENTS:
Consumer exposures were not evaluated. This is
justified by the fact that there are several indoor,
outdoor, and personal monitoring studies
demonstrating low-level concentrations of CC14.
The risk evaluation should include a table and a brief
discussion of these data to provide a more objective
context for its decision not to evaluate risk for
consumers, and for contrasting with occupational
The Consumer Product Safety Commission (CPSC) banned
the use of carbon tetrachloride in consumer products
(excluding unavoidable residues not exceeding 10 ppm
atmospheric concentration) in 1970. As a result of CPSC's
ban, EPA does not consider the use of carbon tetrachloride-
containing consumer products produced before 1970 to be
known, intended, or reasonably foreseen. In accordance with
the CPSC ban, carbon tetrachloride is not identified in the
California Air Resources Board consumer product database
Page 174 of 210
-------
exposures.
Recommendation: The assertion of no significant use in
consumer products should be supported by a more specific
description of the documentation used by EPA to arrive at
its conclusion. Improve the discussion and summarize the
data supporting the decision to exclude consumer
exposures from this evaluation. Tabulate ambient air levels
for perspective in assessing consumer background
exposures.
nor the Washington State Product Testing Data list or the
State of Vermont list of Chemicals in Children's Products and
no consumer uses are listed in the CDR.
As stated in the Problem Formulation, EPA determined after
additional public outreach, literature searches and other
reasonably available information, the consumer uses of
carbon tetrachloride that were initially identified in the Scope
document {i.e., commercially available aerosol and non-
aerosol adhesives/sealants, paints/coatings, and
cleaning/degreasing solvent products) only have the potential
for negligible exposure. Carbon tetrachloride is not a direct
reactant or additive in the formulation of solvents for
consumer use in cleaning and degreasing, adhesives and
sealants or paints and coatings. Trace levels of carbon
tetrachloride in the chlorinated substances used to
manufacture the products are expected to volatilize during the
product manufacturing process.
Risks from background concentrations to carbon tetrachloride
are assessed under the EPA NATA. The 2014 NATA reports
a national ambient carbon tetrachloride concentration of 0.53
|ig/m3 and 3 in a million cancer risk.
https://www.epa.gov/national-air-toxics-assessment/2014-
nata-assessment-results#pollutant
( onK'iit :iikI Organization
Charge Question 6.1: Please provide suggestions for improving the clarity of the information presented in the draft risk evaluation.
Charge Question 6.2: Is the draft risk evaluation narrative presented in an objective and balanced manner and supportive of the risk
characterization? If not, please provide some specific recommendations to improve the draft risk evaluation in this area.
Page 175 of 210
-------
Charge Question 6.3: Is the quality of the data used in the risk characterization appropriate for the purposes of the evaluation? If
not, please provide specific examples and recommendations that may include additional data that EPA could consider in their
assessment.
Charge Question 6.4: Are the uncertainties and assumptions underlying the risk assessment transparently documented? If not, which
uncertainties and assumptions could benefit from additional contextualization and/or clarification?
Charge Question 6.5: What additional analyses might provide useful insight into the sensitivity of the risk characterization
conclusions, including but not limited to the assumptions mentioned in Sections 2, 3, 4, and 5 of the draft risk evaluation?
#
Summary ol'Comments lor Specific Issues Related to
Charge Question 6
KIW/OPPT Response
Clarity and completeness of review
SACC
SACC COMMENTS:
Table E-l does not list ALL facilities reported CC14
releases, only the 21 facilities with largest releases. A
histogram (or estimated probability distribution curve) of
annual releases from all 49 facilities would be useful in
understanding the larger picture of releases.
A graphic has been added to Appendix E.
SACC
SACC COMMENTS:
Appendix F of the draft risk evaluation should include
the specific information it is cited as having rather than
referring the reader to U.S. EPA (2020), which appears
to be incorrectly titled and dated. Appendix F is often
inadequate when referencing important aspects of the
exposure estimation.
Recommendation: Expand Appendix F to include pertinent
material from Supplemental Information on Occupational
Exposure Assessment (U.S. EPA, 2020).
This appendix has been removed from final risk evaluation.
The reader is referred to the supplemental file instead.
SACC
SACC COMMENTS:
A Committee member noted that Tables 4.3 to 4.6 of
the draft risk evaluation are very helpful, although
some Committee members preferred other formats
versus stacked bars (e.g., parallel bars).
A member commented that Figures 4-1 to 4-4 were
very good, and that EPA should do the same type of
EPA appreciates this recommendation and will consider it for
future risk evaluations
Page 176 of 210
-------
graphical representation for dermal exposure.
SACC
SACC COMMENTS:
Recommendation: EPA should clearly describe which
conditions of use pose unacceptable risks in the absence of
PPE and further identify those conditions of use where
assumed PPE use reduces risk to a level that the condition
of use is assessed as having reasonable risks. This should
be clarified in the Executive Summary (tables under
Summary of Risk Determinations, p. 22-23) and in Table
4-13 and Table 5-1 of the draft risk evaluation.
EPA considers the uncertainties associated with each
condition of use, and how the uncertainties may result in a
risk estimate that overestimates or underestimates the risk.
Based on such analysis, EPA determines whether or not the
identified risks are unreasonable. Such consideration carries
extra importance when the risk estimates are close to the
benchmarks for acute, chronic non-cancer risks, and cancer
risks.
EPA's approach for developing exposure assessments for
workers is to use reasonably available information and expert
judgment. EPA considers each condition of use and uses
exposure scenarios with and without PPE that may be
applicable to particular worker tasks on a case-specific basis
for a given chemical. For the purposes of determining
whether or not a condition of use presents unreasonable risks,
EPA incorporates assumptions regarding PPE use based on
reasonably available information and professional judgment
underlying the exposure scenarios. These assumptions are
described in the unreasonable risk determination for each
condition of use, in section 5.2. While EPA has evaluated
worker risk with and without PPE, as a matter of policy, EPA
does not believe it should assume that workers are
unprotected by PPE where such PPE might be necessary to
meet federal regulations, unless it has evidence that workers
are unprotected. Additionally, in consideration of the
uncertainties and variabilities in PPE usage, EPA uses the
high-end exposure value when making its unreasonable risk
determination in order to address those uncertainties. EPA
has also outlined its PPE assumptions in section 5.1.
SACC
SACC COMMENTS:
Recommendation: Consider incorporating additional
EPA is in the process of evaluating the body of reasonably
available literature on AFs in order to determine whether to
Page 177 of 210
-------
studies identified by the Committee as important into the
risk evaluation. The following are some studies considered
by the Committee to be important:
Environmental studies: Johnson et al. (2017); Kienzler
etal. (2017).
Genotoxicity/toxicokinetics/mechanistic studies: Slater
(1987); MAK (2000); Manibusan et al. (2007);
Eastmond (2008); Hernandez et al. (2009); Borgert et
al. (2015); Corthay (2014); Candeias and Gaipl (2016);
Garner and Visser (2020); Sanzgiri et al. (1995 and
1997); Kim et al. (1990); Rao and Rechnagel (1968,
1969); Bruckner et al. (2002); Thrall et al. (2000);
Kappus et al. (1985); WHO (1999); Seifert et al.
(1994); Weber et al. (2003); Manibusan et al. (2007);
Malaguarnera et al. (2012).
Spermatotoxicity studies: Smyth et al. (1936); Adams
et al. (1952); El Faras et al. (2016); Turk et al. (2016).
Epidemiology studies: Hill et al. (2003); Heineman et
al. (1994).
Studies on oxidative stress: Okora et al. (2019);
Altinoz et al. (2018); Ritesh et al. (2015).
It was noted that some of these studies listed above were
initially identified, but not carried forward for evaluation
and this is an example of how the TSCA systematic review
system is not working as expected.
revise standards for application of AF and the acute to
chronic ratio for the 20 high-priority substances undergoing
TSCA risk evaluation. EPA will consider the (Kienzler. 2.017)
study in its assessment. Until the body of scientific evidence
for AFs is evaluated, EPA will continue to use OPPT
methodology as cited in the risk evaluation (
2012b) and applv an AF of 5 for acute and 10 for chronic fish
and aquatic invertebrate data. EPA considers these AFs to be
protective of fish and aquatic invertebrates from acute and
chronic exposures to neutral organic substances such as
carbon tetrachloride, which produce toxicity from narcosis.
EPA does not have reasonably available information that
carbon tetrachloride is a thyroid endocrine disruptor. EPA
consulted ("Johnson et al.. 2017) while examining amphibian
variation in sensitivity and constructing SSDs in the final risk
assessment.
EPA used the approach described in Section 1.5 of the final
risk evaluation to evaluate, extract and integrate carbon
tetrachloride's human health hazard and dose-response
information from the identified studies. After implementation
of this approach and methodology, EPA redesigned the
weight of evidence (WOE) narrative for the identified human
health hazards for carbon tetrachloride to improve clarity and
transparency based on recommendations from SACC.
SACC
SACC COMMENTS:
One Committee member indicated that the word
"benchmark" was used to represent two fundamentally
different concepts in the draft risk evaluation, which
both differ from how benchmark is typically used by
EPA (U.S. EPA, 2012; Davis et al., 2011) and other
organizations such as the European Food Safety
The use of the term benchmark has been clarified and
harmonized with other TSCA risk evaluations. EPA will
consider further clarifications and harmonization in future
risk evaluations, as needed.
Page 178 of 210
-------
Authority (EFSA, 2017; Haber et al., 2018) and may be
a source of confusion.
For cancer effects, the draft risk evaluation defines
benchmark risk as the target risk (10"4) and BMD as the
dose estimated to correspond to that target 10"4 risk.
For noncancer effects, the benchmark MOE is a
unitless factor that is divided into the POD (generally
the NOAEL exposure) to determine a sufficiently safe
exposure. Both uses differ from how EPA has used the
BMD term previously, and also differ from the original
purpose of the BMD.
