EPA Document # EPA-740-R1-8012
December 2020
United States	Office of Chemical Safety and
Environmental Protection Agency	Pollution Prevention
Risk Evaluation for
Asbestos
Part I: Chrysotile Asbestos
December 2020

-------
PREAMBLE
In this preamble, the Agency describes its approach to completing the Risk Evaluation for Asbestos
under TSCA Section 6(a). The risk evaluation will be issued in two parts:
•	Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos (published with this preamble)
•	Risk Evaluation for Asbestos Part 2: Legacy Uses and Associated Disposals of Asbestos
(forthcoming)
Figure P-l shows a timeline for the development of the risk evaluation for asbestos. It starts with the
identification of asbestos as one of the First 10 Chemicals for risk evaluation under the Toxic Substances
Control Act (TSCA) in December of 2016. A Scope document and a Problem Formulation document
were then developed (2017 and 2018, respectively) and a draft risk evaluation (RE) was released to the
public in March of 2020. In late 2019, the court in Safer Chemicals, Healthy Families v. EPA, 943 F.3d
397 (9th Cir. 2019) held that EPA's Risk Evaluation Rule, 82 FR 33726 (July 20, 2017), should not have
excluded "legacy uses" {i.e., uses without ongoing or prospective manufacturing, processing, or
distribution) or "associated disposals" {i.e., future disposal of legacy uses) from the definition of
conditions of use, although the court upheld EPA's exclusion of "legacy disposals" {i.e., past disposal).
Due to the court ruling, in the March 2020 draft risk evaluation, EPA had signaled the inclusion of other
fiber types, in addition to chrysotile, as well as consideration of legacy uses and associated disposal for
the asbestos risk evaluation in a supplemental scope document and supplemental risk evaluation when
these activities are known, intended, or reasonably foreseen. This was supported by both public
comment and the SACC during the SACC Peer Review (virtual) meeting.
Figure P-2 is a text box with definitions for terms and documents important to understanding the shift in
the development of the risk evaluation for asbestos from 2016 to the present (2020).
The Path to Finalizing the Risk Evaluation for Asbestos: Parts 1 and 2
After considering SACC recommendations and public comments on the March 2020 draft risk
evaluation of asbestos, EPA decided to divide the risk evaluation into two parts: Part 1 on chrysotile
asbestos (herein) and Part 2 on legacy uses and associated disposal of asbestos (forthcoming). Together,
the documents will make up the risk evaluation for asbestos under TSCA Section 6.
Part 1, which accompanies this Preamble, completes the evaluation of chrysotile asbestos imported,
processed and distributed for use in the United States. EPA is confident that the chrysotile asbestos
conditions of use (COUs) represent all intended, known, or reasonably foreseen import, processing, and
distribution of chrysotile asbestos; uses of chrysotile asbestos that have been imported, processed, and
distributed; and disposal of such chrysotile asbestos uses.
In finalizing the risk evaluation Part 1 (chrysotile asbestos), EPA made appropriate and necessary
changes to update the document to reflect the best available science (following the standard in TSCA
section 26(h)) to support the risk determination and inform risk management decision for the conditions
of use evaluated in this document based on recommendations from the SACC and public comment.
These changes are reflected in the accompanying Response to Comments document. However, some
recommendations and comments that were identified in the SACC report are more relevant to what EPA
will address in Part 2 of the risk evaluation for asbestos {i.e., for legacy uses, including chrysotile and
other fiber types of asbestos).
Page 2 of352

-------
EPA has initiated the process for Part 2 and is currently identifying the relevant information available.
EPA will describe the COUs and the fiber types to be examined in a scope document that is currently
under development and will be made available for public comment. After review and consideration of
public comments, EPA will revise, where appropriate, and publish a final scope document. The legacy
uses and associated disposals of chrysotile asbestos were excluded from the Scope document for Part 1
and will be included in Part 1. Thus, the COUs included in Part 1 and those to be included in Part 2 will
not overlap. Subsequent to finalizing the Scope, EPA will develop Part 2 of the risk evaluation for
asbestos.
Page 3 of352

-------
Figure P-l: Schematic of the TSCA Risk Evaluation Timeline
Dec 2016
Jun 2017
May! 2018
Jun 2018
Nov 2019
Mar 2020
Jun 2020
Aug 2020
Dec 2020
Spring Summer
2021
Release of Scope Document for Asbestos
SACC Report Received (Aug 28, 2020)
Risk Evaluation for
Asbestos Part 1:
Chrvsotile Asbestos
Asbestos Identified as One of First 10
Chemicals for Risk Evaluation Under TSCA
Public Comment Period End (Jun 2, 2020)
SACC Meeting Held (Jun 8-11, 2020)
Release of Problem Formulation for
Asbestos
9th Circuit Court Decision -
Legacy Use and Associated
Disposals Are Not Excluded
from Conditions of Use
Release of Draft Risk Evaluation for
Asbestos
Dev elopment of Draft Risk Evaluation for
Asbestos
Draft Scope Document for Part 2:
Legacy Uses and Associated Disposals
for Asbestos
l~
PubKc Comment Period on Draft Scope
for Part 2
Risk Management for
Chrysotile Asbestos
Final Scope for Part 2
Development Review of Risk Evaluation
for Asbestos Part 2: Legacy Uses and
Associated Disposals of Asbestos and
Risk Management
Page 4 of 352

-------
Figure P-2: Important Definitions for the Risk Evaluation for Asbestos
Definitions
Asbestos. For the purposes of the Risk Evaluation for asbestos under TSCA Section 6(a), EPA
is using the TSCA Title II (added to TSCA in 1986), Section 202 definition; which is -
"asbestiform varieties of six fiber types - chrysotile (serpentine), crocidolite (riebeckite),
amosite (cummingtonite-grunerite), anthophyllite, tremolite or actinolite." The latter five
fiber types are amphibole varieties. This definition was previously defined in the scope
document and has consistently been applied in this risk evaluation process.
Chrysotile Asbestos. One of the six fiber types of asbestos as defined above. Chrysotile
asbestos is the only fiber type currently being imported, processed, or distributed in the
United States. These activities, along with the ensuing uses and disposals, encompass the
Conditions of Use (COUs) presented in Part 1 of the Risk Evaluation for asbestos.
Draft Risk Evaluation for Asbestos. The title of the March 2020 publicly released draft risk
evaluation. Although the draft was focused on chrysotile asbestos, the title and contents of
the document generated some confusion as was evident by peer review and public
comments received. Throughout this document {i.e., Risk Evaluation for Asbestos Part 1:
Chrysotile Asbestos), the term is used only to refer to the March 2020 draft risk evaluation.
Risk Evaluation for Asbestos. The risk evaluation for asbestos will consist of two Parts:
Part 1 is on chrysotile asbestos (finalized December 2020) and Part 2 will be on legacy uses
and associated disposal, including the five other fiber types of asbestos (scope and risk
evaluation are forthcoming).
Risk Evaluation for Asbestos: Part 1 - Chrysotile Asbestos. The December 2020 risk
evaluation of the asbestos fiber type (chrysotile) currently imported, processed and
distributed for use in the United States. Hereafter, referred to as Part 1 or Part 1 of the risk
evaluation.
Risk Evaluation for Asbestos: Part 2 - Legacy uses and associated disposals of asbestos.
The forthcoming risk evaluation for the legacy uses and associated disposals, including the
Page 5 of352

-------
TABLE OF CONTENTS
pSEAMBLilZIZZZIIZZIZZZIZZZIZZIIZZZIIZZIZZZIZZZIZZZZZIi"
ABBREVIATIONS	20
EXECUTIVE SUMMARY	22
1	INTRODUCTION	35
1.1	Physical and Chemical Properties and Environmental Fate	38
1.2	Uses and Production Volume	41
1.3	Regulatory and Assessment History	41
1.4	Scope of the Evaluation	43
1.4.1	Refinement of Asbestos Fiber Type Considered in the Risk Evaluation for Asbestos Part 1:
Chrysotile Asbestos	43
1.4.2	Conditions of Use Included in the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos
	43
1.4.3	Refinement of Evaluation of Releases to Surface Water	47
1.4.4	Exposure Pathways and Risks Addressed by Other EPA-Administered Statutes	47
1.4.5	Conceptual Models	54
1.5	Systematic Review	57
1.5.1	Data and Information Collection	57
1.5.2	Data Evaluation	64
2	EXPOSURES	65
2.1	Fate and Transport	65
2.2	Releases to Water	66
2.2.1	Water Release Assessment Approach and Methodology	66
2.2.2	Water Releases Reported by Conditions of Use	67
2.2.2.1	Processing and Industrial Use of Chrysotile Asbestos Diaphragms in Chlor-alkali
Industry	67
2.2.2.2	Processing Chrysotile Asbestos-Containing Sheet Gaskets	68
2.2.2.3	Industrial Use of Sheet Gaskets at Chemical Production Plants	68
2.2.2.4	Industrial Use and Disposal of Chrysotile Asbestos-Containing Brake Blocks in Oil
Industry	68
2.2.2.5	Commercial Use, Consumer Use, and Disposal of Aftermarket Automotive Chrysotile
Asbestos-Containing Brakes/Linings, Other Vehicle Friction Products, and Other Chrysotile
Asbestos-Containing Gaskets	69
2.2.3	Summary of Water Releases and Exposures	69
2.3	Human Exposures	70
2,3.1 Occupational Exposures	70
2.3.1.1	Occupational Exposures Approach and Methodology	72
2.3.1.2	Consideration of Engineering Controls and Personal Protective Equipment	72
2.3.1.3	Chlor-Alkali Industry	75
2.3.1.3.1	Process Description - Asbestos Diaphragms	75
2.3.1.3.2	Worker Activities - Asbestos Diaphragms	78
2.3.1.3.3	Number of Sites and Potentially Exposed Workers - Asbestos Diaphragms	80
2.3.1.3.4	Occupational Inhalation Exposures - Chrysotile Asbestos Diaphragms	81
2.3.1.3.5	Exposure Results for Use in the Risk Evaluation for Asbestos: Part 1- Chlor-Alkali
	83
2.3.1.3.6	Data Assumptions, Uncertainties and Level of Confidence	85
2.3.1.4	Sheet Gaskets	85
2.3.1.4.1 Process Description - Sheet Gasket Stamping	85
Page 6 of352

-------
2.3.1.4.2	Worker Activities - Cutting of Asbestos-containing Sheet Gaskets	88
2.3.1.4.3	Number of Sites and Potentially Exposed Workers - Sheet Gasket Stamping	89
2.3.1.4.4	Occupational Inhalation Exposure Results - Sheet Gasket Stamping	89
2.3.1.4.5	Exposure Data for Use in the Risk Evaluation for Asbestos: Part 1 - Chrysotile
Asbestos; Sheet Gasket Stamping	91
2.3.1.4.6	Data Assumptions, Uncertainties and Confidence Level	92
2.3.1.5	Use of Gaskets in Chemical Production	93
2.3.1.5.1	Process Description - Sheet Gasket Use	93
2.3.1.5.2	Worker Activities - Sheet Gasket Use	94
2.3.1.5.3	Number of Sites and Potentially Exposed Workers - Sheet Gasket Use	94
2.3.1.5.4	Occupational Inhalation Exposures - Sheet Gasket Use	94
2.3.1.5.5	Exposure Results for Use in the Risk Evaluation for Asbestos Part 1: Chrysotile
Asbestos - Sheet Gasket Use	96
2.3.1.5.6	Data Assumptions, Uncertainties and Level of Confidence	97
2.3.1.6	Oil Field Brake Blocks	97
2.3.1.6.1	Process Description - Oil Field Brake Blocks	97
2.3.1.6.2	Worker Activities - Oil Field Brake Blocks	99
2.3.1.6.3	Number of Sites and Potentially Exposed Workers - Oil Field Brake Blocks	99
2.3.1.6.4	Occupational Inhalation Exposures - Oil Field Brake Blocks	100
2.3.1.6.5	Exposure Results for Use in the Risk Evaluation for Asbestos: Part 1 Chrysotile
Asbestos - Oil Field Brake Blocks	101
2.3.1.6.6	Data Assumptions, Uncertainties and Level of Confidence	102
2.3.1.7	Aftermarket Automotive Brakes/Linings and Clutches	102
2.3.1.7.1	Process Description - Aftermarket Automotive Brakes/Linings and Clutches	103
2.3.1.7.2	Worker Activities - Aftermarket Automotive Brakes/Linings and Clutches	106
2.3.1.7.3	Number of Sites and Potentially Exposed Workers - Aftermarket Automotive
Brakes/Linings and Clutches	108
2.3.1.7.4	Occupational Inhalation Exposures - Aftermarket Automotive Brakes/Linings and
Clutches	108
2.3.1.7.5	Exposure Data for Use in the Risk Evaluation for Asbestos Part 1: Chrysotile
Asbestos - Aftermarket Auto Brakes/Linings and Clutches	110
2.3.1.7.6	Data Assumptions, Uncertainties and Level of Confidence	112
2.3.1.8	Other Vehicle Friction Products	113
2.3.1.8.1	Installing New Brakes on New Cars for Export Only	113
2.3.1.8.2	Use of Brakes/Frictional Products for a Single, Large Transport Vehicle (NASA
Super-Guppy)	114
2.3.1.9	Other Gaskets-Utility Vehicles (UTVs)	118
2.3.1.9.1	Process Description - UTV Gasket installation/Servicing	118
2.3.1.9.2	Worker Activities - UTV Gasket Installation/Servicing	118
2.3.1.9.3	Number of Sites and Potentially Exposed Workers - UTV Gasket
Installation/Servicing	118
2.3.1.9.4	Occupational Inhalation Exposures for Use in the Risk Evaluation for Asbestos Part
1: Chrysotile Asbestos for UTV Gasket Installation/Servicing	120
2.3.1.9.5	Data Assumptions, Uncertainties and Level of Confidence	122
2.3.1.10	Summary of Inhalation Occupational Exposure Assessment	122
2.3.2 Consumer Exposures	124
2.3.2.1 Consumer Inhalation Exposures of Do-It-Yourself (DIY) Mechanics During Brake
Repair: Approach and Methodology	126
Page 7 of352

-------
2.3.2.1.1	Consumer Exposure Results - Do-It-Yourself (DIY) Mechanics During Brake Repair
	128
2.3.2.1.2	Exposure Data for Do-It-Yourself (DIY) Mechanics During Brake Repair	129
2.3.2.1.3	Exposure Estimates for DIY Brake Repair/Replacement Scenario	132
2.3.2.1.4	Data Assumptions, Uncertainties and Level of Confidence	133
2.3.2.2	Consumer Exposures Approach and Methodology - DIY Gaskets in UTVs	135
2.3.2.2.1	Consumer Inhalation Exposures - DIY Gaskets in UTVs	138
2.3.2.2.2	Exposure Estimates for DIY UTV Exhaust System Gasket Removal/Replacement
Scenario	139
2.3.2.2.3	Data Assumptions, Uncertainties and Level of Confidence	139
2.3.2.3	Summary of Inhalation Data Supporting the Consumer Exposure Assessment	141
2.3.3 Potentially Exposed or Susceptible Subpopulations	142
3 HAZARDS (EFFECTS)	145
3.1	Environmental Hazards	145
3.1.1	Approach and Methodology	145
3.1.2	Hazard Identification - Toxicity to Aquatic Organisms	145
3.1.3	Weight of Scientific Evidence	146
3.1.4	Summary of Environmental Hazard	147
3.2	Human Health Hazards from Inhalation of Chrysotile Asbestos	147
3.2.1	Approach and Methodology	148
3.2.2	Hazard Identification from Inhalation of Chrysotile Asbestos	150
3.2.2.5 Non-Cancer Hazards from Inhalation of Chrysotile Asbestos	150
3.2.2.2	Cancer Hazards from Inhalation of Chrysotile Asbestos	151
3.2.2.3	Mode of Action (MOA) Considerations for Chrysotile Asbestos	151
3.2.3	Derivation of a Chrysotile Asbestos Inhalation Unit Risk	153
3.2.3.1	Considerations in Derivation of a Chrysotile Asbestos Inhalation Unit Risk	153
3.2.3.2	Rationale for Asbestos-Specific Data Evaluation Criteria	153
3.2.3.3	Additional considerations for final selection of studies for exposure-response	155
3.2.3.4	Statistical Methodology	157
3.2.3.4.1	Cancer Risk Models for Asbestos Exposures	157
3.2.3.4.2	Derivation of Potency Factors	158
3.3.3.4.3	Extrapolation from Workers to the General Population to Derive an Inhalation Unit
Risk 159
3.2.3.4.4	Life-Table Analysis and Derivation of Inhalation Unit Risk	160
3.2.3.5	Study Descriptions and Model Fitting Results	161
3.2.3.5.1	Highest quality cohorts with results carried forward for IUR derivation	162
3.2.3.5.2	Other cohorts with results not carried forward for IUR derivation	166
3.2.3.6	Lung Cancer and Mesothelioma Potencies Ranges by Industry	170
3.2.3.7	Summary of Results of North and South Carolina Cohorts	171
3.2.3.8	Derivation of Inhalation Unit Risk of Cancer Incidence: Issues to Consider	171
3.2.3.8.1	Biases in the Cancer Risk Values	171
3.2.3.8.2	Combining Lung Cancer Unit Risk and Mesothelioma Unit Risk	173
3.2.3.9	Derivation of Inhalation Unit Risk of Cancer Incidence	174
3.2.4	Potentially Exposed or Susceptible Subpopulations	175
Page 8 of352

-------
4	RISK CHARACTERIZATION	177
4.1	Environmental Risk	Ill
4.2	Human Health Ri sk	178
4.2.1	Risk Estimation Approach	178
4.2.2	Risk Estimation for Workers: Cancer Effects Following Less than Lifetime Inhalation
Exposures by Conditions of Use	183
4.2.2.1	Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures for Chlor-
alkali Industry	184
4.2.2.2	Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures for Sheet
Gasket Stamping	188
4.2.2.3	Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures for Sheet
Gasket Use in Chemical Production	190
4.2.2.4	Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures for Oilfield
Brake Blocks	192
4.2.2.5	Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures for
Aftermarket Auto Brakes and Clutches	193
4.2.2.6	Risk Estimation for Cancer Effects Following Chronic Exposures for Other Vehicle
Friction Products	196
4.2.2.7	Risk Estimation for Cancer Effects Following Chronic Exposures for Replacing Brakes
on the NASA Large Transport Plane {i.e., Super Guppy)	199
4.2.2.8	Risk Estimation for Cancer Effects Following Inhalation Exposures for Gasket
Installation/Servicing in UTVs	200
4.2.2.8. Summary of Risk Estimates for Cancer Effects for Occupational Inhalation Exposure
Scenarios for All COUs	202
4.2.3	Risk Estimation for Consumers: Cancer Effects by Conditions of Use	204
4.2.3.1	Risk Estimation for Cancer Effects Following Episodic Inhalation Exposures for DIY
Brake Repair/Replacement	204
4.2.3.2	Risk Estimation for Cancer Effects following Episodic Inhalation Exposures for UTV
Gasket Repair/replacement	209
4.2.3.3	Summary of Consumer and Bystander Risk Estimates by COU for Cancer Effects
Following Inhalation Exposures	211
4.3	Assumptions and Key Sources of Uncertainty	213
4.3.1	Key Assumptions and Uncertainties in the Uses of Asbestos in the U.S	213
4.3.2	Key Assumptions and Uncertainties in the Environmental (Aquatic) Assessment	214
4.3.3	Key Assumptions and Uncertainties in the Occupational Exposure Assessment	215
4.3.4	Key Assumptions and Uncertainties in the Consumer Exposure Assessment	216
4.3.5	Key Assumptions and Uncertainties in the Human Health IUR Derivation	217
4.3.6	Key Assumptions and Uncertainties in the Cancer Risk Values	218
4.3.7	Confidence in the Human Health Risk Estimations	219
4.4	Other Risk-Related Considerations	226
4.4.1	Potentially Exposed or Susceptible Subpopulations	226
4.4.2	Aggregate and Sentinel Exposures	227
4.5	Risk Conclusions	228
4.5.1	Environmental Risk Conclusions	228
4.5.2	Human Health Risk Conclusions to Workers	228
4.5.3	Human Health Risk Conclusions to Consumers	228
5	RISK DETERMINATION	229
5.1 Unreasonable Risk	229
Page 9 of352

-------
5.1.1	Overview	229
5.1.2	Risks to Human Health	230
5.1.3	Determining Environmental Risk	231
5.2	Risk Determination for Chrysotile Asbestos	231
5.2.1	Occupational Processing and Use of Chrysotile Asbestos	234
5.2.2	Consumer Uses of Chrysotile Asbestos	243
5.3	Unreasonable Risk Determination Conclusion	247
5.3.1	No Unreasonable Risk Determinations	247
5.3.2	Unreasonable Risk Determinations	248
5.4	Risk Determination for Five other Asbestiform Varieties	249
6 REFERENCES	250
APPENDICES	263
Appendix A Regulatory History	263
A.l Federal Laws and Regulations		..........................			.263
A,2 State Laws and Regulations											..267
A.3 International Laws and Regulations[[[269
Appendix B List of Supplemental Documents	270
Appendix C Conditions of Use Supplementary Information	272
Appendix D Releases and Exposure to the Invironment Supplementary Information.,,,,,,...,.,. 274
Appendix El I^co logical I^ata Fix traction Fairies	2H0
Appendix F Environmental Fate Data Extraction Table	285
Appendix G SAS Codes for Estimating Ki and Km from Grouped Data	291
Appendix H BKIR IV Equations for l ife Table Analysis	297
Appendix I S. V S (.ode late 1 a bit . V o a I \ sis299
Appendix J Resu Its of Modeling for IUR Derivation	329
Appendix K Less Than Lifetime (or Partial lifetime) IUR	332
Appendix L Sensitivity Analysis of Exposures for DIY/Bystander Episodic Exposure Scenarios
334

-------
LIST OF TABLES
Table 1-1. Physical and Chemical Properties of Chrysotile Asbestos Fibersa	39
Table 1-2. Assessment History of Asbestos	42
Table 1-3. Categories Determined Not to be Manufactured (Including Imported), Processed, or
Distributed for the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos	44
Table 1-4. Categories of Conditions of Use Included in this Risk Evaluation for Asbestos Part 1:
Chrysotile Asbestos	45
Table 2-1. EPA OW Six Year Review Cycle Data for Asbestos in Drinking Water, 1998-2011	 67
Table 2-2. Crosswalk of Conditions of Use and Occupational and Consumer Scenarios Assessed in the
Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos	70
Table 2-3. Assigned Protection Factors for Respirators in OSHA Standard 29 CFR 1910.134eg	73
Table 2-4. 30-min Short-Term PBZ Sample Summary*	82
Table 2-5. Full-Shift* PBZ Sample Summary**	82
Table 2-6. Summary of PBZ Sampling Data for All Other Durations*	82
Table 2-7. Summary of ACC Short-Term PBZ Sampling Data by Exposure Group (samples from 2001
to 2016)	83
Table 2-8. Summary of Chrysotile Asbestos Exposures During Processing and Use in the Chlor-Alkali
Industry Used in EPA's Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos	84
Table 2-9. Short-Term PBZ Chrysotile Asbestos Sampling Results (EHM, 2013)	91
Table 2-10. Summary of Asbestos Exposures During Sheet Gasket Stamping Used in EPA's Risk
Evaluation for Asbestos Part 1: Chrysotile Asbestos	92
Table 2-11. Summary of Asbestos Exposures During Sheet Gasket Use Used in the Risk Evaluation for
Asbestos Part 1: Chrysotile Asbestos	97
Table 2-12. Summary of Total Establishments in Relevant Industries and Potentially Exposed Workers
and ONUs for Oilfield Brake Blocks	100
Table 2-13. Summary of Asbestos Exposures During Use in Brake Blocks for the Risk Evaluation for
Asbestos Part 1: Chrysotile Asbestos	101
Table 2-14. PBZ Asbestos Concentrations Measured by OSHA for Workers at Automotive Repair,
Services, and Parking Facilities	109
Table 2-15. Summary of Asbestos Exposures During Replacement of Aftermarket Automotive Parts
Used in the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos	Ill
Table 2-16. Other Vehicle Friction Products Exposure Levels (from Aftermarket Automotive Parts
exposure levels) Used in the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos 113
Table 2-17. Summary of Asbestos Exposures During Replacement of Brake Pads/Blocks in the NASA
Super Guppy Used in the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos.... 117
Table 2-18. Number of Other Motor Vehicle Dealers	118
Table 2-19. Number of ATV and Watercraft Dealers in NAICS 44128 	 119
Table 2-20. Selected Mechanics and Repair Technicians in NAICS 4412 (Other Motor Vehicle Dealers)
	119
Table 2-21. Number of Employees per Establishment in NAICS 4412 in Relevant Occupations	120
Table 2-22. Estimated Number of Sites and Employees for UTV Engine Repair	120
Table 2-23. UTV Gasket Installation/Servicing Exposure Levels for the Risk Evaluation for Asbestos
Part 1: Chrysotile Asbestos	121
Table 2-24. Summary of Occupational Inhalation Exposures	123
Table 2-25. Summary of Studies Satisfying Conditions/Factors for Use in Consumer DIY Brake
Exposure Scenario	128
Table 2-26. Exposure Concentrations from Blake (2003) and Sheehy (1989) Studies to the DIY User
During Various Activities	130
Page 11 of 352

-------
Table 2-27. Estimated Exposure Concentration for DIY Consumer User and Bystander for Risk
Evaluation for Asbestos Part 1: Chrysotile Asbestos - DIY Brake Repair/Replacement
Scenario	132
Table 2-28. Summary of Studies Satisfying Factors Applied to Identified Literature	136
Table 2-29. Summary Results of Asbestos Exposures in Gasket Repair Studies	137
Table 2-30. Estimated Exposure Concentrations for UTV Gasket Repair/Replacement Scenario - DIY
Mechanic and Bystander for Use in the Risk Evaluation for Asbestos Part 1: Chrysotile
Asbestos	139
Table 2-31. Summary of Consumer Inhalation Exposures	141
Table 2-32. Percentage of Employed Persons by Age, Sex, and Industry Sector (2017 and 2018 worker
demographics from BLS)	143
Table 2-33. Percentage of Employed Adolescents by Industry Sector (2017 and 2018 worker
demographics from BLS)	144
Table 3-1. Environmental Hazard Characterization of Chrysotile Asbestos	147
Table 3-2. Study Cohort, Individual studies and Study Quality of Commercial Chrysotile Asbestos
Reviewed for Assessment of Lung Cancer and Mesothelioma Risks	156
Table 3-3. Model Fitting Results for the South Carolina Cohort	163
Table 3-4. Model Fitting Results for the North Carolina Cohort	165
Table 3-5. Model Fitting Results for the Chongqing China Cohort	167
Table 3-6. Model Fitting Results for the Quebec, Canada Cohort	169
Table 3-7. Model Fitting Results for the Qinghai, China Cohort	170
Table 3-8. Comparison of Cancer Potencies (KLand Km) by Industry	171
Table 3-9. Cohorts and Preferred Statistical Models for SC and NC Cohorts	171
Table 3-10. Addressing Underascertainment of Mesothelioma	172
Table 4-1. Use Scenarios and Populations of Interest for Cancer Endpoints for Assessing Occupational
Risks Following Inhalation Exposures to Chrysotile Asbestos	181
Table 4-2. Use Scenarios and Populations of Interest for Cancer Endpoints for Assessing Consumer
Risks Following Inhalation Exposures to Chrysotile Asbestos	181
Table 4-3. Reported Respirator Use by COU for Asbestos Occupational Exposures	182
Table 4-4. Excess Lifetime Cancer Risk for Chlor-alkali Industry Full Shift Workers and ONUs
(Personal Samples) before consideration of PPE and any relevant APF	184
Table 4-5. Excess Lifetime Cancer Risk for Chlor-alkali Industry Workers (Short-Term Personal
Samples from Table 2-4, 8-hour full shift) before consideration of PPE and any relevant
APF	185
Table 4-6. Excess Lifetime Cancer Risk for Chlor-alkali Industry Full Shift Workers and ONUs (from
Table 4-4) after consideration of PPE with APF=10 for all workers (excluding ONUs)186
Table 4-7. Excess Lifetime Cancer Risk for Chlor-alkali Industry Full Shift Workers and ONUs (from
Table 4-4) after consideration of PPE with APF=25 for all workers (excluding ONUs)186
Table 4-8. Excess Lifetime Cancer Risk for Chlor-alkali Industry Short-Term Personal Samples (from
Table 4-5) after consideration of PPE with APF=25 for short-term workers for 0.5 hours
(excluding ONUs)	186
Table 4-9. Excess Lifetime Cancer Risk for Chlor-alkali Industry Short-Term Personal Samples (from
Table 4-5) after consideration of PPE and with APF=10 for full-shift workers and with
APF=25 for short-term workers (excluding ONUs)	187
Table 4-10. Excess Lifetime Cancer Risk for Chlor-alkali Industry Short-Term Personal Samples (from
Table 4-5) after consideration of PPE and with APF=25 for full-shift workers and with
APF=25 for short-term workers (excluding ONUs)	187
Page 12 of 352

-------
Table 4-11. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Full Shift Workers and ONUs
(from Table 2-10, Personal Samples) before consideration of PPE and any relevant APF
	188
Table 4-12. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Short-term Exposures within an 8-
hour Full Shift (from Table 2-10, Personal Samples) before consideration of PPE and any
relevant APF	189
Table 4-13. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Full Shift Workers and ONUs
(from Table 4-11) after consideration of PPE using an APF=10 (excluding ONUs)	189
Table 4-14. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Full Shift Workers and ONUs
(from Table 4-11) after consideration of PPE using an APF=25 (excluding ONUs)	190
Table 4-15. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Short-term Exposures within an 8-
hour Full Shift (from Table 4-12) after consideration of PPE using an APF=10 for both
full-shift and short-term exposures (excluding ONUs)	190
Table 4-16. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Short-term Exposures within an 8-
hour Full Shift (from Table 4-12) after consideration of PPE using an APF=25 for both
full-shift and short-term exposures (excluding ONUs)	190
Table 4-17. Excess Lifetime Cancer Risk for Sheet Gasket Use in Chemical Production (using data from
titanium dioxide production), 8-hour TWA (from Table 2-11., Personal Samples) before
consideration of PPE and any relevant APF	191
Table 4-18. Excess Lifetime Cancer Risk for Sheet Gasket Use in Chemical Production, 8-hour TWA
(from Table 4-6) after consideration of PPE using the APF=10 reflecting the current use
of respirators (excluding ONUs)	191
Table 4-19. Excess Lifetime Cancer Risk for Sheet Gasket Use in Chemical Production, 8-hour TWA
(from Table 4-6) after consideration of PPE using an APF=25 (excluding ONUs)	191
Table 4-20. Excess Lifetime Cancer Risk for Oil Field Brake Block Use, 8-hour TWA (from Table 2-13
before consideration of PPE and any relevant APF	192
Table 4-21. Excess Lifetime Cancer Risk for Oil Field Brake Block Use, 8-hour TWA (from Table 4-20)
after consideration of PPE using an APF=10 (excluding ONUs)	192
Table 4-22. Excess Lifetime Cancer Risk for Oil Field Brake Block Use, 8-hour TWA (from Table 4-20)
after consideration of PPE using an APF=25 (excluding ONUs)	193
Table 4-23. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes and
Clutches in an Occupational Setting, 8-hour TWA Exposure (from Table 2-15.) before
consideration of PPE and any relevant APF	193
Table 4-24. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes and
Clutches in an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift
(from Table 2-15.) before consideration of PPE and any relevant APF	194
Table 4-25. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes and
Clutches in an Occupational Setting, 8-hour TWA Exposure (from Table 4-23) after
consideration of PPE with APF=10 (excluding ONUs)	194
Table 4-26. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes and
Clutches in an Occupational Setting, 8-hour TWA Exposure (from Table 4-24.) after
consideration of PPE with APF=25 (excluding ONUs)	195
Table 4-27. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes and
Clutches in an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift
(from Table 4-24) after consideration of PPE with APF=10 (excluding ONUs)	195
Table 4-28. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes and
Clutches in an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift
(from Table 4-24) after consideration of PPE with APF=25 (excluding ONUs)	195
Page 13 of 352

-------
Table 4-29. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in an
Occupational Setting, 8-hour TWA Exposure (from Table 2-15.) before consideration of
PPE and any relevant APF	196
Table 4-30. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in an
Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from Table
2-15.) before consideration of PPE and any relevant APF	197
Table 4-31. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in an
Occupational Setting, 8-hour TWA Exposure (from Table 4-29) after consideration of
PPE with APF=10 (excluding ONUs)	198
Table 4-32. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in an
Occupational Setting, 8-hour TWA Exposure (from Table 4-24.) after consideration of
PPE with APF=25 (excluding ONUs)	198
Table 4-33. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in an
Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from Table 4-
30) after consideration of PPE with APF=10 (excluding ONUs)	198
Table 4-34. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in an
Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from Table 4-
30) after consideration of PPE with APF=25 (excluding ONUs)	199
Table 4-354-35. Excess Lifetime Cancer Risk for Replacing Brakes on the NASA Large Transport Plane
{i.e., Super Guppy) in an Occupational Setting, 8-hour TWA Exposure (from Table 2-17)
before consideration of PPE and any relevant APF	199
Table 4-364-36. Excess Lifetime Cancer Risk for Replacing Brakes on the NASA Large Transport Plane
{i.e., Super Guppy) in an Occupational Setting, Short-term Exposures Within an 8-hour
TWA (from Table 2-17) before consideration of PPE and any relevant APF	200
Table 4-37. Excess Lifetime Cancer Risk for UTV Gasket Installation/Servicing in an Occupational
Setting, 8-hour TWA Exposure (from Table 2-23.) before consideration of PPE and any
relevant APF	201
Table 4-38. Excess Lifetime Cancer Risk for UTV Gasket Installation/Servicing in an Occupational
Setting, 8-hour TWA Exposure (from Table 4-35) after consideration of PPE with
APF=10 (excluding ONUs)	201
Table 4-39. Excess Lifetime Cancer Risk for UTV Gasket Installation/Servicing in an Occupational
Setting, 8-hour TWA Exposure (from Table 4-35) after consideration of PPE with
API; 25 (excluding ONUs)	201
Table 4-40. Summary of Risk Estimates for Inhalation Exposures to Workers and ONUs by COU .... 202
Table 4-41. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with Compressed
Air Use for Consumers and Bystanders (exposures from Table 2-31 without a reduction
factor) with Exposures at 30% of 3-hour User Concentrations between Brake/Repair
Replacement (Consumers 1 hour/day spent in garage; Bystanders 1 hour/day)	206
Table 4-42. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with Compressed
Air Use for Consumers for 20-year duration (exposures from Table 2-31 without a
reduction factor) with Exposures at 30% of 3-hour User Concentrations between
Brake/Repair Replacement (Consumers 1 hour/day spent in garage)	206
Table 4-43. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with Compressed
Air Use for Consumers and Bystanders (exposures from Table 2-31 without a reduction
factor) with Exposures at 30% of 3-hour User Concentrations between Brake/Repair
Replacement (Consumers 8-hours/day spent in garage; Bystanders 1 hour/day)	207
Table 4-44. Risk Estimate using one brake change at age 16 years with 10 years further exposure.
Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with
Page 14 of 352

-------
Compressed Air Use for Consumers and Bystanders (exposures from Table 2-31 without
a reduction factor) (Consumers 1 hour/day spent in garage; Bystanders 1 hour/day).... 207
Table 4-45. Excess Lifetime Cancer Risk for Outdoor DIY Brake/repair Replacement for Consumers
and Bystanders (5 minutes per day in driveway) (from Table 2-31 with a reduction factor
of 10)	208
Table 4-46. Excess Lifetime Cancer Risk for Outdoor DIY Brake/Repair Replacement for Consumers
and Bystanders (30 minutes per day in driveway) (from Table 2-31 with a reduction
factor of 10)	209
Table 4-47. Risk Estimate using one UTV gasket change at age 16 years with 10 years further exposure.
Excess Lifetime Cancer Risk for Indoor DIY UTV gasket change for Consumers and
Bystanders (exposures from Table 2-31 without a reduction factor) (Consumers 1
hour/day spent in garage; Bystanders 1 hour/day)	210
Table 4-48. Excess Lifetime Cancer Risk for Indoor DIY UTV Gasket /Repair Replacement for
Consumers and Bystanders (exposures from Table 2-31) (Users 1 hour/day spent in
garage; Bystanders 1 hour/day)	210
Table 4-49. Excess Lifetime Cancer Risk for Indoor DIY Gasket/Repair Replacement for Consumers
and Bystanders (exposures from Table 2-31) (Consumers 8-hours/day spent in garage;
Bystanders 1 hour/day)	211
Table 4-50. Summary of Risk Estimates for Inhalation Exposures to Consumers and Bystanders by COU
(Cancer benchmark is 10-6)	212
Table 4-51. Ratios of risks for alternative exposure scenarios using scenario-specific partial lifetime
IURs from Appendix K by age at first exposure and duration of exposure compared to
baseline occupational exposure scenarios (baseline scenario: first exposure at 16 years for
40 years duration)	220
Table 4-52. Ratios of risks for alternative exposure scenarios using scenario-specific partial lifetime
IURs from Appendix K by age at first exposure and duration of exposure compared to
baseline consumer DIY exposure scenarios (baseline scenario: first exposure at 16 years
for 62 years duration)	221
Table 4-53. Ratios of risks for alternative exposure scenarios using scenario-specific partial lifetime
IURs from Appendix K by age at first exposure and duration of exposure compared to
baseline consumer bystander exposure scenarios (baseline scenario: first exposure at 0
years for 78 years duration)	221
Table 4-54. Results of Sensitivity Analysis of Exposure Assumptions for Consumer DIY/Bystander
Episodic Exposure Scenarios	222
Table 4-55. Time Spent (minutes/day) in Garage, Doers Only (Taken from Table 16-16 in EFH, 2011)
	223
Table 4-56. Summary of Estimated Number of Exposed Workers and DIY Consumers3	226
Table 5-1. Risk Determination for Chrysotile Asbestos: Processing and Industrial Use of Asbestos
Diaphragms in Chlor-alkali Industry (refer to section 4.2.2.1 for the risk characterization)
	235
Table 5-2. Risk Determination for Chrysotile Asbestos: Processing Asbestos-Containing Sheet Gaskets
(refer to section 4.2.2.2 for the risk characterization)	237
Table 5-3. Risk Determination for Chrysotile Asbestos: Industrial Use of Asbestos-Containing Sheet
Gaskets in Chemical Production	239
Table 5-4. Risk Determination for Chrysotile Asbestos: Industrial Use and Disposal of Asbestos-
Containing Brake Blocks in Oil Industry (refer to section 4.2.2.4 for the risk
characterization)	240
Table 5-5. Risk Determination for Chrysotile Asbestos: Commercial Use and Disposal of Aftermarket
Automotive Asbestos-Containing Brakes/Linings and Other Vehicle Friction Products
Page 15 of 352

-------
(Commercial Mechanic Brake Repair/Replacement is Representative for both COUs;
refer to section 4.2.2.5 and 4.2.2.6 for the risk characterization)	241
Table 5-6. Risk Determination for Chrysotile Asbestos: Commercial Use and Disposal of Other
Asbestos-Containing Gaskets	242
Table 5-7. Risk Determination for Chrysotile Asbestos: Consumer Use and Disposal of Aftermarket
Automotive Asbestos-Containing Brakes/Linings	243
Table 5-8. Risk Determination for Chrysotile Asbestos: Consumer Use and Disposal of Other Asbestos-
Containing Gaskets	246
Page 16 of 352

-------
LIST OF FIGURES
Figure P-l: Schematic of the TSCA Risk Evaluation Timeline	4
Figure P-2: Important Definitions for the Risk Evaluation for Asbestos	5
Figure 1-1. Chrysotile Asbestos	40
Figure 1-2. Chrysotile Asbestos Life Cycle Diagram	46
Figure 1-3. Chrysotile Asbestos Conceptual Model for Industrial and Commercial Activities and Uses:
Potential Exposures and Hazards	55
Figure 1-4. Chrysotile Asbestos Conceptual Model for Consumer Activities and Uses: Potential
Exposures and Hazards	56
Figure 1-5. Key/Supporting Data Sources for Environmental Fate	60
Figure 1-6. Key/Supporting Data Sources for Engineering Releases and Occupational Exposure	61
Figure 1-7. Key/Supporting Data Sources for Consumer and Environmental Exposure	62
Figure 1-8. Key /Supporting Data Sources for Environmental Hazard	63
Figure 1-9. Key/Supporting Data Sources for Human Health Hazard	64
Figure 2-1. Closeup of a Chrysotile Asbestos Diaphragm Outside of the Electrolytic Cell Photograph
Courtesy of the American Chemistry Council	75
Figure 2-2. Process Flow Diagram of an Asbestos Handling System and Slurry Mix Tank Image
Courtesy of the American Chemistry Council	77
Figure 2-3. Electrolytic Cell Construction	78
Figure 2-4. Typical Gasket Assembly	86
Figure 2-5. Chrysotile Asbestos-Containing Stamping Operation	87
Figure 2-6. Rule Blade for Stamping Machine	87
Figure 2-7. Asbestos Warning Label on Finished Gasket Product	88
Figure 2-8. Photographs of Typical Oil Field Drawworks	98
Figure 2-9. Illustrations of brake assembly components: (a) a brake lining designed to be used with an
internal drum brake and (b) a brake pad designed for use with a disc brake	104
Figure 2-10. Schematic of a clutch assembly. The clutch disc is made of friction material, which may
contain asbestos	105
Figure 2-11. NASA Super Guppy Turbine Aircraft	114
Figure 2-12. Brakes for NASA Super Guppy Turbine Aircraft	115
Figure 2-13. Ventilated Walk-in Booth Where Brakes Pads Are Replaced	116
Figure 3-1. EPA Approach to Hazard Identification, Data Integration, and Dose-Response Analysis for
Chrysotile Asbestos	148
Page 17 of 352

-------
LIST OF APPENDIX TABLES
TableAPX D-l. Summary of Asbestos TRI Production-Related Waste Managed from 2015-2018 (lbs)
	275
Table APX D-2. Summary of Asbestos TRI Releases to the Environment from 2015-2018 (lbs)	276
Table APX E-l. Summary Table On-topic Aquatic Toxicity Studies That Were Evaluated for
Chrysotile Asbestos	280
Table APX F-l. Other Fate Endpoints Study Summary for Chrysotile Asbestos	285
Table_APX F-2. Hydrolysis Study Summary for Chrysotile Asbestos	287
Table APX F-3. Aquatic Bioconcentration Study Summary for Chrysotile Asbestos	289
TableApx K-l. (LTL) Chrysotile Asbestos Inhalation Unit Risk Values for Less Than Lifetime
Condition of Use	332
Table Apx L-l. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with
Compressed Air Use for Consumers for 20 year duration (exposures from Table 2-27
without a reduction factor) (Consumers 1 hour/day spent in garage)	334
Table Apx L-2. Ratios of risk for alternative exposure scenarios compared to DIY User and Bystander
exposure scenario assuming DIY User is first exposed at age 16 years for 62 years
duration and DIY Bystander is exposed from age 0-78 years	335
TableApx L-3. Sensitivity Analysis #1: Summary of Risk Estimates for Inhalation Exposures to
Consumers and Bystanders by COU (Cancer benchmark is 10"6) Comparing the Baseline
Exposure Scenario from Table 4-48 with Risks Assuming DIY Users Are Exposed From
Age 16-36 years and Bystanders Are Exposed Age 0-20 years	336
Table Apx L-4: Results of 24 Sensitivity Analysis of Exposure Assumptions for Consumer
DIY/Bystander Episodic Exposure Scenarios	346
Table_Apx M-l: Estimate of adjustment factor for ovarian cancer	351
Table_Apx M-2: Adjustment factor for laryngeal cancer	352
Page 18 of 352

-------
ACKNOWLEDGEMENTS
This report was developed by the United States Environmental Protection Agency (U.S. EPA), Office of
Chemical Safety and Pollution Prevention (OCSPP), Office of Pollution Prevention and Toxics (OPPT).
Acknowledgements
The OPPT Assessment Team gratefully acknowledges participation and/or input from Intra-agency
reviewers that included multiple offices within EPA, Inter-agency reviewers that included multiple
Federal agencies, and assistance from EPA contractors GDIT (Contract No. CIO-SP3,
HHSN316201200013W), ERG (Contract No. EP-W-12-006), Versar (Contract No. EP-W-17-006), ICF
(Contract No. EPC14001) and SRC (Contract No. EP-W-12-003).
EPA acknowledges three academic epidemiologists (Drs. Leslie Elliott, Dana Loomis and Leslie
Stayner) subcontracted to SRC who contributed to the development of the Inhalation Unit Risk (IUR),
were present at the Science Advisory Committee on Chemicals (SACC) meeting in June of 2020 and
helped address appropriate SACC and public comments. Dr. Leslie Stayner developed Appendix M.
Docket
Supporting information can be found in public docket: IJHJ IQ-OPP'l _.aM y V 36.
Disclaimer
Reference herein to any specific commercial products, process or service by trade name, trademark,
manufacturer or otherwise does not constitute or imply its endorsement, recommendation or favoring by
the United States Government.
Authors/Contributors
Andrew Gillespie (Division Director), Ryan Wallace (Deputy Division Director), Collin Beachum
(Management Lead), Jennifer Nichols (Staff Lead), Juan Bezares-Cruz, Robert Courtnage, Jay Jon,
Emily Nolan, Abhilash Sasidharan, William Silagi, Mitchell Sumner, Kevin Vuilleumier, and Erik
Winchester. Former team members: Sheila Canavan (former Division Director), Stan Barone (former
Deputy Division Director), Louis Scarano (former Management Lead), Andrea Pfahles-Hutchens
(former Staff Lead), Lea Carmichael, Freeborn (Garrett) Jewett, Nathan Mottle, Amelia Nguyen, Heidi
Bethel, Francesca Branch (formerly EPA), Mari Lee (formerly EPA), Richard Fehir (formerly EPA),
Ernest Falke (retried), Sharon Austin (deceased)
Technical Assistance and Support Staff: Hillary Hollinger, Cynthia McOliver, and Khoa Nguyen.
OPPT gratefully acknowledges the contributions of our colleagues from EPA's Office of Research and
Development (ORD): Thomas Bateson, Leonid Kopylev, Ingrid Druwe, and Walter Cybulski. Drs.
Bateson and Kopylev served as primary authors for the hazard, dose-response and risk characterizations
sections of this document (Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos). Dr. Ingrid Druwe
was the lead contributor for the mode of action section.
Page 19 of 352

-------
ABBREVIATIONS
ABPO
1989 Asbestos Ban and Phase Out Rule
ACC
American Chemistry Council
ADC
Average Daily Concentration
AHERA
Asbestos Hazard Emergency Response Act
AIC
Akaike Information Criterion
ASHAA
Asbestos School Hazard Abatement Act
ASHARA
Asbestos School Hazard Abatement Reauthorization Act
ATSDR
Agency for Toxic Substances and Disease Registry
CAA
Clean Air Act
CASRN
Chemical Abstracts Service Registry Number
CBI
Confidential Business Information
CDR
Chemical Data Reporting
CERCLA
Comprehensive Environmental Response, Compensation and Liability Act
COU
Condition of Use
CPSC
Consumer Product Safety Commission
CWA
Clean Water Act
DIY
Do-It-Yourself
DPT
Diffuse Pleural Thickening
EG
Effluent Guideline
ELCR
Excess Lifetime Cancer Risk
EMP
Elongated Mineral Particle
EPA
Environmental Protection Agency
EPCRA
Emergency Planning and Community Right-to-Know Act
EU
European Union
FDA
Food and Drug Administration
f/cc
Fibers per cubic centimeter
FHSA
Federal Hazardous Substance Act
g
Gram(s)
HAP
Hazardous Air Pollutant
HEPA
High-Efficiency Particulate Air
HTS
Harmonized Tariff Schedule
IARC
International Agency for Research on Cancer
IRIS
Integrated Risk Information System
IUR
Inhalation Unit Risk
Ki
Lung cancer potency factor
Km
Mesothelioma potency factor
LADC
Lifetime Average Daily Concentration
lb
Pound
LTL
Less Than Lifetime
LOEC
Lowest Observable Effect Concentration
MAP
Model Accreditation Plan
MCLG
Maximum Contaminant Level Goal
|im
Micrometers
MFL
Million Fibers per Liter
mppcf
million particles per cubic foot of air
mg
Milligram(s)
MPa
Megapascal
MSHA
Mine Safety and Health Administration
Page 20 of 352

-------
mV
Millivolt
NAICS
North American Industry Classification System
ND
Non-detects (value is < analytical detection limit)
NEI
National Emissions Inventory
NESHAP
National Emission Standard for Hazardous Air Pollutants
NIH
National Institutes of Health
NIOSH
National Institute for Occupational Safety and Health
NMRD
Non-Malignant Respiratory Disease
NPL
National Priorities List
NTP
National Toxicology Program
OCSPP
Office of Chemical Safety and Pollution Prevention
OEM
Original Equipment Manufacturer
ONU
Occupational Non-User
OPPT
Office of Pollution Prevention and Toxics
OSHA
Occupational Safety and Health Administration
PCM
Phase Contrast Microscopy
PECO
Population, Exposure, Comparator and Outcome
PEL
Permissible Exposure Limit
PESO
Pathways/Processes, Exposure, Setting and Outcomes
PF
Problem Formulation
POD
Point of Departure
POTW
Publicly Owned Treatment Works
PPE
Personal Protective Equipment
ppm
Part(s) per Million
RCRA
Resource Conservation and Recovery Act
RA
Risk Assessment
RESO
Receptors, Exposure, Setting/Scenario and Outcomes
RfC
Reference Concentration
RIA
Regulatory Impact Analysis
RR
Relative Risk
SACC
Science Advisory Committee on Chemicals
SDS
Safety Data Sheet
SDWA
Safe Drinking Water Act
SMR
Standardized Mortality Ratio
SNUN
Significant New Use Notice
SNUR
Significant New Use Rule
TSFE
Time Since First Exposure
TCCR
Transparent, Clear, Consistent, and Reasonable
TEM
Transmission Electron Microscopy
TRI
Toxics Release Inventory
TSCA
Toxic Substances Control Act
TURA
Toxics Use Reduction Act
TWA
Time Weighted Average
UB
Upper Bound
U.S.
United States
USGS
United States Geological Survey
UTV
Utility vehicle
WHO
World Health Organization
Page 21 of 352

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
EXECUTIVE SUMMARY
This Risk Evaluation for Asbestos, Part 1: Chrysotile Asbestos (hereafter referred to as "Part 1" or "Part
1 of the risk evaluation") for imported, processed and distributed uses of chrysotile asbestos1 was
performed in accordance with the Frank R. Lautenberg Chemical Safety for the 21st Century Act and is
being issued following public comment and peer review. The Frank R. Lautenberg Chemical Safety for
the 21st Century Act amended the Toxic Substances Control Act (TSCA), the Nation's primary
chemicals management law, in June 2016. Under the amended statute, EPA is required, under TSCA
Section 6(b), to conduct risk evaluations to determine whether a chemical substance presents an
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. Also, as required by TSCA
Section 6(b), EPA established, by rule, a process to conduct these Risk Evaluations, Procedures for
Chemical Risk Evaluation Under the Amended Toxic Substances Control Act (82 FR 33726) (Risk
Evaluation Rule). Part 1 of the risk evaluation is in conformance with TSCA Section 6(b) and the Risk
Evaluation Rule and is to be used to inform risk management decisions. In accordance with TSCA
Section 6(b), if EPA finds unreasonable risk from a chemical substance under its conditions of use in
any final Risk Evaluation, the Agency will propose actions to address those risks within the timeframe
required by TSCA. However, any proposed or final determination that a chemical substance presents
unreasonable risk under TSCA Section 6(b) is not the same as a finding that a chemical substance is
"imminently hazardous" under TSCA Section 7. The conclusions, findings, and determinations in Part 1
are for the purpose of identifying whether the chemical substance presents unreasonable risk under the
conditions of use, in accordance with TSCA section 6, and are not intended to represent any findings
under TSCA section 7.
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 with the best
available science and base its decisions on the weight of the scientific evidence2. 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 (U.S. EPA. 2018a). The data collection,
evaluation, and integration stages of the systematic review process are used to develop the exposure, fate
and hazard assessments for the risk evaluations. To satisfy requirements in TSCA Section 26(j)(4) and
40 CFR 702.51(e), EPA has provided a list of studies considered in carrying out Part land the results of
those studies are included in the Systematic Review Data Quality Evaluation Documents (see Appendix
B).
Asbestos is subject to federal and state regulations and reporting requirements. Asbestos is reportable to
the Toxics Release Inventory (TRI) under Section 313 of the Emergency Planning and Community
1	As noted in the PREAMBLE, this document is Part 1 of the final Risk Evaluation for asbestos and is limited to chrysotile
asbestos and the conditions of use (COUs) defined in this document. Part 2 is forthcoming and will be on legacy uses and
associated disposal of asbestos.
2	Weight of the scientific evidence is defined in EPA regulations as a systematic review method, applied in a manner suited to
the nature of the evidence or decision, that uses a pre-established protocol to comprehensively, objectively, transparently, and
consistently identify and evaluate each stream of evidence, including strengths, limitations, and relevance of each study and
to integrate evidence as necessary and appropriate based upon strengths, limitations and relevance. 40 CFR 702.33.

-------
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
Right-to-Know Act (EPCRA) but is only reportable in the friable3 form at concentration levels of 0.1%
or greater. It is designated a Hazardous Air Pollutant (HAP) under the Clean Air Act (CAA), and is a
hazardous substance under the Comprehensive Environmental Response, Compensation and Liability
Act (CERCLA). Asbestos is subject to National Primary Drinking Water Regulations (NPDWR) under
the Safe Drinking Water Act (SDWA) and designated as a toxic pollutant under the Clean Water Act
(CWA) and as such is subject to effluent limitations. Under TSCA, EPA has promulgated several
regulations for asbestos, including the Asbestos Ban and Phase Out rule of 1989, which was then largely
vacated in 1991, and under the Asbestos Hazard Emergency Response Act (AHERA), which requires
inspection of schools for asbestos. On April 25, 2019, EPA finalized an Asbestos Significant New Use
Rule (SNUR) under TSCA Section 5 that prohibits manufacture (including import) or processing of
discontinued uses of asbestos from restarting without EPA having an opportunity to evaluate each
intended use for risks to health and the environment and to take any necessary regulatory action, which
may include a prohibition.
Asbestos has not been mined or otherwise produced in the U.S. since 2002. Although there are several
known types of asbestos, the only form of asbestos known to be imported, processed, or distributed for
use in the United States is chrysotile asbestos. As a naturally occurring mineral, chrysotile can co-occur
with other minerals, including amphibole forms of asbestos. Trace amounts of these minerals may
remain in chrysotile as it is used in commerce. This commercial chrysotile, rather than theoretically
"pure" chrysotile, is therefore the substance of concern for this assessment. Raw chrysotile asbestos
currently imported into the U.S. is used exclusively by the chlor-alkali industry. The total amount of raw
chrysotile asbestos imported into the U.S. in 2019 was 100 metric tons. EPA has also identified the
importation of asbestos-containing products; however, the import volumes of those products are not
fully known. The asbestos-containing products that EPA has identified as being imported and used are
sheet gaskets, brake blocks, aftermarket automotive brakes/linings, other vehicle friction products, and
other gaskets. In Part 1 of the asbestos risk evaluation, EPA evaluated the following categories of
conditions of use (COU): importing; processing; distribution in commerce; occupational and consumer
uses; and disposal.
Approach
EPA used reasonably available information (defined in 40 CFR 702.33 as "information that EPA
possesses or can reasonably generate, obtain, and synthesize for use in risk evaluations, considering the
deadlines specified in TSCA section 6(b)(4)(G) for completing such evaluation "), in a fit-for-purpose
approach, to develop a document that relies on the best available science and is based on the weight of
the scientific evidence. EPA used previous analyses as a starting point for identifying key and
supporting studies to inform the exposure, fate, and hazard assessments. EPA also evaluated other
studies published since the publication of previous analyses. EPA reviewed the information and
evaluated the quality of the methods and reporting of results of the individual studies using the
evaluation strategies described in Application of Systematic Review in TSCA Risk Evaluations (U.S.
EPA. 2018aY
During development of this Part 1 of the risk evaluation for asbestos, the only asbestos fiber type that
EPA identified as imported, processed, or distributed under the COUs in the United States is chrysotile,
the serpentine variety. Chrysotile is the prevailing form of asbestos currently mined worldwide, and so it
is assumed that a majority of commercially available products fabricated overseas that contain asbestos
The TRI listing has the following definition for friable: "This term refers to a physical characteristic of asbestos. EPA
interprets "friable" as being crumbled, pulverized, or reducible to a powder with hand pressure. Again, only manufacturing,
processing, or use of asbestos in the friable form triggers reporting." (40 CFR Part 372).
Page 23 of 352

-------
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
are made with chrysotile. Any asbestos being imported into the U.S. in articles is believed to be
chrysotile. The other five forms of asbestos are now subject to a SNUR4.
EPA evaluated the following categories of COU of chrysotile asbestos in this Part 1 of the risk
evaluation for asbestos: importing; processing; distribution in commerce; occupational and consumer
uses (use of diaphragms in the chlor-alkali industry, sheet gaskets in chemical production facilities,
oilfield brake blocks, aftermarket automotive brakes/linings, other vehicle friction products, and other
gaskets); and disposal. EPA reviewed the court decision in Safer Chemicals Healthy Families v. EPA,
943 F.3d 397 (9th Cir. 2019). This Part 1 of the risk evaluation for asbestos does not reflect
consideration of any legacy uses and associated disposal for chrysotile asbestos or other asbestos fiber
types as a result of that decision. EPA intends to consider legacy uses and associated disposal and other
fiber types in Part 2 of the asbestos risk evaluation.
In the problem formulation (U.S. EPA. 2018d) (PF), EPA identified the COUs and presented three
conceptual models and an analysis plan. These have been carried into this document where EPA has
quantitatively evaluated the risk to human health using monitoring data submitted by industry and found
in the scientific literature through systematic review for the COUs (identified in Section 1.4.3 of this
Part 1 of the risk evaluation for asbestos).
During the PF phase of the Risk Evaluation, EPA was still in the process of identifying potential
chrysotile asbestos water releases for the TSCA COUs to determine the need to evaluate risk to aquatic
and sediment-dwelling organisms. In the draft Risk Evaluation released in March 2020, EPA concluded
that, based on the reasonably available information in the published literature, provided by industries
using asbestos, and reported in EPA databases, there were minimal or no releases of asbestos to surface
water associated with the COUs that EPA is evaluating in Part 1. EPA has considered peer review and
public comments on this conclusion and has retained the finding in the draft Risk Evaluation that there is
low or no potential for environmental risk to aquatic or sediment-dwelling receptors from the COUs
included in this Part 1 of the risk evaluation for asbestos. This is because EPA is confident that the
minimal water release data cannot be attributed to chrysotile asbestos from the COUs in this document.
However, in Part 2 of the Risk Evaluation for Asbestos that will examine legacy uses and associated
disposals of asbestos, EPA expects to address the issue of releases to surface water based on those other
asbestos uses (See Section 4.1).
In occupational settings, EPA evaluated inhalation exposures to workers and occupational non-users, or
ONUs. EPA used inhalation monitoring data submitted by industry and literature sources, where
reasonably available and that met TSCA systematic review data evaluation criteria, to estimate potential
inhalation exposures. In consumer settings, EPA evaluated inhalation exposures to both consumers (Do-
it-Yourselfers or DIY mechanics) and bystanders and used estimated inhalation exposures, from
literature sources where reasonably available and that met data evaluation criteria, to estimate potential
exposures using a range of user durations. These analyses are described in Section 2.3.
EPA evaluated reasonably available information for human health hazards and identified hazard
4 This requires notification to, and review by, the Agency should any person wish to pursue manufacturing, importing, or
processing crocidolite (riebeckite), amosite (cummingtonite-grunerite), anthophyllite, tremolite or actinolite (either in raw
form or as part of articles) for any use (40 CFR 721.11095). Therefore, under the final asbestos SNUR, EPA will be made
aware of manufacturing, importing, or processing for any intended use of crocidolite (riebeckite), amosite (cummingtonite-
grunerite), anthophyllite, tremolite or actinolite (either in raw form or as part of articles). If EPA finds upon review of the
Significant New Use Notice (SNUN) that the significant new use presents or may present an unreasonable risk (or if there is
insufficient information to permit a reasoned evaluation of the health and environmental effects of the significant new use),
then EPA would take action under TSCA section 5(e) or (f) to the extent necessary to protect against unreasonable risk.
Page 24 of 352

-------
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
endpoints for cancer. EPA used the Framework for Human Health Risk Assessment to Inform Decision
Making (U.S. EPA. 2014a) to evaluate, extract, and integrate asbestos' dose-response information. EPA
evaluated the large database of health effects associated with asbestos exposure cited in numerous U.S.
and international data sources. Many authorities have established that there are causal associations
between asbestos exposures and cancer (NTP. 2016; I ARC. 2012b; ATSDR. 2.001a;	38b;
987; U.S. EPA. 1986; IARC. 19771
EPA evaluated inhalation exposures to chrysotile asbestos in occupational and consumer settings in this
Part 1 of the risk evaluation for asbestos. Dermal exposures were identified as a possible route of
exposure in the PF but were not included in the evaluation since the only reported effects were dermal-
specific lesions, and the major hazard concern is development of cancer via inhalation. Chrysotile
asbestos is a fiber and is not expected to be absorbed into the body through the exterior skin surfaces and
be distributed to the lungs. Furthermore, as also described in the PF, non-cancer hazards from inhalation
exposures were identified for consideration at that time, but risks associated with non-cancer effects
were not quantified. Both the SACC and public comments suggested that EPA consider non-cancer
effects in Part 1 of the risk evaluation for asbestos; however, EPA maintains that the evaluation of
cancer effects and subsequent risk determinations, that consider non-cancer risks, are health protective
for the evaluated COUs for chrysotile asbestos. In Part 2 of the risk evaluation for asbestos that will
examine legacy uses and associated disposals, EPA will consider the reasonably available information
for cancer and non-cancer hazards.
Given the well-established carcinogenicity of asbestos for cancer, EPA, in its PF document, decided to
limit the scope of its systematic review to cancer and to inhalation exposures with the goal of updating,
or reaffirming, the existing 1988 EPA inhalation unit risk (IUR) for general asbestos (
1988b). Therefore, the literature was reviewed to determine whether a new IUR needed to be
developed. The IUR for asbestos developed in 1988 was based on 14 epidemiologic studies that
included occupational exposure to chrysotile, amosite, or mixed-mineral exposures [chrysotile, amosite,
crocidolite]. However, EPA's research to identify COUs indicated that only chrysotile asbestos is
currently being imported in the raw form or imported in products. Therefore, studies of populations
exposed only to chrysotile provide the most informative data for developing the TSCA risk estimates
for the COUs presently considered for chrysotile asbestos, and EPA decided to focus on studies where
the exposure was limited to chrysotile asbestos. EPA will consider legacy uses and associated disposals
for all 6 fiber types including in the AHERA Title II definition in Part 2 of the risk evaluation for
asbestos.
As stated in Section 3.2, epidemiological studies on mesothelioma and lung cancer in cohorts of workers
using chrysotile in commerce were identified that could inform the estimation of an exposure-response
function allowing for the derivation of a chrysotile asbestos IUR. EPA could not find any recent risk
values in the literature for chrysotile asbestos following the derivation of the IRIS IUR value from the
1980s.
Cancer potency values were either extracted from published epidemiology studies or derived from the
data within those studies. Once the cancer potency values were obtained, they were adjusted for
differences in air volumes between workers and other populations so that those values can be applied to
the U.S. population, as a whole, in standard EPA life-table analyses. The life-table methodology allows
the estimation of an exposure concentration associated with a specific extra risk of cancer incidence
caused by chrysotile asbestos. The risk of mesothelioma was adjusted to compensate for
underascertainment of mesothelioma. The risk of lung cancer was adjusted to account for the risk of
Page 25 of 352

-------
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
other established cancer endpoints {i.e., cancers of larynx and ovary)5. According to standard
practice, the lifetime unit risks for lung cancer and mesothelioma were estimated separately and then
statistically combined to yield the cancer inhalation unit risk. Less-than-lifetime or partial lifetime unit
risks were also derived for a range of exposure scenarios based on different ages of first exposure and
different durations of exposure {e.g., 20 years old and 40 years of exposure) (Section 3.2: Human Health
Hazards).
Risk Characterization
Environmental Risk: Based on the reasonably available information in the published literature,
provided by industries using chrysotile asbestos, and reported in EPA databases, there is minimal or no
releases of chrysotile asbestos to surface water associated with the COUs that EPA is evaluating in this
Part 1 of the risk evaluation for asbestos. Thus, EPA believes there is low or no potential for
environmental risk to aquatic or sediment-dwelling receptors from the COUs included in this
document because water releases associated with the COUs were not identified and not expected.
Similarly, EPA expects low or no risk to terrestrial species from water pathways, including biosolids,
as discussed in the problem formulation.
Human Health Risks: EPA identified cancer risks from inhalation exposure to chrysotile asbestos.
For workers and ONUs, EPA estimated cancer risk from inhalation exposures to chrysotile asbestos
using IUR values and exposures for each COU. EPA estimated risks using several occupational
exposure scenarios related to the central and high-end estimates of exposure without the use of
personal protective equipment (PPE), and with potential PPE for workers using chrysotile asbestos.
Industry submissions indicated that some workers used respirators for certain tasks, but not others, and
some workers used ineffective respirators (sheet gasket stampers). Sheet gasket stampers using N95
respirators are not protected as OSHA's Asbestos standards prohibit the use of filtering facepiece
respiratory for protection against asbestos fibers (OSHA asbestos standards do not specifically regulate
N95 respirators). Although hypothetical respirator usage with an applied protection factor (APF) of 10
and 25 was calculated for all COUs, actual respirator use was limited to an APF of 10 (the use of sheet
gaskets) and APFs of 10 and 25, in some cases, for chlor-alkali use. No other APFs were indicated for
any other COU. For asbestos, nominal APFs {e.g., 25) may not be achieved for all PPE users. More
information on respiratory protection, including EPA's approach regarding the occupational exposure
scenarios for asbestos, is in Section 2.3.1.2.
For workers, cancer risks in excess of the benchmark of 1 death per 10,000 (or 1 x 10"4) were indicated
for virtually all quantitatively assessed COUs (except the Super Guppy scenario) under high-end and
central tendency exposure scenarios when PPE was not used. Risks were below the benchmark for
chlor-alkali workers (full-shift, central tendency exposure estimate only) and the specialized brake pad
work for the NASA Super Guppy aircraft (both for central and high-end exposure estimates). With the
hypothetical use of PPE at APF of 10 (except for chlor-alkali processing and use [short-term6] and sheet
gasket use), most risks were reduced for central tendency estimates but still persisted for sheet gasket
stamping, auto brake replacement, other vehicle friction products and utility vehicle (UTV use and
disposal) gasket replacement for high-end exposure estimates (both 8-hour and short-term durations).
Although not expected to be worn given the reasonably available information, when PPE with an APF
5	The methodology involved in risk characterization has evolved over time and the existing EPA IURs for other asbestos fiber
types U.S.	114b. 1986") estimated risks of cancer mortality and did not account for the risk of other cancers, and the
1986 IUR did not adjust for mesothelioma underascertainment.
6	Short-term means accounting for higher exposures during short periods of time during the work shift. See Section 2.3.1 for
more information.
Page 26 of 352

-------
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
of 25 was applied, risk was still indicated only for the high-end, short term exposure scenario for the
auto brakes and other vehicle friction products. EPA's estimates for worker risks for each occupational
scenario are presented by each COU in Section 4.2.2 and summarized in Table 4-38.
For ONUs, cancer risks in excess of the benchmark of 1 death per 10,000 (or 1 x 10"4) were indicated for
both central tendency and high-end exposures for sheet gasket use (in chemical production) and UTV
gasket replacement. In addition, cancer risks for ONUs were indicated for high-end exposures only for
chlor-alkali, sheet gasket stamping. ONUs are not assumed to be using PPE to reduce exposures to
chrysotile asbestos used in their vicinity. EPA's estimates for ONU risks for each occupational exposure
scenario are presented by each COU in Section 4.2.2 and summarized in Table 4-38.
For consumers (Do-it-Yourselfers, or DIY) and bystanders of consumer use, EPA estimated cancer
risks resulting from inhalation exposures with a range of user durations, described in detail in Section
4.2.3. EPA assumed that consumers or bystanders would not use PPE.
For consumers and bystanders, cancer risks in excess of the benchmark of 1 death per 1,000,000 (or 1
x 10"6) were indicated for most COUs for consumer exposure scenarios. Risks were indicated for all
high-end exposures for both consumers and bystanders for brake and UTV gasket indoor scenarios;
and the high-end consumer outdoor scenarios (for 30-minute exposures). EPA's estimates for
consumer and bystander risks for each consumer use exposure scenario are presented in Section 4.2.3
and summarized in Table 4-48.
Uncertainties: Uncertainties have been identified and discussed after each section in this Part 1 of the
risk evaluation for asbestos. In addition, Section 4.3 summarizes the major assumptions and key
uncertainties by major topic: uses of asbestos, occupational exposure, consumer exposure,
environmental risk, IUR derivation, cancer risk value and human health risk estimates.
Beginning with the February, 2017 request for information on uses of asbestos (see z tblic
Meeting) and followed by the Scope document (June 2017d), Problem Formulation (June 2018d), and
draft Risk Evaluation (2020). EPA has refined its understanding of the current conditions of use of
chrysotile asbestos in the U.S. Chrysotile asbestos was the only fiber type imported, processed, or
distributed in commerce for use in 2019 (from the latest import records). All the raw asbestos imported
into the U.S. is used by the chlor-alkali industry for use in asbestos diaphragms. The remaining COUs
involve use and disposal of articles that contain chrysotile asbestos. EPA received voluntary
acknowledgement of these uses/disposals from a handful of industries that fall under these COU
categories.
By finalizing the asbestos SNUR on April 25, 2019 to include manufacturing (including import) or
processing of discontinued uses not already banned under TSCA, EPA is highly certain that
manufacturing (including import), processing, or distribution of asbestos is not intended, known or
reasonably foreseen for uses beyond use in the six product categories in this Part 1 of the risk
evaluation for asbestos. EPA will consider legacy uses and associated disposals of asbestos in Part 2 of
the risk evaluation for asbestos.
For occupational exposures, the number of chlor-alkali plants in the U.S. is known and therefore the
number of workers potentially exposed from chlor-alkali activities is reasonably certain. The number of
workers potentially exposed for other COUs is less certain. Only two workers were identified for
stamping sheet gaskets, and two Ti02 manufacturing facilities were identified in the U.S. that use
chrysotile asbestos-containing gaskets. However, EPA is not certain if chrysotile asbestos-containing
sheet gaskets are used in other industries and to what extent. For the other COUs, no estimates of the
Page 27 of 352

-------
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
number of potentially exposed workers were submitted to EPA by industry or its representatives, so
estimates were used and were based on market estimates for that work category. However; no
information on the market share for asbestos containing products, with the exception of aftermarket
automotive brakes/linings, is reasonably available. Based on peer review and public comments received
on the draft Risk Evaluation, EPA adjusted its estimates for the number of potentially affected
individuals who may purchase and use chrysotile asbestos aftermarket automotive brakes/linings (see
Section 4.3.7). Therefore, numbers of workers potentially exposed were estimated and, based on the
COU, these estimations have a range of uncertainty from low (chlor-alkali) to high (sheet gasket use,
oilfield brake blocks, aftermarket automotive brakes/linings, other vehicle friction products and other
gaskets).
Exposures for ONUs can vary substantially. Most data sources do not sufficiently describe the proximity
of these employees to the exposure source. As such, exposure levels for the ONU category will vary
depending on the work activity. It is unknown whether these uncertainties overestimate or underestimate
exposures.
A review of reasonably available literature for consumer exposure estimates related to brake
repair/replacement activities by a DIY consumer was limited and no information for consumer exposure
estimates related to UTV exhaust system gasket repair/replacement activities was found. This absence of
scenario-specific exposure information required EPA to use surrogate monitoring data from
occupational studies to evaluate consumer risk resulting from exposure to asbestos during these two
activities. The surrogate occupational studies tended to be based on older studies that may or may not
reflect current DIY consumer activities, including best practices for removing asbestos containing
materials. In addition, EPA is uncertain about the number of chrysotile asbestos containing brakes that
are being purchased online and installed in cars (classic cars or newer cars) or gaskets that are being
replaced in UTVs.
After the PF was released, EPA continued to search EPA databases and all publicly available literature
and contact industries to shed light on potential releases to water from the COUs in this Part 1 of the risk
evaluation for asbestos for the purpose of evaluating risk to aquatic or sediment-dwelling organisms.
EPA found minimal or no releases of asbestos to surface water associated with the COUs in this risk
evaluation. In addition, there are no reported releases of asbestos to water from TRI. Despite the fact that
the comprehensive efforts put forth have not identified releases of chrysotile asbestos into water from
COUs, EPA acknowledges that uncertainty remains.
Epidemiologic studies are observational and as such are potentially subject to confounding and selection
biases. Most of the studies of asbestos exposed workers did not have information to control for cigarette
smoking, which is an important risk factor for lung cancer in the general population. However, the bias
related to this failure to control for smoking is believed to be small. It is unlikely that smoking rates
among workers in the chosen epidemiology studies differed substantially enough with respect to their
cumulative chrysotile exposures to induce important confounding in risk estimates for lung cancer (see
Section 4.3.7). Mesothelioma is not related to smoking and thus smoking could not be a confounder for
mesothelioma.
Depending on the variations in the exposure profile of the workers/occupational non-users and
consumers/bystanders, risks could be under- or over-estimated for all COUs. The estimates for extra
cancer risk were based on the EPA-derived IUR for chrysotile asbestos. The occupational exposure
assessment made standard assumptions of 240 days per year, 8 hours per day over 40 years starting at
Page 28 of 352

-------
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
age 16 years. 7 This assumes the workers and ONUs are regularly exposed until age 56. If a worker
changes jobs during their career and is no longer exposed to chrysotile asbestos, this may overestimate
exposures. However, if the worker stays employed after age 56, it would underestimate exposures.
EPA's assessments, risk estimations, and risk determinations accounted for uncertainties throughout this
Part 1 of the risk evaluation for asbestos. EPA used reasonably available information, in a fit-for-
purpose approach, to develop a document that relies on the best available science and is based on the
weight of the scientific evidence. For instance, systematic review was conducted to identify reasonably
available information related to chrysotile asbestos hazards and exposures. The consideration of
uncertainties supports the Agency's risk determinations, each of which is supported by substantial
evidence, as set forth in detail in later sections of this Part 1 of the risk evaluation for asbestos.
Potentially Exposed Susceptible Subpopulations (PESS): TSCA ง 6(b)(4) requires that EPA conduct a
risk evaluation to "determine whether a chemical substance presents an unreasonable risk of injury to
health or the environment, without consideration of cost or other non-risk factors, including an
unreasonable risk to a potentially exposed or susceptible subpopulation identified as relevant to the risk
evaluation by the Administrator, under the conditions of use. " TSCA ง 3(12) states that "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, pregnant women, workers, or the elderly. "
EPA identified certain human subpopulations who may be more susceptible to exposure to chrysotile
asbestos than others. Workers exposed to chrysotile asbestos in workplace air, especially if they work
directly with chrysotile asbestos, are most susceptible to the health effects associated with chrysotile
asbestos. Although it is clear that the health risks from chrysotile asbestos exposure increases with
higher exposure and longer exposure time, investigators have found asbestos-related diseases in
individuals with only brief exposures. Generally, those who develop asbestos-related diseases could
show no signs of illness for decades after exposure.
A source of variability in susceptibility between people is smoking history or the degree of exposure to
other risk factors with which asbestos interacts. In addition, the long-term retention of asbestos fibers in
the lung and the long latency period for the onset of asbestos-related respiratory diseases suggest that
individuals exposed earlier in life may be at greater risk to the eventual development of respiratory
problems than those exposed later in life (ATSDR. 2001a). There is also some evidence of genetic
predisposition for mesothelioma related to having a germ line mutation in BAP1 (Testa et al... 2011).
Cancer risks were indicated for all the worker COUs and most of the consumer/bystander COUs. In
addition, several subpopulations (e.g., smokers, genetically predisposed individuals, workers who
change their own asbestos-containing brakes) may be more susceptible than others to health effects
resulting from exposure to asbestos. These subpopulations are discussed in more detail for potentially
exposed or susceptible subpopulations and aggregate exposures in Section 4.4 and Section 4.5.
EPA based its risk determinations on the high-end exposure estimates for workers, consumers, and
bystanders in order to capture individuals who may be PESS.
Aggregate and Sentinel Exposures: Section 6(b)(4)(F)(ii) of TSCA requires the EPA, as a part of the risk
evaluation, to describe whether aggregate or sentinel exposures under the conditions of use were
7 The Fair Labor Standards Act of 1938 allows adolescents to work an unrestricted number of hours at age 16 years.
Page 29 of 352

-------
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
considered and the basis for their consideration. The EPA has defined aggregate exposure as "the
combined exposures to an individual from a single chemical substance across multiple routes and
across multiple pathways (40 CFR ง 702.33)." Exposures to chrysotile asbestos were evaluated by the
inhalation route only. Inhalation, dermal, and oral exposures could occur simultaneously for workers and
consumers. EPA chose not to employ simple additivity of exposure pathways within a COU since the
most critical exposure pathway is inhalation and the target being assessed is combined lung cancer and
mesothelioma. Furthermore, EPA recognizes it is possible that workers exposed to chrysotile asbestos at
work might also be exposed as consumers (e.g., by changing asbestos-containing brakes at home) or
may cause unintentional exposure to individuals in their residence due to take-home exposure from
contaminated clothing or other items. While adding such exposures could increase risks to the worker,
ONU, consumer, or bystander, which already individually exceed the cancer benchmarks in virtually
every scenario evaluated, these additional pathways are not evaluated together because EPA did not
identify or receive information which could inform developing such an exposure scenario and does not
have models which could adequately evaluate and address such combined scenarios.
The EPA defines sentinel exposure as "the exposure to 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 ง 702.33)." In this Part 1 of the risk evaluation for asbestos, the EPA
considered sentinel exposure the highest exposure given the details of the COU and the potential
exposure scenarios. EPA considered sentinel exposures by considering risks to populations who may
have upper bound (e.g., high-end) exposures. EPA's decisions for unreasonable risk are based on high-
end exposure estimates to capture individuals who may receive sentinel exposure.
Risk Determination
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. In making this determination, 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 COU; the population exposed (including any potentially exposed or susceptible
subpopulations); the severity of hazard (including the nature of the hazard, the irreversibility of the
hazard); and uncertainties. EPA also takes into consideration the Agency's confidence in the data used
in the risk estimate. This includes an evaluation of the strengths, limitations, and uncertainties associated
with the information used to inform the risk estimate and the risk characterization. The rationale for the
risk determination is discussed in Section 5.2.
Environmental Risk: As described in the problem formulation (U.S. EPA. 2018d). EPA found that
exposures to terrestrial species may occur from the conditions of use due to releases to air, water or land.
During the course of developing the draft risk evaluation for asbestos, OPPT worked closely with the
offices within EPA that administer and implement regulatory programs under the Clean Air Act (CAA),
the Safe Drinking Water Act (SDWA), and the Clean Water Act (CWA. Through this intra-agency
coordination, EPA determined that exposures to terrestrial species via surface water, ambient air and
disposal pathways fall under the jurisdiction of other environmental statutes administered by EPA, i.e.,
CAA, SDWA, and the CWA. As explained in more detail in Section 1.4.2, 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
Page 30 of 352

-------
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
consistent with 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 deadline for completing risk
evaluations. EPA has therefore tailored the scope of this Part 1 of the risk evaluation for asbestos using
authorities in TSCA sections 6(b) and 9(b)(1). EPA did not evaluate hazards or exposures from
chrysotile asbestos releases to terrestrial pathways for terrestrial organisms, and as such the
unreasonable risk determinations for relevant conditions of use do not account for exposures to
terrestrial organisms.
After the PF was released, EPA continued to search EPA databases as well as the literature and
contacted industries to shed light on potential releases of chrysotile asbestos to water from the TSCA
COUs. Based on the reasonably available information in the published literature, provided by industries
using asbestos, and reported in EPA databases, there is minimal or no releases of chrysotile asbestos to
surface water associated with the COUs in this Part 1 of the risk evaluation for asbestos. EPA has
considered peer review and public comments on this conclusion and EPA is confident that the minimal
water release data available cannot be attributed to chrysotile asbestos from the COUs in this document.
Therefore, EPA concludes there is no unreasonable risk to aquatic organisms (including sediment-
dwelling organisms) from the COUs in this Part 1 of the risk evaluation for asbestos. Details are
provided in Section 4.1.
Risk of Injury to Health: EPA's determination of unreasonable risk for specific COUs of chrysotile
asbestos listed below are based on health risks to workers, occupational non-users, consumers, or
bystanders from consumer use. The health effect driver for EPA's determination of unreasonable risk is
cancer from inhalation exposure. As described below, risks to the general population were not evaluated
for these conditions of use.
There are physical-chemical characteristics that are unique to asbestos, such as insolubility in water,
opportunity for suspension and extended duration in air, transportability and the friable nature of
asbestos-containing products. These attributes allow asbestos fibers to be released, settled, and to again
become airborne ("re-entrainment") under certain conditions of use. Also unique to asbestos is the
impact of the timing of exposure relative to the cancer outcome; the most relevant exposures for
understanding cancer risk were those that occurred decades prior to the onset of cancer. In addition to
the cancer benchmark, the physical-chemical properties and exposure considerations are important
factors in considering risk of injury to health. To account for the exposures for ONUs and, in certain
cases bystanders, EPA derived a distribution of exposure values for calculating the risk for cancer by
using area monitoring data (i.e., fixed location air monitoring results) where available for certain
conditions of use and when appropriate applied exposure reduction factors, using data from published
literature (see Sections 2.3.1 and 2.3.2 for details on ONU and bystander methods, respectively). The
risk determination for each COU in this Part 1 of the risk evaluation for asbestos considers both central
tendency and high-end risk estimates for workers, ONUs, consumers and bystanders. Where relevant
EPA considered PPE for workers. For many of the COUs both the central tendency and high-end risk
estimates exceed the risk benchmark for each of the exposed populations evaluated. However, the risk
benchmarks do not serve as a bright line for making risk determinations and other relevant risk-related
factors were considered. EPA focused on the high-end risk estimates rather than central tendency risk
estimates to be protective of workers, ONUs, consumers, and bystanders. Additionally, EPA's
confidence in the data used in the risk estimate is considered.
Risk to the General Population: As part of the problem formulation for asbestos, EPA found that
exposures to the general population may occur from the conditions of use due to releases to air, water or
land. During the course of developing the draft risk evaluation for asbestos, OPPT worked closely with
Page 31 of 352

-------
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
the offices within EPA that administer and implement regulatory programs under the Clean Air Act
(CAA), the Safe Drinking Water Act (SDWA), and the Clean Water Act (CWA). Through this intra-
agency coordination, EPA determined that exposures to the general population via surface water,
drinking water, ambient air and disposal pathways falls under the jurisdiction of other environmental
statutes administered by EPA, (i.e., CAA, SDWA, and the CWA). As explained in more detail in
Section 1.4.2, 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 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
deadline for completing risk evaluations. EPA has therefore tailored the scope of this Part 1 of the risk
evaluation for asbestos using authorities in TSCA sections 6(b) and 9(b)(1). Therefore, EPA did not
evaluate hazards or exposures to the general population in this document, and as such the unreasonable
risk determinations for the relevant conditions of use do not account for exposures to the general
population.
Risk to Workers: The conditions of use of asbestos that present an unreasonable risk to workers include
processing and industrial use of chrysotile asbestos-containing diaphragms, processing and industrial use
of chrysotile asbestos-containing sheet gaskets and industrial use of chrysotile asbestos-containing brake
blocks, aftermarket automotive chrysotile asbestos-containing brakes/linings, other vehicle friction
products, and other chrysotile asbestos-containing gaskets. A full description of EPA's determination for
each condition of use is in Section 5.2.
Risk to Occupational Non-Users (ONUs): EPA determined that the conditions of use that present
unreasonable risks for ONUs include processing and industrial use of chrysotile asbestos-containing
diaphragms, processing and industrial use of chrysotile asbestos-containing sheet gaskets and industrial
use of chrysotile asbestos-containing brake blocks, other vehicle friction products, and other chrysotile
asbestos-containing gaskets. A full description of EPA's determination for each condition of use is in
Section 5.2.
EPA generally assumes compliance with OSHA requirements for protection of workers, including the
implementation of the hierarchy of controls. In support of this assumption, EPA used reasonably
available information indicating that some employers, particularly in the industrial setting, are providing
appropriate engineering, or administrative controls, or PPE to their employees consistent with OSHA
requirements. While EPA does not have reasonably available information to either support or contradict
this assumption for each condition of use, EPA does not believe that the Agency must presume, in the
absence of such information, a lack of compliance with existing regulatory programs and practices.
Rather, EPA assumes there is compliance with worker protection standards unless case-specific facts
indicate otherwise, and therefore existing OSHA regulations for worker protection and hazard
communication will result in use of appropriate PPE in a manner that achieves the stated APF or PF.
EPA's decisions for unreasonable risk to workers are based on high-end exposure estimates, in order to
account for the uncertainties related to whether or not workers are using PPE. EPA believes this is a
reasonable and appropriate approach that accounts for reasonably available information and professional
judgement related to worker protection practices, and addresses uncertainties regarding availability and
use of PPE.
Risk to Consumers: For consumers, EPA determined that the conditions of use that present an
unreasonable risk are use of aftermarket automotive chrysotile asbestos-containing brakes/linings and
Page 32 of 352

-------
464	other chrysotile asbestos-containing gaskets. A full description of EPA's determination for each
465	condition of use is in Section 5.2.
466	Risk to Bystanders (from consumer uses): EPA determined that the conditions of use that present an
467	unreasonable risk to bystanders are use of aftermarket automotive chrysotile asbestos-containing
468	brakes/linings and other chrysotile asbestos-containing gaskets. A full description of EPA's
469	determination for each condition of use is in Section 5.2.
470	Summary of Risk Determinations for the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos: In
471	conducting risk evaluations, "EPA will determine whether the chemical substance presents an
472	unreasonable risk of injury to health or the environment under each condition of use [] within the scope
473	of the risk evaluation, either in a single decision document or in multiple decision documents..." 40
474	CFR 702.47. Under EPA's implementing regulations, "[a] determination by EPA that the chemical
475	substance, under one or more of the conditions of use within the scope of the risk evaluation, does not
476	present an unreasonable risk of injury to health or the environment will be issued by order and
477	considered to be a final Agency action, effective on the date of issuance of the order." 40 CFR
478	702.49(d).
479	EPA has determined that there are no conditions of use presenting an unreasonable risk to environmental
480	receptors (see details in Section 5.1).
481	EPA has determined that the following conditions of use of chrysotile asbestos present an unreasonable
482	risk of injury to health to workers (including, in some cases, occupational non-users) or to consumers
483	(including, in some cases, bystanders).
484	Pursuant to TSCA section 6(i)(2), the unreasonable risk determinations for these conditions of use are
485	not considered final agency action. EPA will initiate TSCA section 6(a) risk management actions on
486	these conditions of use as required under TSCA section 6(c)(1).8 The details of these determinations are
487	presented in Section 5.2.
Omi|):i(ion;il Conditions of I so 1 hill Present nil I nreiisonsihle Risk
•	Processing and Industrial use of Chrysotile Asbestos Diaphragms in the Chlor-alkali Industry
•	Processing and Industrial Use of Chrysotile Asbestos-Containing Sheet Gaskets in Chemical
Production
•	Industrial Use and Disposal of Chrysotile Asbestos-Containing Brake Blocks in Oil Industry
•	Commercial Use and Disposal of Aftermarket Automotive Chrysotile Asbestos-Containing
Brakes/Linings
•	Commercial Use and Disposal of Other Chrysotile Asbestos-Containing Vehicle Friction
Products
•	Commercial Use and Disposal of Other Asbestos-Containing Gaskets
488
8 Although EPA has identified both industrial and commercial uses here for purposes of distinguishing scenarios in this
analysis, the Agency interprets the authority over "any manner or method of commercial use" under TSCA section 6(a)(5) to
reach both.
Page 33 of 352

-------
C onsumer I ses nnd Disposal tlisit Present sin I nresisonsihle Risk
•	Aftermarket Automotive Chrysotile Asbestos-Containing Brakes/Linings
•	Other Chrysotile Asbestos-Containing Gaskets
489	EPA has determined that the following conditions of use of chrysotile asbestos do not present an
490	unreasonable risk of injury to health. These determinations are considered final agency action and are
491	being issued by order pursuant to TSCA section 6(i)(l), and the TSCA section 6(i)(l) order is contained
492	in Section 5.3.1 of Part 1 of the risk evaluation for asbestos. The details of these determinations are
493	presented in Section 5.2.
494
Conditions of I se tlisit Do Not Present sin I nre:is<>ii;ihie Kisk
•	Import of chrysotile asbestos and chrysotile asbestos-containing products
•	Distribution of chrysotile asbestos-containing products
•	Use of chrysotile asbestos-containing brakes for a specialized, large NASA transport plane
•	Disposal of chrysotile asbestos-containing sheet gaskets processed and/or used in the industrial
setting and asbestos-containing brakes for a specialized, large NASA transport plane
495
Page 34 of 352

-------
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
1 INTRODUCTION
This document presents Part 1 (chrysotile asbestos) of the risk evaluation of asbestos9 under the Frank
R. Lautenberg Chemical Safety for the 21st Century Act which amended the Toxic Substances Control
Act (TSCA), the Nation's primary chemicals management law, in June 2016.
The Agency published the Scope of the Risk Evaluation for Asbestos (U.S. EPA. 2017d) in June 2017,
and the Problem Formulation in June 2018	2018d), which represented the analytical phase of
risk evaluation in which "the purpose for the assessment is articulated, the problem is defined, and a
plan for analyzing and characterizing risk is determined" as described in Section 2.2 of the Framework
for Human Health Risk Assessment to Inform Decision Making. EPA received comments on the
published Problem Formulation for asbestos and has considered the comments specific to asbestos, as
well as more general comments regarding EPA's Risk Evaluation approach for developing the Risk
Evaluations for the first 10 chemicals EPA is evaluating; including this Part 1 of the risk evaluation for
asbestos.
In the problem formulatior	)18d\ EPA identified the conditions of use (COUs) and
presented three conceptual models and an analysis plan. Based on EPA's analysis of the COUs,
physical-chemical and fate properties, environmental releases and exposure pathways, the problem
formulation preliminarily concluded that further analysis was necessary for exposure pathways to
workers (including ONUs), consumers (including bystanders), and surface water, based on a qualitative
assessment of the physical-chemical properties and fate of asbestos in the environment. After the
problem formulation was released, there were two major developments that warranted changes/updates
prior to release of the draft Risk Evaluation. First, EPA continued to search EPA databases as well as the
literature and either engaged in a dialogue with industries or reached out for a dialogue to shed light on
potential releases to water for chrysotile asbestos. It was concluded there were no water releases for the
COUs associated with chrysotile asbestos based on the collected information. Second, a new COU was
discovered within the vehicle friction products category {i.e., the use of brakes/friction products in a
large aircraft operated by NASA). Both of these were included in the subsequently published draft Risk
Evaluation for asbestos for which EPA has taken public and peer review comments.
As per EPA's final rule, Procedures for Chemical Risk Evaluation Under the Amended Toxic
Substances Control Act (82 Fed. Reg. 33726 (July 20, 2017)), the draft Risk Evaluation for asbestos was
subject to both public comment and peer review; which are distinct but related processes. EPA provided
60 days for public comment on any and all aspects of the draft Risk Evaluation, including the
submission of any additional information that might be relevant to the science underlying the document
and the outcome of the systematic review associated with chrysotile asbestos. This satisfies TSCA (15
U.S.C. 2605(b)(4)(H)), which requires EPA to provide public notice and an opportunity for comment on
a draft Risk Evaluation prior to publishing a final Risk Evaluation.
Peer review was conducted in accordance with EPA's regulatory procedures for chemical Risk
Evaluations, including using the EPA Peer Review Handbook (U.S. EPA. 2015b) and other methods
consistent with the science standards laid out in Section 26 of TSCA (See 40 CFR 702.45). As explained
in the Risk Evaluation Rule (I. j J \ _J0Jja), the purpose of peer review is for the independent review
of the science underlying the risk assessment. Peer review addressed aspects of the underlying science as
9 As noted in the PREAMBLE, this is Part 1 of the final Risk Evaluation for asbestos. Part 1 includes the imported,
processed, and distributed uses of chrysotile asbestos in the United States. Part 2 will be on legacy uses and associated
disposals of asbestos. Please see Figure P-2 for definitions and terms used throughout this document and note that
occasionally the term "asbestos" is used (depending on context), but the focus of this Part ldocument is chrysotile asbestos.
Page 35 of 352

-------
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
outlined in the charge to the peer review panel such as hazard assessment, assessment of dose-response,
exposure assessment, and risk characterization.
As explained in the Risk 'Evaluation Rule (82 Fed. Reg. 33726 (July 20, 2017)), it is important for peer
reviewers to consider how the underlying risk evaluation analyses fit together to produce an integrated
risk characterization, which forms the basis of an unreasonable risk determination. EPA believed peer
reviewers were most effective in this role if they received the benefit of public comments on draft risk
evaluations prior to peer review. For this reason, and consistent with standard Agency practice, the
public comment period preceded peer review. In response to public comments received on the draft Risk
Evaluation and/or in response to peer review, the overall approach to finalizing the Risk Evaluation for
asbestos changed as described in the Preamble by dividing the Risk Evaluation into two parts.
Furthermore, EPA responded to public and peer review comments received on the draft Risk Evaluation
in the response to comments document and, where appropriate, made revisions in response to those
comments in Part 1 of the risk evaluation.
The conclusions, findings, and determinations in Part 1 of the risk evaluation are for the purpose of
identifying whether exposure to chrysotile asbestos presents unreasonable risk or no unreasonable risk
under the conditions of use, in accordance with TSCA Section 6, and are not intended to represent any
findings under TSCA section 7.
Asbestos has been regulated by various Offices of EPA for years. The Risk Evaluation for asbestos has
posed some unique challenges to OPPT. Unlike the other nine chemicals that are part of the "First 10"
Risk Evaluations under the Lautenberg Act of 2016, asbestos is a naturally occurring fiber, which poses
its own set of issues, including defining: (1) the COU (by asbestos fiber type); (2) the appropriate
inhalation unit risk (IUR) value to use for the hazard/dose-response process; and (3) the appropriate
exposure assessment measures.
The COUs in this Part 1 of the risk evaluation for asbestos are limited to only a few categories of
ongoing and prospective uses, and chrysotile is the only type of asbestos fiber identified for these COUs.
Ongoing uses of asbestos in the U.S. were difficult to identify despite using an extensive list of
resources. To determine the COUs of asbestos and inversely, activities that do not qualify as COUs,
EPA conducted extensive research and outreach. EPA identified activities that include import of raw
chrysotile asbestos, used solely in the chlor-alkali industry, and import and use of chrysotile asbestos-
containing products. The COUs included in this Part 1 of the risk evaluation for asbestos that EPA
considers to be known, intended, or reasonably foreseen are the import, use, distribution and disposal of
chrysotile asbestos diaphragms, sheet gaskets, other gaskets, oilfield brake blocks, aftermarket
automotive brakes/linings, and other vehicle friction products and the processing, distribution and
disposal of chrysotile asbestos diaphragms and sheet gaskets. Some of these COUs are very specialized.
Since the Problem Formulation, several conditions of use were removed from the scope of the draft Risk
Evaluation based on further investigation (see Section 1.4.4); these COUs pertain to woven products,
cement products, and packings (from "gaskets and packings"). EPA determined that there is no evidence
to indicate manufacture (including import), processing, or distribution of asbestos-containing woven
products, cement products, or packings. These conditions of use were added to the Significant New Use
Rule (SNUR) for asbestos (40 CFR 721.11095). The Asbestos SNUR is an Agency action
complementary to the Risk Evaluation for asbestos and taken under TSCA section 5 to prohibit any
manufacturing (including import) or processing for discontinued uses of asbestos from restarting without
EPA having an opportunity to evaluate them to determine risks to health or the environment and take
any necessary regulatory action, which may include a prohibition. The final asbestos SNUR ensures that
any manufacturing (including import) and processing for all discontinued uses and types of asbestos that
are not already banned are restricted from re-entering the U.S. marketplace without notification to EPA
Page 36 of 352