Recommendation: For cancer risk, the term BMD should
be reserved for PODs that are estimated by BMDs
corresponding to risks (benchmark responses [BMRs]) at
the lower end of the observable range (e.g., 0.1%),
estimated using the methods discussed in U.S. EPA
(2012). For noncancer risk, the "benchmark MOE" should
be appropriately termed instead of the "total uncertainty
factor (UFt)."
SACC
SACC COMMENTS:
Some Committee members found the explanation of
the approach used to calculate HEDs using a
pharmacokinetic model difficult to follow. It is not
always clear if the dose being discussed represents the
dose applied to a rodent, or the HED.
For example, it was not clear on first reading that the
BMDLio of 14.3 mg/m3 (line 4103) refers to the HED.
Recommendation: EPA should adopt a method for
distinguishing exposures to rodents from HEDs and apply
this distinction consistently.
The dose response section in the final risk determination
provides better characterization of POD derivation.
SACC
SACC COMMENTS:
The Committee expects that many readers will likely
focus on risk determination values under conditions
Because carbon tetrachloride is an intermediate and is mostly
used at large facilities, EPA assumes the use of a respirator
with an APF of 50 and gloves with a PF of 20. The risk
Page 179 of 210
-------
with and without PPE use, and not also carefully
consider the background information about PPE
presented in this draft risk evaluation. The risk
evaluation should alert readers to pay attention to this
information, and in particular, alert readers to
conditions of use where the decision of no
unreasonable risk is directly tied to assumptions of PPE
use.
Whenever EPA derives or cites a risk that is not
unreasonable because of the assumption of PPE use, a
modifying phrase should be added to enhance attention
to the limitations in this assumption (e.g., EPA has
determined that Condition of Use x (Scenario x) does
not present an unreasonable risk contingent upon
adherence to OSHA standards on exposure controls
and PPE requirements and recommendations).
Recommendation: Highlight for readers those conditions
of use where the determination of no unreasonable risk is
directly related to the assumption of PPE use.
evaluation also presents estimated risk in the absence of PPE
and does not assume that ONUs use PPE.
EPA must evaluate the conditions of use it expects to
consider under TSCA in the risk evaluation and propose risk
management for any condition of use which the Agency
determines presents unreasonable risk. Risk management
activities will only occur after EPA has completed the risk
evaluation.
See the Executive Summary, updated Risk Characterization
(Section 4), and updated Risk Determination (Section 5) for
more clarification on how these sections support each other
and how new information submitted to or obtained by EPA
following publication of the draft risk evaluation was
incorporated.
SACC
SACC COMMENTS:
In Section 1.3, Regulatory and Assessment History (pp.
26-28), the draft risk evaluation mentions national and
international laws to which CC14 is subject (Subsection
1.3.1, p. 21, and Appendix A) and prior assessments by
other national and international agencies with
regulatory mandates on CC14 (Table 1-3, p. 27).
However, the section is not very informative. It needs
to provide a brief and specific description of the
relevance to the current risk evaluation, or whether the
prior assessments have indeed addressed exposures and
risks that EPA decided not to address in this risk
evaluation (for example, risks to populations in close
vicinity to major point sources). Having a statutory
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation.
Page 180 of 210
-------
mandate for evaluating environmental or human health
risk from a compound was not sufficient to
demonstrate that indeed such evaluation has been done
for all relevant situations.
Recommendation: Provide more specific information
about relevance of other legislation and the specifics of
environmental or human health risk addressed by other
organizations.
SACC
SACC COMMENTS:
The production and releases of CC14 are difficult to
reconcile in terms of mass balance when comparing the
releases reported in TableApx D-l (Appendix D, p.
237; 2018 TRI Data) and the production volume listed
in Table 1-2 (p. 26) for the CDR 2012 to 2015. In
particular, the footer in Table 1-2 [i.e., "The CDR data
for the 2016 reporting period is available via
ChemView..."] raises the question of why it was not
added to the release volumes.
Recommendation: Add explanatory information in Tables
1-2 and D-l describing the differences between the
reporting periods for production and release.
EPA noted in Table D-l that carbon tetrachloride release data
reported by facilities to TRI in 2017 was reviewed and
available in 2018.
Table 1-2 and Table Apx D-l are presenting the most
recently available data. CDR data is collected every four
years and includes data from the previous four years. CDR
data is named for the year it is reported. Therefore, Table 1-2
presents production volumes from the 2016 CDR reporting
period which includes data from 2012-2015. The 2020 CDR
reporting period is in-progress (and will include data from
2016 to 2019) with the reporting period ending on November
30, 2020.
CDR is a collection of basic exposure-related information on
the types, quantities, and uses of chemical substances
manufactured domestically or imported into the United
States. The CDR rule, promulgated under the authority of
Section 8(a) of TSCA, requires chemical substance
manufacturers (including importers) to report manufacturing
and processing data and industrial, commercial, and
consumer use information for certain chemical substances on
the TSCA Inventory.
Meanwhile, TRI tracks the management of certain toxic
Page 181 of 210
-------
chemicals that may pose a threat to human health and the
environment. U.S. facilities in different industry sectors must
report annually how much of each chemical is released to the
environment and/or managed through recycling, energy
recovery and treatment. Under the TRI rule, regulated
facilities must report information on releases and other waste
management for specific chemical substances in accordance
with Section 313 of the Emergency Planning and Community
Right-to-Know Act (EPCRA). TRI data is collected annually
for the previous year and is named after the year of data it
represents (not the reporting year). 2018 TRI data was
collected in 2019 and 2019 data will not be published until
January 2021.
SACC
SACC COMMENTS:
A Committee member considers the following statement
(p. 20, lines 737-739) to not be exactly true: "In each risk
evaluation under TSCA section 6(b), EPA determines
whether a chemical substance presents an unreasonable
risk of injury to health or the environment, under the
conditions of use. The determination does not consider
costs or other non-risk factors." Decisions to set the target
cancer risk for exposed workers at 10"4, set MOE levels at
10 or 100, set BMR levels at 5% or 10%, set expected
working life length, or apply UFs are all policy decisions
that can involve costs or other non-risk factors. EPA
should modify this statement.
EPA applied risk assessment methods tailored to the
requirements of TSCA. TSCA compels EPA to conduct risk
evaluations to determine whether a chemical substance
presents unreasonable risk, without consideration of cost or
other non-risk factors, under the conditions of use. EPA's
decision to use a 10"4 cancer risk benchmark, specific MOEs
and BMRs, and applied UFs are risk factors; the Agency
does not consider these to be non-risk factors.
In addition to assessing the cancer risk using a linear
extrapolation approach and comparing the results to the
standard cancer benchmark of lxlO"4, EPA also assessed
cancer risk using a threshold approach. Based on the
threshold approach, EPA identified MOEs for cancer risks.
EPA used both the risk estimates derived from the linear
extrapolation approach and the MOEs derived from the
threshold approach for the unreasonable risk determinations
for individuals exposed to carbon tetrachloride.
Page 182 of 210
-------
SACC
SACC COMMENTS:
Terminology such as "slope" and "MS-combo model" are
not referred or referenced and are applied in the draft risk
evaluation under the assumption that other people are
familiar with them.
A description of the MS-combo model has been added to the
final risk evaluation.
SACC
SACC COMMENTS:
The draft risk evaluation (p. 28, line 1012) indicates
that "EPA conducted public outreach and literature
searches to collect information about carbon
tetrachloride conditions of use..without providing
any specifics.
Recommendation: Quantify what is entailed in the phrase
"Public outreach and literature searches."
EPA conducted literature searches for reasonably available
information and convened meetings with companies, industry
groups, chemical users and other stakeholders to aid in
identifying conditions of use and verifying conditions of use
for carbon tetrachloride. All public outreach is available in
the docket (EPA-HQ-OPPT-2016-0733). All cited references
are available for public review, subject to limitations under
TSCA section 14.
SACC
SACC COMMENTS:
EPA should look at how small changes in grouping
conditions of use affect the conclusions. Consider
parametrizing some of the qualitative assumptions or
input different assumptions and assess how the risk
conclusions vary.
Recommendation: Consider performing a more robust
sensitivity analysis such as the one proposed in Thabane et
al. (2013) study.
EPA will consider this SACC recommendation in future risk
evaluations.
SACC
SACC COMMENTS:
The risk evaluation should reference environmental
discharges and pathways that were addressed by other
regulations by including hyperlinks that would direct the
reader to the relevant regulations and documentation.
Clarifying language about what pathways are under other
statutes has been added to Section 1.4.3 of the Risk
Evaluation.
SACC
SACC COMMENTS:
Some key information is not located in the body of the
draft risk evaluation, but the reader is referred to an
appendix for detail. Once in the appendix, the reader is
referred to a supplemental document for the
EPA reconsidered the appropriate placement of information
in the final risk evaluation and made necessary changes to
improve the exposure, hazard, and risk discussions in the
body of the risk evaluation.
Page 183 of 210
-------
information. This daisy chain of referrals complicates
reading and comprehension.
The Committee encouraged placing the information
and discussion that is crucial to establishing the
exposure, hazard, and risk findings in the body of the
risk evaluation, and placing the detailed arguments and
computations in appendices or supplemental
documents. Creating a concise and comprehensive
discussion in the body of the risk evaluation is difficult
but important to constituent understanding.
Recommendation: Consider carefully which information
needs to be provided explicitly in the body of the risk
evaluation from the more detailed information available in
the appendices.
SACC
SACC COMMENTS:
Recommendation: Consider reordering the presentation of
materials in the draft risk evaluation to discuss
environmental exposures, hazard, and risk characterization
(Environment; new Section 2) before human health
exposures, hazard, and risk characterization (Human
Health; new Section 3) and followed by PESS exposures,
hazard, and risk characterization (PESS; new Section 4).
A majority of the Committee supported this
recommendation as a way to reduce repetition that
occurs throughout the document and improve clarity
and readability.
The remaining Committee members proposed
presenting the environmental and occupational
exposure, hazard, and risk characterizations as two
distinct sections instead of rotating through these topics
in Section 2 (Exposures), Section 3 (Hazards), and
Section 4 (Risk Characterization).
This is a cross-cutting issue raised on the processes that EPA
is going to be looking at in a more holistic way for the next
20 TSCA risk evaluations and will not be addressed in the
carbon tetrachloride risk evaluation.
SACC
SACC COMMENTS:
This is a cross-cutting issue raised on the processes that EPA
Page 184 of 210
-------
Recommendation: Optimize the use of active links within
the risk evaluation and provide external access to increase
readability and transparency.
It would be very helpful to improving reading
comprehension if links could be provided that tie
directly to the subsection (e.g., the specific table,
section, page) of the document (e.g., appendix or
supplemental document) where the specific
information is located. Currently, links are to a whole
document that require readers to search the document
for the specific information referenced. Specifically,
key values in summary tables (e.g., tables of exposure
estimates, PODs, risk estimates) should be linked
either internally to where they are discussed in the risk
evaluation document or externally to documents where
the value is derived and/or discussed.
is going to be looking at in a more holistic way for the next
20 TSCA risk evaluations. Additional linking has been added
in the carbon tetrachloride risk evaluation.
SACC
SACC COMMENTS:
The Committee noted that the names of EPA staff who
were involved in writing the current document are not
listed. The Committee hopes that that these staff members
will be recognized in the immediate future for their work.
Names of staff have been added.
33
PUBLIC COMMENTS:
The TSCA program should list the individual cancer sites
- including brain and nervous system cancers for CC14 -
as is done by the EPA IRIS program and IARC. This
information is important for researchers wanting to
conduct risk studies, employers wanting to inform and
protect vulnerable workers, insurers wanting to identify
liability, workplace compensation programs wanting to
identify causality, and others.
EPA lists the individual cancer sites in the table of
epidemiologic literature on cancers.
45
PUBLIC COMMENTS:
The dose-response section (3.2.5) and the accompanying
supplemental BMD modeling document are poorly
The information on the dose-response section has been
expanded in the final risk evaluation.
Page 185 of 210
-------
described. It has been customary in other risk evaluations
to provide summary tables listing all of the various model
combinations, with the final selected data set highlighted.
Additional summaries linking the BMD modeling results
to the POD selection process should be provided. This
would provide additional clarity to the POD discussion
section.
45
PUBLIC COMMENTS:
EPA should provide more discussion in these risk
evaluations about the substance of its intra-agency
coordination with program offices about existing
regulatory requirements that protect various media
pathways {i.e., air, water, land).
Clarifying language about what pathways are under the
jurisdiction of other EPA-administered statutes has been
added to Section 1.4.3 of the Risk Evaluation.
Discussions across EPA's program offices occur as the risk
evaluation is conducted and refined. Communication and
coordination between program offices within EPA occurs
regularly on TSCA-related efforts.
Objective presentation of risk findings
26
PUBLIC COMMENTS:
EPA's exclusion of CC14's use in the aerospace industry is
based on an unsubstantiated personal communication,
which the public cannot access to assess its accuracy and
reliability. EPA cannot rely on unverified and potentially
unrepresentative personal communications. EPA should
exercise its authority under TSCA section 8 to obtain
information that could be used to confirm or negate this
personal communication.
EPA's exclusion of aerospace uses from the conditions of use
for carbon tetrachloride is based on communication with
Aerospace Industries Association (AIA) quoted in the risk
evaluation. Specific details on this communication are
described in section 1.4.3.1. As described in this section,
AIA explained that uses previously identified as conditions
of use have been discontinued and EPA determined that the
uses are not intended, known, or reasonably foreseen to
occur. As a result, EPA did not include these uses in the risk
evaluation.
Appropriateness and quality of data used in risk evaluation
SACC
SACC COMMENTS:
The article selection for a systematic review should
follow established guidelines, such as Preferred
Reporting Items for Systematic Reviews and Meta-
Analyses (PRISMA) for observational studies. All
epidemiologic journals currently require a PRISMA
The major components of a PRISMA diagram {i.e.,
identification, screening, and eligibility of final included
articles) are represented in the Literature Flow Diagram for
Human Health Hazard Data Sources. Literature Flow
Diagrams were developed as an overview of the systematic
review process and (see Figures 1.5-1.9). Literature Flow
Page 186 of 210
-------
figure that shows the data identification, screening, and
eligibility of final included articles. The SACC report
includes an example PRISMA 2009 Flow Diagram.
In addition, the exclusion and inclusion criteria should
be defined a priori and applied to the article selection
and identification. This approach has been in place for
over 10 years and should be adopted for TSCA
evaluations for assessing epidemiologic studies.
This is one solution to the issue with understanding the
process for selecting and excluding articles and related
justifications that the Committee has discussed in
previous reviews.
Recommendation: Modify epidemiologic study
identification and selection methodology to comply with
established PRISMA guidelines.
Diagrams in Section 1.5.1 of the risk evaluation.
EPA developed inclusion and exclusion criteria for
epidemiologic studies a priori and applied these criteria
during the screening phase of the systematic review. See the
Strategy for Conducting Literature Searches for Carbon
Tetrachloride (CCL4): Supplemental Document to the TSCA
Scope Document for the initial i n cl u si on/ex cl u si on screening
criteria applied during the title/abstract screening for
relevancy phase of the systematic review process for perc
(see Section 4).The Problem Formulation of the Risk
Evaluation for Carbon Tetrachloride has the chemical-
specific inclusion/exclusion criteria applied during the full-
text screening phase of the systematic review process (see
Appendix F).
SACC
SACC COMMENTS:
The Committee suggests using one of the many
published, validated systems for evaluation of the
quality of the literature, such as the National Institutes
of Health (NIH) assessment tool (NIH Study Quality
Assessment Tools), or others available in the literature.
This approach would allow for a non-biased,
standardized, accepted evaluation, comparable to other
evaluations. In addition, evaluation is usually
performed by two independent reviewers, and any
discrepancy in findings are discussed and consensus is
reached.
Recommendation: Use current best practice methods for
quality review of literature including use of two
independent reviewers.
EPA/OPPT's quality evaluation method was developed
following identification and review of various published
qualitative and quantitative scoring systems when developing
the systematic review process for the first 10 TSCA risk
evaluations (e.g., OHAT Risk of Bias tool, CRED, etc.; see
Appendix A of the Application of Systematic Review in
TSCA Risk Evaluations document and references therein), as
well as soliciting input from scientists based on their expert
knowledge about evaluating various data/information
sources specifically for risk assessment purposes.
The NASEM TSCA Committee will review EPA's
systematic review process and provide recommendations for
improvement. EPA will consider future revisions to the
TSCA data quality evaluation tools after that time.
SACC
SACC COMMENTS:
One Committee member submitted a relatively simple,
standard template for data extraction from
The NASEM TSCA Committee will review EPA's
systematic review process, and EPA will consider revisions to
the process based on their recommendations.
Page 187 of 210
-------
epidemiology studies. When using this template, the
specifics of data extraction must be decided before
evaluating the study. The template also included
objective criteria for quality evaluation of studies, so
both the criteria for quality scoring and data extraction
are decided before looking at the findings.
Recommendation: Continue to improve the systematic
review process.
43, 23,
41, 30
PUBLIC COMMENTS:
The TSCA method represents a deeply flawed and
unscientific approach to systematic review and will
compromise the quality, validity, and protectiveness of
EPA's risk evaluations (e.g., see commentary
published in the American Journal of Public Health).
The method lacks transparency and is not empirically
based, making it likely to have resulted in a biased
evidence base.
EPA should address the SACC's prior comments on
the TSCA method and incorporate the recommended
changes.
The TSCA method departs radically from accepted
scientific principles for systematic review adopted by
the IOM NTP, and EPA's IRIS program and endorsed
by the NAS and other peer review bodies.
EPA should not use the TSCA systematic review
method until it is validated by the NAS. The review by
NAS is not likely to be completed before the risk
evaluations for the first 10 chemicals have gone
through a round of public comment and peer review.
In completing the ongoing risk evaluations, EPA must
adopt a well-established systematic review method,
such as those developed by IOM, NTP (Office of
Health Assessment and Translation [OHAT]), EPA's
EPA published the titie/abstract inclusion/exclusion criteria
for carbon tetrachloride in Appendix E of the Strategy for
Conducting Literature Searches for Carbon Tetrachloride
and inclusion/exclusion criteria statements used during full
text screening in Appendix F to the Problem, Formulation of
the Risk Evaluation for Carbon Tetrachloride. Data ciualitv
criteria used for scoring each discipline are provided in a
separate document titled Application of Systematic Review in
TSCA Risk Evaluations, which also outlines evidence
integration strategies that will be further developed for the
next risk evaluations.
EPA consulted multiple systematic review frameworks and
the IRIS program when developing the systematic review
process.
EPA is reviewing its data quality criteria and will publish a
protocol document for the next TSCA risk evaluations. To
date, EPA has already made changes to the physical
chemical, environmental, epidemiological criteria since the
Application of Systematic Review in TSCA Risk Evaluations
was published. These changes included validation and
improvement efforts to ensure that the most relevant studies
were included in the TSCA risk evaluations. The most up-to-
date data quality evaluation criteria will be available for
Page 188 of 210
-------
IRIS program, or the University of California, San
Francisco (Navigation Guide), that is endorsed by the
NAS and other peer reviewers.
review in the upcoming the Systematic Review Protocol
Supporting the TSCA Risk Evaluations document (under
development).
EPA anticipates feedback from the NASEM TSCA
Committee on its systematic review process, including the
epidemiological data quality criteria and will carefully
review and implement relevant recommendations.
43
PUBLIC COMMENTS:
Appendix A of the Application of Systematic Review in TSCA
The TSCA approach applies a rigid scoring system to
grade the "quality" of studies. This system could result in
many studies being arbitrarily classified as "poor" or
"unacceptable" based on a small number of reporting or
methodology limitations that do not negate their overall
value. Other systematic review methodologies do not use
numerical scoring systems for assessing study quality and
the NAS recommends strongly against such scoring. The
SACC previously noted that "The Agency should provide
justification for using a weighted scoring system and the
rationale for the specific metrics used for differential
weighting in its evaluation of studies."