-------
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
and review and any necessary regulatory action by the Agency. Thus, should any person wish to
manufacture, import, or process asbestos for an activity that is not a COU identified in this document or
subject to an existing ban, then EPA would review the risk of the activity associated with such a use in
accordance with TSCA section 510.
During the investigation of the COUs, EPA also determined that asbestos is no longer mined in the U.S.,
and that only chrysotile asbestos is being imported. The other five forms of asbestos identified for the
Risk Evaluation, including crocidolite (riebeckite), amosite (cummingtonite-grunerite), anthophyllite,
tremolite or actinolite, are no longer manufactured, imported, or processed in the United States and are
also now subject to the SNUR. After EPA confirmed that chrysotile asbestos is the only type of asbestos
still being imported into the U.S. either in raw form or in products, EPA developed a chrysotile IUR11 to
be used in Part 1 of the risk evaluation. The IUR for asbestos developed in 1988 was based on 14
epidemiologic studies that included occupational exposure to chrysotile, amosite, or mixed-mineral
exposures (chrysotile, amosite, crocidolite). As a naturally occurring mineral, chrysotile asbestos can co-
occur with other minerals, including amphibole forms of asbestos. Trace amounts of these minerals may
remain in chrysotile asbestos as it is used in commerce. The epidemiologic studies that are reasonably
available include populations exposed to chrysotile asbestos, which may contain small, but variable
amounts of amphibole asbestos. Because the only form of asbestos imported, processed, or distributed
for use in the United States today is chrysotile asbestos, studies of populations exposed only to
chrysotile asbestos provide the most informative data for the purpose of updating the TSCA risk
estimates for the COUs for chrysotile asbestos in this document. EPA will consider legacy uses and
associated disposals of asbestos in Part 2 of the risk evaluation for asbestos (as noted in the Preamble).
EPA stated in the Problem Formulation that the asbestos Risk Evaluation would focus on
epidemiological data on lung cancer and mesothelioma. The 1988 IUR identified asbestos as a
carcinogen causing both lung cancer and mesothelioma from inhalation exposures and derived a unit
risk to address both cancers and cover all asbestos fiber types (for all TSCA Title II fiber types - see
Section 1.1). Over 24,000 studies were initially identified for consideration during the Systematic
Review process to determine whether the existing IRIS 1988 IUR was appropriate for TSCA purposes.
Once EPA determined that only conditions of use of chrysotile asbestos were going to be evaluated for
the draft risk evaluation, the focus turned to whether a chrysotile-specific IUR could be derived. EPA is
not aware of any other chrysotile asbestos-specific IUR or any other risk-based values having been
estimated for other types of cancer for asbestos fiber types by either EPA or other government agencies.
For the derivation of a chrysotile asbestos IUR, epidemiological studies on mesothelioma and lung
cancer in cohorts of workers using chrysotile asbestos in commerce were identified to inform the
estimation of an exposure-response function.
Related to the focus on chrysotile asbestos is the method of identifying asbestos in studies used to
develop the IUR. The IUR is based on fiber counts made by phase contrast microscopy (PCM) and
should not be applied directly to measurements made by other analytical techniques. PCM
measurements made in occupational environments were used in the studies that support the derivation of
the chrysotile asbestos IUR. PCM detects only fibers longer than 5 |im and >0.4 |im in diameter, while
transmission electron microscopy (TEM), often found in environmental monitoring measurements, can
detect much smaller fibers. In developing a PCM-based IUR in this Part 1 of the risk evaluation for
10	As of December 2020, EPA has not received any SNUNs for asbestos.
11	Inhalation Unit risk (IUR) is typically defined as a plausible upper bound on the estimate of cancer risk per |ig/nv' air
breathed for 70 years. For asbestos, the IUR is expressed as cancer risk per fibers/cc (in units of the fibers as measured by
PCM).
Page 37 of 352

-------
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
asbestos, several TEM papers modeling risk of lung cancer were found, but because there was no TEM-
based modeling of mesothelioma risk, TEM data could not be used to derive a TEM-based IUR.
EPA derived an IUR for chrysotile asbestos using five epidemiological study cohorts analyzing lung
cancer and mesothelioma. EPA derived cancer-specific unit risks using lifetables. Different modeling
choices and combinations of cancer-specific unit risks yielded candidate IUR values ranging from 0.15
to 0.45 per f/cc, indicating low model-based uncertainty. The IUR chosen is 0.16 per f/cc and it was
applied to the COUs to calculate lifetime risks for workers and consumers.
EPA estimated risks for workers, occupational non-users (ONUs), consumers (do-it-yourself [DIY]
mechanics) and bystanders for the COUs identified. Inhalation exposure scenarios were used to estimate
risks for cancer based on the EPA-derived IUR for chrysotile asbestos. This assessment is unique with
respect to the timing of exposure relative to the cancer outcome as the time since first exposure plays a
dominant role in modeling risk. Occupational exposures assumed 240 days/year for 8-hour workdays for
40 years starting at 16 years old; with other starting ages and exposure durations also presented.
Occupational exposures for chlor-alkali and sheet gasket workers and ONUs were based on monitoring
data supplied by companies performing the work. Consumer exposures were based on study data
provided in the literature for gasket replacement and brakes. Consumer exposures assumed that DIY
mechanics for both COUs changed brakes or gaskets once every three years (the task taking three hours)
over a lifetime and that exposures lingered between the episodic exposures.
Section 1 presents the basic physical-chemical characteristics of chrysotile asbestos, as well as a
background on regulatory history, COUs, and conceptual models, with particular emphasis on any
changes since the publication of the draft Risk Evaluation. Section 1 also includes a discussion of the
systematic review process utilized in this document. Section 2 provides a discussion and analysis of the
exposures, both health and environmental, that can be expected based on the COUs for chrysotile
asbestos. Section 3 discusses the environmental and health hazards of chrysotile asbestos. Section 4
presents the risk characterization, where EPA integrates and assesses reasonably available information
on health and environmental hazards and exposures, as required by TSCA (15 U.S.C. 2605(b)(4)(F)).
This section also includes a discussion of any uncertainties and how they impact this document. Section
5 presents the risk determination of whether risks posed by the chemical substance under the COUs are
"unreasonable" as required under TSCA (15 U.S.C. 2605(b)(4)).
1.1 Physical and Chemical Properties and Environmental Fate
Asbestos is a "generic commercial designation for a group of naturally occurring mineral silicate fibers
of the serpentine and amphibole series" (I ARC. 2012b). The Chemical Abstracts Service (CAS)
definition of asbestos is "a grayish, non-combustible fibrous material. It consists primarily of impure
magnesium silicate minerals." The general CAS Registry Number (CASRN) of asbestos is 1332-21-4;
this is the only asbestos CASRN on the TSCA Inventory. However, other CASRNs are available for
specific fiber types.
TSCA Title II (added to TSCA in 1986), Section 202 defines asbestos as the "asbestiform varieties of
six fiber types - chrysotile (serpentine), crocidolite (riebeckite), amosite (cummingtonite-grunerite),
anthophyllite, tremolite or actinolite." The latter five fiber types are amphibole varieties. In the Problem
Formulation of the Risk Evaluation for Asbestos (EPA-HQ-OPPT-2016-0736-0131) (U.S. EPA. 2018d).
physical and chemical properties of all six fiber types were presented. As discussed in more detail in
Section 1.4, the risk evaluation for asbestos Part 1 has focused on chrysotile asbestos given EPA's
Page 38 of 352

-------
665
666
667
668
669
670
671
672
673
674
675
knowledge of the COUs of chrysotile asbestos, and EPA will consider legacy uses and associated
disposals of asbestos in Part 2 of the final Risk Evaluation for asbestos (see Preamble).
Table 1-1. shows the physical and chemical properties for chrysotile asbestos, a hydrated magnesium
silicate mineral, with relatively long and flexible crystalline fibers that are capable of being woven.
Chrysotile asbestos fibers used in most commercial applications consist of aggregates and usually
contain a broad distribution of fiber lengths. Chrysotile asbestos fiber bundle lengths usually range from
a fraction of a millimeter to several centimeters, and diameters range from 0.1 to 100 (.im (Villa. 2002).
Chrysotile asbestos fibers have a net positive surface charge and form a stable suspension in water.
Table 1-1. Physical and Chemical Properties of Chrysotile Asbestos Fibers
Properly
Description
Essential composition
Mg silicate with some water
Color
White, gray, green, yellowish
Surface areab' (m2/g)
13.5-22.4
Hardness (Mohs)
2.5-4.0
Specific gravity
2.4-2.6
Flexibility
High
Texture
Silky, soft to harsh
Spinnability
Very good
Fiber size, median true diameter
(^m)ฐ
0.06ฎ
Fiber size, median true length (|im)d
0.55ฎ
Resistance to Acids
Resistance to Bases
Weak, undergoes fairly rapid attack
Very good
Zeta potential (mV)d
+13.6 to +54
Decomposition temperature (ฐC)
600-850
aBadollef (1951)
b Addison et al. (1966)
0 Hwang (1983)
d Virta (2011)

e The reported values for diameter and length are median values. As reported in Virta (2011). "Industrial chrysotile
fibers are aggregates.. .that usually exhibit diameters from 0.1 to lOOjim: their lengths range from a fraction of a
millimeter to several centimeters, although most chrysotile fibers used are < 1 cm."
Page 39 of 352

-------
676	Figure 1-1 shows two pictures of chrysotile asbestos; one at the "macro" level and one at the
677	microscopic level.
678
679
686	Figure 1-1. Chrysotile Asbestos.
687	Both photographs are from the USGS, A (top and B (bottom)
688
Page 40 of 352

-------
689
1.2 Uses and Production Volume
690	The only form of asbestos manufactured (including imported), processed, or distributed for use in the
691	United States today is chrysotile asbestos. The United States Geological Survey (USGS) estimated that
692	100 metric tons of raw chrysotile asbestos were imported into the U.S. in 2019 (USGS. 2020). This raw
693	asbestos is used exclusively by the chlor-alkali industry and imported amounts tend to range between
694	100 and 800 metric tons during a given year.
695	In addition to the use of raw imported chrysotile asbestos by the chlor-alkali industry, EPA is also aware
696	of imported asbestos-containing products; however, the import volumes of those products are not fully
697	known. The asbestos-containing products that EPA has identified as being imported and used are sheet
698	gaskets, brake blocks, aftermarket automotive brakes/linings, other vehicle friction products, and other
699	gaskets. More information about the uses of chrysotile asbestos and EPA's methodology for identifying
700	COUs is provided in Section 1.4.1 of this document. EPA will consider legacy uses and other types of
701	asbestos fibers in Part 2 of the risk evaluation of asbestos (see Preamble).
702	1.3 Regulatory and Assessment History
703	EPA conducted a search of existing domestic and international laws, regulations and assessments
704	pertaining to asbestos. EPA compiled this summary from data available from federal, state, international
705	and other government sources, as cited in Appendix A (Regulatory History). EPA evaluated and
706	considered the impact of at least some of these existing laws and regulations to determine what, if any
707	further analysis might be necessary as part of the risk evaluation for asbestos. Consideration of the nexus
708	between these regulations and the TSCA COUs evaluated in Part 1 of the risk evaluation for asbestos
709	were developed and described in the Problem Formulation document and are further described in
710	Section 1.4.2.
711	Federal Laws and Regulations
712	Asbestos is subject to federal statutes or regulations, other than TSCA, that are implemented by other
713	offices within EPA and/or other federal agencies/departments. A summary of federal laws, regulations
714	and implementing authorities is provided in Appendix A. 1.
715	State Laws and Regulations
716	Asbestos is subject to statutes or regulations implemented by state agencies or departments. A summary
717	of state laws, regulations and implementing authorities is provided in Appendix A.2.
718	Laws and Regulations in Other Countries and International Treaties or Agreements
719	Asbestos is subject to statutes or regulations in countries other than the United States and/or
720	international treaties and/or agreements. A summary of these laws, regulations, treaties and/or
721	agreements is provided in Appendix A.3.
722	Table 1-2. Assessment History of Asbestos provides assessments related to asbestos conducted by other
723	EPA Programs and other organizations. Depending on the source, these assessments may include
724	information on COU, hazards, exposures and potentially exposed or susceptible subpopulations.
Page 41 of 352

-------
725 	Table 1-2. Assessment History of Asbestos
Authoring Organization
Assessment
EPA assessments
EPA, Integrated Risk Information
System (IRIS)
IRIS Assessment on Asbestos (1988b)
EPA, Integrated Risk Information
System (IRIS)
IRIS Assessment on Libbv Amphibole Asbestos (2014c)

EPA, Region 8
Site-Wide Baseline Ecological Risk Assessment Libbv Asbestos
Suoerfund Site. Libbv Montana U.S. EPA. (2014b)
EPA, Drinking Water Criteria
Document
cine Water Criteria Document for Asbestos (1985)
EPA, Ambient Water Quality
Criteria for Asbestos
Asbestos: Ambient Water Quality Criteri ( ))
EPA, Final Rule (40 CFR Part
763)
Asbestos: Manufacture. Importation. Processing and Distribution
in Commerce Prohibitions (1989)
EPA, Asbestos Modeling Study
Final Report; Asbestos Modeling Studv 8a)
EPA, Asbestos Exposure
Assessment
Revised Report to support. 988)

EPA, Nonoccupational Exposure
Report
Revised Draft Report, Nonoccupational Asbestos Exposure Versar
7)
EPA, Airborne Asbestos Health
Assessment Update
Support document for NESHAP review (1986)
Other U.S.-based organizations
National Institute for Occupational
Safety and Health (NIOSH)
Asbestos Fibers and Other Elongate Mineral Particles: State of the
Science and Roadman for Research ( )
Agency for Toxic Substances and
Disease Registry (ATSDR)
Toxicologic lie for Asbestos (2001a)
National Toxicology Program
(NTP)
Report on Carcinogens. Fourteenth Edition ( )
CA Office of Environmental
Health Hazard Assessment
(OEHHA), Pesticide and
Environmental Toxicology Section
Public Health Goal for Asbestos in Drinking Water (2003)

International
International Agency for Research
on Cancer (IARC)
IARC Monographs on the Evaluation of Carcinogenic Risks to
Humans. Arsenic. Metals. Fibres, and Dusts. Asbestos (Chrvsotile.,
Amosite. Crocidolite. Tremolite. Actinolite. and Anthophvllite)
(2012b)
World Health Organization
(WHO)
World Health Organization (WHO) Chrvsotile Asbestos (2014)

Page 42 of 352

-------
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
Authoring Organization
Assessment
Environment and Climate Change
Canada
Prohibition of Asbestos and Products Containing Asbestos
Regulations
https://www.canada.ca/content/dam/eccc/documents/pdf/pollution-
waste/asbestos-amiante/general%20factsheet%20_EN.pdf
1.4 Scope of the Evaluation
1.4.1 Refinement of Asbestos Fiber Type Considered in the Risk Evaluation for
Asbestos Part 1: Chrysotile Asbestos
EPA determined that the only form of asbestos manufactured (including imported), processed, or
distributed for use in the United States today is chrysotile asbestos. The other five forms of asbestos are
no longer manufactured, imported, or processed in the United States and are now subject to a significant
new use rule (SNUR) that requires notification of and review by the Agency should any person wish to
pursue manufacturing, importing, or processing crocidolite (riebeckite), amosite (cummingtonite-
grunerite), anthophyllite, tremolite or actinolite (either in raw form or as part of articles) for any use (40
CFR 721.11095). Therefore, under the final asbestos SNUR, EPA will be made aware of manufacturing,
importing, or processing for any intended use of the other forms of asbestos. If EPA finds upon review
of the Significant New Use Notice (SNUN) that the significant new use presents or may present an
unreasonable risk (or if there is insufficient information to permit a reasoned evaluation of the health and
environmental effects of the significant new use), then EPA would take action under TSCA section 5(e)
or (f) to the extent necessary to protect against unreasonable risk.
Data from USGS indicates that the asbestos being imported for chlor-alkali plants is all chrysotile
asbestos. Virta (2006) notes that when South Africa closed its amosite and crocidolite mines (in 1992
and 1997 respectively), worldwide production of amosite and crocidolite ceased. Virta (2006) concluded
that almost all of the world's production of asbestos is chrysotile and that "[sjmall amounts, probably
less than a few thousand tons, of actinolite, anthophyllite, and tremolite asbestos are produced for local
use in countries such as India, Pakistan, and Turkey."
Chrysotile asbestos is the prevailing form of asbestos currently mined worldwide, and therefore;
commercially available products fabricated overseas are made with chrysotile asbestos. Any asbestos
being imported into the U.S. in articles for the COUs EPA has identified is believed to be chrysotile
asbestos. Based on EPA's determination that chrysotile asbestos is the only form of asbestos imported
into the U.S. as both raw form and as contained in articles, EPA is performing a quantitative evaluation
for chrysotile asbestos in this Part 1 of the risk evaluation for asbestos. EPA will consider legacy uses
and associated disposals of asbestos in Part 2. Together, Parts 1 and 2 will constitute the final Risk
Evaluation for asbestos.
1.4.2 Conditions of Use Included in the Risk Evaluation for Asbestos Part 1:
Chrysotile Asbestos
TSCA ง 3(4) defines the COU as "the circumstances, as determined by the Administrator, under which
a chemical substance is intended, known, or reasonably foreseen to be manufactured, processed,
distributed in commerce, used, or disposed of." Throughout the scoping ( ), PF (2018d). and risk
evaluation stages, EPA identified and verified the uses of asbestos.
Page 43 of 352

-------
764
765
766
767
768
769
770
111
772
773
774
775
776
777
|778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
To determine the COUs of asbestos and inversely, activities that do not qualify as a COU, EPA
conducted extensive research and outreach. This included EPA's review of published literature and
online databases including the most recent data available from EPA's Chemical Data Reporting (CDR)
and Toxics Release Inventory (TRI) programs, Safety Data Sheets (SDSs), the U.S. Geological Survey's
Mineral Commodities Summary and Minerals Yearbook, the U.S. International Trade Commission's
DataWeb and government and commercial trade databases. EPA also reviewed company websites of
potential manufacturers, importers, distributors, retailers, or other users of asbestos. EPA also received
comments on the Scope of the Risk Evaluation for Asbestos (EPA-HQ-OPPT-2016-0736-0086, (2017c)
that were used to inform the COUs. In addition, prior to the June 2017 publication of the scope
document, EPA convened meetings with companies, industry groups, chemical users, and other
stakeholders to aid in identifying COUs, and verifying COUs identified by EPA.
EPA has removed from this Part 1 of the risk evaluation for asbestos any activities that EPA has
concluded do not constitute COUs - for example, because EPA has insufficient information to find
certain activities are circumstances under which the chemical is actually "intended, known, or
reasonably foreseen to be manufactured, processed, distributed in commerce, used or disposed of."
Since the PF document was published in June 2018 (	18d). EPA has further refined the
COU of asbestos as described in the draft Risk Evaluation. In that document, EPA determined that
packings, woven products, and cement products are not current COUs. Asbestos "packings" are listed
under a broader category of "gaskets, packings, and seals" and more detailed data revealed that only
imported gaskets, not packings, contain asbestos. EPA concluded that "woven and knitted fabrics,"
which are reported in USGS's 2016 Minerals Yearbook under Harmonized Tariff Schedule (HTS) code
6812.99.0004 are misreported (see Appendix C for further explanation). Upon further review, EPA
determined that woven products are not a COU but are precursors to asbestos-containing products or
physical attributes of the asbestos. EPA contacted potential foreign exporters of asbestos woven
products and asbestos cement products, and these foreign companies informed EPA that they do not
have customers in the United States (2018b. c). The Agency has not found any evidence to suggest that
woven products (other than those that are already covered under a distinct COU such as brake blocks
used in draw works) or cement products imported into the United States contain asbestos. Furthermore,
EPA discussed the use of asbestos in cement pipe with a trade organization, who indicated that domestic
production, importation, or distribution for such a use is neither known to be currently ongoing nor
foreseeable (A.WWA. 2019). Based on outreach activity and lack of evidence, EPA does not believe
asbestos packings, asbestos woven products (that are not already covered under a separate and ongoing
COU), or asbestos cement products are manufactured (including imported), processed, or distributed in
the United States, and therefore, packings, woven products, and cement products are no longer under
consideration for this Part 1 of the risk evaluation on asbestos which is focused on chrysotile asbestos
and are now subject to the asbestos SNUR under TSCA section 5. Table 1-3. represents the activities
that have been removed from the scope of this risk evaluation for chrysotile asbestos (Part 1 of the risk
evaluation for asbestos). EPA will consider legacy uses and other asbestos fiber types in Part 2 of the
risk evaluation for asbestos.
Table 1-3. Categories Determined Not to be Manufactured (Including Imported), Processed, or
Product Category
Kxamplc
Asbestos Cement Products
Cement pipe
Asbestos Woven Products
Imported Textiles
Asbestos Packings
Dynamic or mechanical seals
Page 44 of 352

-------
806	EPA has verified that U.S. automotive manufacturers are not installing asbestos brakes on new cars for
807	domestic distribution or use. Therefore, this use will only be evaluated in occupational settings for one
808	use that EPA identified for cars that are manufactured with asbestos-containing brakes in the U.S. but
809	are exported and not sold in the U.S. However, removing and installing asbestos brakes in older vehicles
810	by both professional mechanics and DIY consumers will be evaluated (see Table 1-4. below). The only
811	use that was identified for the "other gaskets" category was for a specific utility vehicle (UTV) that has
f$12	an asbestos-containing gasket in its exhaust system.
813	Based on the above discussion, the COUs that are included in this Part 1 of the risk evaluation for
814	asbestos are described in Table 1-4.
815	The life cycle diagram is presented in Figure 1-2. Chrysotile Asbestos.
816
817	Table 1-4. Categories of Conditions of Use Included in this Risk Evaluation for Asbestos Part 1:
818	Chrysotile Asbestos		
Prodiiel Category
Example
Asbestos Diaphragms
Chlor-alkali Industry
Sheet Gaskets
Chemical Production
Oilfield Brake Blocks
Oil Industry
Aftermarket Automotive Brakes/Linings
Foreign aftermarket brakes sold online
Other Vehicle Friction Products
Brakes installed in exported cars
Other Gaskets
Utility Vehicles
819
Page 45 of 352

-------
MANUFACTURE/IMPORT PROCESSING INDUSTRIAL, COMMERCIAL, CONSUMER USES RELEASES/WASTE DISPOSAL
Manufacture
Asbestos-Con tain ing

Asbestos-Containing
(Non-U.S. Mining)
Diaphragms

Diaphragms

(C hlor-alkali Industry)


Import of Raw
(15 sites)


750 Metric Tons
(2019 USGS)
100 Metric Tons
(2020 USGS)
Import
(Contained
within
Imported
Products)
Volume'1 Amount
Unknown
Asbestos Sheets
(Stamping/Cuttl
ng>
Oilfield Brake
Blocks
Afterniarket Auto
Brakes/ Linings
Other Gaskets
Other Vehicle
Friction Products
A sb est os- Con tain ing
Sheet Gaskets
Disposal
See Appendix D
for Environmental
Releases and
Wastes
= Mamlfactiu iiig'hi)port I	I = Prcices^|n„	I	= Conditions of use. There is no distinction between industrial.
I	'	commercial, and consumer uses in the Life Cycle Diagram
820
821	Figure 1-2. Chrysotile Asbestos Life Cycle Diagram
822	The life cycle diagram depicts the COUs that have been assessed in this risk evaluation. It has been updated to reflect the removal from the PF
823	of woven products, cement products, and packing (see Section 1.4.3) as well as using the 2019 import volume of raw asbestos (reported in
824	2020).

-------
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
1.4.3 Refinement of Evaluation of Releases to Surface Water
EPA did not evaluate the risk to aquatic species from exposure to surface water in its PF. During the PF
phase of the Risk Evaluation, EPA was still in the process of identifying potential asbestos water
releases for the TSCA COUs for chrysotile asbestos. After the PF was released, EPA continued to search
EPA databases as well as the literature and attempted to contact industries to shed light on potential
releases to water. The available information indicated that there were surface water releases of asbestos;
however, it is unclear of the source of the asbestos and the fiber type present. In the draft Risk
Evaluation, EPA concluded that, based on the reasonably available information in the published
literature, provided by industries using asbestos, and reported in EPA databases, there were minimal or
no releases of asbestos to surface water associated with the COUs that EPA is evaluating (see Appendix
D).
EPA has considered peer review and public comments on this conclusion and has decided to keep the
finding made in the draft Risk Evaluation {i.e., that there were minimal or no releases of asbestos to
surface water associated with the COUs that EPA is evaluating in this Part 1 of the risk evaluation for
asbestos). This is because EPA is confident that the minimal water release data available and reported
more fully in the PF - and now presented again in Appendix D - cannot be attributed to chrysotile
asbestos from the COUs in this Part 1 of the risk evaluation for asbestos. Assessing possible risk to
aquatic organisms from the exposures described would not be reasonably attributed to the COUs.
However, based on the decision to develop a scope and risk evaluation for legacy uses and associated
disposals of asbestos (Part 2 of the final Risk Evaluation for asbestos), EPA expects to address the issue
of releases to surface water based on those other uses.
1.4.4 Exposure Pathways and Risks Addressed by Other EPA-Administered Statutes
In its TSCA Section 6(b) risk evaluations, EPA is coordinating action on certain exposure pathways and
risks falling under the jurisdiction of other EPA-administered statutes or regulatory programs. More
specifically, EPA is exercising its TSCA authorities to tailor the scope of its risk evaluations, rather than
focusing on environmental exposure pathways addressed under other EPA-administered statutes or
regulatory programs or risks that could be eliminated or reduced to a sufficient extent by actions taken
under other EPA-administered laws. EPA considers this approach to be a reasonable exercise of the
Agency's TSCA authorities, which include:
•	TSCA Section 6(b)(4)(D): "The Administrator shall, not later than 6 months after the initiation
of a risk evaluation, publish the scope of the risk evaluation to be conducted, including the
hazards, exposures, conditions of use, and the potentially exposed or susceptible subpopulations
the Administrator expects to consider.
•	TSCA Section 9(b)(1): "The Administrator shall coordinate actions taken under this chapter with
actions taken under other Federal laws administered in whole or in part by the Administrator. If
the Administrator determines that a risk to health or the environment associated with a chemical
substance or mixture could be eliminated or reduced to a sufficient extent by actions taken under
the authorities contained in such other Federal laws, the Administrator shall use such authorities
to protect against such risk unless the Administrator determines, in the Administrator's
discretion, that it is in the public interest to protect against such risk by actions taken under this
chapter."
•	TSCA Section 9(e): .. [I]f the Administrator obtains information related to exposures or
releases of a chemical substance or mixture that may be prevented or reduced under another
Federal law, including a law not administered by the Administrator, the Administrator shall

-------
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
make such information available to the relevant Federal agency or office of the Environmental
Protection Agency."
•	TSCA Section 2(c): "It is the intent of Congress that the Administrator shall carry out this
chapter in a reasonable and prudent manner, and that the Administrator shall consider the
environmental, economic, and social impact of any action the Administrator takes or proposes as
provided under this chapter."
•	TSCA Section 18(d)(1): "Nothing in this chapter, nor any amendment made by the Frank R.
Lautenberg Chemical Safety for the 21st Century Act, nor any rule, standard of performance,
risk evaluation, or scientific assessment implemented pursuant to this chapter, shall affect the
right of a State or a political subdivision of a State to adopt or enforce any rule, standard of
performance, risk evaluation, scientific assessment, or any other protection for public health or
the environment that— (i) is adopted or authorized under the authority of any other Federal law
or adopted to satisfy or obtain authorization or approval under any other Federal law..
TSCA authorities supporting tailored risk evaluations and intra-agencv referrals
TSCA Section 6(b)(4)(D)
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.
In the problem formulation documents for many of the first 10 chemicals undergoing risk evaluation,
EPA applied this 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, by
excluding, on a case-by-case basis, certain exposure pathways that fall under the jurisdiction of other
EPA-administered statutes." 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. Consistent with the approach
articulated in the problem formulation documents, and as described in more detail below, EPA is
exercising its authority under TSCA to tailor the scope of exposures evaluated in TSCA risk evaluations,
rather than focusing on environmental exposure pathways addressed under other EPA-administered,
media-specific statutes and regulatory programs.
TSCA Section 9(b)(1)
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." This broad, freestanding
authority provides for intra-agency coordination and cooperation on a range of "actions." In EPA's
view, the phrase "actions taken under [TSCA]" in the first sentence of Section 9(b)(1) is reasonably read
to encompass more than just risk management actions, and to include actions taken during risk
evaluation as well. More specifically, the authority to coordinate intra-agency actions exists regardless
of whether the Administrator has first made a definitive finding of risk, formally determined that such
risk could be eliminated or reduced to a sufficient extent by actions taken under authorities in other
EPA-administered Federal laws, and/or made any associated finding as to whether it is in the public
interest to protect against such risk by actions taken under TSCA. TSCA Section 9(b)(1) therefore
provides EPA authority to coordinate actions with other EPA offices without ever making a risk finding
or following an identification of risk. This includes coordination on tailoring the scope of TSCA risk
Page 48 of 352

-------
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
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).
In a narrower application of the broad authority provided by the first sentence of TSCA Section 9(b)(1),
the remaining provisions of Section 9(b)(1) provide EPA authority to identify risks and refer certain of
those risks for action by other EPA offices. Under the second sentence of Section 9(b)(1), "[i]f the
Administrator determines that a risk to health or the environment associated with a chemical substance
or mixture could be eliminated or reduced to a sufficient extent by actions taken under the authorities
contained in such other Federal laws, the Administrator shall use such authorities to protect against such
risk unless the Administrator determines, in the Administrator's discretion, that it is in the public interest
to protect against such risk by actions taken under [TSCA]." Coordination of intra-agency action on
risks under TSCA Section 9(b)(1) therefore entails both an identification of risk, and a referral of any
risk that could be eliminated or reduced to a sufficient extent under other EPA-administered laws to the
EPA office(s) responsible for implementing those laws (absent a finding that it is in the public interest to
protect against the risk by actions taken under TSCA).
Risk may be identified by OPPT or another EPA office, and the form of the identification may vary. For
instance, OPPT may find that one or more conditions of use for a chemical substance present(s) a risk to
human or ecological receptors through specific exposure routes and/or pathways. This could involve a
quantitative or qualitative assessment of risk based on reasonably available information (which might
include, e.g., findings or statements by other EPA offices or other federal agencies). Alternatively, risk
could be identified by another EPA office. For example, another EPA office administering non-TSCA
authorities may have sufficient monitoring or modeling data to indicate that a particular condition of use
presents risk to certain human or ecological receptors, based on expected hazards and exposures. This
risk finding could be informed by information made available to the relevant office under TSCA Section
9(e), which supports cooperative actions through coordinated information-sharing.
Following an identification of risk, EPA would determine if that risk could be eliminated or reduced to a
sufficient extent by actions taken under authorities in other EPA-administered laws. If so, TSCA
requires EPA to "use such authorities to protect against such risk," unless EPA determines that it is in
the public interest to protect against that risk by actions taken under TSCA. In some instances, EPA may
find that a risk could be sufficiently reduced or eliminated by future action taken under non-TSCA
authority. This might include, e.g., action taken under the authority of the Safe Drinking Water Act to
address risk to the general population from a chemical substance in drinking water, particularly if the
Office of Water has taken preliminary steps such as listing the subject chemical substance on the
Contaminant Candidate List. This sort of risk finding and referral could occur during the risk evaluation
process, thereby enabling EPA to use more a relevant and appropriate authority administered by another
EPA office to protect against hazards or exposures to affected receptors.
Legislative history on TSCA Section 9(b)(1) supports both broad coordination on current intra-agency
actions, and narrower coordination when risk is identified and referred to another EPA office for action.
A Conference Report from the time of TSCA's passage explained that Section 9 is intended "to assure
that overlapping or duplicative regulation is avoided while attempting to provide for the greatest
possible measure of protection to health and the environment." S. Rep. No. 94-1302 at 84. See also H.
Rep. No. 114-176 at 28 (stating that the 2016 TSCA amendments "reinforce TSCA's original purpose of
filling gaps in Federal law," and citing new language in Section 9(b)(2) intended "to focus the
Administrator's exercise of discretion regarding which statute to apply and to encourage decisions that
avoid confusion, complication, and duplication"). Exercising TSCA Section 9(b)(1) authority to
coordinate on tailoring TSCA risk evaluations is consistent with this expression of Congressional intent.
Page 49 of 352

-------
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
Legislative history also supports a reading of Section 9(b)(1) under which EPA coordinates intra-agency
action, including information-sharing under TSCA Section 9(e), and the appropriately positioned EPA
office is responsible for the identification of risk and actions to protect against such risks. See, e.g.,
Senate Report 114-67, 2016 Cong. Rec. S3522 (under TSCA Section 9, "if the Administrator finds that
disposal of a chemical substance may pose risks that could be prevented or reduced under the Solid
Waste Disposal Act, the Administrator should ensure that the relevant office of the EPA receives that
information"); H. Rep. No. 114-176 at 28, 2016 Cong. Rec. S3522 (under Section 9, "if the
Administrator determines that a risk to health or the environment associated with disposal of a chemical
substance could be eliminated or reduced to a sufficient extent under the Solid Waste Disposal Act, the
Administrator should use those authorities to protect against the risk"). Legislative history on Section
9(b)(1) therefore supports coordination with and referral of action to other EPA offices, especially when
statutes and associated regulatory programs administered by those offices could address exposure
pathways or risks associated with conditions of use, hazards, and/or exposure pathways that may
otherwise be within the scope of TSCA risk evaluations.
TSCA Sections 2(c) & 18(d)(1)
Finally, TSCA Sections 2(c) and 18(d) support coordinated action on exposure pathways and risks
addressed by other EPA-administered statutes and regulatory programs. Section 2(c) directs EPA to
carry out TSCA in a "reasonable and prudent manner" and to consider "the environmental, economic,
and social impact" of its actions under TSCA. Legislative history from around the time of TSCA's
passage indicates that Congress intended EPA to consider the context and take into account the impacts
of each action under TSCA. S. Rep. No. 94-698 at 14 ("the intent of Congress as stated in this
subsection should guide each action the Administrator takes under other sections of the bill").
Section 18(d)(1) specifies that state actions adopted or authorized under any Federal law are not
preempted by an order of no unreasonable risk issued pursuant to TSCA Section 6(i)(l) or a rule to
address unreasonable risk issued under TSCA Section 6(a). Thus, even if a risk evaluation were to
address exposures or risks that are otherwise addressed by other federal laws and, for example,
implemented by states, the state laws implementing those federal requirements would not be preempted.
In such a case, both the other federal and state laws, as well as any TSCA Section 6(i)(l) order or TSCA
Section 6(a) rule, would apply to the same issue area. See also TSCA Section 18(d)(l)(A)(iii). In
legislative history on amended TSCA pertaining to Section 18(d), Congress opined that "[t]his approach
is appropriate for the considerable body of law regulating chemical releases to the environment, such as
air and water quality, where the states have traditionally had a significant regulatory role and often have
a uniquely local concern." Sen. Rep. 114-67 at 26.
EPA's careful consideration of whether other EPA-administered authorities are available and more
appropriate for addressing certain exposures and risks is consistent with Congress' intent to maintain
existing federal requirements and the state actions adopted to locally and more specifically implement
those federal requirements, and to carry out TSCA in a reasonable and prudent manner. EPA believes it
is both reasonable and prudent to tailor TSCA risk evaluations in a manner reflective of expertise and
experience exercised by other EPA and State offices to address specific environmental media, rather
than attempt to evaluate and regulate potential exposures and risks from those media under TSCA. This
approach furthers Congressional direction and EPA aims to efficiently use Agency resources, avoid
duplicating efforts taken pursuant to other Agency and State programs, and meet the statutory deadline
for completing risk evaluations.
Page 50 of 352

-------
1003
EPA-administered statutes and regulatory programs that address specific exposure pathways and/or risks
1004	During the course of the risk evaluation process for asbestos, Part 1 (chrysotile asbestos), OPPT worked
1005	closely with the offices within EPA that administer and implement regulatory programs under the Clean
1006	Air Act (CAA), the Safe Drinking Water Act (SDWA), and the Clean Water Act (CWA). Through intra-
1007	agency coordination, EPA determined that specific exposure pathways are well-regulated by the EPA
1008	statutes and regulations described in the following paragraphs.
1009	Ambient Air Pathway
1010	The CAA contains a list of hazardous air pollutants (HAP) and provides EPA with the authority to add
1011	to that list pollutants that present, or may present, a threat of adverse human health effects or adverse
1012	environmental effects. For stationary source categories emitting HAP, the CAA requires issuance of
1013	technology-based standards and, if necessary, additions or revisions to address developments in
1014	practices, processes, and control technologies, and to ensure the standards adequately protect public
1015	health and the environment. The CAA thereby provides EPA with comprehensive authority to regulate
1016	emissions to ambient air of any hazardous air pollutant.
1017	Asbestos was designated as a HAP on March 31, 1971 (36 FR 5931) and remains listed as a HAP under
1018	Section 112 of the CAA. See 42 U.S.C. ง 7412. EPA has issued a number of standards for source
1019	categories that emit pollutants designated as a HAP prior to the 1990 Clean Air Act Amendments as
1020	well as technology-based standards for source categories that emit pollutants listed as a HAP under
1021	Section 112 of the CAA. The National Emission Standard for Asbestos includes standards for multiple
1022	sources or potential sources of asbestos releases to the ambient air including asbestos mills, roadways,
1023	manufacturing, demolition and renovation, insulating materials, waste disposal, and operations that
1024	convert asbestos-containing waste material into non-asbestos (asbestos-free) material, among others. See
1025	40 CFR part 61, subpart M. Because stationary source releases of asbestos to ambient air are addressed
1026	under the CAA, EPA is not evaluating emissions to ambient air from commercial and industrial
1027	stationary sources or associated inhalation exposure of the general population or terrestrial species in
1028	this Part 1 of the risk evaluation for asbestos.
1029	Drinking Water Pathway
1030	EPA has regular analytical processes to identify and evaluate drinking water contaminants of potential
1031	regulatory concern for public water systems under the Safe Drinking Water Act (SDWA). Under
1032	SDWA, EPA must also review existing national primary drinking water regulations every 6 years, and
1033	subsequently revise them as appropriate.
1034	EPA has promulgated National Primary Drinking Water Regulations (NPDWRs) for asbestos under
1035	SDWA. See 40 CFR part 141; Appendix A. EPA has set an enforceable Maximum Contaminant Level
1036	(MCL) as close as feasible to a health based, non-enforceable Maximum Contaminant Level Goal
1037	(MCLG). Feasibility refers to both the ability to treat water to meet the MCL and the ability to monitor
1038	water quality at the MCL, SDWA Sections 1412(b)(4)(D) and 1401(l)(C)(i), Public water systems are
1039	required to monitor for the regulated chemical based on a standardized monitoring schedule to ensure
1040	compliance with the MCL. The MCL for asbestos in water is 7 million fibers/liter, or 7 MFL.
1041	Hence, because the drinking water exposure pathway for asbestos is currently addressed in the NPDWR,
1042	EPA is not evaluating exposures to the general population from the drinking water exposure pathway in
1043	in this Part 1 of the risk evaluation for asbestos.
Page 51 of 352

-------
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
Ambient Water Pathway
EPA develops recommended water quality criteria under Section 304(a) of the CWA for pollutants in
surface water that are protective of aquatic life or human health designated uses. EPA develops and
publishes water quality criteria based on priorities of states and others that reflect the latest scientific
knowledge. A subset of these chemicals is identified as "priority pollutants" (103 human health and 27
aquatic life). The CWA requires states adopt numeric criteria for priority pollutants for which EPA has
published recommended criteria under Section 304(a), the discharge or presence of which in the affected
waters could reasonably be expected to interfere with designated uses adopted by the state. When states
adopt criteria that EPA approves as part of state's regulatory water quality standards, exposure is
considered when state permit writers determine if permit limits are needed and at what level for a
specific discharger of a pollutant to ensure protection of the designated uses of the receiving water. Once
states adopt criteria as water quality standards, the CWA requires that National Pollutant Discharge
Elimination System (NPDES) discharge permits include effluent limits as stringent as necessary to meet
standards. CWA Section 301(b)(1)(C). This is the process used under the CWA to address risk to human
health and aquatic life from exposure to a pollutant in ambient waters.
EPA has identified asbestos as a priority pollutant and EPA has developed recommended water quality
criteria for protection of human health for asbestos which are available for adoption into state water
quality standards for the protection of human health and are available for use by NPDES permitting
authorities in deriving effluent limits to meet state narrative criteria. See, e.g., 40 CFR part 423,
Appendix A; 40 CFR 131.11(b)(1); 40 CFR 122.44(d)(l)(vi); and 40 CFR 131.36(b)(1), 131.38(b)(1),
and 40 CFR part 122, Appendix D, Table V. As such, EPA is not evaluating exposures to the general
population from the surface water exposure pathway in this Part 1 of the risk evaluation for asbestos.
EPA has not developed CWA section 304(a) recommended water quality criteria for the protection of
aquatic life for asbestos, so there are no national recommended criteria for this use available for
adoption into state water quality standards and available for use in NPDES permits.
On-site Releases to Superfund Sites
The Comprehensive Environmental Response, Compensation, and Liability Act - otherwise known as
CERCLA or Superfund - provides EPA with broad authority to address uncontrolled or abandoned
hazardous-waste sites as well as accidents, spills, and other releases of hazardous substances, pollutants
and contaminants into the environment. Through CERCLA, EPA is provided authority to conduct a
response action and seek reimbursement of cleanup costs from potentially responsible parties, or in
certain circumstances, order a potentially responsible party to conduct a cleanup.
CERCLA Section 101(14) defines "hazardous substance" by referencing other environmental statutes,
including toxic pollutants listed under CWA Section 307(a); hazardous substances designated pursuant
to CWA Section 311(b)(2)(A); hazardous air pollutants listed under CAA Section 112; imminently
hazardous substances with respect to which EPA has taken action pursuant to TSCA Section 7; and
hazardous wastes having characteristics identified under or listed pursuant to RCRA Section 3001. See
40 CFR 302.4. CERCLA Section 102(a) also authorizes EPA to promulgate regulations designating as
hazardous substances those substances which, when released into the environment, may present
substantial danger to the public health or welfare or the environment. EPA must also promulgate
regulations establishing the quantity of any hazardous substance the release of which must be reported
under Section 103. Section 103 requires persons in charge of vessels or facilities to report to the
National Response Center if they have knowledge of a release of a hazardous substance above the
reportable quantity threshold.
Page 52 of 352

-------
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
Asbestos is a hazardous substance under CERCLA. Releases of friable asbestos in excess of 1 pound
within a 24-hour period must be reported (40 CFR 302.4, 302.6). This Part 1 of the risk evaluation for
asbestos does not include on-site releases to the environment of asbestos at Superfund sites and
subsequent exposure of the general population or non-human species.
Disposal Pathways
Asbestos is not regulated as a RCRA hazardous waste under RCRA Subtitle C; therefore, asbestos solid
wastes are not required to be disposed of in Subtitle C hazardous waste landfills. However, it is possible
that asbestos wastes could be disposed this way due to other characteristics of an asbestos-containing
solid waste mixture. EPA is not evaluating on-site releases to land from RCRA Subtitle C hazardous
waste landfills or exposures of the general population or terrestrial species from such releases in this Part
1 of the risk evaluation for asbestos. Design standards for Subtitle C landfills require double liner,
double leachate collection and removal systems, leak detection system, run on, runoff, and wind
dispersal controls, and a construction quality assurance program. They are also subject to closure and
post-closure care requirements including installing and maintaining a final cover, continuing operation
of the leachate collection and removal system until leachate is no longer detected, maintaining and
monitoring the leak detection and groundwater monitoring system. Bulk liquids may not be disposed in
Subtitle C landfills. Subtitle C landfill operators are required to implement an analysis and testing
program to ensure adequate knowledge of waste being managed, and to train personnel on routine and
emergency operations at the facility. Hazardous waste being disposed in Subtitle C landfills must also
meet RCRA waste treatment standards before disposal. See 40 CFR part 264; Appendix A.
In addition, landfills have special requirements for handling and securing the asbestos-containing waste
regulated under the National Emission Standard for Asbestos (40 C.F.R. Part 61, Subpart M) to prevent
releases of asbestos into the air. This regulation requires regulated asbestos-containing waste material be
sealed in a leak-tight container while wet, labeled, and disposed of properly in a landfill qualified to
receive asbestos waste. Landfills have special requirements for handling and securing the asbestos
containing waste to prevent releases of asbestos into the air. Transportation vehicles that move the waste
from the point of generation to the asbestos landfill have special labeling requirements and waste
shipment recordkeeping requirements. EPA is not evaluating emissions from the asbestos waste pathway
from the processing and use of chrysotile asbestos diaphragms at chlor-alkali facilities. The National
Emission Standard for Asbestos specifically addresses this asbestos waste pathway. See 40 CFR งง
61.144(a)(9), 61.150. Finally, asbestos fibers (all six types) are not likely to be leached out of a landfill.
EPA is not evaluating on-site releases to land from RCRA Subtitle D municipal solid waste (MSW)
landfills or exposures of the general population or terrestrial species from such releases in this Part 1 of
the risk evaluation for asbestos. While permitted and managed by the individual states, municipal solid
waste landfills are required by federal regulations to implement some of the same requirements as
Subtitle C landfills. MSW landfills generally must have a liner system with leachate collection and
conduct groundwater monitoring and corrective action when releases are detected. MSW landfills are
also subject to closure and post-closure care requirements and must have financial assurance for funding
of any needed corrective actions. MSW landfills have also been designed to allow for the small amounts
of hazardous waste generated by households and very small quantity waste generators (less than 220 lbs
per month). Finally, asbestos fibers (all six types) are not likely to be leached out of a landfill.
EPA is not evaluating on-site releases to land from industrial non-hazardous waste and
construction/demolition waste landfills or associated exposures to the general population or terrestrial
species in this Part 1 of the risk evaluation for asbestos. Industrial non-hazardous and
construction/demolition waste landfills are primarily regulated under authorized state regulatory
Page 53 of 352

-------
1133	programs. States must also implement limited federal regulatory requirements for siting, groundwater
1134	monitoring and corrective action and a prohibition on open dumping and disposal of bulk liquids. States
1135	may also establish additional requirements such as for liners, post-closure and financial assurance, but
1136	are not required to do so. See, e.g., RCRA Section 3004(c), 4007; 40 CFR part 257.
1137	EPA is not evaluating emissions to ambient air from municipal and industrial waste incineration and
1138	energy recovery units or associated exposures to the general population or terrestrial species in this Part
1139	1 of the risk evaluation for asbestos, as these emissions are regulated under Section 129 of the Clean Air
1140	Act. CAA Section 129 requires EPA to review and, if necessary, add provisions to ensure the standards
1141	adequately protect public health and the environment. Thus, combustion by-products from incineration
1142	treatment of asbestos wastes would be subject to these regulations, as would asbestos burned for energy
1143	recovery. See 40 CFR part 60.
1144	EPA is not evaluating on-site releases to land that go to underground injection or associated exposures to
1145	the general population or terrestrial species in this Part 1 of the risk evaluation for asbestos.
1146	Environmental disposal of asbestos injected into Class I hazardous well types are covered under the
1147	jurisdiction of the SDWA and disposal of asbestos via underground injection is not likely to result in
1148	environmental and general population exposures. See 40 CFR part 144.
1149	1.4.5 Conceptual Models
1150	The conceptual models have been modified to reflect the refined COUs of chrysotile asbestos described
1151	in Section 1.4.1. Figure 1-3. Chrysotile Asbestos and Figure 1-4. Commercial Chrysotile Asbestos
1152	present the conceptual models for industrial and commercial uses and consumer uses, respectively-The
1153	chrysotile asbestos conceptual model for environmental releases and wastes from the refined COUs was
1154	removed and is discussed in Releases and Exposure to the Environment Supplementary Information
1155	Appendix D.
Page 54 of 352

-------
1156
1157
INDUSTRIAL AND COMMERCIAL
ACTIVITIES/USES
EXPOSURE PATHWAY
EXPOSURE ROUTE
RECEPTORS3
HAZARDS
Workers,
Occupational
Non-Users
Workers,
Occupational
Non-Users
Sheet Gaskets
Oilfield brake blocks
Other Gaskets
Inhalation
As bestos-Conta i n i ng
Diaphragms
Waste Handling,
Treatment and Disposal
Other Vehicle Friction
Products
Aftermarket Vehicle
Brakes/Linings
Direct Contact with
Dry/Friable Asbestos
Outdoor/Indoor Air
Hazards Potentially Associated with
Asbestos Exposure
See Section 3.2
Emissions to Wastewater,
Liquid Waste
1158
1159
1160
1161
1162
1163
Figure 1-3. Chrysotile Asbestos Conceptual Model for Industrial and Commercial Activities and Uses: Potential Exposures and
Hazards
The conceptual model presents the exposure pathways, exposure routes and hazards to human receptors from industrial and commercial
activities and uses of asbestos.
8 Receptors include PESS.
Page 55 of 352

-------
CONSUMER ACTIVITIES/USES a
EXPOSURE PATHWAY b	EXPOSURE ROUTE
RECEPTORS
HAZARDS
Indoor/Outdoor Air
Inhalation
Other Gaskets
Consumer Handlingof
Disposal and Waste
Aftermarket Auto Brakes/
Linings
Consumers,
Bystanders
Hazards Potentially Associated with
Asbestos Exposure
See Section 3.2
1164
1165	Figure 1-4. Chrysotile Asbestos Conceptual Model for Consumer Activities and Uses: Potential Exposures and Hazards
1166	aWoven products were removed from this model after the PF was published. Utility vehicle gaskets were added.
1167	bProducts may be used during indoor and outdoor activities.
1168	cReceptors include PESS.
Page 56 of 352

-------
1.5 Systematic Review
TSCA requires EPA to use scientific information, technical procedures, measures, methods, protocols,
methodologies and models consistent with the best available science and base decisions under Section 6
on the weight of scientific evidence. Within the TSCA risk evaluation context, the weight of the
scientific evidence is defined as "a systematic review method, applied in a manner suited to the nature
of the evidence or decision, that uses a pre-established protocol to comprehensively, objectively,
transparently, and consistently identify and evaluate each stream of evidence, including strengths,
limitations, and relevance of each study and to integrate evidence as necessary and appropriate based
upon strengths, limitations, and relevance" (40 C.F.R. 702.33).
To meet the TSCA science standards, EPA used the TSCA systematic review process described in the
Application of Systematic Review in TSCA Risk Evaluations document (U.S. EPA. 2018a). The process
complements the risk evaluation process in that the data collection, data evaluation and data integration
stages of the systematic review process are used to develop the exposure and hazard assessments based
on reasonably available information. EPA defines "reasonably available information" to mean
information that EPA possesses, or can reasonably obtain and synthesize for use in risk evaluations,
considering the deadlines for completing the evaluation (40 CFR 702.33).
EPA is implementing systematic review methods and approaches within the regulatory context of the
amended TSCA. Although EPA will make an effort to adopt as many best practices as practicable from
the systematic review community, EPA expects modifications to the process to ensure that the
identification, screening, evaluation and integration of data and information can support timely
regulatory decision making under the aggressive timelines of the statute.
1,5.1 Data and Information Collection
EPA planned and conducted a comprehensive literature search based on key words related to the
different discipline-specific evidence supporting this Part 1 of the risk evaluation for asbestos (e.g.,
environmental fate and transport; engineering releases and occupational exposure; exposure to general
population, consumers and environmental exposure, and environmental and human health hazard). EPA
then developed and applied inclusion and exclusion criteria during the title and abstract screening to
identify information potentially relevant for the risk evaluation process. The literature and screening
strategy as specifically applied to asbestos is described in the Strategy for Conducting Literature
Searches for Asbestos: Supplemental Document to the TSCA Scope Document (EP A-HQ-OPPT-2016-
0736). and the results of the title and abstract screening process were published in the Asbestos (CASRN
1332-21-4) Bibliography: Supplemental File for the TSCA Scope Document, EP A-HQ-OPPT-2016-
0736) (U.S. EPA. 2017b).
For studies determined to be on-topic (or relevant) after title and abstract screening, EPA conducted a
full text screening to further exclude references that were not relevant to this Part 1 of the risk evaluation
for asbestos. Screening decisions were made based on eligibility criteria documented in the form of the
populations, exposures, comparators, and outcomes (PECO) framework or a modified framework.12
12 A PESO statement was used during the full text screening of environmental fate and transport data sources. PESO stands
for Pathways and Processes, Exposure, Setting or Scenario, and Outcomes. A RESO statement was used during the full text
screening of the engineering and occupational exposure literature. RESO stands for Receptors, Exposure, Setting or
Scenario, and Outcomes.
Page 57 of 352

-------
Data sources that met the criteria were carried forward to the data evaluation stage. The inclusion and
exclusion criteria for full text screening for asbestos are available in Appendix D of the Problem
Formulation of the Risk Evaluation for Asbestos (U.S. EPA. 2018d).
Although EPA conducted a comprehensive search and screening process as described above, EPA made
the decision to leverage the literature published in previous assessments13 when identifying relevant key
and supporting data14 and information for developing this Part 1 of the risk evaluation for asbestos. This
is discussed in the Strategy for Conducting '.Literature Searches for Asbestos: Supplemental Document to
the TSCA Scope Document (EPA-HQ-OPPT-2016-073 6). In general, many of the key and supporting
data sources were identified in the comprehensive Asbestos Bibliography: Supplemental File for the
TSCA Scope Document (U.S. EPA... 2017a. b). However, there were instances in which EPA missed
relevant references that were not captured in the initial categorization of the on-topic references. EPA
found additional relevant data and information using backward reference searching, which is a technique
that will be included in future search strategies. This issue is discussed in Section 4 of the Application o f
Systematic Review for TSCA Risk Evaluatior,	,018a). Other relevant key and supporting
references were identified through targeted supplemental searches to support the analytical approaches
and methods in this Part 1 of the risk evaluation for asbestos (e.g., to locate specific information for
exposure modeling) or to identify new data and information published after the date limits of the initial
search.
EPA used previous chemical assessments to quickly identify relevant key and supporting information as
a pragmatic approach to expedite the quality evaluation of the data sources, but many of those data
sources were already captured in the comprehensive literature search as explained above. EPA also
considered newer information on asbestos not taken into account by previous EPA chemical assessments
as described in the Strategy for Conducting Literature Searches for Asbestos: Supplemental Document
to the TSCA Scope Document (EPA-HQ-OPl	736). EPA then evaluated the relevance and
quality of the key and supporting data sources, as well as newer information, instead of reviewing all the
underlying published information on asbestos. A comprehensive evaluation of all of the data and
information ever published for a substance such as asbestos would be extremely labor intensive and
could not be achieved considering the deadlines specified in TSCA Section 6(b)(4)(G) for conducting
risk evaluations.
This pragmatic approach allowed EPA to maximize the scientific and analytical efforts of other
regulatory and non-regulatory agencies by accepting, for the most part, the relevant scientific knowledge
gathered and analyzed by others except for influential information sources that may have an impact on
the weight of the scientific evidence and ultimately the risk findings. The influential information (i.e.,
key/supporting) came from a smaller pool of sources subject to the rigor of the TSCA systematic review
process to ensure that this Part 1 of the risk evaluation for asbestos used the best available science and
the weight of the scientific evidence.
Figure 1-5. to Figure 1-9. depict the literature flow diagrams illustrating the results of this process for
each scientific discipline-specific evidence supporting this Part 1 of the risk evaluation for asbestos.
Each diagram provides the total number of references at the start of each systematic review stage (i.e.,
13	Examples of existing assessments are EPA's chemical assessments (e.g., previous work plan risk assessments, problem
formulation documents), ATSDR's Toxicological Profiles, EPA's IRIS assessments and ECHA's dossiers. This is described
in more detail in the Strategy for Conducting Literature Searches for Asbestos: Supplemental File for the TSCA Scope
Document (EPA-HO-OPPT-2016-Q736V
14	Key and supporting data and information are those that support key analyses, arguments, and/or conclusions in this Part 1
of the risk evaluation for asbestos.
Page 58 of 352

-------
data search, data screening, data evaluation, data extraction/data integration) and those excluded based
on criteria guiding the screening and data quality evaluation decisions.
EPA bypassed the data screening step for data sources that were highly relevant to this Part 1 of the risk
evaluation for asbestos and moved these sources directly to the data quality evaluation step, as described
above. These data sources are depicted as "key/supporting data sources" in the literature flow diagrams.
Note that the number of "key/supporting data sources" were excluded from the total count during the
data screening stage and added, for the most part, to the data evaluation stages depending on the
discipline-specific evidence. The exception was the releases and occupational exposure data sources that
were subject to a combined data extraction and evaluation step as shown in Figure 1-6.
EPA did not have a previous, recent toxicity assessment for general asbestos on which to build;
therefore, initially the Systematic Review included a very large number of papers for all areas. Initially,
studies were limited to those published after 1987, containing at least one of the six fiber types identified
under TSCA. In terms of evaluating human health, only observational human studies were identified for
searching; however, the scope of the risk evaluation was further refined to identify studies pertaining to
only mesothelioma and lung cancer as health outcomes15, as well as studies containing information
specific to chrysotile asbestos only.
As the process proceeded, more data became available and the systematic review was refined. This
included exposure and engineering citations, e.g., correspondences with industry, considered to be on-
topic and used to inform the likelihood of exposure. The nature of these documents is such that the
current framework as outlined in the Application of Systematic Review in TSCA Risk Evaluations (U.S.
EPA. 2018a) is not well suited for the review of these types of references. And as such, these references,
were handled on a case-by-case basis and are cited in the references section of this document.
Information for fate assessment for the first 10 chemical risk evaluations considered the physical
chemical properties of the chemical and environmental endpoints. For the first 10 chemicals, EPA
assessed chemical fate as defined by traditional fate endpoints, for example, solubility, partitioning
coefficients, biodegradation and bioaccumulation - properties that do not apply to asbestos minerals. As
such, there were few discipline-specific papers identified in the fate systematic review of asbestos
literature (Figure 1-5).
15 See Appendix M for an exception in response to peer review and public comments.
Page 59 of 352

-------
Figure 1-5. Key/Supporting Data Sources for Environmental Fate
Note 1: Literature search results for the environmental fate of asbestos yielded 7,698 studies. Of these
studies 7,687 were determined to be off-topic or they did not meet screening criteria (such as non-primary
source data or lacking quantitative fate data). The remaining studies entered full text screening for the
determination of relevance to the risk evaluation. There were three key and/or supporting data sources
identified, the primary literature cited in these sources were passed directly to data evaluation. One
primary study was deemed unacceptable based on the evaluation criteria for fate and transport studies and
the remaining 10 primary studies were carried forward to data extraction/data integration according to
Appendix F in Application of Systematic Review for TSCA Risk Evaluations U.S. EPA (2018a). The data
evaluation and data extraction files are provided in Appendix F.
Note 2: Data sources identified relevant to physical-chemical properties were not included in this
literature flow diagram. The data quality evaluation of physical-chemical properties studies can be found
in the supplemental document, Data Quality Evaluation of Physical-Chemical Properties Studies (U.S.
EPA. 2020j) and the extracted data are presented in Table 1-1.
Data Evaluation (n=ll)
Data Screening (n=7,698)
* Key/Supporting
Data Sources (n=3)
Data Search Results (n=7,698)
Data Extraction/Data Integration (n=10)
Excluded References
(n=7,687)
Excluded: Ref that are
unacceptable based on the
evaluation criteria (n=l)
*These are key and supporting studies from existing assessments (e.g., EPA IRIS assessments, ATSDR assessments, ECHA
dossiers) that were considered highly relevant for the TSCA risk evaluation. These studies bypassed the data screening
step and primary references cited therein were passed directly to the data evaluation step.
Page 60 of 352

-------
Figure 1-6. Key/Supporting Data Sources for Engineering Releases and Occupational Exposure
Note: Literature search results for environmental release and occupational exposure yielded 10,031 data
sources. Of these data sources, 114 were determined to be relevant for this Part 1 of the risk evaluation for
asbestos through the data screening process. These relevant data sources were entered into the data
extraction/evaluation phase. After data extraction/evaluation, EPA identified several data gaps and
performed a supplemental, targeted search to fill these gaps (e.g., to locate information needed for exposure
modeling). The supplemental search yielded six relevant data sources that bypassed the data screening step
and were evaluated and extracted in accordance with Appendix D in Application of Systematic Review for
TSCA Risk Evaluations (U.S. EPA. 2018a). Of the 120 sources from which data were extracted and
evaluated, 39 sources only contained data that were rated as unacceptable based on serious flaws detected
during the evaluation. Of the 81 sources forwarded for data integration, data from 42 sources were
integrated, and 39 sources contained data that were not integrated (e.g., lower quality data that were not
needed due to the existence of higher quality data, data for release media that were removed from scope
after data collection). The data evaluation and data extraction files are provided as separate files (See
Appendix B).
n=114
Key/supporting
data sources
(n=6)
Excluded References (n=9.917)
'Data Sources that were not
integrated (n=39)
Data Search Results (n=10,031)
Excluded: Ref that are
unacceptable based on
evaluation criteria (n=39)
Data Integration (n=42)
Data Extraction/Data Evaluation (n=120)
Data Screening (n=10,031)
"The quality of data in these sources (n=39) were acceptable for risk assessment purposes, but they were ultimately
excluded from further consideration based on EPA's integration approach for environmental release and occupational
exposure data/information EPA's approach uses a hierarchy of preferences that guide decisions about what types of
data/information are included for further analysis, synthesis and integration into the environmental release and
occupational exposure assessments EPA prefers using data with the highest rated quality among those in the higher
level of the hierarchy of preferences (i.e.. data > modeling > occupational exposure limits or release limits). If warranted.
EPA may use data/information of lower rated quality as supportive evidence in the environmental release and
occupational exposure assessments
Page 61 of 352

-------
Figure 1-7. Key/Supporting Data Sources for Consumer and Environmental Exposure
Note: Literature search results for consumer and environmental exposure yielded 1,509 data sources. Of
these data sources, 84 made it through data screening and into data evaluation. These data sources were
then evaluated based on a set of metrics to determine overall relevancy and quality of each data source.
The data evaluation stage excluded an additional 56 data sources based on unacceptability under data
evaluation criteria (6), not considered a primary source of data, no extractable data, or overall low
relevancy to the COUs evaluated (50). The remaining 28 data sources that made it to data evaluation had
data extracted for use within this Part 1 of the risk evaluation for asbestos. The data evaluation and data
extraction files are provided as separate files (See Appendix B).
Excluded References (n = 1425)
Data Evaluation (n = 84)
Data Screening (n = 1509)
Data Search Results (n = 1509)
Unacceptable based on data evaluation criteria (n = 6)
Not primary source, not extractable or
not most relevant (n = 50)
Excluded References (n = 56)
Data Extraction'Data Integration (n = 28)
•The quality of data in these sources were acceptable for risk assessment purposes and considered for
integration. The sources; however, were not extracted for a variety of reasons, such as they contained
only secondary source data, duplicate data, or non-extractable data (i.e., charts or figures). Additionally,
some data sources were not as relevant to the PECO as other data sources which were extracted,
Page 62 of 352

-------
Figure 1-8. Key /Supporting Data Sources for Environmental Hazard
Note: The environmental hazard data sources were identified through literature searches and screening
strategies using the ECOTOX Standing Operating Procedures. Additional details can be found in the
Strategy for Conducting Literature Searches for Asbestos: Supplemental Document to the TSCA Scope
Document, (EPA-HQ-OPPT-2016-0736). During PF, EPA made refinements to the conceptual models
resulting in the elimination of the terrestrial exposure pathways. Thus, environmental hazard data sources
on terrestrial organisms were determined to be out of scope and excluded from data quality evaluation.
The data evaluation file is provided as a separate file (See Appendix B).
Key/Supporting
Studies
(n = 0)
Excluded References due to
ECOTOX Criteria
(n =48)
Excluded References due to
ECOTOX Criteria
(n = 29/6)
Data Extraction I Data Integration (n = 4)
Data Evaluation (n= 10)
Full Text Screening (n = 58)
Excluded References that are
unacceptable based
on evaluation criteria and/or are
out of scope
(n = 6)
Data Search Results (n = 3034)
Title/Abstract Screening (n = 30 34)
Page 63 of 352

-------
Figure 1-9. Key/Supporting Data Sources for Human Health Hazard
Note: Studies were restricted to only mesothelioma and lung cancer as health outcomes16, and further
restricted to studies containing information specific to chrysotile asbestos only. The data evaluation and
data extraction files are provided as separate files (See Appendix B).
n=23 data sources
Key/supporting data
sources
(n = 3 data sources )
Excluded References
(n = 24,012 data sources)
Data Search Results (n= 24,050 data sources)
Excluded: Ref that are
unacceptable based on
evaluation criteria (n = 0 cohorts)
Data Extraction/Data Integration

-------
2 EXPOSURES
This section describes EPA's approach to assessing environmental and human exposures. First, the fate
and transport of chrysotile asbestos in the environment is characterized. Then, releases of chrysotile
asbestos to the environment to evaluate environmental receptors are assessed. Finally, an evaluation of
exposures to humans (occupational and consumers; including potentially exposed or susceptible
subpopulations (PESS)) is presented. For all exposure-related disciplines, EPA screened, evaluated,
extracted and integrated available empirical {i.e., monitoring) data.
Exposure equations and selected values used in the exposure assessment are presented in the following
sections. More specific information is provided in supplemental files identified in Appendix B: List of
Supplemental Documents.
2.1 Fate and Transport
Although Part 1 of the risk evaluation on asbestos is focused on the chrysotile asbestos fiber type, some
of the information in this section is taken from and pertains to asbestos fibers in general. Asbestos is a
persistent mineral fiber that can be found in soils, sediments, lofted in air and windblown dust, surface
water, ground water and biota (ATSDR. 2001b). Asbestos fibers are largely chemically inert in the
environment. They may undergo minor physical changes, such as changes in fiber length or leaching of
surface minerals, but do not react or dissolve in most environmental conditions (Fav era-Lorn go et at..
2005; Gronow. 1987; Schreier et at... 1987; Choi and Smith. 1972).
The reasonably available data/information on the environmental fate of chrysotile asbestos is found in
Appendix F. Those data are summarized below.
Chrysotile asbestos forms stable suspensions in water; surface minerals may leach into solution, but the
underlying silicate structure remains unchanged at neutral pH (Gronow. 1987; Bales and Morgan. 1985;
Choi and Smith. 1972). Small asbestos fibers (<1 |im) remain suspended in air and water for significant
periods of time and may be transported over long distances (Jaenicke. 1980). As stated in the asbestos
PF, once in water it will eventually settle into sediments (or possibly biosolids from wastewater
treatment plants). Chrysotile asbestos fibers will eventually settle to sediments and soil, and movement
therein may occur via erosion, runoff or mechanical resuspension (wind-blown dust, vehicle traffic, etc.)
(ATSDR. 2.001b).
Limited information is available on the bioconcentration or bioaccumulation of asbestos. Aqueous
exposure to chrysotile asbestos (104-108 fibers/liter) results in embedding of fibers in the tissues of
aquatic organisms (Belaneer et at.. 1990; Belanger et al.. 1986c; Bel anger et at.. 1986a. b). In controlled
laboratory experiments, asbestos had a negligible bioconcentration factor (BCF slightly greater than 1)
(Belanger et at.. 1987). Chrysotile asbestos is not expected to bioaccumulate in food webs (ATSDR.
i ).
Chrysotile asbestos, which is the focus of the Risk Evaluation for Asbestos Part 1, may be released to
the environment through industrial or commercial activities, such as processing raw chrysotile asbestos,
fabricating/processing asbestos containing products, or the lofting of friable chrysotile asbestos during
use, disturbance and disposal of asbestos containing products.
Page 65 of 352

-------
2.2 Releases to Water
2,2,1 Water Release Assessment Approach and Methodology
The environmental exposure characterization focuses on aquatic releases of chrysotile asbestos from
facilities that import, process, or use asbestos under industrial and/or commercial COUs as well as the
consumer COUs included in this document. To characterize environmental exposure, EPA assessed
point estimate exposures derived from measured concentrations of asbestos in surface water in the
United States. Measured surface water concentrations were obtained from EPA's Water Quality
Exchange (WQX) using the Water Quality Portal (WQP) tool, which is the nation's largest source of
water quality monitoring data and includes results from EPA's STORage and RETrieval (STORET)
Data Warehouse, the United States Geological Survey (USGS) National Water Information System
(NWIS), and other federal, state, and tribal sources. A literature search was also conducted to identify
other peer-reviewed or authoritative gray sources of measured surface water concentrations in the
United States, but no data were found.
As discussed in the PF document and Exposure Pathways and Risks Addressed by Other EPA-
Administered Statutes (Section 1.4.4) , because the drinking water exposure pathway for asbestos is
currently addressed in the Safe Drinking Water Act (SDWA) via a NPDWR for asbestos, this pathway
(drinking water for human health) was not evaluated in this Part 1 of the risk evaluation for asbestos.
The Office of Water does not have an ambient water quality criterion for asbestos for aquatic life. Thus,
potential releases from industrial and commercial activities associated with the TSCA COUs to surface
water were considered in this Part 1 of the risk evaluation for asbestos. However, identifying or
estimating asbestos concentrations in water to evaluate risk to environmental receptors has been
challenging. During the PF phase of the RE, EPA was still in the process of identifying potential
asbestos water releases for the TSCA COUs. After the PF was released, EPA continued to search other
sources of data including TRI data, EPA environmental and compliance monitoring databases, including
permits, industry responses to EPA questions, and other EPA databases. Details of these investigations
are included in Appendix D and summarized below.
TRI data (Table APX D-2) show that there were zero pounds of friable asbestos reported as released to
water via surface water discharges in 2018. In addition, TRI reports zero pounds of friable asbestos
transferred off-site to publicly owned treatment works (POTWs) or to non-POTW facilities for the
purpose of wastewater treatment. The vast majority of friable asbestos waste management was disposal
to hazardous waste landfills and to non-hazardous waste landfills.
EPA issues Effluent Limitations Guidelines and Pretreatment Standards, which are national regulatory
standards for industrial wastewater discharges to surface waters and POTWs (municipal sewage
treatment plants). EPA issues these guidelines for categories of existing sources and new sources under
Title III of the Clean Water Act (CWA). The standards are technology-based {i.e., they are based on the
performance of treatment and control technologies); they are not based on risk or impacts upon
receiving waters (see Industrial Effluent Guidelines for more information). For most operations covered
by effluent guidelines and standards for the asbestos manufacturing point source category (40 CFR 427),
the discharge of all pollutants is prohibited. For certain asbestos manufacturing operations, the effluent
guidelines establish limits on the allowable levels of total suspended solids (TSS), pH, or chemical
oxygen demand (COD). The regulations do not establish specific limits for asbestos from those
operations where discharges are allowed. Thus, without the requirement to measure asbestos
concentrations in effluent, estimating asbestos levels in effluent or receiving waters is challenging.
Page 66 of 352

-------
EPA investigated industry sector, facility, operational, and permit information regulated by NPDES
(National Pollutant Discharge Elimination System) under the CWA to identify any permit limits,
monitoring and reporting requirements, and any discharge provisions related to asbestos. The CWA
prohibits point source pollutant discharges into waters of the United States unless specifically authorized
under the Act, for example through a permit under section 402 (by EPA or an authorized state) that
establishes conditions for discharge. Available data were accessed through EPA's Envirofacts and
Enforcement and Compliance History Online (ECHO) systems to identify any evidence of asbestos
discharge pertaining to the COUs being evaluated herein. EPA found that no asbestos discharges
pertaining to the COUs were reported, and no specific asbestos violations were reported. None of the
industrial permits pertaining to the COUs (i.e., chlor-alkali and sheet gasket facilities) had requirements
to monitor for asbestos. No violation of TSS standards or pH standards were reported.
EPA reports asbestos levels in drinking water from compliance monitoring data from 1998 through 2011
in two separate six year review cycles (see Table 2-1). However, these data cannot be traced to a specific
COU in this Part 1 of the risk evaluation for asbestos. In addition, the data are from public water
supplies and most likely represent samples from finished drinking water (i.e., tap water) or some other
representation that may not reflect the environment in which ecological organisms exist. For these two
reasons, these data may not be relevant in assessing the environmental release pathway.
Table 2-1. EPA OW Six Year Review Cycle Data for Asbestos in Drinking Water, 1998-2011
Re\iew Cycle
\limber of Systems
Sampled
Number of Systems with
Detections Minimum Reporting
l.exel (MRI. of" 2 Ml 1.)
Number of Systems with
Detections the MCI. of 7
Ml 1.
1998-2005
8,278
268 (3.2%)
14 (<0.2%)
2006-2011
5,785
214(3.7%)
8 (<0.2%)
MRL = Minimum Reporting Level
MFL = million fibers per liter
MCL = maximum contaminant level
2.2.2 Water Releases Reported by Conditions of Use
2.2.2.1 Processing and Industrial Use of Chrysotile Asbestos Diaphragms in Chlor-
alkali Industry
As noted in the PF, EPA staff visited two separate chlor-alkali facilities in March of 2017 to better
understand how chrysotile asbestos diaphragms are used, managed and disposed of. The American
Chemistry Council (ACC) provided a process description of on-site wastewater treatment methods
employed by chlor-alkali facilities to manage and treat wastewater based on their NPDES permits. Some
companies in the chlor-alkali industry are known to collect all used diaphragms, hydroblast the asbestos
off the screen on which the diaphragm is formed, and filter press the asbestos-containing wastewater.
This water in these cases is collected to a sump, agitated, and transferred to a filter press. The filter press
contains multiple filter plates with polypropylene filter elements (8 to 100 |im pore size). After solids
separation, the filters are removed to large sacks for disposal to a landfill that accepts asbestos-
containing waste per federal and state asbestos disposal regulations. The effluent is filtered again and
discharged to the facility's wastewater collection and treatment system (See Attachment B in ACC
Submission). Asbestos releases from chlor-alkali facility treatment systems to surface water and POTWs
are not known. While the treatment technologies employed would be expected to capture asbestos
solids, the precise treatment efficiency is not known. Chlor-alkali facilities are not required to monitor
Page 67 of 352

-------
effluents for asbestos releases, and EPA's broader research into this COU did not find asbestos water
release data. EPA acknowledges there is some uncertainty in this conclusion in the absence of
monitoring data to confirm assumptions; however, EPA believes this uncertainty is low.
Another data source considered for asbestos water releases from chlor-alkali facilities was the TRI.
According to the TRI reporting requirements, facilities are required to disclose asbestos waste
management practices and releases only for the portion of asbestos that is friable. TRI reporting is not
required for other forms of asbestos (e.g., non-friable asbestos, asbestos in aqueous solutions) (U.S.
). Consistent with this qualification in the TRI reporting requirements, no chlor-alkali
facilities reported asbestos surface water discharges to TRI in reporting year 2018. All chlor-alkali
facilities reported zero surface water discharges and zero off-site transfers for wastewater treatment.
2.2.2.2 Processing Chrysotile Asbestos-Containing Sheet Gaskets
Based on reasonably available process information provided during an EPA site visit, sheet gasket
stamping occurs in a warehouse setting with stamping machines (Branham email(s) and observations
during August 2, 2018 plant visit to Gulfport, MS) (Branham. 2018). The warehouse has no industrial
wastewater or water systems, except for potable uses. Housekeeping practices used in relevant work
areas at the facility EPA visited included a weekly "wipe-down" of equipment (e.g., machine presses,
dies) and workstations (e.g., tabletops) with damp rags, which were disposed of with asbestos-
containing gasket scraps. This waste was double bagged, sealed, labeled as asbestos, placed in special
container, and disposed in a landfill permitted to accept asbestos wastes. This company has two sites and
does not report to TRI for friable asbestos and does not have NPDES permits.
EPA attempted to identify other companies that fabricate asbestos-containing sheet gaskets in the United
States but could not locate any. Therefore, it is not known how many sites fabricate imported sheet
gaskets containing asbestos in the United States. If other companies stamp gaskets in the same way that
EPA observed at one facility, it could then be assumed that there will not be water releases. However, it
is not possible to rule out incidental releases of asbestos fibers in wastewater at other fabrication
facilities if different methods are used, but any amounts of release cannot be quantified.
2.2.2.3 Industrial Use of Sheet Gaskets at Chemical Production Plants
Based on reasonably available process information for the titanium dioxide (TiCh) production facility -
the example used in this Part 1 of the risk evaluation for asbestos for chemical production plants -
described by ACC (ACC. ^ ) and EPA knowledge of the titanium manufacturing process, the
purpose of the gasket is to seal equipment components. The information indicates that after maintenance
workers remove a gasket from a flange, he or she will double-bag and seal the gasket and label the bag
"asbestos," and place it in special containers for disposal in a landfill permitted to accept asbestos
wastes. It appears that there are no water releases during use of asbestos gaskets or disposal, and water is
not used as an exposure control method; therefore, releases to water are not anticipated. However, it is
not possible to rule out incidental releases of asbestos fibers in wastewater at other facilities if different
methods are used, but any amounts of release cannot be quantified.
2.2.2.4 Industrial Use and Disposal of Chrysotile Asbestos-Containing Brake Blocks in
Oil Industry
EPA attempted to evaluate potential water releases of asbestos from use in oil field brake blocks. EPA
found no reasonably available data or publications documenting asbestos releases from the use of oil
field brake blocks to water. The only relevant information obtained was an industry contact's remark
Page 68 of 352

-------
that workers wash down drawworks before removing used brake blocks and installing new ones (Popik.
2018) - a comment that suggests some asbestos fibers may be released into water during this practice.
The TRI reporting requirements do not apply to the three NAICS codes believed to best represent the
industries that use oil field brake blocks. No other reasonably available data, such as relevant sampling
data, publications, or other quantitative insights were found to inform the release assessment. The
reasonably available information currently available for this COU is insufficient for deriving water
release estimates.
Regarding solid waste, used brake blocks are replaced when worn down to 0.375-inch thickness at any
point. Because the remaining portions of the used blocks still contain asbestos, they will be handed as
solid waste and are likely handled similarly to used asbestos-containing sheet gaskets: bagged and sent
to landfills permitted to accept asbestos waste. The Safety Data Sheet (SDS) obtained for asbestos-
containing brake blocks includes waste disposal. It suggests associated waste should be sent to landfills
(Stewart & Stevenson. 2000). While asbestos in these brake blocks are generally considered non-friable
when intact, it is unclear if the asbestos in the used brake blocks is friable or remains non-friable.
2,2,2,5 Commercial Use, Consumer Use, and Disposal of Aftermarket Automotive
Chrysotile Asbestos-Containing Brakes/Linings, Other Vehicle Friction Products, and
Other Chrysotile Asbestos-Containing Gaskets
EPA determined that water releases for aftermarket chrysotile asbestos-containing automotive parts
(brakes, clutches, gaskets, utility vehicle (UTV) gaskets) do not involve the use of water during the
removal and clean up. EPA has not identified peer-reviewed publications that measure water releases of
asbestos associated with processing, using, or disposing of aftermarket automotive products.
2.2.3 Summary of Water Releases and Exposures
During the PF phase of the RE, EPA was still in the process of identifying potential asbestos water
releases for the TSCA COUs in this document. After the PF was released, EPA continued to search EPA
databases as well as the literature and attempted to contact industries to shed light on potential releases
to water. Very little information was located that indicated that there were surface water releases of
asbestos; however, it is unclear of the source of the asbestos and the fiber type present. In the draft Risk
Evaluation, EPA concluded that, based on the reasonably available information in the published
literature, provided by industries using asbestos, and reported in EPA databases, there were minimal or
no releases of asbestos to surface water associated with the COUs that EPA is evaluating (see Appendix
D). EPA has considered peer review and public comments on this conclusion and has decided to keep
the finding made in the draft Risk Evaluation {i.e., that there were minimal or no releases of asbestos to
surface water associated with the COUs that EPA is evaluating in this Part 1 of the risk evaluation for
asbestos). This is because EPA is confident that the minimal water release data available cannot be
attributed to chrysotile asbestos from the COUs in this Part 1 of the risk evaluation for asbestos.
Assessing possible risk to aquatic organisms from the exposures described would not be reasonably
attributed to the COUs. However, based on the decision to develop a scope and risk evaluation for
legacy uses and associated disposals of asbestos (Part 2 of the final Risk Evaluation for asbestos), EPA
expects to address the issue of releases to surface water based on those other uses (See Section 4.1).
Page 69 of 352

-------
2.3 Human Exposures
EPA evaluated both occupational and consumer scenarios for each COU. The following table provides a
description of the COUs and the scenario (occupational or consumer) evaluated in this Part 1 of the risk
evaluation for asbestos.
Table 2-2. Crosswalk of Conditions of Use and Occupational and Consumer Scenarios Assessed in
	the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos	
COU
Scenario
Form of Chrysotile Asbestos1
Diaphragms for Chlor-Alkali
Industry (Processing and Use)
Occupational
Imported raw asbestos (used to fabricate
diaphragms)
Brake Block Use (Use)
Occupational
Imported article
Sheet Gaskets
Stamping (Processing)
Occupational
Imported sheets
Sheet Gaskets
In chemical production (Use)
Occupational
Gaskets imported or purchased in US
Brakes
Installation in exported cars (Use)
Occupational
Imported brakes
Brakes
Repair/replacement (Use and
Disposal)
Occupational (repair
shops)
Imported brakes
Brakes
Repair/replacement (Use and
Disposal)
Consumer (DIY)
Imported brakes
UTV Gaskets
Manufacture UTV in US (Use
and Disposal)
Occupational
Imported gaskets
UTV Gaskets
Repair/replacement (Use and
Disposal)
Occupational (repair
shops)
Imported gaskets
UTV Gaskets
Repair/replacement (Use and
Disposal)
Consumer (DIY)
Imported gaskets
1 EPA understands that, with the exception of chlor-alkali and possibly sheet gaskets, these products
could be purchased through the internet.
2.3.1 Occupational Exposures
For the purposes of this assessment, EPA considered occupational exposure of the total workforce of
exposed users and non-users, which include, but are not limited to, male and female workers who are
>16 years of age. This section summarizes the key occupational acute and chronic inhalation exposure
concentrations for asbestos.
Page 70 of 352

-------
EPA only evaluated inhalation exposures to workers and occupational non-users (ONUs) in association
with chrysotile asbestos manufacturing (import), processing, distribution and use in industrial
applications and products in this Part 1 of the risk evaluation for asbestos. The physical condition of
chrysotile asbestos is an important factor when considering the potential human pathways of exposure.
Several of the asbestos-containing products identified as COUs of asbestos are not friable as intact
products; however, the products can be made friable due to physical and chemical wear over time.
Exposures to asbestos can potentially occur via all routes; however, EPA anticipates that the most likely
exposure route is inhalation for workers and ONUs. ONUs do not directly handle asbestos or asbestos-
containing products but are present during their work time in an area where asbestos or an asbestos-
containing product is or may be present.
Where available, EPA used inhalation monitoring data from industry, trade associations, or the public
literature. For each COU, EPA separately evaluates exposures for workers and ONUs. A primary
difference between workers and ONUs is that workers may handle chemical substances and have direct
contact with chemicals, while ONUs are working in the general vicinity but do not handle the chemical
substance. Examples of ONUs include supervisors/managers, and maintenance and janitorial workers
who might access the work area but do not perform tasks directly with chrysotile asbestos or chrysotile
asbestos containing products. For inhalation exposure, in cases where no ONU sampling data are
available, EPA typically assumes that ONU inhalation exposure is comparable to area monitoring results
that may be available or assumes that ONU exposure is likely lower than workers.
EPA considered two issues unique to asbestos, when compared to other chemicals for which EPA has
developed TSCA risk evaluations. One issue is the possibility of asbestos fibers settling to surfaces and
subsequently becoming resuspended into the workplace air. The extent to which this process occurs is
presumably reflected in the sampling data that EPA considered for each COU. The second unique issue
for asbestos is that it can be found in friable and non-friable materials; and the friability of the materials
has direct bearing on asbestos releases to the air. This issue is also presumably reflected in the sampling
data (i.e., asbestos in friable materials has a greater likelihood of being detected in the air samples, as
compared to asbestos in non-friable materials).
Components of the Occupational Exposure Assessment
The occupational exposure assessment of each COU comprises the following components:
•	Process Description: A description of the COU, including the role of asbestos in the use;
process vessels, equipment, and tools used during the COU; and descriptions of the worker
activities, including an assessment for potential points of worker exposure.
•	Worker Activities: Activities in which workers may be potentially exposed to asbestos.
•	Number of Sites and Potentially Exposed Workers: Estimated number of sites that use
asbestos for the given COU; estimated number of workers, including ONUs, who could
potentially be exposed to asbestos for the given COU.
•	Occupational Inhalation Exposure Results: EPA used exposure monitoring data provided by
industry, when it was available, to assess occupational inhalation exposures. EPA also
considered worker exposure monitoring data published in the peer-reviewed literature. In all
cases, EPA synthesized the reasonably available information and considered limitations
associated with each data set. Later in this section, EPA reports central tendency and high-end
estimates for exposure distribution derived for workers and for ONUs for each COU and
Page 71 of 352

-------
acknowledges the limitations associated with these exposure estimates.
• Inhalation Exposure Results for Use in the Part 1 of the risk evaluation for Asbestos:
Central tendency and high-end estimates of inhalation exposure to workers and ONUs.
2,3,1.1 Occupational Exposures Approach and Methodology
EPA reviewed reasonably available information from OSHA, NIOSH, the peer-reviewed literature,
industries using asbestos or asbestos-containing products, and trade associations that represent this
industry (e.g., ACC) to identify relevant occupational inhalation exposure data. The data provided by
OSHA included sampling results in states with federal OSHA oversight; data from "state plan states"
were not included in OSHA's data submittal to EPA. Quantitative data obtained during Systematic
Review were used to build appropriate exposure scenarios when monitoring data were not reasonably
available to develop exposure estimates. For uses with limited available exposure data the assessment
used similar occupational data and best professional judgment to estimate exposures. In these cases,
EPA used assumptions to evaluate risk.
General Inhalation Exposures Approach and Methodology
EPA provided occupational exposure results for each COU that were representative of central tendency
estimates and high-end estimates when possible. A central tendency estimate was assumed to be
representative of occupational exposures in the center of the distribution for a given COU. EPA's
preference was to use the 50th percentile of the distribution of inhalation exposure data as the central
tendency. In cases where other approaches were used, the text describes the rationale for doing so. EPA
provided high-end estimates at the 95th percentile. If the 95th percentile was not available, or if the full
distribution was not known and the preferred statistics were not available, EPA used a reported
maximum value to represent the high-end estimate.
2.3.1.2 Consideration of Engineering Controls and Personal Protective Equipment
OSHA requires employers to utilize the hierarchy of controls to address hazardous exposures in the
workplace. The hierarchy of controls prioritizes the most effective measures to address exposure; the
first of which is to eliminate or substitute the harmful chemical (e.g., use a different process, substitute
with a less hazardous material), thereby preventing or reducing exposure potential. Following
elimination and substitution, the hierarchy prioritizes engineering controls to isolate employees from the
hazard (e.g., source enclosure, local exhaust ventilation systems), followed by administrative controls, or
changes in work practices to reduce exposure potential. Administrative controls are policies and
procedures instituted and overseen by the employer to prevent worker exposures. As the last means of
control, the use of personal protective equipment (PPE) (e.g., respirators, gloves) is required, when the
other feasible control measures cannot reduce workplace exposure to an acceptable level.
OSHA Respiratory Protection and Asbestos Standards
OSHA has standards that are applicable to occupational exposure to asbestos including the Respiratory
Protection Standard (29 CFR ง 1910.134); and the Asbestos Standard for general industry (29 CFR ง
1910.1001) construction (29 CFR ง 1926.1101), and shipyards (29 CFR ง 1915.1001). These standards
have multiple provisions that are highlighted below.
OSHA's Respiratory Protection Standard (29 CFR ง 1910.134) requires employers to provide
respiratory protection whenever it is necessary to protect the health of the employee from contaminated
Page 72 of 352

-------
or oxygen deficient air. This includes situations where respirators are necessary to protect employees in
an emergency. Employers must follow the hierarchy of controls which requires the use of engineering
and work practice controls where feasible. Only if such controls are not feasible or while they are being
implemented may an employer rely on a respirator to protect employees. Respirator selection provisions
are provided in ง 1910.134(d) and require that appropriate respirators be 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. Assigned protection factors (APFs) are provided in Table 1 under
ง 1910.134(d)(3)(i)(A) (see below in Table 2-3.). APFs refer to the level of respiratory protection that a
respirator or class of respirators is expected to provide to employees when the employer implements a
continuing, effective respiratory protection program.
Table 2-3. Assigned Protection Factors for Respirators in OSHA Standard 29 CFR 1910.134eg
Type of Uespiralor11-h
Quarter
Mask
Half Mask
lull
Kacepiece
llelmel/
Mood
l.oose-filling
l-'acepiece
1. Air-Purifying Respirator
5
10 c
50


2. Powered Air-Purifying
Respirator (PAPR)

50
1,000
25/1,000 d
25
3. Supplied-Air Respirator (SAR) or Airline Respirator
• Demand mode

10f
50


• Continuous flow mode

50 f
1,000
25/1,000 d
25
• Pressure-demand or other
positive-pressure mode

50 f
1,000


4. Self-Contained Breathing Apparatus (SCBA)
• Demand mode

10f
50
50

• Pressure-demand or other
positive-pressure mode (e.g.,
open/closed circuit)


10,000
10,000

a Employers may select respirators assigned for use in higher workplace concentrations of a hazardous substance for use at
lower concentrations of that substance, or when required respirator use is independent of concentration.
b The assigned protection factors are only effective when the employer implements a continuing, effective respirator program
as required by 29 CFR ง 1910.134, including training, fit testing, maintenance, and use requirements.
0 This APF category includes filtering facepieces and half masks with elastomeric facepieces.
d The employer must have evidence provided by the respirator manufacturer that testing of these respirators demonstrates
performance at a level of protection of 1,000 or greater to receive an APF of 1,000. This level of performance can best be
demonstrated by performing a workplace protection factor (WPF) or simulated workplace protection factor (SWPF) study or
equivalent testing. Absent such testing, all other PAPRs and SARs with helmets/hoods are to be treated as loose-fitting
facepiece respirators and receive an APF of 25.
e These APFs do not apply to respirators used solely for escape. For escape respirators used in association with specific
substances covered by 29 CFR ง 1910 subpart Z, employers must refer to the appropriate substance-specific standards in that
subpart. Escape respirators for other IDLH atmospheres are specified by 29 CFR ง 1910.134(d)(2)(ii).
These respirators are not common.
8 Respirators with bolded APFs satisfy the OSHA requirements for asbestos and an appropriate respirator should be selected
based on the air concentration. Filtering facepiece respirators do not satisfy OSHA requirements for protection against
asbestos fiber.
Page 73 of 352

-------
OSHA's asbestos standards also include respiratory protection provisions found at 29 CFR ง
1910.1001(g) for general industry, 29 CFR ง 1926.1101(h) for construction, and 29 CFR ง 1915.1001(g)
for shipyards. The respiratory protection provisions in these standards require employers to provide each
employees an appropriate respirator that complies with the requirements outlined in the provision. In the
general industry standard, paragraph (g)(2)(ii) requires employers to provide an employee with a tight-
fitting, powered air-purifying respirator (PAPR) instead of a negative pressure respirator selected
according to paragraph (g)(3) when the employee chooses to use a PAPR and it provides adequate
protection to the employee. In addition, paragraph (g)(3) of the general industry standard states that
employers must not select or use filtering facepiece respirators for protection against asbestos fibers.
Therefore, filtering facepieces (N95), quarter masks, helmets, hoods, and loose fitting facepieces should
not be used. OSHA's 29 CFR ง 1910.1001 (g)(3)(ii) also indicates that high-efficiency particulate air
(HEPA) filters for PAPR and non-powered air-purifying respirators should be provided.
APFs are intended to guide the selection of an appropriate class of respirators to protect workers after a
substance is determined to be hazardous, after an occupational exposure limit is established, and only
when the occupational exposure limit is exceeded after feasible engineering, work practice, and
administrative controls have been put in place. For asbestos, the employee permissible exposure limit
(PEL) is O.lfibers per cubic centimeter (f/cc) as an 8-hour, time-weighted average (TWA) and/or the
excursion limit of l.Of/cc averaged over a sampling period of 30 minutes.
Using the OSHA PEL for asbestos of 0.1 f/cc, a half-mask negative pressure HEPA filtered facepiece
(when fitted properly) can provide protection in atmospheres with up to 1.0 f/cc [0.1 f/cc multiplied by
the APF of 10],
Only the respirator types and corresponding APFs bolded in Table 2-3. meet the OSHA requirements
for asbestos. The specific respiratory protection required in any situation is selected based on air
monitoring data. OSHA specifies that the Maximum Use Concentration (MUC) be calculated to assess
respirator selection. The MUC is the maximum amount of asbestos that a respirator can handle from
which an employee can be expected to be protected when wearing a respirator. The APF of the
respirator or class of respirators is the amount of protection that it provides the worker compared to not
wearing a respirator. The permissible exposure limit for asbestos (0.1 f/cc) sets the threshold for
respirator requirements. The MUC can be determined by multiplying the APF specified for a respirator
by the OSHA PEL, short-term exposure limit, or ceiling limit.
The APFs are not assumed to be interchangeable for any COU, any workplace, or any worker. The use
of a respirator would not necessarily resolve inhalation exposures if the industrial hygiene program in
place is poorly maintained. An inadequate respiratory protection program could lead to inadequate
respirator fit tests and poor maintenance of respirators which could affect AFP. Table 2-3. can be used
as a guide to show the protectiveness of each category of respirator. Based on the APFs specifically
identified for asbestos and presented in Table 2-3, inhalation exposures may be reduced by a factor of 10
to 10,000 assuming employers institute a comprehensive respiratory protection program.
However, for asbestos, nominal APFs in Table 2-3 may not be achieved for all PPE users (Riala and
Riipimem. 1998). investigated performance of respirators and HEPA units in 21 different exposure
abatement scenarios; most involved very high exposures not consistent with COUs identified in this RE.
However, for three abatement scenarios, exposure concentrations were below 1 f/cc, which is relevant to
the COUs in this draft risk evaluation. In the three scenarios, actual APFs were reported as 50, 5, and 4.
The strength of this publication is the reporting of asbestos samples inside the mask, use of worker's
own protective equipment, and measurement in different real work conditions. The results demonstrate
that while some workers have protection above nominal APF, some workers have protection below
Page 74 of 352

-------
nominal APF, so even with every worker wearing a respirator, some of these workers would not be
protected.
2.3.1.3 Chlor-Alkali Industry
This section reviews the presence of chrysotile asbestos in semi-permeable diaphragms used in the
chlor-alkali industry and evaluates the potential for worker exposure to asbestos.
2.3.1.3.1 Process Description - Asbestos Diaphragms
Asbestos (raw chrysotile asbestos) is used in the chlor-alkali industry for the fabrication of semi-
permeable diaphragms, which are used in the production of chlorine and sodium hydroxide (caustic
soda). Because it is chemically inert and able to effectively separate the anode and cathode chemicals in
electrolytic cells (tJSGS. 2017). the incorporation of asbestos can be viewed as vital. Figure 2-1. below
shows a typical diaphragm after it has been formed.	