Risk Evaluations explains the basis for EPA/OPPT's
development of a numerical scoring system to inform the
characterization of the data/information sources during the
data integration phase. The goal is to provide transparency
and consistency to the evaluation process along with creating
evaluation strategies that meet the TSCA science standards
for various data/information streams.
EPA/OPPT's quality evaluation method was developed
following identification and review of various published
qualitative and quantitative scoring systems to inform our
own fit-for-purpose tool. The development process involved
reviewing various evaluation tools/frameworks (e.g., OHAT
Risk of Bias tool, CRED, etc. ; see Table 1 and Appendix A of
the TSCA SR document and references therein), as well as
soliciting input from scientists based on their expert
knowledge about evaluating various data/information sources
for risk assessment purposes. While there are many published
systematic review tools available for human health and
environmental health hazard assessment, no systematic
review tools were identified that encompass either exposure
assessment (e.g. general population exposures, occupational
exposures and industrial releases) or fate and transport
assessment.
Page 189 of 210
-------
In order to ascertain the quality of the reasonably available
data, EPA used a numerical scoring system to assign a
qualitative rating. This approach added consistency and
transparency to the evaluation process. Scores were used for
the purpose of assigning the confidence level rating of High,
Medium, Low, or Unacceptable, and inform the
characterization of data/information sources during the data
integration phase. In all evaluation strategies, professional
judgment was employed to determine the adequacy or
appropriateness of the qualitative rating assigned by the
numerical scoring system.
41,43
PUBLIC COMMENTS:
EPA's Application of Systematic Review in TSCA Risk
EPA fails to use a protocol that outlines the pre-
established methods to be used throughout the
systematic review process. This directly contradicts the
EPA's 2017 framework rules mandating that the
Agency use "a pre-established protocol" to conduct
risk assessments. A protocol for the review needs to be
established in advance of individual evaluations to
eliminate the potential for bias and to assure that
evidence reviews are conducted using consistent, well-
defined criteria.
EPA must immediately implement protocols for all
future draft risk evaluations. The use of pre-established
protocols minimizes biases in the evidence base by
explicitly pre-defining how questions will be
formulated, searches will be conducted, eligibility
criteria will be applied, and quality of the included
studies will be assessed. It allows greater transparency
in the decision-making process throughout the
systematic review and is a fundamental element
required to ensure the integrity of evidence-based
Evaluations document and several supplemental documents
demonstrate how systematic review was conducted for the
first 10 chemicals undergoing risk evaluation under TSCA.
As described in the Application of Systematic Review in
TSCA Risk Evaluations, EPA/OPPT implemented a structured
process of identifying, evaluating and integrating evidence for
both the hazard and exposure assessments developed during
the TSCA risk evaluation process. Because EPA/OPPT
developed and implemented systematic review processes and
procedures in tandem with development of actual TSCA risk
evaluations, EPA/OPPT acknowledged it expected that new
approaches and/or methods would be developed to address
specific assessment needs for the relatively large and diverse
chemical space under TSCA. Thus, EPA/OPPT expected to
document the progress of implementing systematic review in
the draft risk evaluations and through publication of
supplemental documents.
Page 190 of 210
-------
evaluations.
The TSCA systematic review process is undergoing
improvements for the next risk evaluations and includes
updates to better align with the systematic review best
practices that commenters indicated in the public comments.
EPA may need to develop new methods and approaches to
ensure that the systematic review process is sensitive to the
constraints and requirements applicable to risk evaluations
under TSCA including tight statutory deadlines. The body of
information compiled in the data quality and data extraction
supplemental files accompanying each TSCA Risk
Evaluation are the primary pool of studies that were
considered for the first 10 risk evaluations. In addition, other
data sources and information will be considered and possibly
incorporated in the draft risk evaluations based on
information submitted during public comment periods, peer
review comments and targeted supplemental searches (e.g., to
locate specific data for building exposure scenarios and
modeling).
EPA is continuously creating and improving methods for
efficiently evaluating the overall body of evidence and
numerous changes in the methods were due to validation and
improvement efforts to ensure that the most relevant studies
were included in the TSCA risk evaluations. The most up-to-
date data quality evaluation criteria will be available for
review in the upcoming the Systematic Review Protocol
Supporting the TSCA Risk Evaluations document (under
development)
41,43
PUBLIC COMMENTS:
EPA fails to use pre-established methods for evidence
integration. The TSCA approach fails to address
critical elements, including identification and
evaluation of each stream of evidence and integration
When synthesizing and integrating evidence for each human
health hazard endpoint, EPA considered quality, consistency,
relevancy, coherence and biological plausibility as specified
in Application of Systematic Review in TSCA Risk
Evaluations. EPA used an informal framework for most
Page 191 of 210
-------
of evidence as necessary and appropriate based on
strengths, limitations, and relevance. The draft risk
evaluation fails to clearly define how the quality of the
body of evidence has been evaluated for each evidence
stream and it has failed to pre-specify the method for
integrating two or more streams of evidence in
formulating the final conclusions.
EPA should use an approach to evidence integration
that has been recommended and successfully applied
by the IARC, NTP's OHAT, the Navigation Guide, or
the NAS.
The data integration process should consist of:
assigning an overall rating in the confidence of the
body of evidence for each specified outcome using
explicit, predefined criteria; translating the overall
rating into a conclusion on the level of evidence for a
health effect; and then formulating a hazard
identification conclusion. Human and animal evidence,
when available, should be integrated, while
mechanistic data may be used to help inform the final
conclusions.
endpoints but did array the immunological evidence within a
more formal framework to respond to a comment by the
SACC (see Appendix A below and Appendix M in the risk
evaluation).
Sections 3.2.3 and 3.3.4 describe EPA's process of weighing
and integrating scientific evidence for hazard endpoints.
EPA is developing and implementing more formal and
structured data integration strategies for the next set of TSCA
chemical risk evaluations. In addition, EPA anticipates
feedback from the NASEM TSCA Committee on its
systematic review process and will carefully review and
implement relevant recommendations.
41,43
PUBLIC COMMENTS:
EPA continues to use methods that lack transparency to
identify "key/supporting/influential information," and
does not provide the details of the methods for the
"hierarchy of preferences" approach that excludes
relevant studies. The "hierarchy of preferences" is a
new concept that was not part of the original TSCA
systematic review method document, nor in the
scoping or problem formulation documents, and has
not been subject to peer review or public comment.
EPA does not explain why some types of studies
should receive preference over others. There are no
Different lines of evidence are routinely used in TSCA
chemical assessments because of data availability, sources,
underlying documentation, and quality varies. EPA
preferentially relies on a variety of test and analog data. In the
absence of suitable test data, predictive modeling tools may
be used. For environmental hazards, if the modeling tools
cannot provide predictions to an endpoint of interest, then
calculations like acute-to-chronic ratios can be used to fill in
data gaps.
PECO/RESO statements or a modified framework were used
to describe the full-text inclusion and exclusion criteria for
Page 192 of 210
-------
objective criteria for determining which evidence to
rely on and which to exclude, undermining
transparency and consistency and encouraging
subjective judgments. There is a lack of clarity on how
EPA chose and evaluated the key sources, which at
their time of incorporation, outweigh the results from
EPA's screening process. There is also a lack of clarity
on how EPA came to its decisions about which studies
it chose to exclude and which to include in its
supplemental information. This pattern obscures the
evidence base for this draft risk evaluation, potentially
leading to biased results.
selecting relevant references. These criteria are provided in
the TSCA Problem Formulation documents for each chemical
as some criteria reflect chemical-specific issues that are better
discussed in each chemical risk evaluation.
41,43
PUBLIC COMMENTS:
The updated TSCA data quality criteria for
epidemiological studies make it more difficult for
epidemiological studies to be scored as high quality
and thus limit the weight that they receive in TSCA
risk evaluations. The method can exclude a study based
on only one "unacceptable" criterion rather than
considering all relevant science while accounting for
"strengths and limitations" as required by TSCA. EPA
has failed to explain or justify the updated criteria.
The criteria are based on an arbitrary list of metrics
including several scoring metrics not related to bias,
but rather to reporting. In Metric 13 'Statistical power,'
a study can only be scored as 'Medium' or
'Unacceptable.' In fact, with EPA's updated criteria,
epidemiological studies can no longer score high on
seven metrics, but no such change has been made for
the animal or in vitro studies. Further, there is no
empirical justification for these 'scores' on the
different metrics.
The epidemiologic criteria were revised to more stringently
distinguish between High, Medium and Low studies (see
revisions in the supplemental file to the carbon tetrachloride:
Updates to the Data Quality Criteria for Epidemiological
Studies). After additional piloting of the criteria, EPA found
that the initial iteration of the epidemiological data quality
criteria fas published in the Application of Systematic Review
in TSCA Risk Evaluations) was inadvertently skewing aualitv
scores toward the tail ends of the scoring spectrum (High and
Unacceptable). In order for the criteria to represent a more
accurate depiction of the quality levels of the epi literature,
the criteria were revised using two methods.
The first method was to make the unacceptable metrics less
stringent. This was accomplished by either rewording the
metrics to allow for more professional judgment in the
interpretation of the unacceptable criterion, or in some cases,
completely removing the unacceptable bin from metrics that
EPA determined were not influential enough to completely
disqualify a study from consideration (mostly metrics in the
Analysis and Biomonitoring domain). EPA found that these
Page 193 of 210
-------
criteria changes greatly reduced the type one error in the
Unacceptable scoring. No acceptable studies were
inaccurately classified as Unacceptable.
The second method was to reduce the number of studies that
received an overall High rating. The majority of overall
scores in EPA's initial evaluations during piloting tended to
be High. Therefore, EPA strived to revise the criteria to
provide more degradation in the scoring to more accurately
and objectively distinguish studies of the highest quality from
medium and low-quality studies. To do this, EPA removed
the High criterion from some metrics, particularly in
dichotomous metrics (High/Low or High/Unacceptable) that
were primarily being binned as High by reviewers across the
majority of the studies. These dichotomous metrics were
contributing to the overall quality scores being skewed
towards High. To address this, EPA shifted some of the
dichotomous metrics such that the highest metric score
possible (for all studies) is a Medium. The change led to the
dichotomous metrics having less significant impact to the
numerical scoring and the overall quality rating for each
study.