Figure 2-1. Closeup of a Chrysotile Asbestos Diaphragm Outside of the Electrolytic Cell
Photograph Courtesy of the American Chemistry Council
Chlor-alkali industry representatives have stated that three companies own a total of 15 chlor-alkali
facilities in the United States that use asbestos-containing semi-permeable diaphragms onsite. Some of
these facilities fabricate diaphragms onsite from asbestos, and other facilities receive fabricated
diaphragms from other chlor-alkali facilities and send them back when the diaphragms reach the end of
service life. EPA does not expect exposures to occur when handling fabricated diaphragms. Based on
information provided by ACC, the management of asbestos in the chlor-alkali industry is performed in a
closely controlled process from its entry into a port in the United States through all subsequent uses.
ACC reports that engineering controls, PPE, employee training, medical surveillance, and personal
monitoring are all used to monitor and mitigate worker exposures ACC submission, see Enclosure C).
The remainder of this section is based on a description of the chlor-alkali diaphragm manufacturing
process and associated asbestos controls. ACC provided this information to EPA, and it is included in
the docket (ACC Submission). Unless otherwise specified, all process details presented in the following
paragraphs are based on this docket submission. In addition, in 2017 EPA engineers conducted site visits
to two chlor-alkali facilities. During these site visits, the observations by EPA engineers were generally
consistent with details of the process descriptions provided by industry and described below. Other
citations are included in the following paragraphs only for specific details not covered in the main
docket reference ( ;PA-HO-OPPT-2016-0763-0Q52).
Page 75 of 352

-------
After arriving at the plant, the shipping container with raw chrysotile asbestos is inspected, and any
damaged containers are shipped back to the sender. Port and warehouse workers manage and remediate
any damaged containers in conformance with OSHA's asbestos standard for general industry, which
includes requirements for PPE and respiratory protection (as described above in Section 2.3.1.2).
Chrysotile asbestos within the containers is sealed in bags, and workers' first task after opening the
containers is to inspect bags for leaks. If bags are broken or loose asbestos is evident, the area is
controlled to prevent accidental exposure, the bags are repaired, and the location is barricaded and
treated as an area requiring cleanup; workers involved in this activity wear PPE and use respiratory
protection, per requirements in OSHA's asbestos standard. Plastic-wrapped pallets are labeled per
OSHA's hazard communication and asbestos standards. Any loose chrysotile asbestos from punctured
bags inside the container is collected using HEPA-filtered vacuum cleaners or wetted with water and
cleaned up before unloading can proceed. Damaged bags are repaired or placed in appropriately labeled,
heavy-duty plastic bags. Workers not involved in cleanup are prohibited from entering the area until
cleanup is complete. When moving the chrysotile asbestos bags into storage locations, care is taken to
ensure that bags are not punctured, and personnel moving the bags wear specific PPE, including
respirators. Storage areas are isolated, enclosed, labeled, secured and routinely inspected. Any area or
surface with evidence of chrysotile asbestos is cleaned by a HEPA-filtered vacuum or wetted and
cleaned up by trained employees wearing PPE.
To create chrysotile asbestos-containing diaphragm cells, sealed bags of chrysotile asbestos are opened,
and the asbestos is transferred to a mixing tank. At some plants, this process is fully automated and
enclosed, in which the sealed bags of chrysotile asbestos are placed on a belt conveyor. The conveyor
transfers the sealed bag to an enclosure above a mixing vessel. Mechanical knives cut open the bag, and
the asbestos and bag remnants fall via a chute into the mixing vessel. In other cases, opening of the
sealed bags takes place in glove boxes. Empty bags are placed into closed and labeled waste containers,
either through a port in the glove box or during the automated process. The glove boxes are sealed
containers with gloves built into the side walls, which allow workers to manipulate objects inside while
preventing any exposure from occurring. Glove boxes also allow workers to open sealed bags and
transfer chrysotile asbestos to a mixing tank via a closed system maintained under vacuum.
Once in the mixing vessel, the raw chrysotile asbestos used to create a diaphragm is blended with a
liquid solution of weak caustic soda and salt, thus forming a chrysotile asbestos slurry. Modifiers (e.g.,
Halarฎ, Teflonฎ) are added to the slurry. Figure 2-2. shows a process flow diagram of an example
glove-box-based asbestos handling system and slurry mix tank.
Page 76 of 352

-------
Representative Asbestos Handling Process Flow Diagram
PRE
FILTER
HEPA
FILTER
INLET
AIR
CHAMBER nc
sic
BLOWER
H20
ADDITIVE
ADDITIVE
MIX
CHAMBER
HEPA
FILTER
WEIGH SCALE
GLOVE BOX
GLOVE BOX
WASTE
BAG
LIQUOR
ADDITION
SLURRY
TANK
ASBESTOS . | R-
MIX TANK i1^1 F
ASBESTOS CLEAN ROOM
OVERFLOW TO
CELL RENEWAL
WASH WATER SUMP
ASBESTOS HANDLING SYSTEM
Figure 2-2. Process Flow Diagram of an Asbestos Handling System and Slurry Mix Tank Image
Courtesy of the American Chemistry Council
Source: EPA-HQ-QPPT-2016-0736-0106
The chrysotile asbestos slurry is deposited onto a metallic screen or perforated plate to form the
diaphragm, using a vacuum to evenly apply the slurry across the screen or plate. The diaphragm is
drained to remove unbound (free) water and then placed in an oven to dry and harden the asbestos. The
modifiers sinter and fuse to the asbestos, and the asbestos fuses to the screen or plate; and the product
materia] is non-friable. After cooling, the diaphragm is installed in the electrolytic cell.
The amount of asbestos used for each diaphragm ranges from 50 to 250 pounds (depending on cell size)
and a typical chlor-alkali facility will use about 5 to 25 tons of raw asbestos per year. Industry
representatives stated during meetings with EPA that a standard-sized manufacturing cell has a surface
Page 77 of 352

-------
area of 70 rrr and each cell typically has 20 chrysotile asbestos diaphragms within it, although cell sizes
vary EPA Preliminary Information).
The chlor-alkali chemical producti on process involves the separation of the sodium and chloride atoms
of salt in saltwater (brine) via electricity to produce sodium hydroxide (caustic soda), hydrogen, and
chlorine. This reaction occurs in an electrolytic cell. The cell contains two compartments separated by a
semi-permeable diaphragm, which is made mostly of chrysotile asbestos. The diaphragm prevents the
reaction of the caustic soda with the chlorine and allows for the separation of both materials for further
processing.
The cell will typically operate for one to three years before it must be replaced due to a loss of
conductivity. Many factors can determine the life of a cell, including the brine quality and the cell size.
During the March 2017 site visit, EPA learned that at least one facility bags and discards the whole
diaphragm apparatus. However, other chlor-alkali facilities reuse parts of the electrolytic cell, including
the screen or plate on which the chrysotile asbestos diaphragm was formed. The spent asbestos
diaphragm is not reusable and must be hydroblasted off the screen in a cleaning bay (remaining in a wet
state) in order for the screen to be reused. The excess water used during this process is filtered prior to
discharge to the facility's wastewater collection and treatment system. The filtered waste is placed into
containers, sealed, and sent to a landfill that accepts asbestos-containing waste per federal and state
asbestos disposal regulations EPA Preliminary Information). Figure 2-3. illustrates components and
construction of an electrolytic cell.
Top Gasket
Diaphragm
Cathode
Base Assembly
Figure 2-3. Electrolytic Cell Construction
Image courtesy of the American Chemistry Council
Source: (See Enclosure B)
2.3.1.3.2 Worker Activities - Asbestos Diaphragms
Workers may be potentially exposed to asbestos during various activities associated with constructing,
using, and deconstructing asbestos diaphragms, including:
•	Inspecting or handling broken bags
•	Remediating loose asbestos inside the shipping container
Page 78 of 352

-------
•	Opening the bag and handling raw asbestos
•	Preparing the diaphragm using asbestos slurry
•	Installing the diaphragm in an electrolytic cell (assembly)
•	Maintaining the electrolytic cells
•	Removing, dismantling, and hydroblasting diaphragms
Based on information provided by industry, when receiving and unloading bags at the facility, workers
may be protected through the use of PPE, including respiratory protection (e.g., half-mask respirator
with HEPA filters), work gloves, and disposable particulate suits (See Enclosure C). Industry reports
that chlor-alkali facilities rarely receive damaged bags of chrysotile asbestos. According to Occidental,
the last time the company's facilities reported receiving a broken bag was between 4 and 10 years ago
and the range in this estimate reflects observations from different Occidental facilities (Occidental.
Volume 2. p. 27).
As noted previously, some facilities have fully automated and enclosed systems for transferring sealed
bags of asbestos to mixing vessels. However, some chlor-alkali facilities transfer materials to a glovebox
for weighing operations, during which workers typically wear PAPRs, gloves, and disposable particulate
suits (See Enclosure €). The specific practices for loading dry asbestos from 40-kg bags into the
glovebox have not been provided to EPA and likely vary depending on the facility and the glovebox
configuration. While some gloveboxes are designed to form a seal with drum-sized product containers,
others may require open handling to load the material from the bulk bag into the glovebox.
Slurry preparation involves enclosed processes and wet methods, which minimize airborne exposure
potential. Because this is a wet process, workers typically wear gloves and boots with disposable
particulate suits, but do not wear respirators even though the short-term (15-minute sampling time)
ambient air concentrations were reported to be 0.02 f/cc at 50th percentile and as high as 0.04 f/cc (See
Enclosure €).
For preparing diaphragms, wet asbestos slurry is deposited onto diaphragm screens. One facility stated
that the wetted diaphragms are vacuum-dried before being placed in ovens to set (Axiall-Westlake.
2017). While forming the diaphragms, workers typically wear gloves and boots with disposable
particulate suits but do not wear respirators (See Enclosure €).
For cell assembly, the diaphragm is reported to be non-friable (See Enclosure C) , thereby eliminating
exposure potential. Workers typically wear impermeable gloves and boots but do not wear respirators
(See Enclosure €). Following cell assembly, the diaphragm is inspected and then joined with other parts
to complete the electrolytic cell. The short-term (15-minute sampling time) ambient air concentrations
for this process were reported to be as high as 0.154 f/cc (See Enclosure €).
Once the diaphragm is in the cell for use in the electrolytic chlor-alkali production process, asbestos
exposure from the diaphragms is not expected to occur because the cells are sealed throughout
production.
Chlor-alkali facilities use different practices for handling used diaphragms. Some facilities recondition
their own diaphragms; some facilities send their used diaphragms to other facilities for reconditioning;
and other facilities dispose of used diaphragms and do not recondition them. At the facilities that do
perform reconditioning, worker cell repair activities involve disassembling cells and then hydroblasting
diaphragms to remove the asbestos coating. For disassembly, workers typically wear impermeable
gloves, boots, goggles, and disposable particulate suits but do not wear respirators even though the short
Page 79 of 352

-------
term (15-minute sampling time) ambient air concentrations were reported to be 0.016 f/cc at 50th
percentile and as high as 0.45 f/cc (See Enclosure €). For hydroblasting, workers wear a supplied air
respirator hood, a waterproof suit, impermeable gloves, and boots (See Enclosure C). This activity
occurs in blasting rooms, and workers (while wearing PPE) may be present in these rooms during
hydroblasting activity (Axiall-Westlake., 2017).
For one site EPA visited, the hydroblasting itself was not enclosed but was conducted in a dedicated
area. The asbestos handling area (slurry mixing, oven, diaphragm disassembly, and hydroblasting area)
was walled off on three sides with a series of giant pull down doors. The fourth side wall did not extend
to the ceiling. The layout of such areas may be different at other sites. The frequency with which
workers conduct hydroblasting varies from one facility to the next. Some facilities do not hydroblast
spent diaphragms at all; while others may conduct this activity up to five times per week, with each
hydroblasting event lasting up to 90 minutes.
Wastewater from hydroblasting is filter pressed to remove asbestos before discharge from the facility.
Workers who perform this task typically wear impermeable gloves, boots, and disposable particulate
suits but do not wear respirators even though the short term (15-minute sampling time) ambient air
concentrations were reported to be 0.0275 f/cc at 50th percentile and as high as 0.2 f/cc (See Enclosure €
). Filters with filter cakes are then removed from the plate press and bagged for disposal. Additionally,
two specific practices are expected to minimize workers' asbestos exposures while completing this
disposal activity: (1) all workers who handle wastes wear PPE, including respirators (PAPR) and (2)
workers wet solid waste before double-bagging the waste, sealing it, and placing it in roll-off containers
for eventual transfer to an asbestos landfill (EPA-HQ-OPPT-201 * *•ซ ' 53-0478).
2.3.1.3.3 Number of Sites and Potentially Exposed Workers - Asbestos
Diaphragms
During a meeting with EPA in January 2017, industry representatives stated that in the United States,
three companies own a total of 15 chlor-alkali plants that continue to fabricate and use semipermeable
diaphragms that contain chrysotile asbestos (EPA-HQ-OPPT-2 '36-0069). These three companies
are Olin Corporation, Occidental Chemical Corporation, and Westlake Corporation. A fourth company,
Axiall Corporation, previously operated chlor-alkali facilities in the United States, but Westlake
Corporation acquired this company in 2016. Throughout this section, the companies are referred to as
Olin, Occidental, and Axiall-Westlake, with the latter referring to chlor-alkali facilities currently owned
by Westlake, which includes some facilities that were previously owned by Axiall.
To confirm this facility count, EPA reviewed two other data sources. First, EPA reviewed Chemical
Data Reporting (CDR) data. Only Olin and Axiall-Westlake reported importing asbestos in 2015. Each
company reported using asbestos at fewer than 10 sites. Second, EPA reviewed the 2017 TRI data and
identified a total of 11 chlor-alkali facilities reporting information on friable asbestos: three Olin
facilities; one Axiall-Westlake facility; and seven Occidental facilities. However, it is possible that some
of the existing chlor-alkali facilities did not have asbestos usage characteristics in 2017 that would have
triggered TRI reporting in that year. These two data sources are consistent with the finding that 15 chlor-
alkali facilities fabricate or use asbestos-containing diaphragms onsite.
In 2016 CDR, Olin reported a total of at least 25 and fewer than 50 workers who are likely exposed to
asbestos across all of the company's chlor-alkali facilities, and Axiall-Westlake reported a total of at
least 50 and fewer than 100 workers who are likely exposed to asbestos across all of the company's
chlor-alkali facilities. This results in an estimate of at least 75 (25 plus 50) and fewer than 148 (49 plus
Page 80 of 352

-------
99) workers likely exposed, although this estimate does not include Occidental facilities. As noted
previously, Occidental facilities did not report to CDR.
ACC has indicated that approximately 100 workers nationwide in the chlor-alkali industry perform daily
tasks working with and handling dry chrysotile asbestos. ACC's estimate is within the range derived
from 2016 CDR and includes Occidental facilities.
Regarding potential ONU exposure, EPA considered the fact that area restrictions and other safety
precautions adopted by the chlor-alkali industry help ensure that no ONU (other than directly exposed
workers) are near the chrysotile asbestos diaphragm fabrication processes and use (EPA-HQ-OPPT-
2016-0763-0052). However, EPA's observations during site visits suggest that chrysotile asbestos
exposure might occur to workers outside these processes. Additionally, some ONUs (e.g., janitorial
staff) may work near the asbestos diaphragm fabrication processes. For purposes of this assessment,
EPA assumes an equal number of ONUs (100) may be exposed to asbestos released from diaphragm
fabrication processes and use.
2.3.1.3.4 Occupational Inhalation Exposures - Chrysotile Asbestos
Diaphragms
To identify relevant occupational inhalation exposure data, EPA reviewed reasonably available
information from OSHA, NIOSH, the peer-reviewed literature, the chlor-alkali industry, and trade
associations that represent this industry (e.g., ACC).
Analysis of Exposed Workers
This section focuses on personal breathing zone (PBZ) data for chlor-alkali workers exposed to
chrysotile asbestos. EPA first considered the 2011 to 2016 nationwide exposure data provided by OSHA
and the history of NIOSH Health Hazard evaluations (HHEs). The OSHA data did not include any
observations from the chlor-alkali NAICS codes (i.e., 325181 for 2011 and 325180 for 2012 to 2016).
Of the NIOSH HHEs reviewed, only two were conducted at chlor-alkali facilities, but these evaluations
focused on chlorine and mercury exposures, not asbestos exposure. One NIOSH HHE considered a
facility that received disassembled diaphragms for servicing (Abundo et a |). NIOSH found that
the anodes contained 80 to 90 percent chrysotile asbestos, but the settled dusts from the electrode-
servicing facility did not have detectable asbestos. The quantitation limit for the dust sampling was not
specified. Finally, the peer-reviewed literature did not include recent quantitative reports of worker
asbestos exposures in the chlor-alkali industry.
To assess occupational inhalation exposures, EPA used exposure monitoring data provided by industry.
Data were provided by the three companies that currently use chrysotile asbestos in the United States
chlor-alkali industry. Occidental provided exposure monitoring data for six facilities for 1996 to 2016
(Oceider a. see Volume 2); Axiall-Westlake provided data for 2016 from a single facility (Axiall.
Attachments 1 and 2); and Olin provided data for 2012 to 2019 from three chlor-alkali facilities and a
fourth facility that reprocesses anodes (Olin Corp. 2017). ACC also provided data for 1996 to 2016
(ACC Data) but those data were duplicative of the data submissions from the individual companies..
EPA also reviewed information published by European Union (EU) agencies (EC. 2014; ECHA. 2014).
The limitation with these publications is that exposure data from EU facilities may not be representative
of the U.S. manufacturing environment, due to differences in process design, production levels,
ventilation practices, regulatory frameworks, and other factors.
Page 81 of 352

-------
The following tables summarize occupational exposure results of different exposure durations for the
fabrication, use, and disposal of chrysotile asbestos diaphragms in the chlor-alkali industry. The
exposure durations considered are full-shift samples, 30-minute average samples, and additional samples
of other durations. The tables summarize 759 personal breathing zone sampling results based on the
combined data from Axiall-Westlake, Occidental, and Olin; which included a numerical sample duration
for each sample. EPA designated samples with durations between 420 and 680 minutes as "full-shift,
samples," as these durations characterize workers with either 8-hour or 10-hour shifts.
For samples with results less than the limit of detection (LOD) or limit of quantitation (LOQ), surrogate
values were used based on statistical analysis guidelines for occupational exposure data that were
developed for EPA (U.S. EPA. 1994). These guidelines call for replacing non-detects with the LOD or
LOQ divided by two or divided by the square root of two, depending on the skewness of the data
distributions. EPA notes that more than half of the samples were non-detectable.
Table 2-4 and Table 2-5 provide both full-shift and short-term sample summaries. Table 2-6 summarizes
PBZ data for all other sampling durations, and Table 2-7 summarizes all short-term samples by exposure
group, with additional breakdown by task. (Note: The data in Table 2-7 were provided by ACC. These
data were not included in the tallies in the other tables, because ACC informed EPA that the data it
provided were duplicates of data from the three companies.)
Table 2-4. 30-min Short-Term PBZ Sample Summary*
Sample
Typo
Dale Range of
Samples
Number of
Samples
.Maximum
Result (f/cc)
50th Percentile
(I/CC)
95th Percentile
(f/cc)
PBZ
2001 to 2017
58
2 2**
0.024
0.512
*Data from Olin and Occidental, 47 percent of these samples were non-detects
**Note: The maximum concentration in this table (11 f/cc) was originally reported as being an "atypical result." The
employer in question required respirator use until re-sampling was performed. The follow-up sample found an exposure
concentration (0.019 f/cc) more than 500 times lower.

Table 2-5. Full-Shift* PBZ Sample
Summary**

Sample
Type
Dale Range of
Samples
Number of
Samples
.Maximum
Result (f/cc)
50th Percentile
(I/CC)
95th Percentile
(I/CC)
PBZ
1996 to 2017
357
0.41
0.0049
0.034
* Includes both 8-hr and 10-hr TWA sample results.
"Data from Axiall-Westlake, Occidental, and Olin. 57 percent of these were non-detects


Table 2-6. Summary of PBZ Sampling Data for All Other Durations*
Sample
Tvpc
Dale Range of
Samples
.Number of
Samples
.Maximum
Result (f/cc)
50"' Percentile
(I/CC)
95"' Percentile
(I/CC)
PBZ
2004 to 2019
344
1.78
0.024
0.514
'Data from Axiall-Westlake, Occidental, and Olin, 53 percent of these were non-detects
Page 82 of 352

-------
Table 2-7. Summary of ACC Short-Term PBZ Sampling Data by Exposure Group (samples from
	 2001 to 2016)			
Kxposure Group / Task .Name(s)
.Number of
Samples
.Maximum
Res u II (l/cc)
501 It
Percentile
(f/cc)
95th
Percentile
(f/cc)
Asbestos Unloading/Transport
8
0.12
0.01
0.09865
Glovebox Weighing and Asbestos




Handling
150
1.7
0.0295
0.44
Asbestos Slurry *
5
0.04
0.02
0.036
Depositing *
27
0.1
0.0125
0.0601
Cell Assembly *
31
0.077
0.012
0.0645
Cell Disassembly *
49
0.45
0.016
0.0732
Filter Press *
36
0.2
0.0275
0.1315
Hydroblasting
20
0.51
0.14
0.453
* Task-specific PPE does not include respirators (See Enclosure O
Analysis of ONUs
At chlor-alkali facilities, ONU exposures to chrysotile asbestos are expected to be limited because most
asbestos handling areas are likely designated regulated areas pursuant to the OSHA asbestos standard,
with access restricted to employees with adequate PPE. However, EPA considered the possibility of
ONU exposure when employees not engaged in asbestos-related activities work near or pass through the
regulated areas and may be exposed to asbestos fibers released into the workplace. These employees
may include maintenance and janitorial staffs.
EPA considered area monitoring data {i.e., fixed location air monitoring results) as an indicator of this
exposure potential. Across the four sampling data sets provided by industry, only the data provided by
01 in included area sampling results (Olin Corp. 2017). The area monitoring data came from Olin's
facilities located in Arkansas and Louisiana. These data include 15 full-shift asbestos samples collected
at fixed locations. The asbestos concentration levels are reported as either 0.004 f/cc [N= 11] or 0.008
f/cc [N=4], EPA has reason to believe these are all non-detect observations. The notes fields in the
sample results identified as 0.008 f/cc state "detection limit was 0.008 f/cc."
EPA followed the same approach as noted above for non-detect observations, which in this case is
replacing the observation by the limit of detection (LOD) divided by two. Therefore, for deriving
exposure estimates, the 15 area samples were assigned numerical values of 0.002 f/cc [N=l 1] and 0.004
f/cc [N=4], The central tendency ONU concentration used in EPA's analysis was 0.0025 f/cc {i.e., the
arithmetic mean of the 15 data points), and the high end ONU concentration used in EPA's analysis was
<0.008 f/cc.
2.3.1.3.5 Exposure Results for Use in the Risk Evaluation for Asbestos: Part
1- Chlor-Alkali
Table 2-8 presents chrysotile asbestos exposure data that EPA used in this Part 1 of the risk evaluation
for asbestos for workers and ONUs in the chlor-alkali industry. EPA's basis for selecting the data points
appears after the table.
Page 83 of 352

-------
Table 2-8. Summary of Chrysotile Asbestos Exposures During Processing and Use in the Chlor-
Alkali Tndustrv Fsed in F.PA's Risk Evaluation for Asbestos Part 1: Chrvsotile Asbestos
Occupational
Kxposurc Scenario
Workers
Central
Tendency
lligh-end
(95,h
percentile)
Kxposu
Confidence
Ualing
•e Levels (17
OMs
Central
Tendency
cc)
lligh-end
Confidence
Ualing
Producing,
handling, and
disposing of
asbestos
diaphragms: full-
shift TWA
exposure
0.0049
0.034
High
0.0025
0.008
Medium
Producing,
handling, and
disposing of
asbestos
diaphragms: short-
term TWA
exposure (30 mins)
0.024
0.512
High
—
—
—
"—" indicates no data reported
The data in Table 2-8 provide a summary of exposure values among workers and ONUs who produce,
handle, and dispose of chrysotile asbestos diaphragms at chlor-alkali facilities. These data represent a
complex mix of worker activities with varying asbestos exposure levels. It should be noted that not all
activities include use of respirators (Table 2-7). The data points in Table 2-8 were compiled as follows
(details presented in Supplemental File: Occupational Exposure Calculations (Chlor-Alkali) (
2.02.0b V
•	Table 2-8 lists the full-shift TWA exposure levels that EPA used in this Part 1 of the risk
evaluation for asbestos. The central tendency value for workers (0.0049 f/cc) is the median value
of the full-shift exposure samples provided by Axiall-Westlake, Olin, and Occidental, and the
high-end value (0.034 f/cc) is the calculated 95th percentile (see Table 2-5).
•	For ONU exposure estimates area samples were used. Two chlor-alkali facilities provided a total
of 15 area samples that were all below the limit of detection (LOD). There were two different
detection limits in the two submissions. Central tendency exposure concentrations were
calculated as the arithmetic mean of the individual observations, using one-half the detection
limit for individual samples; and the high-end concentration is the highest detection limit
provided.
•	The central tendency short-term TWA exposure value for workers was based on short-term (30-
minute) sampling data provided by industry. The value in Table 2-5 (0.024 f/cc) is the median
value of all 30-minute personal samples submitted. The high-end short-term TWA exposure
value for workers (0.512 f/cc) is the calculated 95th percentile value for the compiled industry
short-term exposure data. These values are based on all employee tasks combined. Refer to Table
Page 84 of 352

-------
2-7 for specific employee tasks (e.g., asbestos handling, filter press operation) with higher short-
term exposure levels.
2.3.1.3.6 Data Assumptions, Uncertainties and Level of Confidence
The exposure data shown in Table 2-8 are based on monitoring results from the chlor-alkali industry.
Worker exposure sampling data are available from all three companies (i.e., Axiall-Westlake,
Occidental, Olin) that currently operate the entire inventory of chlor-alkali facilities nationwide and the
overall confidence ratings from systematic review for these data were all rated high. Tables 2-4 through
2-7 summarize 759 individual exposure sampling results, which represent extensive coverage of the
estimated 100 directly exposed workers. Each company submission of monitoring data includes a
variety of worker activities. Therefore, this collection of monitoring data likely captures the variability
in exposures across the different chlor-alkali sites and likely captures the variability in exposures during
normal operations within a single site.
EPA notes the monitoring data cover all of the chlor-alkali companies that use chrysotile asbestos.
However, it is uncertain if some infrequent and high-exposure activities are captured in this dataset, such
as exposures when cleaning spilled asbestos within a container from damaged bags. The high-end
estimates presented in the table 2-8 are applicable to an unknown fraction of the workers.
EPA considered the quality and uncertainties of the data to determine a level of confidence for the
assessed inhalation exposures for this COU. The primary strength of this assessment is the use of
monitoring data from multiple sites, which is the highest level of the inhalation exposure assessment
approach hierarchy. One notable limitation is the considerable portion of non-detectable observations.
EPA investigated different approaches to evaluating the non-detect observations (e.g., substitution with
zero, substitution with the full detection limit) and continues to base its estimated concentrations on the
non-detect substitution methods discussed earlier in this section.
Based on these strengths and limitations of the data, the overall confidence for EPA's assessment of
occupational inhalation exposures for this scenario is high both for 8-hour and short-term durations.
For the ONU data, all of the area monitoring results showed non-detectable levels. In addition, it may be
that ONUs may be exposed at less than a full shift, every workday. Overall, there is medium confidence
for this set of data.
2.3.1,4 Sheet Gaskets
This section describes how chrysotile asbestos-containing rubberized sheeting is processed into gaskets.
2.3.1.4.1 Process Description - Sheet Gasket Stamping
Gaskets are commonly used in industry to form leakproof seals between fixed components (e.g., pipes).
Figure 2-4. shows an asbestos-containing gasket and depicts a typical gasket installation for pipe fittings.
While many asbestos-free gaskets are commercially available and widely used, asbestos-containing
gaskets continue to be the material of choice for industrial applications where gasket material is exposed
to extreme conditions such as titanium dioxide manufacturing (e.g., high temperature, high pressure,
presence of chlorine). Based on correspondence from ACC, gaskets made from non-asbestos materials
reportedly do not provide an adequate seal under these extreme conditions (ACC, 2018).
Page 85 of 352

-------
ฃ


Figure 2-4. Typical Gasket Assembly
From left to right: photograph of a gasket; illustration of a flange before gasket installation; and
illustration of a pipe and flange connection after gasket installation.
Photograph taken by EPA; Illustrations from Wikipedia.
One known company in the United States (Branham Corporation) processes (or fabricates) gaskets from
asbestos-containing rubberized sheeting. This stamping activity occurs at two Branham facilities: one in
Gulfport, Mississippi and the other in Calvert City, Kentucky. Branham imports the sheeting from a
Chinese supplier, and the sheets contain 80 percent (minimum) chrysotile asbestos encapsulated in 20
percent styrene-butadiene rubber (EPA-HQ-QPPT-2016-0736-0067). Branham supplies its finished non-
friable asbestos-containing gaskets to several customers, primarily chemical manufacturing facilities in
the United States and abroad (see Section 2.3.1.5). It is unknown if other U.S. companies import
asbestos-containing sheet material to stamp gaskets.
EPA communicated with industry to understand how Branham typically processes gaskets from
asbestos-containing sheeting. This communication includes an October 2017 meeting between EPA and
industry representatives, written communications submitted by industry representatives and ACC, and
an August 2018 EPA site visit to the Branham gasket stamping facility in Gulfport Mississippi (EPA-
HO-OPPT-2016-0736-0119). An overview of the manufacturing process follows.
Rolls of imported asbestos-containing rubberized sheeting are transported inside bolt-locked, sealed
containers from the port of entry to the Branham facilities. Branham then stores these rolls in the
original inner plastic film wrapping until use. Incoming sheets are typically 1/16-inch thick and weigh
0.6167 pounds per square foot (ACC. 2018). Branham employees stamp and cut gaskets to customer
size specifications in a production area. Various other operations occur simultaneously at the Branham
facilities to include stamping of non-asbestos gaskets using similar stamping machines. These other
operations occur approximately 20 feet away from the stamping machines used to make asbestos-
containing gaskets (EHM. 2013). As noted later in this section, EPA considers the workers supporting
other nearby operations to be ONUs for this risk evaluation.
At the Branham facility visited by EPA, workers used three stamping machines to cut the imported
asbestos-containing sheets into desired sizes. The facility reportedly does not saw gasket material
(Branham. 2018). and EPA did not see evidence of this practice during its site visit. The stamping
machines can be adjusted to make products of varying diameters, from 4 inches to 4 feet. Figure 2-5.
shows a worker wearing a face mask while operating one of the stamping machines, which uses round
Page 86 of 352

-------
headed dies attached to a blade. Blades are not removed from the dies, and the dies are seldom "re-
ruled" (where the rule blade would be pressed back into the wooden die frame).
Figure 2-5. Chrysotile Asbestos-Containing Stamping Operation
Photographs courtesy of Branham Corporation and used with Branham's permission
Figure 2-6 shows a photograph of the rule blade, which is approximately 0.010 inches thick.
Figure 2-6. Rule Blade for Stamping Machine
Photographs courtesy of Branham Corporation and used with Branham's permission
After stamping the sheet, workers place the finished gasket in individual 6-mm thick resealable bags.
These are double-bagged with a warning label and ultimately placed in a container for shipping to
customers. Figure 2-7 shows the warning label that Branham applies to asbestos-containing gasket
products.
Page 87 of 352

-------

Figure 2-7. Asbestos Warning Label on Finished Gasket Product
Photograph taken by EPA and used with Branham's permission
An important consideration for worker exposure is the extent to which sheet gasket stamping releases
asbestos-containing fibers, dusts and particles. Industry representatives have informed EPA that the
stamping process creates no visible dust, due in part to the fact that the sheet gasket material is not
friable; they also noted that asbestos fibers are and encapsulated in rubberized sheet material (ACC.
2018). This statement is consistent with EPA's observations during the site visit, in which no significant
dust accumulations were observed on or near Branham's stamping machines. However, EPA's
observations are based on a single, announced site visit. More importantly, sampling data reviewed for
this operation do indicate the presence of PCM-detected airborne asbestos. This suggests that the
stamping releases some asbestos into the workplace air.
The principal cleanup activity during the stamping operation is collection of unused chrysotile asbestos-
containing scrap sheeting, also referred to by the facility as "lattice drops." Workers manually collect
this material and place it in 6-mm thick polyethylene bags, which are then sealed in rigid containers and
shipped to the following landfills permitted to receive asbestos-containing waste (ACC, 2018. 2017b):
•	Asbestos-containing waste from Branham's Kentucky facility are transported by Branham to the
Waste Path Sanitary Landfill at 1637 Shar-Cal Road, Calvert City, Kentucky.
•	Asbestos-containing waste from Branham's Mississippi facility are transported by Team Waste
to the MacLand Disposal Center at 11300 Highway 63, Moss Point, Mississippi.
No surface wipe sampling data are available to characterize the extent of settled dust and asbestos fibers
present during this operation. The Branham facilities informed EPA that they do not use water,
including to wash away scrap or other debris or perform wet mopping, and EPA confirmed this during
the site visit. Once per week, however, workers use a damp cloth to wipe down the stamping machine
area. Spent cloths from this wiping are bagged and placed in the same rigid containers with the unused
scrap material for eventual disposal.
2.3.1.4.2 Worker Activities - Cutting of Asbestos-containing Sheet Gaskets
Worker activities most relevant to potential asbestos exposure include receiving asbestos-containing
rubber sheeting, processing gaskets by stamping, packaging finished gaskets for shipment, and
collecting asbestos containing scrap waste.
The amount of time that workers conduct cutting asbestos-containing sheets varies with production
demand and other factors. EPA received one month of worker activity data for Branham's Mississippi
Page 88 of 352

-------
facility, and these data indicated that, in May 2018, the worker spent no more than 70 minutes per day
processing asbestos-containing gaskets (Branham. 2018). Branham informed EPA that the worker at the
Kentucky facility perform asbestos-containing gasket stamping activity two to three days per week
(Branham. 2018). The worker exposure levels from the Kentucky facility will be used in Part 1 because
Branham officials informed EPA that they do not anticipate considerable increases or decreases in
production demand for asbestos-containing sheet gaskets.
Information on worker PPE use was based on photographs provided by Branham, information in facility
documents, and observations that EPA made during its site visit. When handling and stamping asbestos-
containing sheeting and when collecting scraps for disposal, the worker wears safety glasses, gloves, and
N95 disposable facepiece masks, consistent with Branham procedures (ACC. 2017a). A 2013 industrial
hygiene evaluation performed by consultants from Environmental Health Management (EHM)
concluded that measured asbestos exposures at Branham's Kentucky facility were not high enough to
require respiratory protection (EHM. 2013); however, the worker uses the N95 masks to comply with
Branham procedures.
2.3.1.4.3 Number of Sites and Potentially Exposed Workers - Sheet Gasket
Stamping
Branham operates two facilities that process asbestos-containing gaskets, with one worker at each
facility who stamps the asbestos-containing sheet gaskets. During its site visit to one facility, EPA
observed that stamping of asbestos-containing sheeting occurs in a 5,500 square foot open floor area.
Other employees work in this open space, typically at least 20 feet away from where asbestos-containing
gaskets are processed. EPA considers these other employees to be ONUs. The facility also included a
fully-enclosed air-conditioned office space, where other employees worked; but those office workers
were not considered to be ONUs.
EPA received slightly varying estimates of the number of workers at Branham's facilities and the
specific locations where they work (ACC. 2018; Branham. 2018). Based on these estimates, EPA
assumes that both facilities have one worker who processes asbestos-containing gaskets, two workers
who process other non-asbestos containing gaskets in the same open floor area (and are considered to be
ONUs), and at least two workers in the office space. Therefore, EPA assumes that asbestos-containing
gasket stamping at this company {i.e., at both facilities combined) includes two directly exposed workers
(one at each facility) and four ONUs (two at each facility).
These estimates are based on the one company known to stamp asbestos-containing sheet gaskets. It is
unknown if other U.S. companies perform this same stamping activity. EPA attempted to identify other
companies that cut/stamp asbestos-containing sheet gaskets in the United States but could not locate
any. Therefore, based on reasonably available information, EPA concluded that there are no other
additional facilities that cut or stamp imported asbestos-containing sheet gaskets.
2.3.1.4.4 Occupational Inhalation Exposure Results - Sheet Gasket Stamping
To identify relevant occupational inhalation exposure data, EPA reviewed reasonably available
information from OSHA, NIOSH, the published literature, and industry. All research steps are
documented below, with more detailed discussion on the most relevant data source, which EPA
determined was the monitoring results conducted at a Branham facility.
EPA considered the 2011 to 2016 nationwide exposure data provided by OSHA and the history of
NIOSH HHEs. EPA also considered the published literature on asbestos exposures associated with sheet
Page 89 of 352

-------
gasket stamping. This search identified two studies that presented original worker exposure monitoring
data. One was a 1998 study of sheet gasket production in Bulgaria (Strokova et a 0. The other was
a 2000 publication as part of litigation support that examined exposures in a simulated work
environment (Fowler. 2000).
EPA determined that a worker exposure monitoring study conducted at one of the Branham facilities
provides the most relevant data for this COU. Branham hired EHM consultants to conduct this study,
which involved a day of PBZ monitoring at the Kentucky facility in December 2012. The EHM
consultants measured PBZ concentrations for one worker - the worker who operated the stamping
machine to process asbestos-containing gaskets - and issued a final report of results in 2013 (EHM.
2013). The EHM consultants measured worker inhalation exposures associated with a typical day of
processing asbestos-containing gaskets and reported that samples were collected "during work periods
when the maximum fiber concentrations were expected to occur" (EHM. 2013). The EHM consultants
did not measure or characterize ONU exposures, although EPA believes that two ONUs are present at
each Branham facility where asbestos-containing sheet gaskets are processed. This determination was
based on observations that EPA made during a site visit to the Branham facility.
The EHM consultants measured worker inhalation exposure during asbestos-containing gasket stamping
operations. Ten short-term samples, all approximately 30 minutes in duration, were collected from one
worker throughout an 8-hour shift. Samples were analyzed by PCM following NIOSH Method 7400.
The short-term exposures ranged from 0.008 f/cc to 0.059 f/cc. Table 2-9. lists the individual
measurement results and corresponding sample durations. Based on the short-term results, the EHM
consultants calculated an 8-hour TWA exposure of 0.014 f/cc, which assumed no exposure during
periods without sampling. (Note: The periods without sampling appear to be the worker's break and
lunch, when exposure would be expected to be zero.)
The EHM consultants' study report includes a data summary table, which indicates that the primary
worker activity covered during the sampling was "cutting gaskets" {i.e., operation of the stamping
machines); however, the EHM consultants also acknowledged that the worker who was monitored
collected scrap material while PBZ sampling occurred (EHM.: ). EPA infers from the document that
the sampling represents conditions during a typical workday and covers multiple worker activities.
Page 90 of 352

-------
Table 2-9. Short-Term PBZ Chrysotile Asbestos Sampling Results (EHM, 2013)
Duration (minutes)
Result (f/cc)
3u
U.U59
27
0.031
36
0.020
32
0.026
29
0.028
35
0.010
40
0.018
29
0.008
30
0.008
25
0.033
2.3.1.4.5 Exposure Data for Use in the Risk Evaluation for Asbestos: Part 1 -
Chrysotile Asbestos; Sheet Gasket Stamping
Table 2-10 presents the asbestos exposure data that EPA used in this Part 1 of the risk evaluation for
asbestos for evaluating risks to workers and ONUs for the COU of processing asbestos-containing sheet
gaskets. Given the small number of sampling data points available to EPA, only central tendency and
high-end estimates are presented and other statistics for the distribution are not calculated. The
following assumptions were made in compiling these data:
•	The central tendency 8-hour TWA exposure value reported for workers (0.014 f/cc) was taken
from the single calculated value from the personal exposure monitoring study of a Branham
worker (EHM. 2013). The calculated value was derived from the ten sampling points shown in
Table 2-9., assuming no exposure occurred when sampling was not conducted.
•	The high-end 8-hour TWA exposure value for workers (0.059) is an estimate, and this full-shift
exposure level was not actually observed. This estimate assumes the highest measured short-term
exposure of the gasket stamping worker could persist for an entire day.
•	The central tendency short-term exposure value for workers (0.024 f/cc) is the arithmetic mean
of the ten short-term measurements reported in the EHM study report on the Branham worker
(EHM. 20131
•	The high-end short-term exposure value for workers (0.059 f/cc) is the highest measured short-
term exposure of the Branham worker. This exposure value occurred during a 30-minute sample
(EHM. 20131
•	EPA did not identify any ONU exposure measurements for this COU. However, the literature
includes "bystander" exposure studies that EPA could use to estimate ONU exposures.
Specifically, one publication (Mangold et al.. 2006) measured "bystander" exposure during
asbestos-containing gasket removal. The "bystander" locations in this study were between 5 and
10 feet from the gasket removal activity, and asbestos concentrations were between 2.5 and 9
Page 91 of 352

-------
times lower than those measured for the worker. Based on these observations, EPA assumes that
ONU exposures for this COU are a factor of 5.75 (i.e., the midpoint between 2.5 and 9) lower
than the directly exposed workers. This concentration reduction factor is consistent with
concentration reduction data reported in other studies in the peer-reviewed literature (e.g.,
Donovan et at..
Table 2-10. Summary of Asbestos Exposures During Sheet Gasket Stamping Used in EPA's Risk
	Evaluation for Asbestos Part 1: Chrysotile Asbestos	
Occupational Kxposure
Scenario
l-'ull-Shifl Kxposures (I'/cc)
Workers
OMs
Central
Tendency
High-
end
Confidence
Ualing
Central
Tendency
lligh-
end
Confidence
Ualing
Sheet gasket stamping: 8-hr
TWA exposure
0.014
0.059
Medium
0.0024
0.010
Medium
Sheet gasket stamping:
Short-term exposures
(approximate 30-minute
duration)
0.024
0.059
Medium
0.0042
0.010
Medium
2.3.1.4.6 Data Assumptions, Uncertainties and Confidence Level
The exposure data shown in Table 2-10 are based on 10 PBZ samples collected from one worker
performing sheet gasket stamping on a single day at a single facility. EPA used the data from this study
because it was the only study available that provided direct observations for chrysotile asbestos-
containing sheet gasket stamping operations in the United States. EPA considered the quality and
uncertainties of the data to determine a level of confidence for the assessed inhalation exposures for this
COU. The primary strength of this assessment is the use of monitoring data, which is the highest level of
the inhalation exposure assessment approach hierarchy. The overall confidence rating from systematic
review for these data was high. These monitoring data were provided to EPA by a single company that
processes chrysotile asbestos-containing sheet gaskets with data representing one of its two facilities.
However, it is not known how many companies and facilities in total process chrysotile asbestos-
containing sheet gaskets in the United States. Therefore, EPA is uncertain if these monitoring data are
representative of the entire U.S. population of workers that are potentially exposed during asbestos-
containing sheet gasket processing. The monitoring data were sampled throughout the day of the worker
performing the sheet gasket stamping; therefore, these data likely capture the variability in exposures
across the various sheet gasket stamping activities. However, it is uncertain if the single sampling day is
representative of that facility's sheet gasket stamping days throughout the year.
The ONU exposure estimate is less certain because no relevant ONU concentration estimates were
available for the Branham facilities. EPA used a concentration reduction factor approach to fill this gap.
As a result, the ONU exposure concentration estimate has greater uncertainty. In addition, ONUs may
not be exposed at full shift, every workday.
Based on these strengths and limitations of the data, the overall confidence for EPA's assessment of
occupational and ONU inhalation exposures for this scenario is medium.
Page 92 of 352

-------
2,3,1,5
Use of Gaskets in Chemical Production
2.3.1.5.1 Process Description - Sheet Gasket Use
Chrysotile asbestos-containing gaskets are used primarily in industrial applications with extreme
operating conditions, such as high temperatures, high pressures, and the presence of chlorine or other
corrosive substances. Such extreme production conditions are found in many chemical manufacturing
and processing operations. These include: the manufacture of titanium dioxide and chlorinated
hydrocarbons; polymerization reactions involving chlorinated monomers; and steam cracking at
petrochemical facilities. EPA has attempted to identify all industrial uses of asbestos-containing gaskets,
but the only use known to the Agency is among titanium dioxide manufacturing facilities.
EPA communicated with the titanium dioxide industry to understand typical industrial uses of chrysotile
asbestos-containing gaskets. This communication includes an October 2017 meeting between EPA and
industry representatives and written communications submitted by industry representatives and ACC.
An overview of chrysotile asbestos-containing gasket use in the titanium dioxide manufacturing industry
follows.
Branham supplies chrysotile asbestos-containing gaskets to at least four titanium dioxide manufacturing
facilities worldwide. Two are Chemours facilities located in DeLisle, Mississippi and New Johnsonville,
Tennessee; and the other two are located outside the United States (Mingis. 2018). The manufacture of
titanium dioxide occurs at process temperatures greater than 1,850 degrees Fahrenheit and pressures of
approximately 50 pounds per square inch, and it involves multiple chemicals, including chlorine,
toluene, and titanium tetrachloride (ACC. 2017b). Equipment, process vessels, and piping require
durable gasket material to contain these chemicals during operation. The Chemours facilities use the
Branham products - sheet gaskets composed of 80 percent (minimum) chrysotile asbestos, fully
encapsulated in styrene-butadiene rubber - to create tight chemical containment seals for these process
components (ACC. 2017b). One of these facilities reports replacing approximately 4,000 asbestos-
containing gaskets of various sizes per year, but any given year's usage depends on many factors (e.g.,
the number of major turnarounds) (ACC. 2017b).
Installed gaskets typically remain in operation anywhere from a few weeks to three years; the time-
frame before being replaced is largely dependent upon the temperature and pressure conditions (ACC.
2.018). whether due to detected leaks or as part of a routine maintenance campaign. Used asbestos-
containing gaskets are handled as regulated non-hazardous material. Specifically, they are immediately
bagged after removal from process equipment and then placed in containers designated for asbestos-
containing waste. Containerized waste (volume not known) from both Chemours domestic titanium
dioxide manufacturing facilities is eventually sent to the following landfills, which are permitted to
receive asbestos-containing waste (ACC. 2017b):
•	Asbestos-containing waste from Chemours' Tennessee facility is transported to the West
Camden Sanitary Landfill at 2410 Highway 70 West, Camden, Tennessee.
•	Asbestos-containing waste from Chemours' Mississippi facility is transported to the Waste
Management Pecan Grove Landfill at 9685 Firetower Road, Pass Christian, Mississippi.
Though Chemours has an active program to replace asbestos-containing gaskets with asbestos-free
alternatives and this program has resulted in considerable decreases in asbestos-containing gasket use
(EPA-HQ-QPPT-2016-0736-0067). gaskets formulated from non-friable chrysotile asbestos-containing
sheeting continue to be the only product proven capable of withstanding certain extreme operating
Page 93 of 352

-------
conditions and therefore provide a greater degree of process safety and integrity than unproven
alternatives according to industry (ACC. 2017b). A single titanium dioxide manufacturer can have
approximately 4,000 gaskets of various sizes distributed throughout the plant which are periodically
replaced during facility shutdowns.
2.3.1.5.2 Worker Activities - Sheet Gasket Use
Worker activities most relevant to chrysotile asbestos exposure include receiving new gaskets, removing
old gaskets, bagging old gaskets for disposal, and inserting replacement gaskets into flanges and other
process equipment. Chrysotile asbestos-containing gaskets are received and stored in individual
resealable 6-mm thick plastic bags. Trained maintenance workers wear leather gloves when handling the
gaskets for insertion into a flange. When removing old gaskets for replacement, trained maintenance
workers wear respiratory protection—either an airline respirator (also known as a supplied air respirator)
or cartridge respirator with P-100 HEPA filters, although the APF for this respiratory protection was not
specified (ACC. 2017a). Respiratory protection is used during this task to protect workers in cases
where the original sheet gasket material has become friable over the service life (ACC. ^ ).
2.3.1.5.3 Number of Sites and Potentially Exposed Workers - Sheet Gasket
Use
As noted previously, EPA is aware of two Chemours titanium dioxide manufacturing facilities that use
chrysotile asbestos-containing gaskets in the United States. However, no estimates of the number of
potentially exposed workers were submitted to EPA by industry or its representatives. As gaskets are
replaced during plant shutdowns, the number of potentially exposed workers would be low as some
workers would be off site during the shutdown.
To estimate the number of potentially exposed workers and ONUs at these two facilities, EPA
considered 2016 data from the Bureau of Labor Statistics for the NAICS code 325180 (Other Basic
Inorganic Chemical Manufacturing). These data suggest an industry-wide aggregate average of 25
directly exposed workers per facility and 13 ONUs per facility. EPA therefore estimates that the two
Chemours facilities combined have approximately 50 directly exposed workers and 26 ONUs.
These estimates are based on the one company known to use asbestos-containing gaskets at its titanium
dioxide manufacturing facilities. Other titanium dioxide manufacturing plants that operate under similar
conditions in the United States are thought to use asbestos-containing gaskets to prevent chlorine leaks,
but EPA does not have information to confirm this (Mingis. 2018).
2.3.1.5.4 Occupational Inhalation Exposures - Sheet Gasket Use
To identify relevant occupational inhalation exposure data, EPA reviewed reasonably available
information from OSHA, NIOSH, the published literature, and industry. All research steps are
documented below, with more detailed discussion on the most relevant data source, which EPA
determined was the monitoring results submitted by ACC for a Chemours titanium dioxide
manufacturing facility.
EPA first considered the 2011 to 2016 nationwide exposure data provided by OSHA and the history of
NIOSH HHEs, but neither resource included asbestos exposure data for the titanium dioxide
manufacturing industry.
Page 94 of 352

-------
EPA also considered the published literature on worker asbestos exposure attributed to gasket removal.
This search did not identify publications that specifically addressed asbestos-containing gasket use in the
titanium dioxide manufacturing industry. However, two peer-reviewed publications measured worker
exposures of gasket removal in settings like those expected for this industry:
•	One publication was a 1996 study of maintenance workers who removed braided gaskets and
sheet gaskets at a chemical plant in the Netherlands (Spence and Rocchi. 1996). The study
considered two types of sheet gasket removal activity: gaskets that could be easily removed with
a putty knife without breaking, and gaskets that required more intensive means (and longer
durations) for removal. Among the data for sheet gasket removal, the highest worker exposure
concentration—with asbestos presence confirmed by TEM analysis—was 0.02 f/cc for a 141-
minute sample. A slightly higher result was reported in a different sample, but TEM analysis of
that sample found no detectable asbestos. The overall representativeness of a study more than 20
years old to today's operations is unclear.
•	The other publication was a 2006 study that used a simulated work environment to characterize
worker and ONU exposure associated with gasket removal onboard a naval ship or at an onshore
site (Mangold et at.. 2006). The simulations considered various gasket removal scenarios (e.g.,
manual removal from flanges, removal requiring use of a knife, removal requiring use of power
wire brushes). The 8-hour TWA PBZ exposures that were not conducted on marine vessels and
therefore considered most relevant to the sheet gasket removal ranged from 0.005 to 0.023 f/cc.
The representativeness of these simulations to an industrial setting is unclear. However, the study
provides useful insights on the relative amounts of asbestos exposure between workers and
ONUs. The simulated gasket removal scenarios with detected fibers suggested that exposure
levels decreased by a factor of 2.5 to 9 between the gasket removal site and the "area/bystander"
locations, approximately 5 to 10 feet away.
Other peer-reviewed publications were identified and evaluated but not considered in this assessment
because they pertained to heavy-duty equipment (Boelter et al. 2011). a maritime setting with confined
spaces (Madl et al.. 2014). and braided packing (Boelter et al.. 2002). Further, EPA compiled and
reviewed a large number of additional studies that characterized worker exposures during gasket
removal. These studies reported a broad range of worker asbestos exposure levels. However, EPA
ultimately chose to base this COU's worker exposure estimates on data provided by industry, given that
the one company known to use the chrysotile asbestos-containing sheet gaskets provided exposure data
(through ACC) for its gasket servicing workers. EPA viewed these direct observations as most
representative for this COU, rather than using surrogate values based on workers in other industries who
may use different gasket removal practices.
EPA determined that worker exposure data submitted by ACC for one of the Chemours titanium dioxide
manufacturing facilities provide the most relevant data for this COU. ACC stated that only trained
Chemours mechanics remove asbestos-containing gaskets, and they use respiratory protection when
doing so (either an atmosphere-supplying respirator or an air-purifying respirator) (ACC. 2017a).
According to the information provided to EPA, 34 worker exposure samples have been collected since
2009 during removal of asbestos-containing gaskets, but the number of workers that were evaluated is
not known (based on discussions with Chemours during a visit to EPA in October 2017). The samples
evidently were collected to assess compliance with OSHA occupational exposure limits, suggesting that
they were analyzed using PCM. Asbestos levels in these samples ranged from 0.0026 to 0.094 f/cc, with
an average of 0.026 f/cc (ACC. 2017a). The documentation provided for these sampling events does not
Page 95 of 352