With the aforementioned changes to the criteria, EPA
observed fewer studies with Unacceptable ratings and more
studies shifting from High to Medium, with only the highest
quality studies receiving a High overall rating. Out of the
-200 relevant epidemiologic studies and cohorts evaluated for
data quality for the first 10 TSCA chemicals, the majority
(-80%) still scored as High or Medium. The remaining -20%
of studies scored Low or Unacceptable. EPA is confident that
no studies of acceptable quality were inappropriately assigned
as Unacceptable. EPA is also confident that the revised
Page 194 of 210
-------
criteria bins the quality levels of these epi studies more
appropriately than the previous iteration. Additional
refinements to the epidemiologic data evaluation criteria are
likely to occur as EPA's validation and process improvement
efforts continue.
EPA anticipates feedback from the NASEM TSCA
Committee on its systematic review process, including the
epidemiological data quality criteria, and will carefully
review and implement relevant recommendations.
41
PUBLIC COMMENTS:
EPA fails to transparently apply predefined eligibility
criteria to the references in the literature search. The
Populations, Exposures, Comparators, and Outcomes
(PECO) statement (framework) should shape the entire
review process, including the search strategy to be used,
the study eligibility criteria to be applied, how the data will
be extracted from the included studies, the strategy for
synthesizing the evidence, and how the results will be
reported. The PECO statement should be designed to
"minimize the risk of researcher biases influencing the
ultimate results of the SR."
EPA/OPPT developed and applied inclusion and exclusion
criteria during title/abstract and full text screening to identify
information potentially relevant for the risk evaluation
process. This step also classifies the references into useful
categories (e.g., on-topic versus off-topic, human versus
animal hazard) to facilitate the sorting of information through
the systematic review process.
The results of initial title/abstract screening for each of the
first 10 chemical risk evaluations are available in an EPA
Dage for Chemicals Undergoing Risk Evaluation under
TSCA.
A summary of the Full Text Screening conducted for the first
10 TSCA risk evaluations is described in Section 3.2.2.2.1 of
the draft risk evaluation and summarized here. The full text
screening was conducted while EPA/OPPT refined the scope
of the TSCA risk evaluations and developed the problem
formulation documents for the first 10 chemical substances.
PECO statements or a modified framework were used to
describe the full-text inclusion and exclusion criteria for
selecting relevant references. These criteria are provided in
Page 195 of 210
-------
the TSCA Problem Formulation documents for each chemical
as some criteria reflect chemical-specific issues that are better
discussed in each chemical assessment.
Each article was generally screened by two independent
reviewers using specialized web-based software {i.e.,
DistillerSR)^. Screeners were assigned batches of references
after conducing pilot testing. Screening forms facilitated the
reference review process by asking a series of questions
based on pre-determined eligibility criteria. DistillerSR was
used to manage the workflow of the screening process and
document the eligibility decisions for each reference. The
screeners resolved conflicts by consensus, or consultation
with an independent individual(s).
As indicated in section 3.2.2.1 of the TSCA SR document,
EPA/OPPT used the infrastructure of the ECOTOX
knowledgebase ( .) to identify single chemical
toxicity data for aquatic life and terrestrial life. It uses a
comprehensive chemical-specific literature search of the open
literature that is conducted according to Standard Operating
Procedures (SOPs), including specific SOPs to fit the needs
of the TSCA risk evaluations^. Due to its well-established
methods to gather high quality data, ECOTOX processes and
data are widely accepted and used by a variety of domestic
and international organizations and researchers. The
ECOTOX literature search strategy is documented in the
Strategy for Conducting Literature Searches documents for
each of the ten TSCA risk evaluations and the data screening
and extraction protocols are described ECOTOX SOPs^.
m In addition to using DistillerSR, EPA/OPPT is exploring automation
and machine learning tools for data screening and prioritization activities
(e.g., SWIFT-Review, SWIFT-Active Screener, Dragon, DoCTER).
Page 196 of 210
-------
SWIFT is an acronym for "Sciome Workbench for Interactive Computer-
Facilitated Text-mining".
^ The ECOTOX SOPs can be found at
httos://cfDub.era.eov/ecotox/helacfm?hetatabs=tab4.
131 The ECOTOX SOPs for TSCA work can be found al
hltDs://cfDub.cDa.aov/ccolox/blackbox/hclD/OPP' adinaGuidclincs
SOP.Ddf and
litlDs://cfDub.e0a.aov/ecotox,/bIackbox/lieto/OPP 3Dort.s50P.Ddi.
41
PUBLIC COMMENTS:
EPA does not provide a method for how to determine
the "adequacy" of the statistical power of a study and
fails to provide any rationale for excluding studies with
<80% statistical power. In Metric 13 'Statistical
power' of the epidemiological criteria, EPA has
confused bias with imprecision, as individual primary
studies that are "underpowered" are still valuable to
decision-making. Small studies may be imprecise, but
that does not mean they should be confused with a
study that is biased.
Importantly, when combined in a meta-analysis that
increases the statistical power of the body of evidence,
small studies that are underpowered can demonstrate
an effect between an exposure and health outcomes.
For example, in a 2017 systematic review by Lam et al.
entitled "Developmental PBDE Exposure and
IQ/ADHD in Childhood: A Systematic Review and
Meta-analysis," none of the four high-quality studies
included in the meta-analysis reported a power
calculation, and therefore would have been considered
'unacceptable' by EPA.
EPA acknowledges that this metric needs further refinement
and agrees that poorly powered studies can still be useful
when combined in meta-analysis.
Additional refinements to the epidemiologic data evaluation
criteria are likely to occur as EPA's validation and process
improvement efforts continue.
EPA has requested feedback from the NASEM TSCA
Committee on its systematic review process, including the
epidemiological data quality criteria, and will carefully
review and implement relevant recommendations.
41
PUBLIC COMMENTS:
Rather than exclude a study based on a lack of reporting,
EPA should instead attempt to request the missing
information required to make the determination from the
The TSCA evaluation strategies consider methodological
design and implementation and reporting within the existing
domains and metrics. Since it is difficult to have high
confidence in data where the underlying methods are
Page 197 of 210
-------
study authors. If EPA is not able to retrieve this missing
information from the study authors, a potential bias (if the
metric being assessed relates to bias and not reporting)
may then be considered in the study. However, the study
should not be excluded from the body of evidence due to
this one criterion.
unreported or poorly reported, EPA assesses reporting and
methodological quality simultaneously. However, EPA
recognizes the challenge of discerning between a deficit in
reporting and a problem in the underlying methodological
quality of the data/information source. Developing a reporting
checklist, guidance document or a separate reporting quality
domain may be possible in the future as EPA uses and
optimizes the evaluation strategies. EPA also designed
evaluation criteria that consider risk of bias and Bradford Hill
aspects when assessing the quality of animal toxicity and
epidemiological studies. Refer to Appendices F, G and H of
the Application of Systematic Review in TSCA Risk
Evaluations document for more information.
SACC
SACC COMMENTS:
EPA relies heavily on prior CC14 assessments done by
other agencies. There is a potential for missing key or
supporting studies if prior assessments did not adhere
to systematic review. The risk evaluation should
include clear statements as to whether principles of
systematic review were applied in the prior
assessments and, if so, to what extents.
There is still an issue of insufficient transparency in the
application of systematic review for selecting data
sources and references used in support of the risk
evaluation and risk characterization.
One Committee member proposed to develop a "key"
to the reference section of the risk evaluation to make it
easy to identify key and supporting sources, identify
the section of the evaluation where they are pertinent,
know whether the citations were subject to data quality
evaluation (including, if applicable, the specific prior
review in which it was included), and identify sources
that were not subject to data quality evaluation and the
This is a cross-cutting issue raised on the processes and the
science and methods that EPA is going to be looking at in a
more holistic way for the next 20 TSCA risk evaluations. All
data used in the carbon tetrachloride risk evaluation were
evaluated under the TSCA systematic review process.
Page 198 of 210
-------
reason why.
Recommendation: Develop and display a "key" for the
reference section that facilitates identifying and tracing
sources throughout the process of systematic peer review
and data source evaluation/validation.
SACC
SACC COMMENTS:
The quality of several studies was described as
unacceptable but they were used in the risk evaluation,
nonetheless. This appears to undermine the goal of using
the best quality studies. An alternative descriptor such as
"poor" could be used to differentiate these studies from
those that are completely unacceptable. The term
"unacceptable" should be restricted to a study deemed of
unacceptable quality for a reason (i.e., unacceptable for...).
The systematic review needs to provide the context for
which a publication may be acceptable or not. This context
is not always clear because the names assigned to criteria
for selection in the systematic review process do not
always reflect clearly the criteria.
Unacceptable dermal studies are no longer used in the
derivation of PODs, an alternate approach is used instead.
31
PUBLIC COMMENTS:
EPA did not complete a data quality review of every cited
genotoxicity study. The supplemental review file for
human health hazard studies only includes four in vitro
studies. Yet, Appendix I summarizes a number of other
studies (excerpted from the EPA IRIS assessment) that do
not appear to have undergone a data quality review
according to the TSCA systematic review protocol.
Data quality review for every cited genotoxicity study is
presented in final risk evaluation.
41
PUBLIC COMMENTS:
There is inconsistency in the reporting of the included
studies in the draft risk evaluation and the accompanying
supplementary files. In 'Carbon tetrachloride
Bibliography: Supplemental File for the TSCA Scope
Document,' there are 107 pages of "On Topic" references
A review of the on topic human health references after the
title and abstract screening revealed a large number of animal
studies that were likely to be of limited use for the following
reasons: (1) The aim of the study was to induce a disease state
in an animal (e.g., cirrhosis, fibrosis, organ damage: liver,
kidney, testes and others) rather than evaluate the effects of
Page 199 of 210
-------
following title and abstract screening for human health
hazard with approximately 2,782-2,996 references.
However, in Figure 1-8 of the CC14 draft risk evaluation,
EPA states that: "The literature search strategy used to
gather human health hazard information for carbon
tetrachloride yielded 6,489 studies... Of the 6,489 studies
identified for carbon tetrachloride 6,454 were excluded as
off topic during the title and abstract screening phase."
Therefore, according to EPA after title and abstract
screening, there were only 35 "On Topic" studies included
in the draft risk evaluation. This is inconsistent with the
bibliography supplemental file for the TSCA Scope
Document, which demonstrates there are >2,500 "On
Topic" references following the title and abstract
screening. EPA has not accounted for or screened these
>2,500 references in the draft risk evaluation.
carbon tetrachloride exposure in animals and/or (2) Exposure
was via injection. In order to refine the search results for full-
text screening, the inclusion/exclusion criteria were revised to
remove these studies from the "on topic" pool. Appendix B in
the Problem Formulation describes the process used to re-
screen the references identified as "on topic" in the first
screening round, including prioritizing the literature for
screening and the re-categorization criteria applied during the
re-screening and tagging.
27,41
PUBLIC COMMENTS:
The draft risk assessment dismissed 99.45% of the 6,489
studies, found when searching for CC14 hazards, at the
"title/abstract screening" stage without any
characterization. The criteria used to dismiss so many
findings were not provided. Although EPA states that
"Because systematic review is an iterative process,
EPA/OPPT expects that some references may move from
the on-topic to the off-topic category and vice versa," this
does not justify the exclusion of 2,500-3,000 "On Topic"
references for Human Health Hazards without explanation.
The SACC should charge EPA to go back to the literature
screening stage and apply the logic that there is no reason
to dismiss a relevant toxicity finding, short of any obvious
irrelevancy.
EPA published the titie/abstract inclusion/exclusion criteria
for carbon tetrachloride in Appendix E of the Strategy for
Conducting literature Searches for Carbon Tetrachloride
and inclusion/exclusion criteria statements used during full
text screening in an appendix to the problem formulation
document for carbon tetrachloride. Data quality criteria used
for scoring each discipline are provided in a separate
document titled Application of Systematic Review in TSCA
Risk Evaluations, which also outlines evidence integration
strategies that will be further developed for the next risk
evaluations.
41
PUBLIC COMMENTS:
A review of the on topic human health references after the
title and abstract screening revealed a large number of animal
Page 200 of 210
-------
The numbers shown in the flow diagram Figure 1-8 do not
accurately reflect the numbers at each step and do not
account for all of the 6,489 references identified from the
'Data Search Results.' As shown, in the 'Data Screening
Step,' of the 6,471 studies, 6,454 studies were excluded.
Therefore, 17 studies should have moved to the 'Data
Evaluation Step,' not 15 as shown here, with 18
'Key/supporting data sources' being added, for a total of
35 studies entering the 'Data Evaluation,' not 33 as shown
here.
studies that were likely to be of limited use for the following
reasons: (1) The aim of the study was to induce a disease state
in an animal (e.g., cirrhosis, fibrosis, organ damage: liver,
kidney, testes and others) rather than evaluate the effects of
carbon tetrachloride exposure in animals and/or (2) Exposure
was via injection. In order to refine the search results for full-
text screening, the inclusion/exclusion criteria were revised to
remove these studies from the "on topic" pool. Appendix B in
the Problem Formulation describes the process used to re-
screen the references identified as "on topic" in the first
screening round, including prioritizing the literature for
screening and the re-categorization criteria applied during the
re-screening and tagging.
41
PUBLIC COMMENTS:
In Figure 1-5, there are 150 data sources included at the
'Data Extraction/Data Evaluation Step' and 141 of these
are excluded without any justification. Studies that make it
to 'Full text screening' but are excluded thereafter should
only be excluded with an explicit justification.
The sources used to collect data were all subjected to data
quality evaluations based on metrics presented in the
Application of Systematic Review in TSCA Risk Evaluations
document, and the full data quality assessments are presented
in a supplemental file.
41
PUBLIC COMMENTS:
The way in which EPA developed and applied the
eligibility criteria for references is deeply concerning. The
literature and screening strategy is described in the Scope
Document, which was published in June 2017. The results
of the screening of literature search were published in
'Carbon tetrachloride (CASRN 56-23-5) Bibliography:
Supplemental File for the TSCA Scope Document'
(webpage 'last updated on June 22, 2017'). As highlighted
in the draft risk evaluation, for studies determined to be
'on-topic' after title and abstract screening, EPA conducted
full text screening to further exclude references that were
not relevant to the risk evaluation. The inclusion and
exclusion criteria for full text screening were published in
EPA designed evaluation criteria that consider risk of bias
and Bradford Hill aspects when assessing the quality of
animal toxicity and epidemiological studies. Refer to
Appendices F, G and H of the Application of Systematic
Review in TSCA Risk Evaluations document for more
information.
EPA is continuously creating and improving methods for
efficiently evaluating the overall body of evidence and
numerous changes in the methods were due to validation and
improvement efforts to ensure that the most relevant studies
were included in the TSCA risk evaluations. The most up-to-
date data quality evaluation criteria will be available for
review in the upcoming the Systematic Review Protocol
Page 201 of 210
-------
the problem formulation for CC14 (published in May
2018), after the searches and initial screening had been
completed. The timing of this is very concerning as the
PECO framework was developed after the studies had
already been identified in the literature search and screened
at the title and abstract stage and therefore could have been
developed to include/exclude studies that would support a
pre-defined health hazard conclusion. EPA's failure to
predefine the study eligibility criteria, applied to the 'on
topic' references in the draft risk evaluation, introduces
significant researcher bias that most likely impacted the
results of the draft risk evaluation.
Supporting the TSCA Risk Evaluations document (under
development).
41
PUBLIC COMMENTS:
In Figure 3-1 of the draft risk evaluation, EPA conflates
data quality evaluation and evidence integration in the
'Human Health Hazard Assessment' and does not clearly
outline how these two critically important steps were
completed. In section 3.2.4, EPA describes how they
conflate both an evaluation of the quality of the body of
evidence and the evidence integration steps during the
'weight of the scientific evidence' process: "Factors
considered in weighing the scientific evidence included
consistency and coherence among human and animal
studies, quality of the studies (such as whether studies
exhibited design flaws that made them unacceptable) and
biological plausibility." EPA does not rate the confidence
in the body of evidence or follow a predefined evidence
integration process that transparently demonstrates how it
arrived at is its final conclusion. Therefore, it is unclear
how EPA translated the available evidence into its final
conclusion. EPA must immediately implement an evidence
integration method that is consistent with best practice in
The sources used to collect human health data for carbon
tetrachloride were all subjected to data quality evaluations
based on metrics presented in the Application of Systematic
Review in TSCA Risk Evaluations document, and the full data
quality assessments are presented in a supplemental file. EPA
is developing and implementing more formal and structured
data integration strategies for the next set of TSCA chemical
risk evaluations. In addition, EPA requested feedback from
the NASEM TSCA Committee on its systematic review
process and will carefully review and implement relevant
recommendations.
Page 202 of 210
-------
systematic review and transparently present how the
conclusions were reached.
41
PUBLIC COMMENTS:
EPA's draft risk evaluation references Klimisch scores (or
European Chemicals Agency [ECHA] reliability scores)
when considering dermal and inhalation risks. These
scores are invoked particularly when discussing studies in
EPA's IRIS assessment for CC14, but they are not present
in the IRIS assessment and only seem to appear behind
studies that score poorly. It is deeply concerning that EPA
is invoking a potentially biased and non-empirically
validated instrument when outlining dermal and inhalation
risks from CC14, as it may present issues with regard to
internal validity and external generalizability.
Klimisch scores are not presented in final risk evaluation.
Editorial comments
SACC
SACC COMMENTS:
Table E-l releases were in pounds/year and Table E-2
releases were in kg/day. One or the other unit of
measurement should be used.
EPA has revised the Risk Evaluation to provide a summary
of the releases in the body of the document in kg/year. The
appendix Table E-l summarizes data as reported to EPA and
found in the Pollutant Reporting Tool. Table E-2 converts the
lb/year to kg for use as model inputs for E-FAST2014.
SACC
SACC COMMENTS:
On p. 22, line 828, the quote regarding the CPSC ban
for CC14 is stated as: "excluding unavoidable residues
not exceeding 10 ppm atmospheric concentration."
This is not correct. The proper quote is provided on
lines 1073-1076 (p. 30) of the draft risk evaluation.
Recommendation: Fix the quote regarding the CPSC ban
for CC14 on p. 22.
The regulation is quoted correctly in the final risk evaluation.
The regulation is also correctly paraphrased (without
quotation marks) throughout the document.
SACC
SACC COMMENTS:
There are formatting issues with Table 4-13 (pp. 160-
163) that make it difficult to read, and there is a lack of
correspondence between some of the row across
columns.
EPA considered many of the editorial suggestions and
comments provided by the SACC and the public and revised
the risk evaluation for clarity. EPA is also considering
improving the cancer risk figures in future risk evaluations.
Page 203 of 210
-------
Figures 4-1 to 4-4 (pp. 156-157) could be made clearer
by using stacked bars rather than parallel bars.
Recommendation: Correct the formatting issues with Table
4-13 and improve clarity of Figures 4-1 to 4-4.
SACC
SACC COMMENTS:
The SACC provided several editorial comments:
On p. 24, there is a "Section 0".
The title of U.S. EPA, 2019b (i.e., "Risk Evaluation for
Carbon Tetrachloride, Supplemental Information on
Releases and Occupational Exposure Assessment")
should be changed to "Draft Risk Evaluation for
Carbon Tetrachloride Supplemental File: Occupational
Exposure Assessment" as shown in HERO.
Table 3-10: According to Nagano et al. (2007a), 31
male mice in the 125-ppm group had a
pheochromocytoma, not 32.
Line 4177: It seems that the Nagana et al. (2007b)
reference should be Nagana et al. (2007a).