-------
indicate the sampling duration or the amount of time that workers performed gasket removal activity,
nor were the raw data provided.
2.3.1.5.5 Exposure Results for Use in the Risk Evaluation for Asbestos Part 1:
Chrysotile Asbestos - Sheet Gasket Use
Table 2-11. presents the worker exposure concentrations that EPA is using in this Part 1 of the risk
evaluation for asbestos for use of chrysotile asbestos-containing gaskets at titanium dioxide
manufacturing facilities. The following assumptions were made in compiling these data:
•	The central tendency 8-hour TWA exposure value for workers (0.026 f/cc) is based on the
average asbestos exposure measurement reported for gasket removal at titanium dioxide
manufacturing facilities (ACC. 2017a). Though the supporting documentation does not specify
sample duration, EPA assumes, based on discussions with Chemours, the average reported
concentration can occur throughout an entire 8-hour shift (e.g., for workers removing gaskets
throughout a day during a maintenance campaign).
•	The high-end 8-hour TWA exposure value for workers (0.094 f/cc) is based on the highest
exposure measurement reported for gasket removal activity at titanium dioxide manufacturing
facilities (ACC. 2017a). Again, the sample duration for this measurement was not provided and
so this concentration represents a high-end by extrapolating the value to represent an entire shift.
•	Because the documentation for the 34 worker exposure samples does not include sample
duration, EPA cannot assume the central tendency and high-end values apply to short-term
exposures. More specifically, if the original data were for full-shift exposures, then assuming
full-shift data points apply to short-term durations would understate the highest short-term
exposures. That is because short-term data within a shift generally span a range of
concentrations, and the corresponding full-shift concentration for that shift would fall within that
range (and be lower than the highest short-term result). Therefore, EPA has determined that this
COU has no reasonably available data for evaluating worker short-term exposures.
•	EPA considered multiple options for estimating ONU exposure concentrations. First, EPA
revisited existing data sources in an attempt to identify direct measurements; however, the data
from facilities that stamp and use sheet gaskets do not have any information relevant to ONUs.
Second, EPA considered assuming ONU exposures are the same as worker exposures. EPA did
not pursue this option, given that ONU exposures are likely less than worker exposures for the
gasket-related conditions of use (i.e., EPA found no instances where ONUs are in very close
proximity to process areas where asbestos-containing gaskets are removed). The third option was
to derive ONU exposures based on a calculated "decay factor." EPA is using this third approach
to estimate ONU exposures. Specifically, the literature includes "bystander" exposure studies
that EPA used to estimate ONU exposures. One publication (Mangold et ai. 2006) measured
"bystander" exposure during asbestos-containing gasket removal. The "bystander" locations in
this study were between 5 and 10 feet from the gasket removal activity, and asbestos
concentrations were between 2.5 and 9 times lower than those measured for the worker. Based
on these observations, EPA assumes that ONU exposures for this COU are a factor of 5.75 (i.e.,
the midpoint between 2.5 and 9) lower than the directly exposed workers.
Page 96 of 352

-------
Table 2-11. Summary of Asbestos Exposures During Sheet Gasket Use Used in the Risk
	 Evaluation for Asbestos Part 1: Chrysotile Asbestos		
Occupational
Kxposure Scenario

8-hr TWA Kxposurc l.eyels (I'/cc)

Workers

OMs

Central
Tendency
lligh-end
Confidence
Rating
Central
Tendency
lligh-end
Confidence
Rating
Shccl gasket use K-
hr TWA exposure
0.026
0.094
Medium
0.005
0.016
Medium
2.3.1.5.6 Data Assumptions, Uncertainties and Level of Confidence
The exposure data shown in Table 2-11. are based on observations from a single reference that presents
worker exposure monitoring data for a single company, and documentation for this study is incomplete.
EPA estimates that using the 34 direct observations for gasket removal workers likely offers the most
representative account of actual exposures, rather than relying on data from the published literature
taken from other occupational settings and based on other worker practices. Moreover, the central
tendency concentration shown in Table 2-11.1 falls within the range of results from the relevant
literature that EPA reviewed,, suggesting that the data source considered (ACC. 2 ) does not
understate exposures.
EPA considered the quality and uncertainties of the data to determine a level of confidence for the
assessed inhalation exposures for this COU. The primary strength of this assessment is the use of
monitoring data, which is the highest level of the inhalation exposure assessment approach hierarchy.
The overall confidence rating from systematic review for these data was rated medium. These
monitoring data were provided to EPA by industry and represent actual measurements made during
asbestos-containing sheet gasket removal at a titanium dioxide manufacturing facility in the United
States. However, based on reasonably available information, EPA is not aware of any additional
facilities that use asbestos-containing sheet gaskets, and EPA could not determine if the industry-
provided monitoring data are representative of all U.S. facilities that use asbestos-containing sheet
gaskets. The monitoring data were collected from 2009 through 2017; therefore, the data likely capture
temporal variability in the facility's operations.
The ONU exposure estimates are based on "decay factors" observed for gasket removal operations.
These ONU estimates are therefore uncertain. The uncertainty cannot be reduced with the data currently
available to EPA. In addition, ONU may not be exposed at full shift, every workday.
Based on these strengths and limitations of the data, the overall confidence for EPA's assessment of
occupational and ONU inhalation exposures for this scenario is medium.
2.3.1.6 Oil Field Brake Blocks
This section reviews the presence of chrysotile asbestos in oil field brake blocks and evaluates the
potential for worker exposure to asbestos during use.
2.3.1.6.1 Process Description - Oil Field Brake Blocks
The rotary drilling rig of an oil well uses a drawworks hoisting machine to raise and lower the traveling
blocks during drilling. The drawworks is a permanently installed component of a mobile drilling rig
package, which can be either "trailerized" or self-propelled. Therefore, there is no on-site assembly of
Page 97 of 352

-------
the drawworks. Except for initial fabrication and assembly prior to installation on a new rig, the
drawworks is not set or installed in an enclosed building (Popik, 2018).
This section focuses on oil field operations involving asbestos-containing brake blocks. EPA
acknowledges that many of today's rigs use electromagnetic braking systems that reportedly do not
contain asbestos and some use braking systems that do not use friction pads during normal operation.
However, a precise count of the rigs with and without asbestos-containing brakes is not available. The
remainder of this section summarizes information EPA compiled on the operations involving the
asbestos-containing brakes.
The drawworks consists of a large-diameter steel spool, a motor, a main brake, a reduction gear, and an
auxiliary brake. The drawworks reels the drilling line over the traveling block in a controlled fashion.
This causes the traveling block and its hoisted load to be lowered into or raised out of the wellbore
(Schlumberger. 2018). The drawworks components are fully enclosed in a metal housing. The brake
blocks, which ride between an inner brake flange and an outer metal brake band, are not exposed during
operation of the drawworks (Popik. 2018).
The brake of the drawworks hoisting machine is an essential component that is engaged when no motion
of the traveling block is desired. The main brake can have several different designs, such as a friction
band brake, a disc brake, or a modified clutch. The brake blocks are a component of the braking system
(Schlumberger. 2018). According to product specification sheets, asbestos-containing brake blocks are
most often used on large drilling drawworks and contain a wire backing for added strength. They are
more resistant than full-metallic blocks, with good flexibility and a favorable coefficient of friction. The
asbestos allows for heat dissipation and the woven structure provides firmness and controlled density of
the brake block. Workers in the oilfield industry operate a drilling rig's brakes in an outdoor
environment and must periodically replace spent brake blocks (Popik. 2018).
Figure 2-8. Photographs of Typical Oil Field Drawworks
Photograph courtesy of Stewart & Stevenson and used with Stewart & Stevenson's permission
Drawworks can have either one or two drums, with each drum usually containing two bands, and each
band usually containing 10 brake blocks, resulting in a total of 20 to 40 brake blocks per drawworks.
The configuration can vary depending on the size of the drawworks. An industry contact specified brake
block dimensions of 8 to 12 inches wide by 12 inches long by 0.75 to 1.125 inches thick and weighing
six to seven pounds per block. The percent chrysotile asbestos composition of the brake blocks is
unknown (Popik. 2018).
Page 98 of 352

-------
Brake blocks do not require maintenance other than replacement when worn down to a 0.375-inch
thickness at any point in the block. The brake blocks typically last between 2 and 3 years under daily
operation of the drawworks. Due to the heterogeneous pressure distribution inherent in the mechanics of
the brake band design, the brake blocks wear differently depending on their position within the band.
However, efforts are made to equalize the tapering pressure distribution by grading the brake block
material in order to achieve a more uniform friction at all points along the brake band (Popik. 2018).
The brake blocks are enclosed in the drawworks, so it is not necessary to clean off brake dust under
normal operations. The drawworks is washed down prior to removal and installation of brake blocks—a
task that could lead to water releases of asbestos dust. Brake block servicing typically takes place
outdoors or in a large service bay inside a shop (Popik. 2018).
EPA obtained a safety data sheet (SDS) from Stewart & Stevenson Power Products, LLC for "chrysotile
woven oilfield brake blocks, chrysotile woven plugs, and chrysotile molded oilfield brake blocks." The
SDS recommends avoiding drilling, sanding, grinding, or sawing without adequate dust suppression
procedures to minimize air releases and inhalation of asbestos fibers from the brake blocks. The SDS
recommends protective gloves, dust goggles, and protective clothing. The SDS also specifies that used
brake block waste should be sent to landfills (Stewart & Stevenson. 2000).
At least one U.S. company imports and distributes non-metallic, asbestos-woven brake blocks used in
the drawworks of drilling rigs. Although the company no longer fabricates brake blocks using asbestos,
the company confirmed that it imports asbestos-containing brake blocks on behalf of some clients for
use in the oilfield industry. It is unclear if any other companies fabricate or import asbestos-containing
brake blocks, or how widespread the continued use of asbestos brake blocks is in oilfield equipment.
However, EPA understands from communications with industry that the use of asbestos containing
brake blocks has decreased significantly over time and continues to decline (Popik. 2018).
2.3.1.6.2 Worker Activities - Oil Field Brake Blocks
Worker activities include receipt of chrysotile asbestos-containing brake blocks, removing old brake
blocks, bagging old brake blocks for disposal, and installing new brake blocks into drawworks
machinery. The activities that may result in asbestos exposure include installing and servicing brake
blocks (which may also expose workers in the vicinity). Additionally, workers at the drawworks may be
exposed to asbestos fibers that are released as the brake blocks wear down over time. EPA has not
identified PPE and industrial hygiene practices specific to workers removing and installing asbestos-
containing brake blocks. EPA notes that workers in the vicinity of brake replacement activity may be
exposed due to brake block wear; and these workers were considered to be ONUs.
2.3.1.6.3 Number of Sites and Potentially Exposed Workers - Oil Field Brake
Blocks
EPA identified one U.S. facility that imports chrysotile asbestos-containing brake blocks (Popik. 2018).
It is unknown how many other facilities import asbestos-containing brake blocks. It is also unknown
how many customers receive brake blocks from the sole facility identified by EPA. Unlike some of the
other COUs, for which extensive information is available to estimate numbers of potentially exposed
workers, EPA found no direct accounts of the number of workers who use asbestos-containing oil field
brake blocks. The lack of information necessitated the use of other established methods to estimate the
number of potentially exposed workers.
Page 99 of 352

-------
To derive these estimates, EPA used 2016 Occupational Employment Statistics data from the Bureau of
Labor Statistics (BLS) and 2015 data from the U.S. Census' Statistics of U.S. Businesses. EPA used
BLS and Census data for three NAICS codes: 211111, Crude Petroleum and Natural Gas Extraction;
213111, Drilling Oil and Gas Wells; and 213112, Support Activities for Oil and Gas Operations. Table
2-13 summarizes the total establishments, potentially exposed workers, and ONUs in these industries.
EPA does not have an estimate of the number of establishments in these industries that use asbestos-
containing brake blocks. Therefore, EPA presents these results as high-end estimates of the number of
establishments and potentially exposed workers and ONUs. The actual number of potentially exposed
workers and ONUs is likely lower that EPA's estimates.
For each of the three NAICS codes evaluated, Table 2-12. presents EPA's estimates of industry-wide
aggregate averages of directly exposed workers per site and ONUs per site. EPA estimates an upper
bound of 21,670 sites, 61,695 directly exposed workers, and 66,108 ONUs.
Table 2-12. Summary of Total Establishments in Relevant Industries and Potentially Exposed
Workers and ONUs for Oilfield Brake Blocks


Toi;il (I'mire
Indnsln Seclon
Workers with KcIcmiiiI Occupations
NAICS
Codes
NAICS
Description
Tol;il
liiins
To(;il
l-'.sliihlish-
ineiils
Tohil
r.inplovees
A\er;i!ic
Km |)l(>\cos
per
l.s(;ihlish-
\\ orkcrs in
Relet iinl
Oeen pil-
lions
Oeenpii-
lioiiiil Non-
l scrs
\\ orkcrs
per Sile
ONI s
per
Sile





m en 1




Crude









Petroleum








211111
and Natural
Gas
Extraction
6,270
7,477
124,847
17
15,380
32,704
2
4
213111
Drilling Oil
and Gas
Wells
1,973
2,313
89,471
39
10,256
7,397
4
3

Support
Activities








213112
for Oil and
Gas
Operations
9,591
11,880
314,589
26
36,059
26,007
3
2
All NAICS
17,834
21,670
528,907
27
61,695
66,108
3
3
2.3.1.6.4 Occupational Inhalation Exposures - Oil Field Brake Blocks
EPA did not identify any studies that contain exposure data related to asbestos-containing brake blocks
but did identify one published study that contains limited air sampling data for asbestos-containing brake
bands (Steimsvae et ai. 2007). In the absence of any other exposure data, the limited data provided in
this study were used to estimate exposures to workers from brake block installation, servicing, and
removal. The study references stationary samples of asbestos fibers taken in 1988 from the drilling floor
at an unnamed facility in Norway's offshore petroleum industry. Use of asbestos was generally banned
in Norway in late 1984, but asbestos brake bands were used in the drilling drawworks on some
installations until 1991. The study notes: ".. .the design of the drilling area might have led to migration
of fibers from the brake bands into the drilling cabin or down one floor to the shale shaker area"
(Steimsvae et at.. 2007).
Page 100 of 352

-------
Stationary samples were taken at two locations: "above brake drum" and "other samples, brake dust."
Reported arithmetic mean concentrations of asbestos fibers for both locations were 0.03 and 0.02 f/cc,
respectively. However, because the publication does not indicate what activities workers performed
during sample collection, no inferences can be made regarding whether the results pertain to brake
installation, removal, servicing, or repair. The study involved an unknown number of measurements
made over an unknown duration of time. While the study does not identify the sample collection
methods or the fiber counting algorithms, some text suggests that the presence of asbestos in the samples
was confirmed by electron microscope. The study reports the following additional details about the
asbestos content of the brake lining: "The composition of the brake lining was: 41% asbestos, 28%
rayon and cotton, 21% binding agent, 9% brass chip" (Steinsvag et at... 2007).
The sample measurements were made over an unknown duration of time, and EPA is assuming
measurements are representative of an 8-hr TWA. EPA assumes the measurements taken above the
brake drum are most relevant to worker exposures, as workers are likely to work nearest the brakes, such
as operating a brake handle to control the speed of the drawworks or replacing the brake blocks. EPA
assumes the other brake dust samples are relevant to ONU exposures as their exact sampling location is
not specified but the arithmetic mean concentration is lower than that of the samples taken above the
brake drum. Since these two results are both arithmetic means, EPA assumed the values were 0.03 and
0.02 f/cc for 8-hour TWA, for workers and ONUs, respectively. This study was rated "low" in
systematic review (Steimsvag et at.. 2007).
2.3.1.6.5 Exposure Results for Use in the Risk Evaluation for Asbestos: Part 1
Chrysotile Asbestos - Oil Field Brake Blocks
The information available to EPA confirms that some brake blocks used in domestic oilfields contain
chrysotile asbestos, as demonstrated by an SDS provided by a supplier. It is reasonable to assume that
wear of the brake blocks over time will release some asbestos fibers to the workplace air. However, the
magnitude of these releases and resulting worker exposure levels is not known. In an effort to provide a
risk estimate for this COU, the exposure scenario described in the previous section will be used. Table
2-13 presents the exposure data used for the risk estimates for brake block usage.
As noted previously, ONUs for this COU include workers in the vicinity of brake blocks, but whose job
duties do not involve repair or servicing of the brake blocks, EPA has not identified specific data on
potential ONU inhalation exposures from brake block use. It is assumed that ONUs do not directly
handle brake blocks and drawworks machineries, and it is also assumed that drawworks are always used
and serviced outdoors close to oil wells. Given the limited information identified above, the lower of the
two reported values in the Norway study will be used to represent ONU exposures for this COU.
Table 2-13. Summary of Asbestos Exposures During Use in Brake Blocks for the Risk Evaluation
	for Asbestos Part 1: Chrysotile Asbestos	

8-hr TWA Kxposurc Levels (I'/cc)
Occupational Kxposurc Scenario
Workers
OMs
(on (nil
Confidence
Central
Confidence

Tendency
Rating
Tendency
Rating
Brake Blocks:
8-hr TWA exposure
0.03
Low
0.02
Low
Page 101 of 352

-------
2.3.1.6.6 Data Assumptions, Uncertainties and Level of Confidence
The extent of brake block usage and associated worker exposures are highly uncertain. EPA was not
able to identify the volume of imported asbestos-containing brake blocks, the number of brake blocks
used nationwide, nor the number of workers exposed as a result of installation, removal, and disposal
activities. Further, the study reviewed in this section examined asbestos exposures in 1988 in Norway's
offshore petroleum industry and is of unknown relevance to today's use of oil field brake blocks in the
United States. No other data for brake blocks could be located.
EPA considered the quality and uncertainties of the data to determine a level of confidence for the
assessed inhalation exposures for this condition of use. The primary strength of this assessment is the
use of monitoring data, which is the highest approach of the inhalation exposure assessment approach
hierarchy. However, the monitoring data are limited to a single offshore oil platform in Norway in 1988.
It is unknown if these data capture current-day U.S. oil field or offshore platform operations. It is also
unknown if the monitoring data capture the variabilities in the day-to-day operations of the single
offshore platform sampled in the study. For this COU, ONU may not be exposed at full shift, every
workday.
These are significant uncertainties in the assessment, but the uncertainties cannot be reduced through
review of other available information. EPA is not aware of published accounts of worker exposure
concentrations in the United States to chrysotile asbestos from oil field brake blocks.
EPA considered asbestos sampling data from hoist crane operations as a surrogate for this COU
(Spencer et al.. 1999). but ultimately believes the one study of brake blocks on an oil rig is more
representative of this COU than measurements from a hoist crane in an industrial setting. EPA believes
the values in Table 2-13 represent the best available information, but there is also reason to believe these
values might overstate actual exposures.
Based on these strengths and limitations of the data, the overall confidence for EPA's assessment of
occupational and ONU inhalation exposures for this scenario is low.
2.3.1.7 Aftermarket Automotive Brakes/Linings and Clutches
The use of chrysotile asbestos in automotive parts has decreased dramatically in the last 30-40 years.
Several decades ago, virtually all vehicles had at least some asbestos-containing components. Currently,
information indicates asbestos containing automobile components are used in a single vehicle which is
manufactured domestically, but only exported and sold outside of the United States. However, the
potential remains for some older vehicles to have asbestos-containing parts and for foreign-made
aftermarket parts that contain asbestos to be imported and installed by consumers in cars when replacing
brakes or clutches.
EPA is aware of one car manufacturer that imports asbestos-containing automotive friction products for
new vehicles, but those vehicles are then exported and not sold in the United States. This COU is
categorized as "other vehicle friction products" in Table 1-4. of Section 1.4.2 of this risk evaluation.
This section reviews the presence of chrysotile asbestos in aftermarket automotive parts and evaluates
the potential for worker exposure to asbestos. The section focuses on asbestos in light-duty passenger
vehicles, including cars, trucks, and vans.
Note that for occupational exposure for this COU, the use of compressed air as a work practice will not
be considered because, in addition to the EPA current best practice guidance (EPA-747-F-04-004). there
Page 102 of 352

-------
is a provision in the OSHA Asbestos Standard: 29 CFR ง 1910.1001(f)(l)(ix): Compressed air shall not
be used to remove asbestos or materials containing asbestos unless the compressed air is used in
conjunction with a ventilation system which effectively captures the dust cloud created by the
compressed air.
2.3.1.7.1 Process Description - Aftermarket Automotive Brakes/Linings and
Clutches
Based on the long history of the use of asbestos in automobile parts, and because aftermarket automotive
parts may still be available for purchase, the Agency believes this COU is still ongoing. Over the past
few decades, automobile weights, driving speeds, safety standards, and applicable environmental
regulations have changed considerably. These and other factors have led to changes in materials of
choice for automobile parts. Asbestos was previously a component of many automobile parts, including
brakes, clutches, gaskets, seam sealants, and exhaust systems (Blake et at.. 2008; Rohl et at.. 1976); and
older model vehicles still in operation may have various asbestos-containing parts. Additionally,
aftermarket automotive parts made from asbestos can be purchased from online retailers, and it is
possible that such products exist in older stockpiles. This section focuses on asbestos in brakes/linings
and clutches because repairs for these parts - and hence potential occupational exposure to asbestos - are
more likely than repairs for other vehicle components that were known to previously contain asbestos
(e.g., seam sealants). For the purpose of this risk evaluation, EPA generally refers to brakes in the
following sections, but this term also includes brake linings, brake pads, and clutches.
Automobile Brakes
Chrysotile asbestos fibers offer many properties (e.g., heat resistance, flexibility, good tensile strength)
that are desired for brake linings and brake pads (Paustenbach et at.. 2004). Up through the 1990s many
new automobiles manufactured in the United States had brake assemblies with asbestos-containing
components. However, by 2000, asbestos was no longer used in the brakes of virtually all automobiles
sold domestically (Paustenbach et at.. 2004). NIOSH reported in the late 1980s that friction materials in
drum brakes typically contained 40 to 50 percent asbestos by weight (OS! 06). Other researchers
reported that some brake components during these years contained as much as 73 percent asbestos, by
weight (Stake et at.. 2003).
The two primary types of automobile brakes are drum brakes and disc brakes, and chrysotile asbestos
has been found both in linings for drum brake assemblies and pads in disc brake assemblies (see Figure
2-9.). Drum brakes were more prevalent than disc brakes in older vehicles. When the vehicle operator
engages drum brakes, the brake shoes (which contain friction materials) contact the rotating brake drum,
and this contact slows the vehicle. Disc brakes are much more common today than drum brakes, and
they function by applying brake pads (which contain friction materials) to the surface of the revolving
brake disc, and this contact slows the vehicle. Richter et at. (2009) state that by the mid-1990s, material
and design improvement led to most cars being manufactured with disc brakes, effectively phasing out
drum brakes in passenger automobiles. However, further investigation online by EPA into the use of
disc/drum brakes indicate that while front brakes appear to mostly have been converted to disc brakes in
front wheel drive vehicles, many passenger vehicles have a combination of disc brakes for the front
wheels and drum brakes for the rear wheels.
Page 103 of 352

-------
Brake
Lining
Shoe
Brake
Pad
Figure 2-9. Illustrations of brake assembly components: (a) a brake lining designed to be used
with an internal drum brake and (b) a brake pad designed for use with a disc brake.
Source: Paustenbach et al. (2004).
Use of asbestos-containing braking systems began to decline in the 1970s due to many factors, including
toxicity concerns, rising insurance costs, regulatory scrutiny, challenges associated with disposing of
asbestos-containing waste, and availability of asbestos-free substitutes (Paustenbach et al.. 2004). In
1989, EPA issued a final rule that banned the manufacturing and importing of many asbestos-containing
products, including automobile brake pads and linings (54 FR 29460). While the United States Court of
Appeals for the Fifth Circuit vacated most of this ban17 in 1991, many manufacturers had already begun
to phase out asbestos-containing materials and develop alternatives, including the non-asbestos organic
fibers that are almost universally used in automobile brake assemblies today (Paustenbach et al.. 2004).
By 2000, domestic manufacturers had eliminated asbestos from virtually all brake assemblies in
automobiles (Paustenbach et al.. 2004). EPA is not aware of any automobile manufacturers that
currently use asbestos products in brake assemblies for U.S. vehicles. In fact, the Agency received
verification from five major vehicle manufacturers that asbestos-containing automotive parts are no
longer used and import data has been misreported under the wrong Harmonized Tariff Schedule (HTS)
code. However, the Agency knows of at least one company that imports asbestos-containing friction
products for use in cars assembled in the U.S., but those vehicles are exported for sale and are not sold
domestically. The COU identified for this scenario is specified as "other vehicle friction products" in
Table 1-34, and the exposure values are based on aftermarket auto brakes (see Section 2.3.1.8).
The history of asbestos in aftermarket brake products has followed a similar pattern. For decades,
asbestos was found in various aftermarket brake replacement parts (e.g., pads, linings, and shoes); but
the same factors listed in the previous paragraph led to a significant decline in the use of asbestos in
aftermarket vehicle friction products. Nonetheless, the literature indicates that asbestos-containing
replacement brake materials continued to be available from parts suppliers into the 2000s; researchers
were able to purchase these materials in 2008 from a vintage auto parts facility (Madl et al.. 2008).
Today, individual consumers can find aftermarket automotive products marketed as containing asbestos
through online retailers.
In more recent years, state laws and regulations have limited sales of asbestos-containing aftermarket
brake parts, even among existing stockpiles. In 2010, for instance, the state of Washington passed its
"Better Brakes Law," which prohibits manufacturers, retailers, wholesalers, and distributors from selling
brake friction material that contains more than 0.1 percent asbestiform fibers (Washington State. 2010).
17 Federal Register notice - https://www.govinfo.gov/content/pkg/FR-1994-06-28/html/94-15676.htm
Page 104 of 352

-------
In the same year, the state of California passed legislation with similar requirements. The not-to-exceed
limit of 0.1 percent asbestiform fibers in aftermarket brake parts now essentially extends nationwide,
due to a memorandum of understanding between EPA and multiple industry stakeholders {e.g., Motor
and Equipment Manufacturers Association, Automotive Aftermarket Suppliers Association, Brake
Manufacturers Council) (U.S. EPA. 2015a).
Despite this trend, asbestos in automotive parts is not banned at the federal level, and foreign suppliers
face no restrictions (other than those currently in place in the states of California and Washington) when
selling asbestos-containing brake products to business establishments and individuals in the United
States. The Motor and Equipment Manufacturers Association informed EPA that approximately $2.2
million of asbestos-containing brake materials were imported into the United States in 2014 (MEM A.
2016). In 2018, the U.S. Geological Survey indicated that "an unknown quantity of asbestos was
imported within manufactured products," such as brake linings (USGS, 2019).
Based on this context, asbestos is currently found in automobile brakes in the United States due to two
reasons: (1) vehicles on the road may have asbestos-containing brakes, whether from original
manufacturers (primarily for older and vintage vehicles) or aftermarket parts; and (2) vehicles may have
new asbestos-containing brakes installed by establishments or individuals that use certain imported
products.
Automobile Clutches
In a manual transmission automobile, which currently accounts for less than 5 percent of automobiles
sold in the United States, the clutch transfers power generated by the engine to the drive train. The
schematic in Figure 2-10. shows a typical clutch assembly. Because it lies at the interface between two
rotating metallic surfaces, the clutch disc typically contains friction materials. Decades ago, the friction
material of choice was chrysotile asbestos, which previously accounted for between 30 and 60 percent of
the friction material in clutch discs (Jiang et al., 2008).
Figure 2-10. Schematic of a clutch assembly. The clutch disc is made of friction material, which
may contain asbestos.
Source: Jiang et al.. (2008).
Transmission
Clutch Disc
t
Pressure Plate Flywheel Engine
Page 105 of 352

-------
Consistent with the history for brakes, friction materials in clutches moved from asbestos-containing to
asbestos-free designs over recent decades. By the 1980s, automobile manufacturers began using various
asbestos-free substitutes in clutch assemblies (Jiang et at... 2008); and by 2000, most automobiles in the
United States were no longer made with asbestos-containing clutches (Cohen and Van Orden. 2008).
EPA is not aware of any car manufacturers that currently import asbestos-containing clutch assemblies.
However, aftermarket clutch parts may contain asbestos. As evidence of this, Jiang et al. (2008) reported
purchasing 27 boxes of asbestos-containing clutch discs that had been stockpiled at a parts warehouse
(Jiang et al.. 2008). suggesting that stockpiles of previously manufactured asbestos-containing clutch
assemblies could be available.
Asbestos-containing aftermarket clutches may be found as imports from foreign suppliers, although the
extent to which this occurs is not known. No barriers currently exist to these imports, as asbestos in
automotive clutches is not banned at the federal level and the brake laws passed in 2010 in the state of
California and the state of Washington do not apply to clutches.
2.3.1.7.2 Worker Activities - Aftermarket Automotive Brakes/Linings and
Clutches
This section describes worker activities for repair and replacement of both brakes and clutches,
including the types of dust control measures that are typically used. For both types of parts, asbestos
exposure may occur during removal and disposal of used parts, while cleaning the assemblies, and
during handling and installation of new parts.
Automobile Brake Repair and Replacement
For both drum brakes and disc brakes, maintenance, repair, inspection, and replacement jobs typically
involve several basic steps. Workers first need access to the brake assembly, which is typically
accomplished by elevating the vehicle and removing the wheel. They then remove dust and debris from
the brake apparatus using methods described below. Replacement or repair of parts follows, during
which workers use various mechanical means to remove old parts and install new ones.
Two critical issues for exposure assessment are the work practices used to remove dust and debris from
the brake assembly and the asbestos content of this material:
1. Work practices for automobile brake repair have changed considerably over the years. In the
1970s, use of compressed air to clean brake surfaces was commonplace (Rohl et al.. 1976).
While effective at quickly preparing surfaces for repair, this practice caused brake dust and other
material to become airborne, leading to potential asbestos exposures among workers and ONUs.
The practice also caused asbestos-containing dust to move to locations throughout the
workplace.
Concerns about asbestos exposure during brake repair led NIOSH to perform a series of
industrial hygiene evaluations in the late 1980s to investigate the effectiveness of different dust
control strategies. Additionally, OSHA amended its asbestos standards in 1994. Major revisions
in these standards included a reduced time-weighted-average permissible exposure limit (PEL)
of 0.1 fiber per cubic centimeter (f/cc) for all asbestos work in all industries, a new classification
scheme for asbestos construction and shipyard industry work which ties mandatory work
practices to work classification, a presumptive asbestos identification requirement for "high
hazard" asbestos containing building materials, limited notification requirements for employers
Page 106 of 352

-------
who use unlisted compliance methods in high risk asbestos abatement work, and mandatory
methods of control for brake and clutch repair (Federal Register, 1994). The requirements
specific to brake and clutch repair are in Appendix F of the general industry standard (see:
https://www.osha.eov/laws-rees/regiilations/standardmtm.ber/191C	?pF). The
updated standards are an important consideration for interpreting worker exposure studies
because observed exposure levels prior to promulgation of OSHA's amended asbestos standard
may not be representative of exposures at establishments that currently comply with OSHA
requirements.
2. The second important consideration for exposure assessment is the asbestos content in brake
dust. Due to the high friction environment in vehicle braking, asbestos fibers in the brake
material degrade both chemically and physically. While brake linings and pads at installation
may contain between 40 and 50 percent chrysotile asbestos (i.e., fibers longer than 5
micrometers) (OSI 36), brake dust is largely made up of particles and fibrous structures
less than 5 micrometers in length, which would no longer be measured as asbestos by PCM. In
1989, NIOSH reviewed brake dust sampling data and concluded "the vast majority of samples"
reviewed contained less than 5 percent asbestos (OSHA, 2006). Other researchers have reported
lower values, indicating that brake dust typically contains less than 1 percent asbestos
(Paustenbach et at.. 2003). Chemical changes also occur, such as transformation into forsterite (a
deformation product of chrysotile), or to transition series fibers (chrysotile/forsterite), but
chemical changes are thought as less important than physical changes for biological outcomes
(OSI )6).
The amount of time that workers repair and replace automobile brakes depends on many factors. The
literature suggests that a typical "brake job" for a single vehicle takes between 1 and 2 hours
(Paustenbach et at.. 2003). While most automotive mechanics perform various repair tasks, some
specialized mechanics work exclusively on brakes. The literature also suggests that the number of brake
repair jobs performed by automotive service technicians and mechanics range from 2 to 40 per week
(Madl et at.. 2008).
Automobile Clutch Repair and Replacement
Repairing and replacing asbestos-containing clutch assemblies could also result in asbestos exposure.
Workers typically elevate vehicles to access the clutch assembly, remove dust and debris, and perform
repair and replacement tasks accordingly. Like asbestos in brakes, asbestos in clutch discs degrades with
use. (Cohen and Van Orden. 2008) evaluated clutch assemblies from a vehicle salvage yard and found
that clutch plates, on average, contained 43 percent asbestos, while the dust and debris in clutch
housings, on average, contained 0.1 percent asbestos (Cohen and Van Orden. 2008).
However, clutch repair and replacement differ from brake work in two important ways. First, clutches
generally do not need to be repaired as frequently. By estimates made in 2008, clutches typically last
three times longer than brake linings (Cohen and Van Orden. 2008). Second, a common clutch repair
method is to remove and replace the entire clutch assembly, rather than replacing the clutch disc
component (Cohen, and Van. Orden. 2008). Finally, vehicles have only one clutch assembly and up to
four brakes; therefore, clutch servicing only involves repair of one apparatus, while brake servicing
involves multiple components. These three and other factors likely result in clutch repair asbestos
exposures being lower than comparable brake repair asbestos exposures.
Page 107 of 352

-------
2.3.1.7.3 Number of Sites and Potentially Exposed Workers - Aftermarket
Automotive Brakes/Linings and Clutches
EPA considered several data sources when estimating the number of workers directly exposed to
asbestos when working with aftermarket automotive products. In the late 1980s, NIOSH conducted a
series of industrial hygiene surveys on brake repair facilities, and the Agency estimated that 155,000
brake mechanics and garage workers in the United States were potentially exposed to asbestos (OSHA.
2006). In 1994, OSHA estimated as part of its updated asbestos rulemaking that 676,000 workers
performed automotive repair activities, and these workers were found in 329,000 establishments {i.e.,
approximately two workers per establishment) (59 FR 40964). Additionally the Bureau of Labor
Statistics estimated that 749,900 workers in the United States were employed as automotive service
technicians and mechanics in 2016 (	S. 2019). This includes workers at automotive repair and
maintenance shops, automobile dealers, gasoline stations, and automotive parts and accessories stores.
ONU exposures associated with automotive repair work are expected to occur because automotive repair
and maintenance tasks often take place in large open bays with multiple concurrent activities. EPA did
not locate published estimates for the number of ONUs for this COU. However, consistent with the
industry profile statistics from OSHA's 1994 rulemaking, EPA assumes that automotive repair
establishments, on average, have two workers who perform automotive repair activities and therefore
EPA assumes an equal number of exposed workers and ONUs for this COU.
EPA estimated the number of potential individuals exposed to asbestos using the limited available
information on the potential market share of asbestos brakes. Details are provided in Section 4.3.7:
Confidence in the Human Health Risk Estimations. EPA assumes that asbestos brakes may represent
only approximately 0.05% of aftermarket automotive brakes. By applying this factor (0.05%) to the
universe of automotive service technicians and mechanics (749,900), EPA's estimate of potentially
exposed workers is 375. For the same reasons noted above, EPA assumes an equal number (375) of
ONUs for this COU.
2.3.1.7.4 Occupational Inhalation Exposures - Aftermarket Automotive
Brakes/Linings and Clutches
To identify relevant occupational inhalation exposure data, EPA reviewed reasonably available
information from OSHA, NIOSH, and other literature. All research steps are documented below, with
more detailed discussion on the most relevant data sources, which EPA determined to be the post-1980
studies conducted by NIOSH and the post-1980 publications in the peer-reviewed literature.
Automobile Brake Repair and Replacement
EPA first considered worker exposure data from OSHA compliance inspections. EPA reviewed data that
OSHA provided for 2011 to 2016 inspections, but these data did not include any PBZ asbestos
measurements for the automotive repair and maintenance industry. For additional insights into OSHA
sampling results, EPA considered the findings published by Cowan et al. (2015). These authors
summarized OSHA workplace compliance measurements from 1984 to 2011, which included 394 PBZ
samples obtained from workers at automotive repair, services, and parking facilities (Cowan et al..
2015). Because the samples were taken for compliance purposes, all measurements were presumably
made using OSHA-approved methods {i.e., PCM analyses of filters). Table 2-14. summarizes these data,
which suggest that asbestos exposures for this COU decreased from the mid-1980s to 2011.
Page 108 of 352

-------
Table 2-14. PBZ Asbestos Concentrations Measured by OSHA for Workers at Automotive Repair,
		Services, and Parking Facilities		
l ime l-'rame
Number of
Samples
Number of
Samples Noil-
Detect for
Asbestos
Number of
Samples with
Delected
Asbestos
Uange of Detected
Asbestos
Concentrations (I7cc)
1984-1989
274
241
33
0.0031 -35.6
1990-1999
101
101
0
N/A
2000-2009
17
17
0
N/A
2010-2011
2
2
0
N/A
Total
394
361
33
0.0031 -35.6
Data from Cowan et at (2015).
Data are personal breathing zone (PBZ) concentrations of unknown duration.
EPA then considered relevant NIOSH publications, focusing on those published since 1980, because
earlier publications evaluated work practices (e.g., compressed air blowdown of brake dust) that are no
longer permitted. Specifically, EPA considered five NIOSH in-depth survey reports published in 1987
and 1988 (Cooper et at.. 1988. 1987; Godbev etaL 1987; Sheehv et al. 1987a; Sheehv etaL 1987b)
and a 1989 NIOSH publication that reviewed these findings (OSHA. 2006). The NIOSH studies
investigated PBZ asbestos exposures among workers who employed various dust removal methods
while servicing brakes. These methods included use of vacuum enclosures, HEPA-filtered vacuums, wet
brushing, and aerosol sprays. In three of the NIOSH studies, the average (arithmetic mean) asbestos
concentration over the 2-hour duration of brake repair jobs was below the detection limit (0.004 f/cc).
The other two studies reported average (arithmetic mean) asbestos concentrations over the brake job
duration of 0.006 f/cc and 0.007 f/cc. NIOSH's summary of the five studies concluded that "exposures
can be minimal" provided workers use proper dust control methods (OSHA. 2006).
EPA also considered the published literature on asbestos exposures associated with automobile brake
repair. This review focused on post-1980 publications that reported original asbestos PBZ measurements
for business establishments in the United States. While EPA is aware of and thoroughly reviewed
studies of asbestos exposure among brake mechanics in various foreign countries (e.g., Australia,
Colombia, Iran, Norway), EPA focused on U.S. business establishments due to the availability of
measurements and the fact that OSHA's asbestos standard mandates controls and other safe work
practices that do not apply in other countries. Further, the profile of brakes encountered in U.S. vehicles
differs from what is seen in other countries.
The following peer-reviewed publications met EPA's selection criteria (and all were given a high rating
in the data evaluation; see supplemental file (U.S. EPA. 2020D:
• The first study was published in 2003, but it evaluated asbestos exposure for brake repair jobs
conducted on vehicles with model years 1965-1968. The study considered work practices
commonly used during the 1960s, such as compressed air blowdowns and arc grinding and
sanding of surfaces (Blake et al.. 2003). PBZ samples were collected during seven test runs, and
measured asbestos concentrations ranged from 0.0146 f/cc to 0.4368 f/cc, with the highest level
observed during arc grinding operations. This range of measurements was for sample durations
ranging from 30 minutes to 107 minutes. These observations were considered in the occupational
exposure evaluation even though they likely represent an upper-bound estimate of today's
exposures. While arc grinding during brake replacement is not believed to be a common practice
today, EPA conducted web searches that identified recent vides showing individuals using arc
Page 109 of 352

-------
grinding during brake repair. Given the evidence of the ongoing activity, even if uncommon,
EPA retained this study in the exposure assessment.
•	The second study, conducted in 2008, measured worker asbestos exposure during the unpacking
and repacking of boxes of asbestos-containing brake pads and brake shoes (Madl et al. 2008).
The asbestos-containing brake materials were originally manufactured for 1970-era automobiles,
and the authors obtained the materials from vintage parts suppliers and repair facilities. The
study evaluated how exposure varied with several parameters, including type of brake material
(e.g., drum, shoe) and worker activity (e.g., packing, unpacking, cleaning). The range of personal
breathing zone concentrations observed across 70 short-term samples was 0.032 f/cc to 0.836
f/cc, with the highest exposure associated with unpacking and packing 16 boxes of asbestos-
containing brake pads over approximately 30 minutes. EPA acknowledges that this study did not
characterize actual brake repair or servicing activities. However, workers must handle
aftermarket parts (i.e., open and close boxes) as part of their overall repair jobs. For this reason,
EPA continued to include this study in its estimates of worker and ONU exposures.
•	The third study examined asbestos exposures during brake repair operations, considering various
worker activities (Weir et al.. 2001). EPA did not use this study's measurements in the
occupational exposure evaluation because the publication lacked details necessary for a thorough
review. For instance, this study (in contrast to all others considered for this COU) did not report
on the complete data set, the time-weighted average exposure values did not include an exposure
duration, and the TEM metrics were qualitative and vague. For these and other reasons, the study
was considered for contextual information, but not quantitatively in the exposure assessment.
Automobile Clutch Repair and Replacement
EPA considered the same automotive brake repair and replacement information sources when assessing
asbestos exposure during automobile clutch repair and replacement but did not identify relevant data
from OSHA monitoring data or NIOSH publications. EPA identified three peer-reviewed publications
(Blake et al.. 2008; Cohen and Van Orden. 2008; Jiang et al.. 2008) that measured worker asbestos
exposure during automotive clutch repair. Though the clutch repair data are limited in comparison to
brake repair exposure data, the three studies suggest that personal breathing zone asbestos
concentrations while repairing or replacing asbestos-containing clutches are comparable to the
concentrations for brake repair and replacement activity. However, the frequency of workers performing
this task is expected to be lower than the brake. As noted earlier, EPA used the available brake repair
data as its basis for deriving exposure estimates for the entire COU of working with aftermarket
automotive parts even though it is clear that the brake-related exposure concentrations may overstate
exposures that occur during clutch repair.
2.3.1.7.5 Exposure Data for Use in the Risk Evaluation for Asbestos Part 1:
Chrysotile Asbestos - Aftermarket Auto Brakes/Linings and Clutches
Table 2-15. presents the asbestos exposure data that EPA used in this Part 1 of the risk evaluation for
asbestos for working with asbestos-containing aftermarket automotive parts. EPA's basis for selecting
the data points appears after the table.
Page 110 of 352

-------
Table 2-15. Summary of Asbestos Exposures During Replacement of Aftermarket Automotive
Parts Used in the Risk Evaluation for Asbestos Part 1: Chrvsotile Asbestos
	,	*L	

Kxposurc Levels (I'/cc)
Occupational
Workers
OMs
Kxposurc Scenario
Central
Tendency
Nigh-end
Confidence
Rating
Central
Tendency
lligh-end
Confidence
Rating
Repairing or replacing
brakes with asbestos-


Medium


Medium
containing aftermarket
0.006
0.094

0.001
0.002

automotive parts: 8-hour






TWA exposure






Repairing or replacing
brakes with asbestos-


Medium


Medium
containing aftermarket
0.006
0.836

0.001
0.002

automotive parts: short-






term exposure






Worker Exposures
•	The central tendency short-term TWA exposure value for workers is based on the seven studies
found to include relevant measurements (Madl et at.. 2.008; Blake et at.. 2003; Cooper et at..
1988. 1987; Godbev etal. 1987; Sheehv et at." 1987a- Sheehv et at.fl 987b). For each study,
EPA identified the central tendency short-term exposure, which was either reported by the
authors or inferred from the range of data points, and the value in Table 2-15. (0.006 f/cc) is the
median of those central tendencies from those seven studies. Thus, three of the studies reported
central tendency concentrations lower than 0.006 f/cc, one reported a central tendency
concentration of 0.006 f/cc, and the other three studies reported higher exposure concentrations.
•	Most of the studies selected for review do not present 8-hour TWA exposure values. They
instead typically report "brake job TWA exposures"—or exposures that occur over the duration
of a single brake repair activity. EPA selected a central tendency 8-hour TWA exposure value for
workers (0.006 f/cc) by assuming the median short-term exposure level could persist for an
entire workday. This is a reasonable assumption for full-time brake repair mechanics, who may
conduct 40 brake repair jobs per week, and a protective assumption for automotive mechanics
who do not repair brakes throughout their shifts.
•	The high-end short-term TWA exposure value for workers (0.836 f/cc) is the highest short-term
personal breathing zone observation among the seven studies that met the review criteria (Madl
et at.. 2008). The highest concentration was from a 15-minute average sample and therefore
might overstate (by no more than a factor of two) the 30-minute concentration. The high-end 8-
hour exposure value for workers (0.094 f/cc) is based on a study (Stake et at.. 2003) that used arc
grinding during brake repair with no exposure controls, which is a representation of a high-end
exposure scenario of today's work practices.
ONU Exposures
EPA used area sampling results from the five NIOSH studies cited above to derive ONU exposure
estimates for this condition of use. In each study, NIOSH collected area samples at the fender and at the
axle of the vehicle as its brakes were being serviced. EPA considered these area samples to be
representative of ONU exposures, because other workers may conduct other tasks at these locations
during brake servicing. The duration of the area sample was the time needed to replace a vehicle's
Page 111 of 352

-------
brakes or two hours, whichever was longer. Across the five studies, more than 70 area samples were
collected at these locations. The area samples were tested for asbestos using PCM, and all were non-
detect. NIOSH reported arithmetic mean concentrations for these samples as <0.002 f/cc. Based on these
data, EPA assumed the ONU central tendency exposure concentration to equal one-half the detection
limit, or 0.001 f/cc; and EPA assumed the ONU high-end exposure concentration to equal the detection
limit for most samples, or 0.002 f/cc. These values were applied to both 8-hour TWA exposure and
short-term exposure. It is possible that ONU may not be exposed at full shift, every workday.
2.3.1.7.6 Data Assumptions, Uncertainties and Level of Confidence
The universe of automotive repair establishments in the United States is expected to have large
variability in the determinants of exposure to asbestos during brake repair. These exposure determinants
include, but are not limited to, vehicle age, type of brake assembly (disc vs. drum), asbestos content of
used and replacement parts, dust control measures used, brake servicing techniques (e.g., use of arc
grinding), number of vehicles serviced per day, and duration of individual repair jobs. It is uncertain if
the studies EPA cited and used fully capture the distribution of determinants of exposure of current
automotive brake jobs, and some of the studies reviewed for this Part 1 of the risk evaluation for
asbestos are based on practices that are not widely used today.
PCM-based personal exposure measurement in an automotive repair facility may overstate asbestos
exposures, which some studies have demonstrated through TEM analyses of filter samples (Blake et at..
2003; Weir et ai. 2001). PCM measurements are based entirely on dimensional criteria and do not
confirm the presence of asbestos, as can be done through supplemental analyses by TEM or another
confirmatory method. Automotive repair facilities involve many machining operations that can release
non-asbestos airborne fibers, such as cellulose fibers from brushes and metal and plastic fragments from
body repair (Blake et al. 2.008).
EPA considered the quality and uncertainties of the data to determine a level of confidence for the
assessed inhalation exposures for this condition of use. The primary strength of this assessment is the
use of monitoring data, which is the highest level of the inhalation exposure assessment approach
hierarchy. The overall confidence ratings from systematic review for these data were high. The
monitoring data were all collected from U.S.-based vehicle maintenance and repair shops. While these
studies were conducted after the implementation of the OSHA rule, many of the studies were conducted
in the late 1980s and may not be representative of current operations. This is particularly true for the
study that evaluated arc grinding. However, that study's results are directly reflected only in the high-
end exposure estimate. (Note: The central tendency value in the table comes from one of the NIOSH
studies.) EPA believes it is appropriate to consider arc grinding in the high-end exposure category, given
evidence that this work practice continues today, albeit uncommonly.
The ONU exposure estimates are based on a dataset comprised entirely of non-detect observations and
therefore are uncertain. The uncertainty cannot be reduced using other sampling results that EPA
considered for this analysis. EPA assigns a "medium" confidence factor for these exposure
concentration estimates.
Based on these strengths and limitations of the data, the overall confidence for EPA's assessment of
occupational and ONU inhalation exposures for this scenario is medium.
Page 112 of 352