Paragraph beginning on line 4202: How is "slope"
defined in this paragraph?
Line 4321: The "MS-combo model" is not defined.
Perhaps the approach in Chiu and Crump (2012) could
prove useful in combining the risk from liver and
adrenal tumors.
It is recommended that the entire document be
reviewed to ensure that all notation has been clearly
defined and is used properly.
Line 800: The same sentence is repeated in this spot.
Table 2.3 needs a reference to Cherrie et al. (2004) (if
that is what was used in developing the table).
On line 2143, "Cherrie" is misspelled, and a complete
citation is needed.
Page 53, lines 1687-1695: The calculations for dermal
EPA considered many of the editorial suggestions and
comments provided by the SACC and the public and revised
the risk evaluation for clarity.
Page 204 of 210
-------
occupational exposure without PPE are not explained,
but it is said that "Dermal exposure assessment is
described in more detail in Appendix E of the
document Risk Evaluation for Carbon Tetrachloride
Supplemental Information on Releases and
Occupational Exposure Assessment' (US EPA,
2019b). When a reviewer tried to access this
information, Appendix E was not found in that
supplemental document. The reference for the dermal
assessment needs to be corrected.
Table 4-13, p. 161 of the draft risk evaluation contains
several formatting issues.
30
PUBLIC COMMENTS:
There appears to be a typographical error in lines 7020,
7023-7025, which state, "Therefore, the amphibian 9-day
lowest LCso of 0.09 mg/L and LC10 of 0.037023 mg/L
were used to derive an acute COC in Appendix Section
G.5 and chronic COC in Appendix Section G.6." In
reviewing the original literature (HERO ID 3616521), it
appears the LC50 values were reported over the range of
0.90-2.83 mg/L for CC14. Given that 0.90 mg/L is the
lowest reported value from that range and used by EPA in
developing the acute COC, the appropriate acute LC50 for
the most sensitive species [Pickerel Frog] is 0.90 mg/L or
900 (J,g/L. That value divided by an AF of 10 results in an
acute COC of 90 (J,g/L, which seems to be appropriately
used in the rest of the risk evaluation. In Appendix Section
G.5, lines 7066-7067, the acute COC appears to use the
correct value, i.e., the "acute COC = (0.9 mg/L)/(AF of 10)
= 0.09 mg/L x 1,000 = 90 |ig/L or 90 ppb."
The error in Appendix F (Formerly Appendix G.3) has been
corrected. 0.90 mg/L was the lowest reported acute toxicity
value and was used by EPA to derive the acute COC (acute
COC = 0.9 mg/L/AF of 10 = 0.09 mg/L x 1,000 = 90 ng/L or
90 ppb).
Page 205 of 210
-------
References
ACC. (2018). Properties of Phosgene. https://www.americanchemistrv.com/ProductsTechnologv/Phosgene/Phosgene~Safe~Practice~
Guidelines/PDF-Prope rties~of~Phosgene.pdf
Barrows. ME; Petrocelli, SR; Mace 311,1), (1980). Bioconcentration and elimination of selected water pollutants by bluegill
sunfish (Lepomis macrochirus). In R Haque (Ed.), Dynamics, exposure and hazard assessment of toxic chemicals (pp. 379-
392). Ann Arbor, Ml: Ann Arbor Science.
Bauder. MB: Palace. VP: Hodson. PV. (2005). Is oxidative stress the mechanism of blue sac disease in retene-exposed trout larvae?
Environ Toxicol Chem 24: 694-702. http://dx.doi.Org/10.1897/04~23R.l
Benson. JM; Springer. PL. (1999). Improved risk estimates for carbon tetrachloride. Final report. (DE-FC04-96AL76406).
Albuquerque, New Mexico: U.S. Department of Energy.
Birge, WJ; Black. JA; Kuehne, RA. (1980). Effects of organic compounds on amphibian reproduction. (PB80147523. KWRRI Research
Report 121). Lexington, KY: Kentucky Water Resources Research Institute, http://dx.doi.org/10.13023/kwrri.rr.121
Black. JA: Birge. WJ: McDonnell. WE: Westerman. AG: Ramey, BA: Bruser. DM. (1982). The aquatic toxicity of organic compounds to
embryo-larval stages offish and amphibians (pp. 61). (KWRRI Research Report No. 133). Lexington, KY: University of
Kentucky. Kentucky Water Resources Research Institute, http://dx.doi.org/10.13023/kwrri.rr.133
Brack. W: Rottler. H. (1994). Toxicity testing of highly volatile chemicals with green algae: A new assay. Environ Sci Pollut Res Int 1:
223-228.
Chen. CY: Wooster. GA: Bowser. PR. (2004). Comparative blood chemistry and histopathology of tilapia infected with Vibrio
vulnificus or Streptococcus iniae or exposed to carbon tetrachloride, gentamicin, or copper sulfate. Aquaculture 239: 421-
443. http://dx.doi.Org/lQ.lQ16/i.aauaculture.2004.05.033
Cheremisinoff. NP; Rosenfeld. R. (2009). Handbook of Pollution Prevention and Cleaner Production Vol. 1: Best Practices in the
Petroleum Industry (1st ed.). Burlington, MA: Elsevier.
Colby, HP: Puree!!. H; Kominami, S. (1994). Adrenal activation of carbon tetrachloride: role of microsomal P450 isozymes. .
Toxicology 94: 31-40.
de Best. JH; Salminen. E; Poddema. HJ; Janssen, PB: Harder. W. (1997). Transformation of carbon tetrachloride under sulfate
reducing conditions. Biodegradation 8: 429-436. http://dx.doi.Org/10.1023/A:1008262225760
de Bias, M; Uj ia-Tellaetxe, i; Carmen Gomez. M: Navazo, M: Alonso. L; Antonio Garcia. J: Purana. N: Iza, J. on: Perley Ramon, J.
(2016). Atmospheric carbon tetrachloride in rural background and industry surrounded urban areas in Northern Iberian
Peninsula: Mixing ratios, trends, and potential sources. Sci Total Environ 562: 26-34.
http://dx.doi.Org/10.1016/i.scitotenv.2016.03.177
Page 206 of 210
-------
de Vera, MP; Pocsidio, GN. (1998). Potential protective effect of calcium carbonate as liming agent against copper toxicity in the
African tilapia Oreochromis mossambicus. Sci Total Environ 214:193-202. http://dx.doi,org/10.1016/50048-9697(93)00065-5
El-Faras, AA: Sadek, IA; Ali, YE: Khalil, M; Mussa, EB, (2016). Protective effects of Vitamin E on CCI4-induced testicular toxicity in male
rats. 103: 157-168. http://dx.doi.Org/10.1556/036.103.2016.2.3
Hansch. C; Leo. A: Hoekman. D. (1995). Exploring QSAR: Hydrophobic, electronic, and steric constants. In C Hansch; A Leo; DH
Hoekman (Eds.), Exploring QSAR: Hydrophobic, Electronic, and Steric Constants. Washington, DC: American Chemical Society.
Heck. JE; Park. AS: Qiu, J: Cockburn. M; Ritz. B. (2013). An exploratory study of ambient air toxics exposure in pregnancy and the risk
of neuroblastoma in offspring. Environ Res 127:1-6. http://dx.doi.Org/10.1016/j.envres.2013.09.002
Heineman, EF; Cocco. P; Gomez. Mb ฆ iVsemeci, M; Stewart. PA: Hayes, RB: Zahm, SH. 1 Uomas, TL; Blair. A. (1994). Occupational
exposure to chlorinated aliphatic hydrocarbons and risk of astrocytic brain cancer. Am J Ind Med 26: 155-169.
http://dx.doi.ore/10.1002/aiim.4700260203
Jia, R; Cao, L; Du. J: Xu. P; Jene\ < >n, G. (2013). The protective effect of silymarin on the carbon tetrachloride (CCI4)-induced liver
injury in common carp (Cyprinus carpio). In Vitro Cellular and Developmental Biology 49: 155-161.
http://dx.doi.org/10.1007/sll626-013-9587-3
Johnson. MS: Aubee, C: Saiice, CJ; Leigh. KB: Liu. E; Pott uj 0 (2017). A review of ecological risk assessment methods for
amphibians: Comparative assessment of testing methodologies and available data [Review]. Integr Environ Assess Manag 13:
601-613. http://dx.doi.ore/10.1002/ieam.1881
Katafjn, X Nagano. K; Aiso, S: Senoh. H; Sakui v, v I J-auciu !. Pkudrui ^ J\_L (1998). A pathological study on spontaneous hepatic
neoplasms in BDF1 mice. J Toxicol Pathol 11: 21-25. http://dx.doi.org/10.1293/tox.ll.21
K, ius. T. (2011). Prevention of CCI(4)-induced oxidative damage in adrenal gland by Digera muricata extract in rat. Pak J
Pharm Sci 24: 469-473.
Kienzler, AHaMWoA. (2017). Waiving chronic fish tests: possible use of acute-to-chronic relationships and interspecies correlations.
Toxicol Environ Chem 99: 1129-1151.
Koskinen. H; Pehkonen, P; Vehniainen, E; Krasnov, A: Rexroad. C: Afanasyev, S: Molsa. H; Qikari, A. (2004). Response of Rainbow
Trout Transcriptome to Model Chemical Contaminants. Arch Biochem Biophys 320: 745-753.
http://dx.doi.Org/10.1016/i.bbrc.2004.06.024
Kotsanis. N: Metcalfe, CD. (1988). Accelerating an in vivo trout carcinogenesis assay with carbon tetrachloride and partial
hepatectomy. Paper presented at 15th Annual Aquatic Toxicity Workshop, November 28-30, 1988, Montreal, Quebec,
Canada.
Page 207 of 210
-------
Lee. SM; Lee, SB; Park, CH; Choi, J. (2006). Expression of heat shock protein and hemoglobin genes in Chironomus tentans (Diptera,
chironomidae) larvae exposed to various environmental pollutants: A potential biomarker of freshwater monitoring.
Chemosphere 65:1074-1081. http://dx.doi.Org/10.1016/i.chemosphere.2006.02.042
Leighton, DT. ir; Cab. JM. (1981). Distribution coefficients of chlorinated hydrocarbons in dilute air-water systems for groundwater
contamination applications. Journal of Chemical and Engineering Data 26: 382-585. http://dx.doi.ore/10.1021/ie00026a010
Lide. DR. (1999). CRC handbook of chemistry and physics: A ready-reference book of chemical and physical data (80th ed.). Boca
Raton, FL: CRC Press.
Lipscomb. JC; Garrett. CM: Snawder. JE. (1997). Cytochrome P450-dependent metabolism of trichloroethylene: Interindividual
differences in humans. Toxicol Appl Pharmacol 142: 311-318. http://dx.doi.ore/10.1006/taap.1996.8040
Li iQ. L; Du. J: Jia. R; Wane. J: Xu. P; Yin, G. (2015). Protective effects of Lycium barbarum polysaccharides against carbon
tetrachloride-induced hepatotoxicity in precision-cut liver slices in vitro and in vivo in common carp (Cyprinus carpio L.).
Comp Biochem Physiol C Toxicol Pharmacol 169: 65-72. http://dx.doi.ore/10.1016/j.cbpc.2014.12.005
Martins. J: Soares. ML: Saker. ML: Olivateles. L; Vasconcelos, VM. (2007). Phototactic behavior in Daphnia magna Straus as an
indicator of toxicants in the aquatic environment. Ecotoxicol Environ Saf 67: 417-422.
http://dx.doi.ore/10.1016/i.ecoenv.2006.11.00
Mccollister. DP: Beamer. WH; Atchison. GJ; Spencer. HC. (1951). The absorption, distribution and elimination of radioactive carbon
tetrachloride by monkeys upon exposure to low vapor concentrations. J Pharmacol Exp Ther 102: 112-124.
Merck. (1996). The Merck index: An encyclopedia of chemicals, drugs, and biologicals. In S Budavari (Ed.), (12th ed.). Rahway, NJ:
Merck & Co., Inc.
Naeano. K: Sasaki, T; Umeda V; Nishizawa. ! ฆ M-awa. N: Ohbayashi. H; Arito. H; Yamamato, S: S. F. (2007). Inhalation carcinogenicity
and chronic toxicity of carbon tetrachloride in rats and mice. Inhal Toxicol 19: 1089-1103.
http://dx.doi.ore/10.1080/08958370701628770
Nelson. JS; Burchfiel. CM: Fekeduleen, D; Andrew. ME. (2012). Potential risk factors for incident glioblastoma multiforme: the
Honolulu Heart Program and Honolulu-Asia Aging Study. J Neurooncol 109: 315-321. http://dx.doi.ore/10.1007/sll060-012-
0895-3
Neta. G; Stewart. PA: Raiaraman. P; Hein, MJ; Waters. MA: Purdue. MP: Samanic, C: Coble. JB: Linet, MS: inskip, PP. (2012).
Occupational exposure to chlorinated solvents and risks of glioma and meningioma in adults. Occup Environ Med 69: 793-
801. http://dx.doi.org/10.1136/oemed-2012-100742
Noweir, MH; Pfitzer, EA: Hatch. TF. (1973). Thermal decomposition of carbon tetrachloride vapors at its industrial threshold limit
concentration. Am Ind Hyg Assoc J 34: 25-37.
Page 208 of 210
-------
NRC. (2014). Acute exposure guideline levels for selected airborne chemicals: Volume 17. Washington, DC: National Academies
Press, http://dx.doi.org/10.17226/18796
OECD. (2011). SIDS initial assessment profile: Carbon tetrachloride [OECD SIDS]. (CoCAM 1,10-12 October 2011). Paris, France.
http://webnet.oecd.org/Hpv/UI/handler.axd?id=cada8da2~6884~48fl~bf42~470f28728
OxyChem. (2014). Chlorinated Organics Handbook.
https://www.oxv.com/OurBusinesses/Chemicals/Products/Documents/ChlorinatedQrgancis/ChlorinatedOrganicsHandbook.
pdf
OxyChem. (2018). Chlorinated Organics Closed Loop Unloading of Solvent Tank Trailers.
https://www.oxy.com/OurBusinesses/Chemicals/Products/Documents/ChlorinatedOrgancis/ClosedLoopUnloadingSolventTa
nkTrailers.pdf
Paustenbad as, ML: Andersen. ME. (1988). A physiologically based pharmacokinetic model for inhaled carbon
tetrachloride. Toxicol Appl Pharmacol 96:191-211.
Rao. KS: Recknagel, RO. (1969). Early incorporation of carbon-labeled carbon tetrachloride into rat liver particulate lipids and
proteins. Exp Mol Pathol 10: 219-228.
Ruder. AM: Yiin. JH; Waters. MA: Carreon, T; Hein. MJ; Butler. MA: Calvert. GM; Davis-King, KE: Schulte. PA: Mandel. JS; Morton. RF;
Reding, DJ; Rosenman, KD: Stewart, PA: Grp. BCCS. (2013). The Upper Midwest Health Study: gliomas and occupational
exposure to chlorinated solvents. Occup Environ Med 70: 73-80. http://dx.doi.org/10.1136/oemed-2011-10Q588
Sanzgiri, UY; Bruckner, JV. (1997). Effect of Emulphor, an emulsifier, on the pharmacokinetic and hepatotoxicity of oral carbon
tetrachloride in the rat. Fundam Appl Toxicol 36: 54-61. http://dx.doi.org/10.1006/faat.1997.2290
Snawder, JE; Lipscomb, JC. (2000). Interindividual variance of cytochrome P450 forms in human hepatic microsomes: correlation of
individual forms with xenobiotic metabolism and implications in risk assessment. Regul Toxicol Pharmacol 32: 200-209.
http://dx.doi.org/10.1006/rtph.200Q.1424
Ten Berge. WF; Zwart, A: Appelman, LM, (1986). Concentration-time mortality response relationship of irritant and systemically
acting vapours and gases. J Hazard Mater 13: 301-309. http://dx.doi.org/10.1016/0304-3894(86)85003-8
Thrall. KD: Vucelick, ME: Gies, RA: Benson. JM. (2000). Comparative metabolism of carbon tetrachloride in rats, mice, and hamsters
using gas uptake and PBPK modeling. J Toxicol Environ Health A 60: 531-548.
[u.L m, > > i ibasi. S: Sortmez, M: Ciftci. M: Yuce, A: Guvenc, M: Kaya. S _<\ M - sakal. M. (2016), Ameliorating effect of
pomegranate juice consumption on carbon tetrachloride-induced sperm damages, lipid peroxidation, and testicular
apoptosis. Toxicol Ind Health 32:126-137. http://dx.doi.org/10.1177/0748233713499600
U.S. Coast Guard. (1985). CHRIS hazardous chemical data: Volume II. Washington, DC: Government Printing Office.
Page 209 of 210
-------
U.S. EPA. (1994). Guidelines for Statistical Analysis of Occupational Exposure Data: Final. United States Environmental Protection
Agency :: U.S. EPA.
U.S. EPA. (2005). Guidelines for carcinogen risk assessment. Risk Assessment Forum, Washington, DC; EPA/630/P-03/001B. U.S. EPA.
http://www.epa.gov/iris/backgr-d.htiii
U.S. EPA. (2010). Toxicological review of carbon tetrachloride (CAS No. 56-23-5) in support of summary information on the
Integrated Risk Information System (IRIS) [EPA Report]. (EPA/635/R-08/005F). Washington, DC.
U.S. EPA. (2012a). Estimation Programs Interface Suite for Microsoftฎ Windows, v 4.11 [Computer Program]. Washington, DC.
Retrieved from https://www.epa.gov/tsca-screening-tools/epi-suitetm-estimation-program-interface
U.S. EPA. (2012b). Sustainable futures P2 framework manual [EPA Report]. (EPA-748-B12-001). Washington DC.
http://www.epa.gov/sustainable~futures/sustainable~futures-p2~framework~manual
U.S. EPA. (2013). Interpretive assistance document for assessment of discrete organic chemicals. Sustainable futures summary
assessment [EPA Report]. Washington, DC. http://www.epa.gov/sites/production/files/2015~Q5/documents/05~
iad discretes iune2013.pdf
U.S. EPA. (2017). Preliminary information on manufacturing, processing, distribution, use, and disposal: Carbon tetrachloride
[Comment]. Washington, DC. https://www.regulations.gov/document?D=EPA~HQ~OPPT~2 33-0003
U.S. EPA. (2018b). Problem formulation of the risk evaluation for carbon tetrachloride (methane, tetrachloro-). (EPA-740-R1-7020).
Washington, DC: Office of Chemical Safety and Pollution Prevention, United States Environmental Protection Agency.
https://www.epa.gov/sites/production/files/2018~06/documents/ccl4 problem formulation 05~31~18.pdf
U.S. EPA. (2019). Draft Risk Evaluation for Carbon Tetrachloride, Systematic Review Supplemental File: Data Quality Evaluation of
Ecological Hazard Studies.
Van Eekert, MHA; Schrodei ! Iv Stams, AJM; Schraa u 'Id. JA. (1998). Degradation and fate of carbon tetrachloride in unadapted
methanogenic granular sludge. Appl Environ Microbiol 64: 2350-2356.
Weber, U; Gingerich, WH; Pfeifer, KF. (1979). Alterations in rainbow trout liver function and body fluids following treatment with
carbon tetrachloride or monochlorobenzene. In MAQ Khan; JJ Lech; JJ Menn (Eds.), ACS Symposium Series, vol 99 (pp. 401-
413). Washington, DC: American Chemical Society, http://dx.doi.org/10.1021/bk~1979~0099.ch024
Whittaker, C: RL^ I McKernan, L; Dankovic. D; l.entz, T; Macmahon, K: Kuempei. E; Zumwalde, R; Schulte. P. (2016). Current
Intelligence Bulletin 68: NIOSH Chemical Carcinogen Policy. US Department of Health and Human Services.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/PB2017101413.xhtml
Page 210 of 210
------- |