-------
2,3,1,8 Other Vehicle Friction Products
While EPA has verified that U.S. automotive manufacturers are not installing asbestos brakes on new
cars for domestic distribution, EPA has identified a company that is importing asbestos-containing
brakes and installing them in their cars in the United States. These cars are exported and not sold
domestically. In addition, there is a limited use of asbestos-containing brakes for a special, large
transport plane (the "Super-Guppy") by the National Aeronautics and Space Administration (NASA).
2.3.1.8.1 Installing New Brakes on New Cars for Export Only
EPA did not identify any studies that contain exposure data related to installation of asbestos-containing
brakes at an Original Equipment Manufacturer (OEM). As a result, the exposure assessment approach
used for the aftermarket automotive brakes/linings and clutches described in Section 2.3.1.7 was also
used for this COU and is reported here in Table 2-16.
Most, if not all, of the literature that EPA reviewed pertained to servicing vehicles that were already
equipped with asbestos-containing brakes and clutches. This servicing requires the removal of asbestos-
containing parts and installation of non-asbestos-containing replacement parts. When removing an
asbestos-containing part, one of the main sources of exposure is the dust and debris that must be
removed from the brake housing, which is not the case for installing OEM asbestos-containing
components on new vehicles. Therefore, the aftermarket auto brakes/linings and clutches exposure value
used to assess this COU may be an overestimate. The actual exposure for OEM installation is likely to
be lower.
Table 2-16. Other Vehicle Friction Products Exposure Levels (from Aftermarket Automotive
Occupational
l-.xposure
Scenario


Kxposuri
l.eyels (I'/cc)


Workers
OMs
Central
Tendency
Nigh-end
Con fidcncc
Kill inซ
Central
Tendency
High-end
Confidence
Rating
Installing brakes
with asbestos-






containing
automotive parts:
0.006
0.094
Low
0.001
0.002
Low
8-hour TWA






exposure






Installing brakes
with asbestos-






containing
automotive parts:
0.006
0.836
Low
0.001
0.002
Low
short-term






exposure






Data Assumptions. Uncertainties and Level of Confidence
The assumptions and uncertainties described above under Section 2.3.1.7.6 apply here. In addition, the
procedure for installing asbestos containing brakes/friction products into a new vehicle does not involve
removing of old asbestos-containing brakes/friction products. Thus, although we do not have the data,
the actual exposure could be lower than estimated here. For this COU, ONU may not be exposed at full
shift, every workday.
Page 113 of 352

-------
Based on these strengths and limitations of the underlying data described above and in Section 2.3.1.7.6,
the overall confidence for EPA's assessment of occupational and ONU inhalation exposures for this
scenario is low.
2.3.1.8.2 Use of Brakes/Frictional Products for a Single, Large Transport
Vehicle (NASA Super-Guppy)
This section evaluates asbestos exposures associated with brake block replacement for the Super Guppy
Turbine (SGT) aircraft, which is operated by the National Aeronautics and Space Administration
(NASA). The SGT aircraft (Figure 2-11) is a specialty cargo plane that transports oversized equipment,
and it is considered a mission-critical vehicle (NASA. 2020b). The aircraft brake blocks contain
chrysotile asbestos, and this section evaluates potential worker exposures associated with servicing the
brakes. This section is based on information provided by NASA.
Figure 2-11. NASA Super Guppy Turbine Aircraft
Photograph courtesy of NASA
Aircraft and Brake Description
Only one SGT aircraft is in operation today, and NASA acquired it in 1997. The SGT aircraft averages
approximately 100 flights per year (NASA. 2020a). When not in use, it is hangered at the NASA
Aircraft Operating Division's (AOD) El Paso Forward Operating Location in El Paso, Texas. This is
also where the aircraft is serviced (VASA. 2020b).
The SGT aircraft has eight landing gear systems, and each system has 32 brake blocks. The individual
blocks (Figure 2-12) contain 43 percent chrysotile asbestos; and they are 4 inches long, 4 inches wide,
and 1 inch thick (NASA,. 2020b). Each brake block weighs approximately 12.5 ounces.
Page 114 of 352

-------
Figure 2-12. Brakes for NASA Super Guppy Turbine Aircraft
Photograph courtesy of NASA
Worker Activities
Replacing asbestos-containing brake blocks is the principal worker activity potentially associated with
asbestos exposure, and this task is performed by four certified technicians. According to NASA, the
brake blocks are not replaced due to excessive wear; rather, they are typically replaced because they
have become separated from the brake system or because they have become covered with hydraulic
fluid or other substances (NASA 2020a). This is an important observation, because in EPA's judgment,
worn brake blocks would be more likely to contain dusts to which workers would be exposed.
In materials provided to EPA, NASA described the process by which workers replace brake blocks. This
process begins by removing the brakes from the landing gear. To do so, the SGT aircraft is raised at the
axle pads, and the landing gear is opened to allow workers access to the individual brake systems. The
workers remove the brakes from the aircraft and clean the brakes at an outdoor wash facility.
The certified technicians then take the brakes into a ventilated walk-in booth (Figure 2-13), which is
where brake block replacement occurs. According to a NASA job hazard analysis, workers use wet
methods to control release of asbestos dust during this task (NASA. 2020a). The workers use spray
bottles containing a soap-water mixture to keep exposed surfaces damp when replacing brake blocks.
Waste dusts generated during this activity are collected using a high-efficiency particulate air vacuum;
and all asbestos-containing wastes, including vacuumed waste, are double-bagged (NASA Occupational
Health, 2020) and disposed of according to waste management regulations for asbestos (NASA, 2020b).
Page 115 of 352

-------
Figure 2-13. Ventilated Walk-in Booth Where Brakes Pads Are Replaced
Photograph courtesy of NASA
The four certified technicians for SGT aircraft brake replacement receive annual training on asbestos.
The training course addresses asbestos health hazards, work practices to reduce generation of airborne
asbestos dust, and information on how PPE can reduce exposures (NASA Occupational Health. 2020).
The training also indicates that brake replacement workers who follow proper methods for controlling
asbestos dust releases are not required to use respiratory protection ( -IASA Occupational Health. 2020).
Respirator usage is also not required because measured exposures were below applicable occupational
exposure limits (NASA. 2020a). Despite respiratory protection not being required, NASA informed
EPA that some certified technicians choose to use half mask air-purifying respirator with P-100
particulate filters when replacing brake blocks ( MASA. 2020a).
Brake pad replacement for the one SGT aircraft occurs infrequently, approximately four times per year
(NASA. 2020a). According to NASA, the four certified technicians who service the aircraft spend
approximately 12 hours per year replacing brake pads.
Number of Sites and Potentially Exposed Workers
Brake pad replacement for the SGT aircraft occurs at only one site nationwide: a NASA facility located
in El Paso, Texas (MAS A. 2020b).
Over the course of a year, only four certified technicians at this location perform brake pad replacement;
and one or two of these technicians will perform individual brake pad replacements (NASA, 2020b).
Because the brake replacement work occurs in a ventilated walk-in booth, asbestos fibers likely are not
released into the general workspace where ONUs may be exposed.
Therefore, for this condition of use, EPA assumes four workers may be exposed, and no ONUs are
exposed.
Worker Inhalation Exposures
EPA's estimate of occupational inhalation exposures for this condition of use are based on five worker
exposure samples that NASA collected in 2014 (NASA. (020a). The sampling was conducted according
to NIOSH Method 7400, and asbestos was not found above the detection limit in any of the samples.
EPA estimated worker exposure levels for the risk evaluation as follows:
Page 116 of 352

-------
ฆ	Three of the five sampling results that NASA provided were labeled as "8-hour TWA"
observations, and EPA considered these to be representative of full shift exposures. The three
results, based on sampling durations of 83, 17, and 85 minutes, were: <0.003 f/cc, <0.006 f/cc,
and <0.0089 f/cc (NASA. 2.020a). To calculate the central tendency for full shift exposure, EPA
replaced the three observations with one-half the detection limit and calculated the arithmetic
mean of those three values. By this approach, EPA calculated a central tendency concentration of
<0.003 f/cc. For the high-end full shift exposure estimate, EPA used the highest detection limit
across the three samples.
ฆ	Two of the five sampling results that NASA provided were labeled as being evaluated for "30-
minute excursion limits"; and EPA considered these to be representative of short-term exposures.
The two results, based on sampling durations of 30 and 35 minutes, were: <0.044 f/cc and
<0.045 f/cc. Following the same approach that was used for full shift exposures, EPA estimated a
central tendency short-term exposure of <0.022 f/cc and a high-end short-term exposure of
<0.045 f/cc.
ฆ	According to NASA (NASA, 2020c), records from a recent 36-month period indicate 3.6 brakes
were changed each year with an average time of 3.3 hours per brake change.
Based on these assumptions, EPA will use the exposure values in Table 2-17.
Table 2-17. Summary of Asbestos Exposures During Replacement of Brake Pads/Blocks in the
NASA Super Guppy Used in the Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos
Occupational
Kxposurc
Scenario
Kxposurc Levels (f/cc)
Workers
OMs
Central
Tendency
lligh-end
Confidence
Rating
( cntral 	
Iligh-cnd
1endenev
Confidence
Rating
Replacing brake pads:
8-hour TWA exposure
<0.003
<0.0089
High
Not expected
High
Replacing brake pads:
short-term exposure
(30 minutes)
<0.022
<0.045
High
EPA assigned a confidence rating of "high" for these exposure data. This rating was based on the fact
that monitoring data are available from the one site where this condition of use occurs. Further,
replacement of SGT aircraft brake blocks occurs approximately 12 hours per year, and the five available
sampling events spanned more than 4 hours. Therefore, the available data, which were collected using
an appropriate NIOSH method, represent almost one-third of the worker activity over an entire calendar
year. The spatial and temporal coverage of these data are greater than those for any other condition of
use in this Part 1 of the risk evaluation for asbestos.
ONI J Inhalation Exposures
As noted previously, EPA assumes no ONU exposures occur, because the worker activity with the
highest likelihood of releasing asbestos occurs in a walk-in ventilated booth, where ONUs are not
present.
Page 117 of 352

-------
2,3,1,9 Other Gaskets-Utility Vehicles (UTVs)
2.3.1.9.1 Process Description - UTV Gasket installation/Servicing
EPA has identified the use of chrysotile asbestos-containing gaskets in the exhaust system of a specific
type of utility vehicle available for purchase in the United States. This COU is identified as "other
gaskets" in Table 1-4. of Section 1.4.2. It is known that these UTVs are manufactured in the United
States, so EPA expects that there is potential for exposures to workers who install the gaskets during
assembly and workers who may repair these vehicles.
To derive occupational exposure values for this Part 1 of the risk evaluation for asbestos, EPA is
drawing on a review of several studies in the literature which characterize exposure scenarios from
asbestos-containing gasket replacement in different types of vehicles.
2.3.1.9.2 Worker Activities - UTV Gasket Installation/Servicing
The UTV manufacturers receive the pre-cut gaskets which are then installed during manufacture of the
UTV. The gaskets may be removed during servicing of the exhaust system.
Thirty studies relating to gasket repair/replacement were identified and reviewed as part of the
systematic review process for the consumer exposure scenario (see Section 2.3.2.2); resulting in
identifying three studies as being relevant to gasket installation and replacement in vehicles (see Table
2-29).
2.3.1.9.3 Number of Sites and Potentially Exposed Workers - UTV Gasket
Installation/Servicing
EPA estimated the number of UTV service technicians and mechanics potentially exposed to asbestos
by assuming that asbestos-containing gaskets are most likely to be replaced at UTV dealerships that sell
these vehicles.18 However, no NAICS codes are specific to UTV dealers. These establishments are
classified under the 4-digit NAICS 4412, "Other Motor Vehicle Dealers." Table 2-188. lists the specific
industries included in that 4-digit NAICS. The industry most relevant to UTV dealers is the 7-digit
NAICS code 4412281, "Motorcycle, ATV, and personal watercraft dealers." The 2012 Economic
Census reports 6,999 establishments in this industry.
Table 2-18. Number of Other Motor Vehicle Dealers
2012 NAICS code
2012 NAICS Code Description
Number of
Establishments
4412
Other motor vehicle dealers
14,249
44121
Recreational vehicle dealers
2,605
441222
Boat dealers
4,645
441228
Motorcycle, ATV, and all other motor vehicle dealers
6,999
4412281
Motorcycle, ATV, and personal watercraft dealers
5,098
4412282
All other motor vehicle dealers
1,901
Source: ("U.S. Census Bureau. 2016a).
18 While UTV owners may have their vehicles serviced at repair and maintenance shops that are not part of dealerships, the
total number of sites and workers exposed may not necessarily change from the estimates in this analysis. More vehicles
being repaired in other types of repair shops would mean fewer vehicles being repaired (and fewer workers exposed) in
dealerships. This analysis simplifies the estimates by assuming that engine repairs all occur at dealerships.
Page 118 of 352

-------
The Economic Census also reports the product and service line statistics for retail establishments down
to the 6-digitNAICS code level. Product and service code 20593 represents "All-terrain vehicles
(ATVs) and personal watercraft." Out of the 6,999 establishments in the 6-digitNAICS code 441228,
Table 2-199. shows that 2,989 of them deal in ATVs and personal watercraft. For purposes of this
assessment, EPA assumes that approximately half of them (1,500 establishments) sell and repair UTVs
and ATVs, and that the other half specialize in personal watercraft.
Table 2-19. Number of ATV and Watercraft Dealers in NAICS 44
1128
2012 NAICS
Code
2012 NAICS Code
Description
Products and
Services Code
Products and Services
Code Description
Number of
Establishments
441228
Motorcycle, ATV,
and all other motor
vehicle dealers
20593
All-terrain vehicles
(ATVs) & personal
watercraft
2,989
Source: (U.S. Census Bureau. 2016b).
The next step in estimating potentially exposed workers is to determine the number of workers engaged
in UTV repairs. This number had to be estimated because the Bureau of Labor Statistics does not
provide employment data by occupation for NAICS 4412281 and because Standard Occupational
Classification (SOC) codes are not specific to workers engaged in UTV repairs. Reasonably available
information to estimate potentially exposed workers is SOCs at the 4-digit NAICS level (NAICS 4412),
which includes dealers in recreational vehicles, boats, motorcycles and ATVs. Table 2-20. presents
SOCs that reflect the types of workers that may repair engines and identifies 41,930 workers in relevant
occupations in NAICS 4412.19
Table 2-20. Selected Mechanics and Repair Technicians in NAICS 4412 (Other Motor Vehicle
	Dealers)		
Occupation (SOC code)
Employment
First-Line Supervisors of Mechanics, Installers, and Repairers (491011)
4,140
Aircraft Mechanics and Service Technicians (493011)
120
Automotive Service Technicians and Mechanics (493023)
3,360
Motorboat Mechanics and Service Technicians (493051)
9,800
Motorcycle Mechanics (493052)
13,250
Recreational Vehicle Service Technicians (493092)
11,260
Total
41,930
Source: (U.S. BLS. 2019).
Based on the estimates for NAICS 4412 in Table 2-18. and Table 2-20., Table 2-21. calculates that
across all entities in NAICS 4412, approximately 3 employees per dealership engage in occupations
potentially relevant to UTV repairs.
19 This count excludes occupations in NAICS 4412 that are less likely to engage in engine repair involving gaskets similar to
those found in UTVs. These would be occupations such as Electrical and Electronic Equipment Mechanics, Installers, and
Repairers (SOC 492000), Automotive Body and Related Repairers (SOC 493021), Mobile Heavy Equipment Mechanics,
Except Engines (SOC 493042), Tire Repairers and Changers (SOC 493093) and Outdoor Power Equipment and Other Small
Engine Mechanics (SOC 493053). The latter covers workers who repair items such as lawn mowers, chain saws, golf carts,
and mobility scooters, which do not generally have engines similar to UTVs.
Page 119 of 352

-------
Table 2-21. Number of Employees per Establishment in NAICS 4412 in Re
evant Occupations
Description
Number
Number of other motor vehicle dealers (NAICS 4412) (see Table 2-17.)
14,429
establishments
Number of mechanics and repair technicians in NAICS 4412 that may repair
engines in recreational vehicles, boats, motorcycles, ATVs, etc. (see Table
2-20.)
41,930 employees
Estimated average number of employees per establishment that may repair
motor vehicle engines (calculated as 41,930 divided by 14,429)
~3 employees per
establishment
Assuming that the average number of mechanic and service technicians across NAICS 4412 is
applicable to NAICS 4412281, Table 2-22. combines the estimate of 1,500 dealerships repairing and
maintaining UTVs/ATVs with the estimated average of 3 employees per establishment from Table 2-21.
to generate an estimate of 4,500 total employees that may repair UTV engines.
Table 2-22. Estimated Number of Sites and Employees for UTV Engine Repair
Description
Number of
establishments
Estimated number of dealerships repairing and maintaining UTVs/ATVs
1,500
Estimated average number of employees per establishment that may repair
motor vehicle engines (see Table 2-21.)
3
Estimated total number of employees that may repair UTVs
4,500
2.3.1.9.4 Occupational Inhalation Exposures for Use in the Risk Evaluation
for Asbestos Part 1: Chrysotile Asbestos for UTV Gasket
Installation/Servicing
No information from OSHA, NIOSH, or the scientific literature was available on occupational
exposures to asbestos associated with installing and servicing gaskets in UTVs. EPA therefore
considered studies of similar worker exposure scenarios to use as a surrogate. Multiple publications (see
Section 2.3.2.2) report on occupational exposures associated with installing and servicing gaskets in
automobiles. However, EPA located only one study (Paustenbach et at.. 2006) that examined exposures
associated with replacing vehicle exhaust systems, which is the UTV component where asbestos-
containing gaskets are found. Therefore, EPA based its occupational inhalation exposure assessment for
UTV gasket installation and servicing on this study.
Worker Exposures
EPA's estimate of occupational inhalation exposures is based on a 2006 study (Paustenbach et al.
2006). in which workers at a muffler shop removed exhaust systems from 16 vehicles. The vehicle
model years ranged from 1946 to 1970; and 12 of the 16 vehicles were found to have asbestos in some
combination of the mufflers, manifold gaskets, and exhaust pipe gaskets. The measured asbestos content
in these components ranged from 9.5 to 80.1 percent, with only chrysotile asbestos fibers detected.
The study considered multiple types of exhaust system projects, including removal of different
combinations of mufflers, exhaust pipes, and exhaust manifolds and conversion from single to dual
exhaust systems. The time needed to remove an exhaust system and install a new one lasted up to 4
hours, but according to the study (Paustenbach et al.. 2006). workers reportedly spent less than one
minute for removal of each gasket. It often took the worker more time to access the gasket due to rusted
Page 120 of 352

-------
bolts than to remove the gasket. Workers reportedly spent less than one minute on the removal of each
gasket. It often took the worker more time to access the gasket due to rusted bolts than to remove the
gasket. All jobs were performed indoors at the muffler shop, with service bay doors closed, and no other
vehicle repair work occurring at the same time.
Personal breathing zone measurements were taken using sampling materials consistent with NIOSH
Method 7400. Overall, 23 valid personal breathing zone samples were collected from mechanics and
tested with PCM. Some additional samples were taken, but they were overloaded with particulate
material and could not be analyzed. Among the 23 valid samples, 17 were non-detect for asbestos by
PCM analysis; and 6 samples contained asbestos at concentrations up to 0.0505 f/cc. The TEM analyses
identified asbestos fibers in 7 of the sampling filters.
Overall, based on the PCM analysis of the 23 valid samples, the study authors reported an average
worker asbestos concentration of 0.024 f/cc and a maximum concentration of 0.066 f/cc. (Note: 1) The
authors reported an average "PCM-adjusted" concentration that is 18 percent lower than the un-adjusted
result. The adjustment accounts for the amount of fibers confirmed by TEM as being asbestos. 2) This
appears to be a detection level 0.132 f/cc divided by two, contrary to more standard division by square
root of two (approximately 1.4), thus underestimating the maximum concentration. The average and
maximum concentrations pertain to the times when sampling occurred, and sampling durations ranged
from 9 to 65 minutes. The study authors calculated an 8-hour TWA exposure concentration of 0.01 f/cc,
based on a worker performing four exhaust system removal tasks in one shift.
EPA used the personal breathing zone (PBZ) values for the worker as follows: the last row in Table 2-30
shows the maximum concentration calculated from the information within the study (Paustenbach et at..
2006) as the high-end estimated concentration for the worker and the mean concentration calculated
from the information within the study as the central tendency concentration (see Table 2-23 below).
ONU Exposures
The same publication (Paustenbach et at.. 2.006) includes area sampling results that EPA found
appropriate for ONU exposures (rather than what the paper defines as a bystander). These samples were
collected at breathing zone height at locations near the ends of the muffler shop bays where the exhaust
system work was performed. The area sample durations ranged from 25 to 80 minutes, and these
samples were collected during exhaust system work. Overall, 21 area samples from these locations were
analyzed by PCM; and 16 of these samples were non-detects for asbestos. Among the PCM data from
this subset of area samples, the authors report that the average asbestos concentration was 0.005 f/cc and
the maximum asbestos concentration was 0.015 f/cc. The study authors did not report 8-hour TWA
concentrations for the area sample locations. EPA used these average and maximum asbestos
concentrations to characterize ONU exposures.
Table 2-23. UTV Gasket Installation/Servicing Exposure Levels for the Risk Evaluation for
		Asbestos Part 1: Chrysotile Asbestos	
Occupational
Kxposure
Scenario


8-hr TWA I'.xposure l.e\els (f/cc)

Asbestos \\ orkcr

OM
Cent nil
Tendency
lli
LIN
U.U24
u.uoo
Modi mil
U.UU5
U. U15
Medium
Page 121 of 352

-------
2.3.1.9.5 Data Assumptions, Uncertainties and Level of Confidence
A principal assumption made in this assessment is that worker asbestos exposures for removing
automobile exhaust systems are representative of worker asbestos exposures associated with installing
and servicing gaskets found in UTV exhaust systems. Further, this assessment assumes that data from
one publication (Paustenbach et at... 2006) are representative of exposures for this condition of use.
However, the job activities and exposure scenarios considered in the publication differ from the UTV-
related exposures in at least two ways.
First, the publication used in this analysis (Paustenbach et at.. 2006) considered older automobiles. This
focus was intentional, because newer vehicles generally do not have asbestos-containing exhaust
systems. However, all vehicles considered in the study were more than 35 years old at the time the
research was published. According to the study, the highest concentrations of asbestos in the removed
gasket was 35.5 to 48.9 percent. It is unclear if the asbestos and other chemicals in the gaskets used in
the automobile exhaust systems from pre-1970 automobiles are representative of the asbestos content in
today's UTV exhaust systems.
Second, because the study considered vintage automobiles that presumably contained older parts, it is
likely that the asbestos-containing gaskets in the exhaust systems had worn down with use and time.
These older gaskets presumably would be more prone to release fibrous asbestos into the air, as
compared to newer gaskets (which typically are pre-formed with the asbestos encapsulated in a binding
agent or some other matrix) (Paustenbach et at.. 2006). Therefore, the asbestos concentrations measured
during the study may overstate the concentrations that might occur during UTV exhaust system
servicing.
Additionally, EPA identified two sources of uncertainty pertaining to the data analysis. One pertains to
the uncertainties associated with non-detect observations. For the average worker exposure
concentration, 74 percent of the samples were non-detects; and the study authors replaced these
observations with one-half the detection limit when calculating average concentrations. Similarly, for
the area sampling results used for ONU exposures, 76 percent of the samples were non-detects.
Moreover, five of the personal breathing zone samples collected from mechanics had filters overloaded
with particulate, and these samples were not analyzed. The authors noted that the overloaded filters may
have resulted from particulate matter released while mechanics used torches to cut and weld exhaust
pipes; but EPA cannot rule out the possibility that these overloaded filters might have contained elevated
levels of asbestos.
It is possible that ONUs may not be exposed at a full shift, every workday. Based on these strengths and
limitations of the data, the overall confidence for EPA's assessment of occupational and ONU inhalation
exposures for this scenario is medium.
2.3.1.10 Summary of Inhalation Occupational Exposure Assessment
Table 2-24. summarizes the inhalation exposure estimates for all COUs that EPA evaluated. Where
statistics can be calculated, the central tendency estimate represents the 50th percentile exposure level of
the available data set, and the high-end estimate represents the 95th percentile exposure level. The
central tendency and high-end exposures for ONU are derived separately from workers, often by using
either a reduction factor or the analytical limit of detection. See the footnotes for an explanation of the
concentrations used for each COU.
Page 122 of 352

-------
Table 2-24. Summary of Occupational Inhalation Exposures
Condition of I so
Diimtion
Type
T\\
Con trill
Tendency
A Kxposures, 1"
(see footnotes)
1 liปh-cnil
'cc
Con lidence
Riilinji
Diaphragms for Chlor-Alkali Industry
(Processing and Use)
Full Shift
Worker
0.0049 (a)
0.034 (a)
High
ONU
0.0025 (b)
0.008 (b)
High
Short-term
Worker
0.024 (a)
0.512(a)
Medium
ONU
No data
No data
-
Sheet gaskets - stamping (Processing)
Full Shift
Worker
0.014(c)
0.059 (c)
Medium
ONU
0.0024 (d)
0.010 (d)
Medium
Short-term
Worker
0.024 (c)
0.059 (c)
Medium
ONU
0.0042 (d)
0.010 (d)
Medium
Sheet gaskets - use
Full Shift
Worker
0.026 (e)
0.094 (e)
Medium
ONU
0.005 (d)
0.016 (d)
Medium
Short-term
Worker
No data
No data
-
ONU
No data
No data
-
Oilfield brake blocks - Use
Full Shift
Worker
0.03 (f)
No data
Low
ONU
0.02 (f)
No data
Low
Short-term
Worker
No data
No data
-
ONU
No data
No data
-
Aftermarket automotive brakes/linings,
clutches (Use and Disposal)
Full Shift
Worker
0.006 (g)
0.094 (g)
Medium
ONU
0.001 (h)
0.002 (h)
Medium
Short-term
Worker
0.006 (g)
0.836 (g)
Medium
ONU
0.001 (h)
0.002 (h)
Medium
Other Vehicle Friction Products (brakes
installed in exported cars) (Use)
Full Shift
Worker
0.006 (g)
0.094 (g)
Medium
ONU
0.001 (h)
0.002 (h)
Medium
Short-term
Worker
0.006 (g)
0.836 (g)
Medium
ONU
0.001 (h)
0.002 (h)
Medium
Replacing brake pads in NASA Super
Guppy
Full Shift
Worker
<0.003
<0.0089
High
ONU
Not Expected
Not Expected
High
Short-Term
Worker
<0.022
0.045
High
ONU
Not Expected
Not Expected
High
Other gaskets - UTVs (Use and Disposal)
Full Shift
Worker
0.024 (i)
0.066 (i)
Low
ONU
0.005 (i)
0.015 (i)
Low
Short-term
Worker
No data
No data
-
ONU
No data
No data
-
Page 123 of 352

-------
(a)	Full-shift exposure concentrations for the chlor-alkali industry are based on worker exposure monitoring data.
Central tendency concentrations are 50th percentile values and high-end concentrations are 95th percentile values.
The central tendency short-term TWA exposure value for workers was based on short-term (30-minute) sampling
data provided by industry. The value in Table 2-5 (0.024 f/cc) is the median value of all 30-minute personal samples
submitted. The high-end short-term TWA exposure value for workers (0.512 f/cc) is the calculated 95th percentile
value for the compiled industry short-term exposure data.
(b)	ONU exposure concentrations for the chlor-alkali industry are based on area monitoring data with all samples being
non-detect observations that were replaced with surrogate values. Central tendency exposure concentrations were
calculated as the arithmetic mean of the individual observations, using one-half the detection limit for individual
samples; and the high-end concentration is the highest detection limit provided.
(c)	Concentrations for sheet gasket stampers are based on worker exposure monitoring data (10 samples). Central
tendency is the single full-shift TWA data point available; and high-end assumes the highest observed short-term
exposure persists over an entire shift. For short-term exposures, central tendency is the median concentration
observed, and high-end is the highest concentration observed.
(d)	Concentrations for ONUs at sheet gasket stamping facilities and sheet gasket use facilities were estimated by EPA
using a concentration-decay factor for bystander exposures derived from the literature.
(e)	Concentrations for sheet gasket use are based on descriptive statistics provided to EPA of 34 worker exposure
monitoring samples. The central tendency concentration is the arithmetic mean and the high-end concentration is the
highest measured value.
(f)	Concentrations for oil field brake blocks are based on two data points—arithmetic mean exposure for different
worker activities—reported in the scientific literature.
(g)	Concentrations for aftermarket automotive parts are based on worker exposure monitoring data documented in seven
studies. For full shift exposures, the central tendency concentration is the median of the arithmetic mean exposure
values reported across the seven studies; and the high-end concentration is the highest TWA exposure concentration
reported. For short-term exposures, the same data set was used but data were summarized for individual
observations, not the full-shift TWA values.
(h)	Concentrations for ONUs at auto repair facilities were estimated using more than 70 area samples that NIOSH
collected at bystander sampling locations.
(i)	Asbestos air measurements from (Paustenbach et al., 2006): Removal and replacement of exhaust system gaskets
from vehicles manufactured before 1974 with original and old exhaust systems.
2.3.2 Consumer Exposures
This section summarizes the data used for estimating consumer inhalation exposures to chrysotile
asbestos for two potential do-it-yourself (DIY) scenarios: (1) automobile brake repair/replacement and
(2) gasket repair/replacement in Utility Vehicles (UTVs). Specifically, the brake repair/replacement
scenario involves repair or installation of imported aftermarket automobile brake pads (disc brakes) or
brake shoes (drum brakes) containing asbestos. The gasket repair/replacement in the UTV scenario
involves removal or installation of aftermarket gaskets for UTV exhaust systems containing asbestos. In
response to peer review and public comments received on the draft Risk Evaluation, EPA recognizes
brake repair/replacement work and gasket repair/replacement work may occur on other vehicle types
(i.e., motorcycles, snowmobiles, tractors). However, EPA did not identify or receive data which could
either inform exposures during such "other vehicle" repair/replacement activities or inform methodology
to extrapolate from automobile specific data to such "other vehicle" activities. Therefore, EPA did not
evaluate other vehicle repair/replacement activities. Additionally, EPA recognizes exposure to
bystanders may occur via take-home/take-in (from garage) exposures, depending on personal hygiene
practices of a DIY consumer, however, EPA did not identify or receive data which could inform
exposures or methodologies to extract such data. Therefore, EPA did not evaluate take-home/take-in
(from garage) exposures to chrysotile asbestos for either brake repair/replacement work or gasket
repair/replacement work.
Inhalation exposures are evaluated for both automobile brake repair/replacement and UTV exhaust
system gasket repair/replacement activities for the individual doing the repair/replacement work and a
Page 124 of 352

-------
potential bystander observing the work within the immediate area. For each scenario, it is assumed that
consumers and bystanders will not be wearing any personal protective equipment.
Dermal exposures are not assessed for consumers in this Part 1 of the risk evaluation for asbestos. The
basis for excluding this route is the expected state of asbestos being only solid/fiber phase. While
asbestos may deposit on open/unprotected skin, it will not absorb into the body through the protective
outer skin layers. Therefore, a dermal dose resulting from dermal exposure is not expected.
The DIY consumer brake assessment and UTV gasket replacement assessment rely on qualitative and
quantitative data obtained during the data extraction and integration phase of Systematic Review to build
appropriate exposure scenarios and develop quantitative exposure estimates using personal inhalation
monitoring data in both the personal breathing zone and the immediate area of the work. The literature
search resulted in very little information specific to consumer exposures, thus the consumer assessment
relies heavily on the review of occupational data, and best professional judgment. Many of the studies in
existing literature are older (dating back to late 1970s) which could add some uncertainty to current
practices used by consumers. When possible, EPA used the most recent studies available and also
considered data quality and adequacy of the data. Targeted literature searches were conducted as
appropriate to augment the initial data obtained and to identify supplemental information such as activity
patterns and exposure factors specific to consumers.
EPA has found no reasonably available information to suggest that chrysotile asbestos-containing brakes
are manufactured in the United States, and based on stakeholder outreach, the Agency does not believe
that any domestic car manufacturer installs asbestos-containing brakes in new cars sold domestically.20
However, general online searches have indicated brakes and gaskets identified to contain asbestos are
available for consumers to purchase as aftermarket replacement parts for cars and UTVs. EPA
recognizes that while an aftermarket product may be labelled to contain asbestos (in particular products
manufactured outside the United States) such labelling is not a guarantee the product actually contains
asbestos. Similarly, it should be recognized that even though a product is not labelled to contain asbestos
(in particular products manufactured outside the United States) such products may contain asbestos but
have no requirement to label as containing asbestos. Based on these possibilities, and to ensure potential
exposure to asbestos during brake repair/replacement activities or UTV exhaust system gasket
repair/replacement activities is adequately evaluated, EPA assumes DIY consumers do these
repair/replacement activities with aftermarket products containing asbestos. This assumption does have
some uncertainties which are discussed in the uncertainties section.
The number of consumers impacted by these COUs is unknown because EPA did not identify or receive
data which can inform the actual number of individuals doing DIY repair/replacement activities
(including potential shade mechanics21 or consumers working on more than one car), the actual number
of those doing the repair/replacement activities with products containing asbestos, and the actual number
of products which contain asbestos purchased for consumer use. This is discussed in more detail in the
uncertainties section (Section 4.3.7: Confidence in the Human Health Risk Estimations).
20	EPA is aware of one car manufacturer who imports asbestos-containing automotive friction products for new vehicles, but
those vehicles are then exported and not sold in the United States.
21	A term used for hobbyist mechanics; or one who works on their own vehicle.
Page 125 of 352

-------
2,3,2.1 Consumer Inhalation Exposures of Do-It-Yourself (DIY) Mechanics During
Brake Repair: Approach and Methodology
This consumer assessment addresses potential scenarios in which a DIY consumer installs, repairs or
replaces existing automobile brakes with imported aftermarket brake pads or shoes which may contain
asbestos; including brake linings. While peer-reviewed literature indicates much of the asbestos brake
pad or shoe use has been phased out and the majority of existing cars on the road do not have asbestos
brakes (Finlev et at.. 2.007). asbestos-containing brakes and shoes can still be purchased in the United
States and from sources outside the United States. This scenario evaluates potential consumer inhalation
exposure to asbestos during removal of the old brakes or shoes containing asbestos, cleaning of the
brake housing, shoes, and wheel assembly, as well as installation and grinding of the newly installed
brakes or shoes containing asbestos. While grinding of brakes or shoes may not be common to all DIY
consumers, there is readily available grinding equipment which consumers can purchase for DIY
projects which fit in a residential garage. Additionally, certain DIY consumers (in particular classic cars
hobbyists but also others) may be required to grind brakes or shoes in order for the aftermarket product
to properly fit the brake assembly. Considering these possibilities, EPA includes grinding activity as part
of its evaluation of asbestos exposure to the DIY consumer during brake repair/replacement activities
while acknowledging the associated uncertainties (discussed in Section 2.3.2.1.4 Data Assumptions,
Uncertainties and Level of Confidence).
Brake repair and replacement typically involve several basic steps. For both drum brakes and disc
brakes, the first step is to access the brake assembly by elevating the vehicle and removing the wheel.
The next step is to remove the old brake pads or shoes followed by cleaning the brake apparatus using
various cleaning equipment such as dry or wet brush, wet rag, brake cleaning fluid, or compressed air.
Although EPA does not recommend the work practice of blowing brakes with compressed air (U.S.
EPA. 20071 there is insufficient information indicating such practice has been fully discontinued by the
consumer. Therefore, EPA includes the use of compressed air to blow brakes in this evaluation to ensure
the potential use of compressed air is considered. After the brake apparatus is cleaned, new pads or
shoes are installed. In some situations, installation of new pads may require additional work such as
brake shoe arc grinding. This additional work may be more likely when consumers are working on
vintage vehicles and brake shoes do not fit exactly inside the brake drum.
Systematic review of the reasonably available literature on brake repair and replacement resulted in
insufficient inhalation personal/area monitoring studies specifically for DIY consumer brake repair.
Therefore, the DIY consumer brake repair/replacement exposure assessment uses surrogate monitoring
data from occupational brake repair studies. EPA recognizes that brake repair/replacement by a
professional mechanic may involve the use of different equipment and procedures. Consumer exposure
during DIY brake repair is expected to differ from occupational brake repair in four ways ("Versar.
1987): (1) consumers generally do not have a fully equipped professional garage to perform auto repairs
(in some cases, the repairs would occur in an enclosed garage); (2) consumers would not wear
respirators, mitigate dust emissions, or have available the professional equipment found in commercial
repair shops; (3) consumers have limited experience, and thus the time required to make repairs would
be longer; and (4) consumers are unlikely to perform more than one brake job per year and it was
assumed that only one consumer would perform the task of replacing asbestos brakes or shoes.
Considering the expected differences between brake repair/replacement work conducted by a
professional mechanic and a DIY consumer, EPA identified several factors to consider during the
systematic review process for using professional mechanic information as a surrogate for the DIY
consumer. The goal was to examine the activity patterns monitored in the various occupational studies
and only select those studies which are expected to represent a DIY consumer scenario.
Page 126 of 352

-------
Specifically, EPA only considered activity patterns within the various occupational studies
representative of expected DIY consumer activity patterns and work practices. EPA also considered only
those studies with information related to typical passenger vehicles (automobiles, light duty trucks,
mini-vans, or similar vehicle types); it is not expected that a typical DIY consumer would perform brake
repair/replacement work on heavy duty trucks, tractor trailers, airplanes, or buses. Furthermore,
consideration was given to reasonably available literature which had monitoring data in the personal
breathing zone of the potential DIY consumer and area monitoring within a garage. Lastly, EPA
considered those studies where the work was performed without localized or area engineering controls
as it is unlikely a DIY consumer will have such controls (e.g., capture hoods, roof vents, industrial
exhaust fans baghouses, etc.) within their residential garage.
The following assumptions are used to assess consumer inhalation exposure to asbestos during DIY
brake repairs:
•	Location: EPA presents an indoor and an outdoor scenario for brake repair and
replacement work. The indoor scenario assumes the DIY brake repair/replacement is
performed in the consumer's residential garage with the garage door closed. It also
assumes the additional work associated with this brake work is arc grinding and occurs
within the garage with the garage door closed. The outdoor scenario assumes the DIY
brake repair/replacement work is performed in the consumer's residential driveway. It
also assumes the additional work associated with this brake work is brake filing and
occurs in the residential driveway.
•	Duration of Activity: Available literature indicates a typical "brake job" for a
professional brake mechanic for a single vehicle takes between one and two hours
(Paustenbach et at.. 2003). No data were found in existing literature on the length of
time needed for a DIY consumer to perform a brake job. EPA assumes a consumer
DIY brake repair/replacement event could take twice as long as a professional
mechanic, or about three hours (double the mean of time found in the literature for
professional mechanics).
•	Cleaning methods: EPA assumes, for the indoor scenario, a consumer may use
compressed air to clean brake assemblies since it was historically utilized, is still
readily available to consumers (canned air or air compressor systems), and nothing
prohibits consumers from using compressed air. EPA assumes, for the outdoor
scenario, a consumer does not use compressed air.
•	Possible additional work during repair/installation of brakes: EPA assumes a consumer
may perform additional work on brakes, like arc grinding, hand filing, or hand sanding
of brake pads as part of the brake repair/replacement work. EPA assumes the
consumer performs arc grinding for the indoor scenario and assumes the consumer
performs hand filing for the outdoor scenario. Concentrations resulting from brake
work including this additional work is utilized as the high-end estimate for consumer
exposure. The central tendency is based on changing out brakes only with no
additional work.
•	Frequency of brake repair jobs: EPA assumes the average consumer performs a single
brake repair/replacement job about once every three years for this evaluation. Brakes
in cars and small trucks are estimated to require replacement approximately every
Page 127 of 352

-------
35,000 to 60,000 miles (Advance Auto Parts, website accessed on November 12,
2018). The three-year timeline is derived by assuming the need to replace brakes every
35,000 miles, and an average number of annual miles driven per driver in the United
States of 13,476 miles/year (	1118). There are several factors which can
affect this assumption which are discussed in Section 2.3.2.1.4 Data Assumptions,
Uncertainties and Level of Confidence but include driving patterns, driving frequency,
distances driven, a DIY consumer is a shade tree mechanic, owns and works on more
than one car within a family, and works on vintage cars.
• Brake type: EPA assumes exposure to asbestos is similar during the replacement of
disc brake pads and drum brake shoes.
2.3.2.1.1 Consumer Exposure Results - Do-It-Yourself (DIY) Mechanics
During Brake Repair
Utilizing the factors and the assumptions discussed above, EPA identified five relevant studies which
could be applied to the expected DIY consumer brake repair/replacement scenario. These references as
well as the data quality scores are provided in the following table:
Table 2-25. Summary of Studies Satisfying Conditions/Factors for Use in Consumer DIY Brake
	Exposure Scenario	
Reference
Occupational
Kxposures?
Consumcr/DI Y
Kxposures?
Data Quality Rating (Score)
Sheehy et al. (1989)
Yes
Yes
Medium (1.7)
Blake et al. (2003)
Yes
No
Medium (1.8)
Paustenbach et al. (2003)
Yes
No
High (1.0)
Yeung et al. (1999)
Yes
No
Medium (2.0)
Kakooei et al. (2011)
Yes
No
Medium (2.0)
Monitoring data from two of the five studies, Sheehy et 89) and Blake et al. (2003) were used to
evaluate consumer inhalation exposure to asbestos resulting from brake repair/replacement work. These
studies were U.S. studies which used standard sampling and analysis methods (including both PCM and
TEM analyses) for asbestos. Sheehy et al. (1989) provided DIY consumer exposure data for work
conducted outdoors (although limited to two samples). Although professional mechanics were
conducting the brake repair/replacement work in the Blake et al. (2003) study, the work practices
utilized by the professional mechanics were comparable to historical DIY consumer practices (including
use of compressed air and other cleaning practices along with potential grinding activities) and neither
engineering controls nor personal protective equipment were used. The third U.S. study, Paustenbach et
al. (2003). was a supplemental study used to inform the length of time it takes a DIY consumer to
complete brake repair/replacement work. The final two studies were non-U.S. studies. Yeung et al.
i) was a secondary study and did not provide supplemental/raw data. Additionally, all breathing
zone and area samples from this study were below the PCM detection limits. Kakooei et al. (2011) had a
limited description of the exposure scenario and therefore may not be representative of the expected
DIY consumer activity. Neither of these non-US studies will be further described in this risk evaluation.
Page 128 of 352

-------
A brief summary of the two monitoring studies used for this evaluation is provided below.
Sheehy et at (1989) measured air concentrations during servicing of rear brakes on a full-size van. The
work was performed outdoors, on a driveway, by a DIY consumer. The DIY consumer wet the drum
brake with a spray can solvent to dissolve accumulated grease and dirt. The mechanic then used a garden
hose to flush the surfaces with water. The duration of the monitoring activity was not provided.
Blake et al. (2003) measured air concentrations in the personal breathing zone of professional mechanics
performing brake repair/replacement work. Blake et al. (2003) evaluated asbestos exposure for brake
repair jobs conducted on passenger vehicles from model years 1965-1968. The study sought to use tools
and practices common to the mid-1960s for cleaning, repairing, and replacing the brakes. In six separate
tests, brake shoe change-outs were conducted on all four wheels of a car which had already been fitted
with new asbestos containing brake shoes and then driven for 1,400 miles prior to the monitoring. The
monitoring began with driving the test car into the service bay and ended upon return from a test drive
after the brake-change out. The total brake change-out monitoring period was 85 to 103 minutes in
duration. In general, all tests involved removing the wheel and tire assemblies, followed by the brake
drum. The drum was then placed on the concrete floor creating a shock which broke loose the brake
dust. Each brake assembly was then blown out using compressed shop air. For two baseline tests, no
additional manipulation of the brake shoes (such as filing, sanding, or arc grinding) was conducted. The
remaining four tests involved additional manipulation of the brake shoes as follows:
1)	arc grinding of the new shoes to precisely match each shoes' radius to that of its companion
brake drum (n = 2), and
2)	sanding to bevel the edges and remove the outermost wear surfaces on each shoe (n=l), and 3)
filing to bevel the square edges of the shoe friction material prior to installation (n=l).
These activities encompassed approximately 12.5 minutes, 4.1 minutes, and 9.7 minutes of the
monitoring period, respectively. An additional test was conducted during cleaning only (sweeping) for a
total of 30 minutes by the mechanic after four brake change test runs. The tests were conducted in a
former automobile repair facility (7 bays, volume of 2,000 m3) with the overhead garage doors closed.
An exhaust fan equipped with a filter was installed 16 meters away from the brake changing area and
operated during all brake changes to ventilate the building. However, smoke testing showed no air
movements toward the exhaust fans suction beyond 8 meters from the fan. PCM and TEM analyses
were conducted on all samples except for the seventh test; which was cleaning the work area after all
brake changes were complete and for which only PCM analysis was conducted.
Blake et al. (2003) included area sampling collected from seven locations within the building during
each test run, including four samples within 3 meters of the vehicle, one sample within 3 meters of the
arc grinding station, and two samples >3 meters from the automobile. Background samples were not
collected.
2.3.2.1.2 Exposure Data for Do-It-Yourself (DIY) Mechanics During Brake
Repair
Consumer inhalation exposure to asbestos for the DIY brake repair/replacement scenario was assessed
for both the consumer user (individual doing the brake repair/replacement work) and a bystander
(individual observing the brake work or present within the garage during the brake work). Consumer
inhalation exposure was evaluated for two conditions for the consumer user and bystander.
Page 129 of 352

-------
1)	All brake work conducted indoors
2)	All brake work conducted outdoors
The monitoring data extracted from the Blake et al. (2003) and Sheehy et al. (1989) studies are
presented in Table 2-26. A discussion of this information follows Table 2-26.
Table 2-26. Exposure Concentrations from Blake (2003) and Sheehy (1989) Studies to the DIY
	User During Various Activities	
St ud v
Activity
Duration
(hours)
Co ncc
PIJZ
lit rat ion (l'/cc)
<3 in IVoni auto
Location
Confidence
Rating
ke et al.
(2003)
Brake shoe removal/
replacement
1.5
0.0217
0.00027
Indoors
Medium
1.4
0.0672
0.0258
Indoors
Medium
Filing brakes
1.7
0.0376
0.0282
Indoors
Medium
Hand sanding
Brakes
1.6
0.0776
0.0133
Indoors
Medium
Arc-grinding
Brakes
1.7
0.4368
0.0296
Indoors
Medium
1.6
0.2005
0.0276
Indoors
Medium
Cleaning facility
0.5
0.0146
0.0069
Indoors
Medium
Sheehy et al.
9)
Brake shoe removal/
replacement
Unknown a
0.007
Not monitoredb
Outdoors
Medium
11 No monitoring duration was provided within the study.
b This study did not include outdoor area monitoring which could be applied to the bystander
For purposes of utilizing the information provided in Table 2-26 within this evaluation, EPA applied the
personal breathing zone (PBZ) values to the DIY consumer user for the indoor and outdoor scenarios
under the assumption that hands on work would result in exposure within the PBZ of the individual.
EPA assumes exposure to asbestos resulting from brake repair/replacement work occurs for the entire
three-hour period it takes the DIY consumer to conduct the work.
EPA applied the area monitoring data obtained less than 3 meters from the automobile for the DIY
bystander for the indoor scenario under the assumption that the bystander could be an observer closely
watching the work being performed, an individual learning how to do brake repair/replacement work, or
even a child within the garage while the brake work is being performed. EPA assumes the bystander
remains within 3 meters of the automobile on which the work is being done for the entire three-hour
period it takes for the DIY consumer to conduct the work.
EPA evaluated consumer bystander exposure for the DIY brake outdoor scenario by applying a
reduction factor of 10 to the PBZ value measured outdoors for the consumer user. The reduction factor
of 10 was chosen based on a comparison between the PBZ and the < 3meter from automobile values
measured indoors across all activities identified in the study data utilized from Blake (a ratio of 6.5). The
ratio of 6.5 was rounded up to 10, to account for an additional reduction in concentration to which a
Page 130 of 352

-------
bystander may be exposed in the outdoor space based on the high air exchange rates and volume in the
outdoor22.
DIY Consumer User
Indoor Scenario
The highest concentration values reported in Blake et al. (2003) occurred during arc grinding of the
brake shoes. While this activity may not be common practice for all brake repair/replacement activities,
affordable grinding machines are readily available to those DIY consumers interested in purchasing and
utilizing such equipment. Additionally, such equipment is also available for rental from various stores.
Because such equipment is readily available to the consumer, EPA utilized the average of the two arc-
grinding values from Blake et al. (2003) as the high-end concentration for the indoor environment under
this exposure scenario.
For this Part 1 of the risk evaluation for asbestos, EPA used the average of the two-brake shoe
removal/replacement values within the Blake et al. (2003) study as the central tendency value for the
indoor scenario. These values were measured during brake repair/replacement activities only (no
additional work like grinding/filing) and do include the use of compressed air. However, compressed air
was only used to blow out residual dust from brake drums after the majority of residual dust is broken
out by placing the brakes on the floor with a shock to knock off loose material. While the use of
compressed air is not a recommended practice, no reasonably available information was found that
surveyed actual cleaning methods used or preferred by DIY consumers for this scenario. Additionally,
compressed air systems (either cans or mechanical air compressors) are readily available and used by
consumers for multiple DIY activities. EPA therefore utilized these values to evaluate consumer
inhalation exposure with the understanding that they may represent a more conservative exposure
concentration value.
Outdoor Scenario
EPA utilized the personal breathing zone concentration from the e et al. (2003) study obtained
during filing of brakes for the high-end exposure concentration for the consumer user under the outdoor
scenario. Although this value was obtained in an indoor environment it is a potential additional work
activity that could also be performed outside and more readily expected to occur outdoors than arc
grinding. Additionally, it is expected that filing of brakes would place a consumer's personal breathing
zone very close to the brakes being filed. Such close proximity is expected to minimize potential impact
of the higher air exchange rates and outdoor volumes on exposure to asbestos in the personal breathing
zone and therefore using the indoor measurements for an outdoor scenario is a feasible exposure
condition.
EPA used the average monitored outdoor concentration measured in the personal breathing zone from
the Sheefay et al. (1989) study to represent the central tendency value for the consumer user under the
outdoor scenario. The Sheehy et a )) study is the only study identified through the systematic
review process which included PBZ monitoring data for a DIY consumer user during outdoor brake
repair/replacement work. The duration of the monitoring in Sheehy et al. 11 ฐ89) was not specified for
the outdoor work, EPA assumes monitoring occurred for the entire expected duration for the DIY
consumer user to complete the work. As the study describes, the DIY consumer user utilized various
22 Although exposures would be very low and are not quantified here, an assumption is made in Section 4.2.3.1 to allow for
cancer risk estimation for bystanders from outdoor brake replacement.
Page 131 of 352

-------
wetting techniques on the brakes to clean grease, dirt, and flush the surface of the drums. Considering
these methods were utilized, EPA assumes compressed air was not used for the outdoor scenario.
Bystander
Indoor Scenario
EPA utilized the Blake et al. (2003) area sampling data obtained within three meters from the
automobile on which the work is being performed to represent exposure concentrations for the bystander
under the indoor scenario. These values are expected to be representative of bystander exposure under
the assumptions described above in that individuals who may remain within the garage during brake
repair/replacement work would be in close quarters within a typical consumer garage for the entire
three-hour period. The high-end value utilized was the highest area concentration monitored within three
meters from the automobile. This value occurred during arc-grinding of the brake shoe. The central
tendency value utilized was the average of the two area sampling concentrations monitored within three
meters from the automobile during brake shoe removal/replacement activities.
Outdoor Scenario:
There were no area monitoring data for the outdoor work in Sheehy et al. (1989) which could be
representative of potential bystander exposure. As a surrogate, EPA used the analysis of reduction
factors (RFs) based on available data for the gasket ONU exposure scenario. Those data showed people
5-10 feet away from the user had measured values from 2.5 to 9-fold lower than the exposure levels
measured for the user. For that COU, EPA used the mean of 5.75 as the RF; which was in the range of
RFs from other COUs. Because there were no such measured data available to estimate an RF for
outdoor bystander work, EPA selected an RF of 10 that was greater than the range of RFs for other
COUs, but still allowed evaluation of potential bystander exposure in an outdoor scenario even though
such exposure is expected to be low due to high air exchange rates and the volume of the outdoor space.
EPA therefore applied a reduction factor of 10 to the data utilized for consumer users to represent the
concentration to which the bystander is exposed under the outdoor scenario. This reduction factor was
applied to both the central tendency and high-end estimates to represent potential exposure of the
bystander.
2.3.2.1.3 Exposure Estimates for DIY Brake Repair/Replacement Scenario
Table 2-27 provides a summary of the data utilized for this evaluation.
Table 2-27. Estimated Exposure Concentration for DIY Consumer User and Bystander for Risk
Evaluation for Asbestos Part 1: Chrysotile Asbestos - DIY Brake Repair/Replacement Scenario
Condition of I so
llsliniiilod ( oi
l)IY I so
('onlml Tondonc\
isiinior l'l\|>
lliiji-ond
isiiro (o neon trillion
IS> sliintlo
(on 1 nil TcihIoiio
(l'/00)
llilih-oiid
Confidence Killing
Aftermarket Automotive
Parts-Brakes (Indoor)
0.0445
0.4368
0.01303
0.02963
Medium
Aftermarket Automotive
Parts-Brakes (Outdoor)
0.007
0.0376
0.0007b
0.0038b
Medium (DIY)
Medium-Low (Bystander)
aBased on area samples, see section 2.3.2.1.2.
b Reduction factor of 10 used, see section 2.3.2.1.2 .
Page 132 of 352

-------
EPA assessed chronic exposures for the DIY brake repair/replacement scenarios based on the exposure
concentrations, assumptions, and exposure conditions described above. Because reasonably available
information was not found to characterize exposure frequencies and lifetime durations, EPA made the
following assumptions:
•	Exposure frequency of 3 hours on 1 day every 3 years or 0.04 days per year. This considers car
maintenance recommendations that brakes be replaced every 35,000 miles, and the average
annual miles driven per driver in the United States is 13,476 miles/year (	2018).
•	Exposure duration of 62 years. This assumes exposure for a DIY consumer user starts at 16 years
old and continues through the average adult lifetime (78 years). EPA also used a range of
exposures (for both age at first exposure and duration of exposure); these are further described in
Section 4.2.3 of the Risk Characterization.
•	To address potential uncertainties surrounding EPA's use of 78-year lifetime and ongoing DIY
brake repair work every 3 years for the entire 62 years, EPA also estimated exposure for a single
brake repair/replacement activity within a lifetime.
2.3.2.1.4 Data Assumptions, Uncertainties and Level of Confidence
Due to lack of reasonably available information on DIY consumer exposures, the consumer assessment
relies on reasonably available occupational data obtained under certain conditions expected to be more
representative of a DIY consumer user scenario (no engineering controls, no PPE, residential garage).
However, the studies utilized have uncertainties associated with the location where the work was done.
In Blake et al. (20031 worker exposures were measured at a former automobile repair facility which had
an industrial sized and filtered exhaust fan unit to ventilate the building during testing while all doors
were closed. A residential garage is not expected to have a filtered exhaust fan installed and operating
during DIY consumer brake repair/replacement activities. While this presents some uncertainty, the
study Blake et al. (2003) performed smoke testing and found that air movement was limited to within
eight meters of the installed and operating exhaust fan. Based on this testing, it is reasonable to assume
that the existence of the exhaust fan would have limited effect on the measured concentrations within the
PBZ of the DIY consumer and limited effect on the measured concentrations at the area monitors which
were within three meters of the automobile being worked on because both locations (automobile and
area monitoring stations) were more than eight meters from the exhaust fan.
The volume of a former automobile repair facility is considerably larger than a typical residential garage
and will have different air exchange rates. While this could raise some uncertainties related to the
applicability of the measured data from Blake et al. (2003) to a DIY consumer user environment, the
locations of the measurements utilized for this evaluation minimize that uncertainty. The PBZ values are
very near the work area and should not be affected by the facility volume or air exchange rates. The area
samples utilized for bystander estimates were obtained within three meters from the automobile on
which the work was being done, so while affected more by volume and air exchange rates, the effects
should still be limited as air movement appeared to be minimal based on the smoke testing conducted in
the Blake et al. (2003) study.
There is some uncertainty associated with the assumed length of time the brake repair/replacement work
takes. EPA assumes it takes a DIY consumer user about three hours to complete brake
repair/replacement work. This is two times as long as a professional mechanic. While it is expected to
take a DIY consumer longer, it is also expected DIY consumer users who do their own brake
Page 133 of 352

-------
repair/replacement work would, over time, develop some expertise in completing the work as they
continue to do it every three years.
There is also some uncertainty associated with the assumption that a bystander would remain within
three meters from the automobile on which the brake repair/replacement work is being conducted for the
entire three-hour period EPA assumes it takes the consumer user to complete the work. However,
considering a residential garage with the door closed is relatively close quarters for car repair work, it is
likely anyone observing (or learning) the brake repair/replacement work would not be able to stay much
farther away from the car than three meters. Remaining within the garage for the entire three hours also
has some uncertainty, although it is expected anyone observing (or learning) the brake
repair/replacement work would remain for the entire duration of the work or would not be able to
observe (or learn) the task.
The assumptions and uncertainties associated with a consumer's use of compressed air to clean brake
drums/pads are discussed above. While industry practices have drifted away from the use of compressed
air to clean brake drums/pads, no reasonably available information was found in the literature indicating
consumers have followed the same discontinuation of such work practices. To consider potential
consumer exposure to asbestos resulting from brake repair/replacement activities, EPA uses data which
included use of compressed air. However, EPA recognizes this may be a more conservative estimate
because use of compressed air typically could cause considerable dust/fibers to become airborne if it is
the only method used. The Blake et al. (2003) study notes that compressed air was used to clean residual
dust from brake drums, but it was only used after "shocking" dust free by placing the brake drums on the
ground to knock dust free. As a result, the bulk of the dust would be on the ground and a limited portion
would be removed through the use of compressed air.
EPA did not purchase brake products which may contain asbestos, or are advertised to contain asbestos,
from online vendors or conduct testing of such products to confirm whether available brake
repair/replacement products did or did not contain asbestos (due largely to cost and resource
constraints). Instead, EPA assumes DIY consumers purchase and use aftermarket brake products which
contain asbestos for this evaluation. However, there is some uncertainty associated with whether
purchased aftermarket brake products installed by DIY consumers for brake repair/replacement work
contain or do not contain asbestos. While some products manufactured and purchased outside of the
United States may be labelled as containing asbestos, the product may not actually contain asbestos as
such labelling could be intended to encourage purchase of the products based on a belief that asbestos
containing products are better than non-asbestos containing products. Similarly, certain products
manufactured and purchased from outside the United States may not be labelled as containing asbestos,
due to the absence of labeling requirements, but may contain asbestos. Finally, if some products do
contain asbestos, there is additional uncertainty that consumers purchase and use those specific asbestos
containing products.
As mentioned earlier, EPA did not evaluate asbestos exposure resulting from brake repair/replacement
work on "other vehicles" like motorcycles, snowmobiles or tractors. The reason these "other vehicles"
were not evaluated is the absence of data to inform asbestos exposure resulting from such "other
vehicle" activities or inform methodology to extrapolate from automobile specific activities to such
"other vehicles". Considering the wide variation in size and accessibility of "other vehicles" and the
absence of data to inform "other vehicle" analyses, the uncertainties could be considerable across
multiple factors including, but not limited to, frequency, duration of work, and exposures resulting from
"other vehicle" brake repair/replacement activities.
Page 134 of 352

-------
EPA recognizes the uncertainty associated with identifying the actual number of consumers and
bystanders receiving an exposure to asbestos from automobile brake repair/replacement activities. In the
draft Risk Evaluation, EPA did not identify data which could inform the actual number of DIY
consumers within the United States involved with DIY brake repair replacement activities that use brake
products containing asbestos. However, EPA provided an estimate of the number of brake repairs
conducted by DIY consumers in the United States as approximately 31 million.
Both peer review and public commenters questioned this estimate; which was based on assuming that
100% of DIYers replace brake pads and that it is likely that the asbestos-containing brakes is a much
lower percentage of the available brakes on the market. EPA agreed and updated the estimated number
of individuals exposed by the limited available information on the potential market share of asbestos
brakes. Details are provided in Section 4.3.7. EPA's updated estimate for number of DIYers assumes
that asbestos brakes may represent approximately 0.05% of aftermarket automotive brakes. By applying
this factor (0.05%) to the universe of DIYers (over 31,000,000), EPA's estimate of potentially exposed
DIYers is a little over 15,900.
EPA has an overall medium confidence rating for the literature, studies, and data utilized for the
Consumer DIY Brake Repair/Replacement COU. This is based on the existence of monitoring data in
both the personal breathing zone and area sampling associated directly with brake repair/replacement
activities. The studies utilized are also representative of expected consumer working conditions for a
DIY consumer. Both factors would indicate a high confidence in the studies and data used. However,
since the data utilized is based on a professional mechanic performing the brake repair/replacement
work rather than a DIY consumer, the overall confidence is medium.
EPA has an overall medium confidence rating for the exposure results associated with the consumer
user under the Consumer DIY Brake Repair/Replacement COU for both indoor and outdoor work. This
is based on the use of direct monitored personal breathing zone data for the individual doing the work in
an indoor and outdoor location.
EPA has an overall medium confidence rating for the exposure results associated with the bystander
indoor location under the Consumer DIY Brake Repair/Replacement Scenario. This is based on the
existence of area monitoring data obtained in the immediate vicinity of the brake repair/replacement
work in an indoor location which is representative of where a bystander may reside during brake
repair/replacement work within a residential garage.
EPA has an overall medium-low confidence rating for exposure results associated the bystander
outdoor location under the Consumer DIY Brake Repair/Replacement Scenario. This is based on the
absence of area monitoring data in an outdoor work location resulting in the need to utilize indoor
measurements and apply an adjustment factor to estimate bystander exposure concentrations in for an
outdoor scenario.
2.3.2.2 Consumer Exposures Approach and Methodology - DIY Gaskets in UTVs
This exposure assessment looks at a potential consumer exposure scenario where a DIY consumer
removes, cleans, handles, and replaces gaskets associated with exhaust systems on UTVs which may
contain chrysotile asbestos. This scenario falls under the "other gaskets" COU in Table 1-4 of this risk
evaluation. Asbestos exposure is estimated for the DIY consumer user (the individual performing the
gasket repair work) as well as a bystander who may observe the gasket work. This scenario also assumes
all the work is conducted indoors (within a garage) and both the consumer and bystander remain in the
garage for the entirety of the work.
Page 135 of 352

-------
There was no reasonably available information found in the published literature related to DIY
consumer exhaust system gasket repair/replacement activities on UTVs. As a result, EPA expanded the
search to include information on occupational gasket repair/replacement for automobiles and identified
several studies with relevant information. The gasket repair/replacement scenario relies on monitored
values obtained in an occupational setting and considers only those environments and working
conditions that may be representative of a DIY consumer scenario.
Thirty studies relating to gasket repair/replacement were identified and reviewed as part of the
systematic review process for exposure. These studies were compared against a series of criteria to
evaluate how representative the studies are for DIY consumer exhaust system gasket repair/replacement
activity. The first two criteria involved identifying whether the studies were automotive in nature and
whether there was enough information about automotive gaskets within the study. EPA also focused on
primary sources of data and not secondary or supplemental sources. The final criterion was to review the
studies to ensure they were consistent with an expected DIY consumer scenario of removal, cleaning,
and replacing gaskets. For example, studies involving machining or processing of gaskets were not
considered as it is unlikely a DIY consumer gasket repair/replacement activity would involve machining
and gasket processing. When compared to these criteria, three of the thirty studies were fully evaluated;
a 2006 study by Blake et al. (2006). a 2005 study by Liukonen and Weir (2.005). and a 2006 study by
Paustenbach et al. (2006). as shown in Table 2-28.
Table 2-28. Summary of Studies Satisfying Factors Applied to Identified Literature
Reference
Occupational
Consumer
Data Quality Rating (Score)
Blake et al. (2006)
Yes
No
Medium (2.1)
Liukonen and Weir (2005)
Yes
No
Medium (2.0)
Paustenbach et al. (2006)
Yes
No
Medium (1.7)
The BI;ike et al. (2006) study measured worker asbestos exposure during automotive gasket
removal/replacement in vintage car engines. The Liukonen and Weir (2005) study measured worker
asbestos exposure during automotive gasket removal/replacement on medium duty diesel engines. The
Paustenbach et al. (2006) study measured worker asbestos exposure during gasket removal/replacement
on automobile exhaust systems of vintage cars (ca. 1945-1975). All three studies were conducted in the
United States and used air sampling methods in compliance with NIOSH methods 7400 and 7402 for
PCM and TEM, respectively. All three studies demonstrate that the highest exposure to asbestos occurs
during removal of old gaskets and cleaning of the area where the gasket was removed. All three studies
received a medium-quality rating through EPA's systematic review data evaluation process.
Relevant data from each of the three studies identified in Table 2-28 were extracted. Extracted data
included vehicle or engine type, sampling duration, sample size, exposure concentrations, and units of
measurement. The extracted data were transcribed into Microsoft Excel for further analysis to calculate
minimum, maximum, and mean concentrations by study, activity type, and sample type. All the
extracted data and calculated values are included in Supplemental File: Consumer Exposure
Calculations (U.S. EPA. 2020a). All analysis and calculations for the three studies were performed
based on the raw data rather than summary data provided by each study due to differences in the
summary methodologies across the studies. For non-detectable samples reported within a study at their
respective sensitivity limits, statistics were calculated based on the full sensitivity value for that sample.
For non-detectable samples reported within a study below their respective sensitivity limits, statistics
were calculated based on one-half the sensitivity limit for that sample. For non-detectable samples
Page 136 of 352

-------
reports at levels greater than their respective sensitivity limits, statistics were calculated based on one-
half the reported non-detectable value. Table 2-29 summarizes the data based on the methodologies
described here.
Table 2-29. Summary Results of Asbestos Exposures in Gasket Repair Studies
St ml v
Kngine W ork
Sample Type
_<
Sample
Size
tir Sample L
Non-
Detectable
Samples
lata
Mean
Sam pie
Duration
(Minutes)
Air Sam
Minimum
)le Concenti
(l'/cc)
.Maximum
ations
Mean
Confidence
Rating
Blake et al. (2006)
28
14
140
0.002
0.027
0.007
Medium
Engine Dissembly
15
4
128
0.003
0.027
0.009
Medium
Area
9
2
135
0.003
0.008
0.005
Medium
Personal
6
2
117
0.007
0.027
0.015
Medium
Engine Reassembly
13
10
153
0.002
0.008
0.003
Medium
Area
9
9
154
0.002
0.008
0.003
Medium
Personal
4
1
153
0.003
0.008
0.005
Medium
Liukonen and Weir
(2005)







Engine Dissembly
29
26
53
0.004
0.060
0.018
Medium
Area
10
10
58
0.004
0.059
0.016
Medium
Observer
3
3
43
0.004
0.057
0.026
Medium
Outdoor
2
2
112
0.006
0.006
0.006
Medium
Personal
14
11
44
0.011
0.060
0.019
Medium
Paustenbach et al.







Engine Dissembly
94
61
39
0.002
0.066
0.014
Medium
Area
22
15
46
0.002
0.015
0.005
Medium
Bystander
44
29
40
0.004
0.030
0.012
Medium
Personal
28
17
32
0.006
0.066
0.024
Medium
After review and consideration of all the information within each of the three studies, EPA used the
Paustenbach et al. (2006) study to evaluate DIY consumer exposure to asbestos resulting from
removal/replacement of exhaust system gaskets for this risk assessment. This study was used because it
was specific to exhaust system work involving asbestos-containing gaskets. It also includes information
applicable to a DIY consumer user (the individuals] doing the gasket work) and the bystander (the
individuals] observing the gasket work).
Page 137 of 352

-------
The Paustenbach et al. (2006) study was conducted in two phases in Santa Rosa, CA during 2004 at an
operational muffler shop that has been open since 1974 and specializes in exhaust repair work. The
repair facility was about 101 feet by 48 feet with five service bay doors. The vehicles studied were
located near the center of the garage. During the study, the bay doors were closed, and no heating, air
condition, or ventilation systems were used.
The Paustenbach et al. (2006) study looked at 16 vehicles manufactured before 1974 with original or old
exhaust systems likely to have asbestos containing gaskets at either the flanges of the muffler system or
the manifold of the engine where the exhaust system connects. The study looked at four different types
of muffler work: 1) removal of exhaust system up to the flange; 2) removal of exhaust system including
manifold gaskets; 3) conversion from single to dual exhaust system; and 4) removal of muffler system
up to the manifold with installation of an asbestos donut gasket. Two mechanics performed the exhaust
repair work and neither mechanic wore respiratory protection. The mechanics removed the gaskets with
either their fingers or by prying with a screwdriver, and any residual gasket material was scraped off
with the screwdriver or pulled off by hand.
All airborne samples were collected using MCE filters consistent with NIOSH method 7400. Personal
breathing zone air samples were collected from the right and left lapel of the mechanic, and area air
samples were collected at four locations about four feet from the vehicle. Background and ambient air
samples were also collected both indoors and outdoors. A total of 134 air samples were collected, but
some samples could not be analyzed due to overloaded filters. Other samples were excluded because
they were taken during work on vehicles with non-asbestos gaskets. Ultimately, 82 air samples (23
personal, 38 area, and 21 background) were analyzed by PCM, and 88 air samples (25 personal, 41 area,
and 22 background) were analyzed by TEM. Samples below the analytical sensitivity limit were
included in the statistical analysis by substituting a value of one-half the sensitivity limit.
2.3.2.2.1 Consumer Inhalation Exposures - DIY Gaskets in UTVs
Consumer inhalation exposure to asbestos for the DIY exhaust system gasket removal/replacement
scenario was assessed for both the DIY consumer (individual doing the exhaust system gasket
removal/replacement work) and a bystander (individual observing the exhaust system gasket
removal/replacement work within the garage).
DIY Consumer
EPA used the PBZ values from Paustenbach et al. (2006) identified in Table 2-29 for the DIY consumer.
The maximum concentration was used as the high-end estimated concentration for the consumer and the
mean concentration was used as the central tendency concentration.
Bystander
EPA used the bystander values from Paustenbach et al. (2006) identified in Table 2-29 for the bystander.
The bystander values from Paustenbach et al. (2006) represent area monitoring obtained within four feet
of the automobile on which the exhaust system work was being performed. EPA believes this distance is
a reasonable distance at which a bystander observing gasket work may be located within a residential
garage during the gasket work. The maximum concentration from Table 2-29 was utilized as the high-
end estimated concentration for the bystander and the mean concentration was utilized as the central
tendency concentration.
Page 138 of 352

-------
2.3.2.2.2 Exposure Estimates for DIY UTV Exhaust System Gasket
Removal/Replacement Scenario
EPA assessed exposures for the DIY UTV exhaust system gasket removal/replacement scenario based
on the exposure concentrations, assumptions, and exposure conditions described above. There was no
reasonably available information found within the literature providing specific information about the
length of time it would take for a DIY consumer to complete an exhaust system gasket
removal/replacement activity on a UTV. The studies from which data was extracted have sample periods
ranging from 32 minutes to 154 minutes to complete various gasket work for a professional mechanic
(assuming the sampling time within these studies was equal to the time it took to complete the gasket
work). Therefore, EPA assumes, for this evaluation, the exhaust system work would take the DIY
consumer three hours to complete which is approximately two times the average sample periods across
the studies extracted.
There was no reasonably available information found within the literature providing specific information
about the frequency of gasket change-out. EPA recognizes that frequency can vary depending on a
variety of factors including the location of the gasket and the number of gaskets needing change-out at
any one time. Additional variability may occur based on the consumer use patterns for a given UTV in
that limited frequency and duration of use may affect the frequency at which a gasket needs to be
changed. Some gasket work may not be needed but performed by a DIY consumer to increase speed or
other factors related to a UTV's performance. The exhaust system gasket on the engine manifold may be
exposed to more extreme temperature fluctuations than one on the muffler and therefore experience
more wear and tear requiring replacement more frequently. Since UTV specific data was neither
identified and evaluated as part of EPA's systematic review process nor provided as part of comments,
EPA assumes, for this evaluation, one or more gaskets will be replaced once every three years.
Exposure durations were assumed to be 62 years. This assumes exposure for the DIY consumer starts at
16 years old and continues through the average adult lifetime of 78 years. Uncertainties associated with
this assumption are discussed in Section 2.3.2.2.3. Table 2-30 provides a summary of the data utilized
for this evaluation. Additional exposure durations were evaluated for reference or comparison. These are
presented in Section 4.2.3.2.
Table 2-30. Estimated Exposure Concentrations for UTV Gasket Repair/Replacement Scenario -
DIY Mechanic and Bystander for Use in the Risk Evaluation for Asbestos Part 1: Chrysotile
Asbestos
Condition of I se
Type
Kxposure Concent
Central Tendency
rations I'Vcc
Nigh-end
Confidence Rating
LTV gasket
Repair/replacement
Paustenbach et al. (2006)
DIY Consumer
0.024
0.066
Medium
Bystander
0.012
0.030
Medium
2.3.2.2.3 Data Assumptions, Uncertainties and Level of Confidence
There was no reasonably available information identified through systematic review providing consumer
specific monitoring for UTV exhaust system gasket repair/replacement activities. Therefore, this
evaluation utilized published monitoring data obtained in an occupational setting of professional
mechanics, as a surrogate for estimating consumer exposures associated with UTV gasket
Page 139 of 352

-------
removal/replacement activities. There is some uncertainty associated with the use of data from an
occupational setting for a consumer environment due to differences in building volumes, air exchange
rates, available engineering controls, and the potential use of PPE. As part of the literature review, EPA
considered these differences and utilized reasonably available information representative of the expected
consumer environment. The Paustenbach et al. (2006) study was conducted in an occupational setting,
but no engineering controls were utilized. Additionally, no additional heating, ventilation, and air
condition systems were utilized during the study. The monitored values used were the PBZ data which
are not expected to be impacted by differences in the ventilation rates, work area volume, or air
exchange rates. Similarly, the area monitoring data utilized for bystander exposure were obtained four
feet from the automobile on which the work was being performed where differences in the ventilation
rates, work area volume, or air exchange rates should have minimal effect on the concentrations to
which the bystander is exposed.
There is some uncertainty associated with the use of an automobile exhaust system gasket
repair/replacement activity as a surrogate for UTV exhaust system gasket repair/replacement activity
due to expected differences in the gasket size, shape, and location. UTV engines and exhaust systems
are expected to be smaller than a full automobile engine and exhaust system, therefore the use of an
automobile exhaust system gasket repair may slightly overestimate exposure to the consumer. At the
same time, the smaller engine and exhaust system of a UTV could make it more difficult to access the
gaskets and clean the surfaces where the gaskets adhere therefore increasing the time needed to clean
and time of exposure resulting from cleaning the surfaces which could underestimate consumer
exposure.
There is some uncertainty associated with the assumption that UTV exhaust system gasket
repair/replacement activities would take a consumer a full three hours to complete. An internet search
revealed some videos suggesting gasket replacement would take a DIY consumer 30 minutes to
complete. This value mirrors the sampling timeframes within the Paustenbach et al. (2006) study.
However, the time needed for a DIY consumer to complete a full UTV exhaust system gasket
repair/replacement activity can vary depending on several factors including location of gaskets, number
of gaskets, size of gasket, and adherence of the gasket and residual material once the system is opened
up and the gasket is removed.
There is uncertainty associated with the assumption that UTV exhaust system gasket repair/replacement
activities would be necessary and performed by a consumer once every three years. A general internet
search ("google") did not identify how often certain gaskets associated with the exhaust systems of
UTVs would last or need to be replaced. Some information was found on ATV maintenance including
repacking the exhaust silencer of ATVs annually on machines that are frequently used or every few
years on machines used seasonally. Other information found online suggested whenever you do exhaust
system maintenance, you should also replace gaskets to ensure an ongoing effective seal for safety and
efficiency.
There is uncertainty associated with the assumption that an individual would be associated with UTV
use or UTV exhaust system gasket repair/replacement activities for the entire average adult lifetime of
78 years beginning at 16 years of age. It is possible certain individuals may be involved with UTV work
prior to 16 years of age. While older individuals may not be associated with their personal UTV and
related gasket work up to age 78, they may provide assistance on gasket work or perhaps change from a
consumer to a bystander.
The EPA has an overall medium confidence rating for the literature, studies, and data utilized for the
Consumer DIY UTV Exhaust System Gasket Repair/Replacement COU. This is based on the existence
Page 140 of 352

-------
of monitoring data in both the personal breathing zone and area sampling associated directly with gasket
repair/replacement activities. The studies utilized are also representative of expected consumer working
conditions for a DIY consumer. Both factors would indicate a high confidence in the studies and data
used. However, since the data utilized is based on a professional mechanic performing the gasket
repair/replacement work rather than a DIY consumer, and the use of a study involving automobile
exhaust system gasket repair/replacement activities as a surrogate for UTV exhaust system work, the
overall confidence is medium.
The EPA has an overall medium confidence rating for the exposure results associated with the consumer
user and bystander under the Consumer DIY Exhaust System Gasket Repair/Replacement COU. This is
based on the use of direct monitored personal breathing zone data for the individual doing the work and
the existence of area monitoring data obtained in the immediate vicinity of the gasket repair/replacement
work in an indoor location which is representative of where a bystander may reside during gasket
repair/replacement work within a residential garage.
The EPA has an overall low confidence rating for the frequency of gasket repair/replacement activities
(once every 3 years). This is based on the absence of data specific to frequency of UTV exhaust system
gasket repair/replacement work. Additionally, the need for such repair/replacement work is expected to
be heavily reliant on the frequency an individual uses the UTV, and the degree to which the UTV is
pushed during use (heavy use, in extreme conditions could require more frequent work while limited
use, in relatively tranquil conditions could require less frequent work).
The EPA has an overall low confidence rating for the lifetime association of an individual with UTV
exhaust system gasket repair/replacement work (16-78 years of age). This is based on the absence of
data on the age distribution of UTV ownership and self-repair/replacement work of exhaust system
gaskets on UTVs. As discussed in the uncertainties, however, while a particular DIY consumer may not
own a UTV for their entire lifetime, they could be involved with UTV exhaust system gasket
repair/replacmenet work in different ways throughout their life (learning how to do the work early in
life, then doing the work, then observing others/or training others to do such work).
2,3,2.3 Summary of Inhalation Data Supporting the Consumer Exposure
Assessment
Table 2-31 contains a summary of the consumer inhalation exposure data used to calculate the risk
estimates in Section 4.2.3.
Table 2-31. Summary of Consumer Inhalation Exposures
Condition of I se
Duration
Type
Kxposure Concenlrs
( cnlnil Tendency
lions. I'/cc
lligh-end
Confidence
Ualing
Brakes
Repair/Replacement
(Indoors)
3 hours once
every 3 years
DIY Consumer
0.0445
0.4368
Medium
Bystander
0.0130
0.0296
Medium
Brakes
Repair/Replacement
(Outdoors)
3 hours once
every 3 years
DIY Consumer
0.007
0.0376
Medium
Bystander
0.0007
0.0038
Medium-Low
UTV gasket
Repair/replacement
3 hours once
every 3 years
DIY Consumer
0.024
0.066
Medium
Bystander
0.012
0.030
Medium
Page 141 of 352

-------
2.3.3 Potentially Exposed or Susceptible Subpopulations
TSCA requires that a risk evaluation "determine whether a chemical substance presents an unreasonable
risk of injury to health or the environment, without consideration of cost or other non-risk factors,
including an unreasonable risk to a potentially exposed or susceptible subpopulation identified as
relevant to the risk evaluation by the Administrator, under the conditions of use." TSCA ง 3(12) states
that "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, pregnant women, workers, or the elderly."
During problem formulation (U.S. EPA. 2018d). EPA identified potentially exposed and susceptible
subpopulations for further analysis during the development and refinement of the life cycle, conceptual
models, exposure scenarios, and analysis plan. In this section, EPA addresses the potentially exposed or
susceptible subpopulations identified as relevant based on greater exposure. EPA addresses the
subpopulations identified as relevant based on greater susceptibility in Section 3.245
In developing this Part 1 of the risk evaluation for asbestos, the EPA analyzed the reasonably available
information to ascertain whether some human receptor groups may have greater exposure than the
general population to the hazard posed by asbestos. Exposures of asbestos would be expected to be
higher amongst groups living near facilities covered under the COUs in this Part 1 of the risk evaluation
for asbestos, workers who use asbestos as part of their work, and groups who have higher age- and
route-specific inhalation intake rates compared to the general population.
Of the human receptors identified in the previous sections, EPA identifies the following as potentially
exposed or susceptible subpopulations due to their greater exposure to chrysotile asbestos and
considered them in this Part 1 of the risk evaluation for asbestos:
•	Workers and occupational non-users for the COUs in this Part 1 of the risk evaluation for
asbestos (chlor-alkali. sheet gaskets, oilfield brake blocks, aftermarket automotive brakes and
linings, other friction products, and other gaskets lUTVsl). EPA reviewed monitoring data
found in published literature and submitted by industry including both personal exposure
monitoring data (direct exposure) and area monitoring data (indirect exposures). Exposure
estimates were developed for users (males and female workers who age >16 years of age
exposed to chrysotile asbestos as well as non-users or workers exposed to chrysotile asbestos
indirectly by being in the same work area of the building. Also, adolescents were considered
as a potentially exposed or susceptible subpopulations.
•	Consumers and bystanders associated with consumer (DIY) use. Chrysotile asbestos has been
identified as being used in products (aftermarket automotive brakes and linings and other
gaskets in UTVs) available to consumers; however, only some individuals within the general
population may use these products {i.e., DIYers or DIY mechanics). Therefore, those who do
use these products are a potentially exposed or susceptible subpopulation due to greater
exposure.
•	Other groups of individuals within the general population who may experience greater
exposures due to their proximity to conditions of use identified in Section 1.4.3 that result in
releases to the environment and subsequent exposures {e.g., individuals who live or work
near manufacturing, processing, use or disposal sites).
Page 142 of 352

-------
Table 2-32 presents the percentage of employed workers and ONUs who may be susceptible
subpopulations within select industry sectors relevant to the chrysotile asbestos COUs. The percentages
were calculated using Current Population Survey (CPS) data for 2017. CPS is a monthly survey of
households conducted by the Bureau of Census for the Bureau of Labor Statistics and provides a
comprehensive body of data on the labor force characteristics. Statistics for the following
subpopulations of workers and ONUs are provided: adolescents, adult men and women. As shown in
Table 2-32, men make up the majority of the workforce in the chrysotile asbestos COUs. In other
sectors, women (including those of reproductive age and elderly women) make up a larger portion of
wholesale and retail trade.
Table 2-32. Percentage of Employed Persons by Age, Sex, and Industry Sector (2017 and 2018
			worker demographics from BLS)		
Age (iroup
Sex
Mining, quarrying, and
oil and gas extraction
.Manufacturing
Wholesale
and retail
trade
( Ol : Oilfield Brake
Block
( Ol : ( hlor-Alkali:
Casket stamping:
(iaskcl use in chemical
plants
( Ol : Auto
brake:
I TV
Adolescent23
(16-19 years)
Male
0.4%
0.8%
3.0%
Female
0.0%
0.4%
3.2%
Adults
(20-54 years)
Male
68.2%
52.9%
42.8%
Female
9.2%
22.2%
35.4%
Elderly (55+)
Male
19.4%
17.5%
12.3%
Female
3.3%
7.3%
9.6%
Manufacturing - The Manufacturing sector comprises establishments engaged in the mechanical,
physical, or chemical transformation of materials, substances, or components into new products.
Establishments in the sector are often described as plants, factories, or mills. For asbestos, this sector
covers the COUs that occur in an industrial setting, including processing and using chlor-alkali
diaphragms, gasket stamping, and gasket use in chemical plants.
Wholesale and retail trade - The wholesale trade sector comprises establishments engaged in
wholesaling merchandise, generally without transformation, and rendering services incidental to the sale
of merchandise. Wholesalers normally operate from a warehouse or office. This sector likely covers
facilities that are engaged in the handling of imported asbestos-containing articles {i.e., aftermarket
automotive parts, other vehicle friction products and other gaskets).
Adolescents, or persons between 16 and 19 years in age, are generally a small part of the total
workforce. Table 2-33 presents further breakdown on the percentage of employed adolescents by
industry subsectors. As shown in the table, they comprise less than 2 percent of the workforce, with the
exception of wholesale and retail trade subsector where asbestos may be used in UTV gaskets and auto
brakes.
23 Note that while BLS defines adolescents as 16-19 years old, EPA defines adolescents as 16 to < 21 years old.
Page 143 of 352

-------
Table 2-33. Percentage of Employed Adolescents by Industry Sector (2017 and 2018 worker
demographics from BLS)
Sector
cor
Adolescents
(16-19 years)
Mining, quarrying, and oil and gas
extraction
Oilfield Brake Block
0.89%
Manufacturing
Chlor- Alkali;
Gasket cut;
Gasket use in chemical plants
1.50%
Wholesale and retail trade
Auto brake;
UTV
6.13%
For consumer exposures, EPA assessed exposures to users and bystanders. EPA assumes, for this
evaluation, consumer users are male or female adults (>16 years of age). Bystanders could be any age
group ranging from infants to adults. EPA estimates bystander risks, including infants, by applying a
specific IUR for age at first exposure and duration of exposure and provides these calculations in
Section 4.2.3.
Page 144 of 352

-------
3 HAZARDS (Effects)
3.1 Environmental Hazards
3.1.1 Approach and Methodology
EPA conducted comprehensive searches for data on the environmental hazards of asbestos, as described
in Strategy for Conducting Literature Searches for Asbestos: Supplemental File for the TSCA Scope
Document ffiPA-HO-OPP'	3(>00831
Only the on-topic references listed in the Ecological Hazard Literature Search Results were considered
as potentially relevant data/information sources for this risk evaluation. Inclusion criteria were used to
screen the results of the ECOTOX literature search (as explained in the Strategy for Conducting
Literature Searches for Asbestos: Supplemental File for the TSCA Scope Document). Since the
terrestrial pathways were eliminated in the PF, EPA only reviewed the aquatic information sources
following problem formulation using the data quality review evaluation metrics and the rating criteria
described in the Application of Systematic Review in TSCA Risk Evaluations (U.S. EPA. 2018a). Data
from the evaluated literature are summarized below and in Table 3-1. in a supplemental file (
2020d) and in Appendix E (data extraction table). Following the data quality evaluation, EPA
determined that of the six on-topic aquatic toxicity studies, four of these studies were acceptable for use
in risk assessment while the two on-topic aquatic plants studies were rated as unacceptable based on the
evaluation strategies described in (	). The studies rated as unacceptable were not used in
this risk evaluation. EPA also identified the following documents sources of environmental hazard data
for asbestos: 45 FR 79318, 1980; AT SDR (2001a):} ^ ^	L_v .v ^ '-014b): WHO (20141:
I ARC (2012b) and Site-Wide Baseline Ecological Risk Assessment, Libby Asbestos Superfund Site,
Libby Montana (	)14b).
3.1.2 Hazard Identification - Toxicity to Aquatic Organisms
Reasonably available information indicated that the hazards from chronic exposure to fish and aquatic
invertebrates following exposure to asbestos at concentrations ranging from 104- 108 fibers/L (which is
equivalent to 0.01 - 100 Million Fibers/Liter (MFL)) resulted in significant effects to development and
reproduction. Sublethal effects were observed following acute and chronic exposure to asbestos at
concentrations lower than 0.01 MFL; for example, reduction in siphoning abilities in clams. As
summarized below and in Appendix Table APX E-l: On-topic Aquatic Toxicity Studies Evaluated for
Chrysotile Asbestos, four citations were determined to be acceptable in quality and relevance for this
risk evaluation. All four citations received a rating of high quality following the data quality evaluation
process.
Bel anger (1986c) exposed larval coho salmon (Oncorhynchus kisutch) and juvenile green sunfish
(Lepomis cyanellus) to chrysotile asbestos at concentrations that were environmentally relevant during
the time of the study and reported behavioral and pathological stress caused by chrysotile asbestos. No
treatment related increases in mortality were detected. Coho were exposed for 40 days at 3.0 MFL and
86 days at 1.5 MFL, while sunfish were exposed for 52 days at 3 MFL and 67 days at 1.5 MFL.
According to the study, coho larvae exposed to 1.5 MFL were significantly more susceptible to an
anesthetic stress test, becoming ataxic and losing equilibrium faster than control fish. Juvenile green
sunfish developed behavioral stress effects in the presence of 1.5 and 3.0 MFL. Specifically, the coho
and green sunfish exposed to 3.0 MFL had sublethal effects, which include the following: epidermal
hypertrophy superimposed on hyperplasia, necrotic epidermis, lateral line degradation, and lesions near
Page 145 of 352

-------
the branchial region. Lateral line abnormalities were associated with a loss of the ability to maintain
normal orientation in the water column.
In addition, Bel anger (1986b) and Bel anger (1986a) investigated the effects of chrysotile asbestos
exposure on larval, juvenile, and adult Asiatic clams (Corbicula sp.). Exposure to 0.01 MFL caused a
significant reduction in release of larva by brooding adults as well as increased mortality in larvae.
Reduced siphoning activity and fiber accumulation in clams were observed in the absence of food after
96-hr of exposure to 0.0001 and 0.1 MFL chrysotile asbestos, respectively Bel anger et a 5b).
Sublethal and reproductive effects observed following 30 days of exposure to 0.0001 to < 100 MFL
chrysotile asbestos include the following: 1) fiber accumulation in gill and visceral tissues, 2) decreased
siphoning activity and shell growth of adult clams, 3) decreased siphoning activity, shell growth, and
weight gain of juveniles, 4) reduction of larva releases, and 5) larva mortality.
Lastly, Belanger (1990) studied the effects of chrysotile asbestos at concentrations of 0, 0.0001, 0.01, 1,
100 or 10,000 MFL on all life stages of Japanese Medaka (Oryzias latipes), including egg development,
hatchability, and survival; reduction in growth of larval to juvenile fish; reproduction performance; and
larval mortality. Eggs were exposed to chrysotile asbestos until hatching for 13-21 days, larvae-juvenile
fish were exposed to chrysotile asbestos for 13 weeks, and juvenile-adult fish were exposed to chrysotile
asbestos for 5 months. Asbestos did not substantially impair egg development, hatchability or survival.
At concentrations of 1 MFL or higher, hatching of eggs was delayed, larval Medaka experienced growth
reduction, and fish developed thickened epidermal tissue. Juvenile fish exposed to 10,000 MFL suffered
98% mortality by 42 days and 100% mortality by 56 days.
3.1.3 Weight of Scientific Evidence
During the data integration stage of systematic review EPA analyzed, synthesized, and integrated the
reasonably available information into Table 3.1. This involved weighing scientific evidence for quality
and relevance, using a weight of scientific evidence (WoE) approach, as defined in 40 CFR 702.33, and
noted in TSCA 26(i) (U.S. EPA. 2018a).
During data evaluation, EPA reviewed on-topic environmental hazard studies for data quality and
assigned studies an overall quality level of high, medium, or low based on the TSCA criteria described
in the Application of Systematic Review in TSCA Risk Evaluations (	i). While integrating
environmental hazard data for asbestos, EPA gave more weight to relevant information that were
assigned an overall quality level of high or medium.
The ten on-topic ecotoxicity studies for chrysotile asbestos included data from aquatic organisms {i.e.,
vertebrates, invertebrates, and plants) and terrestrial species {i.e., fungi and plants). Following the data
quality evaluation, EPA determined that four on-topic aquatic vertebrate and invertebrate studies were
acceptable while the two on-topic aquatic plants studies were unacceptable based on the evaluation
strategies described in the Application of Systematic Review in TSCA Risk Evaluations (U.S. EPA.
2.018a). Since the terrestrial pathways were eliminated in the PF, EPA excluded three studies on
terrestrial species as terrestrial exposures were not expected under the COUs for chrysotile asbestos.
One amphibian study was excluded from further review and considered out of scope because it was not
conducted on chrysotile asbestos. Ultimately the four aquatic toxicity studies were rated high in quality
and used to characterize the adverse effects of chrysotile asbestos to aquatic vertebrate and invertebrate
organisms from chronic exposure, as summarized in Table 3-1. Any information that EPA assigned an
overall quality of unacceptable was not used. The gray literature EPA identified for asbestos had
minimal or no information about environmental hazards and were consequently not used. EPA
Page 146 of 352

-------
determined that data and information were relevant based on whether they had biological,
physical/chemical, and environmental relevance (U.S. EPA. 1998):
•	Biological relevance: correspondence among the taxa, life stages, and processes measured or
observed and the assessment endpoint.
•	Physical/chemical relevance: correspondence between the chemical or physical agent tested and
the chemical or physical agent constituting the stressor of concern.
•	Environmental relevance: correspondence between test conditions and conditions in the
environment.
Table 3-1. Environmental Hazard Characterization of Chrysotile Asbestos
Duration
li'Sl
Organism
Knilpoint
1 lazartl
Value
I nil
KITect KiulpoinUs)
References
Aquatic Organisms
Acute
Aquatic
invertebrates
96-hr LOEC
0.0001-100
MFLd
Reduction in siphoning
activity; Fiber accumulation
Bel anger et al.
!6b)(High)
Chronic
Fish
13-86 day
NOEC3
0.01-1.5
MFL
Behavioral stress (e.g.,
aberrant swimming and loss of
equilibrium); Egg
development, hatchability, and
survival; Growth; Mortality
Bel anger et al.
(High);
13-86 day
LOECb
1-3
Bel anger et al.
:6c)
(High);
Aquatic
invertebrates
30-day
LOEC
0.0001-100
MFL
Reduction in siphoning
activity; Number of larvae
released; Alterations of gill
tissues; Fiber accumulation in
tissues; Growth; Mortality
Bel anger et al.
ป i'ปs b)
(High);
Bel anger et al.
56a)(Hish)
aNOEC, No Observable Effect Concentration.
bLOEC, Lowest Observable Effect Concentration.
0Values in the tables were reported by the study authors and combined in ranges (min to max) from different effect endpoints. The
values of the NOEC and LOEC can overlap because they may be based on different effect endpoints. For example, fish NOEC =1.5
MFL was based on behavioral stress (e.g., aberrant swimming and loss of equilibrium) and fish LOEC = 1 MFL was based on
significant reduction in growth of larval individuals. See Table APX E-l for more details.
dMFL, Million Fibers/Liter.
"Data quality evaluation scores for each citation are in the parenthesis.
3,1,4 Summary of Environmental Hazard
A review of the high-quality aquatic vertebrate and invertebrate studies indicated that chronic exposure
to waterborne chrysotile asbestos at a concentration range of 104-108 fibers/L, which is equivalent to
0.01 to 100 MFL, may result in reproductive, growth and/or sublethal effects to fish and clams. In
addition, acute exposure of waterborne chrysotile asbestos at a concentration range of 102-108 fibers/L to
clams demonstrated reduced siphoning activity.
3.2 Human Health Hazards from Inhalation of Chrysotile Asbestos
Many authorities have established that there are causal associations between asbestos exposures and
lung cancer and mesotheliomas fNTP. 2016: IARC. 2012b: AT SDR. < f,1 v ^ ! >8b; IARC.
1ฎ - / ' I ^ \ s >86; IARC. 1977). EPA also noted in the scope that there is a causal association
Page 147 of 352

-------
between exposure to asbestos and cancer of the larynx and cancer of the ovary ( \RC. 2012b).
Additionally, there is a causal association between asbestos exposures and non-cancer health effects
including respiratory effects {e.g., asbestosis, non-malignant respiratory disease, deficits in pulmonary
function, diffuse pleural thickening and pleural plaques) as well as some evidence of immunological and
lymphoreticular effects (ATS DR.. 2001a). Given the well-established carcinogenicity of asbestos for
lung cancer and mesothelioma and the existence of an IUR for asbestos, EPA, in its PF document,
decided to limit the scope of its systematic review to these two specific cancers and to inhalation
exposures with the goal of updating, or reaffirming, the existing EPA IUR for general asbestos (U.S.
EPA. 1988b). As explained in Section 1.4.1, EPA has determined that the asbestos fiber associated with
the COUs in this Part 1 of the risk evaluation for asbestos is chrysotile asbestos. Thus, the EPA-derived
chrysotile asbestos IUR described in Section 3.2.4 is used to calculate risk estimates.
3,2.1 Approach and Methodology
EPA used the approach described in Figure 3-1 to evaluate, extract and integrate asbestos human health
hazard and dose-response information. This approach is based on th e Application of Systematic Review
in TSCA Risk Evaluations (U.S. EPA. 2018a) and the Framework for Human Health Risk Assessment to
Inform Decision Making (US. EPA. 2014a).
Systematic
Review
Stage
Cancer risk
estimates and
discussion of
WOE
Narrative by
Adverse
Endpoint
(Sections 3.2.1
and 3.2.3)
Data
Summaries for
Adverse
Endpoints
(Appendix B)
IUR derivation
(Sections 3.2.4)
considerations
(Sections 4.2 and
. 4.3) .
Summary
of Fiadiugs
Data
Extraction
Extract data from
key, supporting
and new studies
Data Evaluation
After foil-text screening,
apply pre-determined data
quality evaluation criteria
to assess the confidence of
key and supporting studies
identified from previous
assessments as well as
new studies not
considered in the previous
assessments
Risk Characterization
Analysis
Determines the quantitative
human health risks and
includes, as appropriate, a
discussion of
Considerations regarding
uncertainty and variability
Considerations of
alternative assumptions
Risk
Characterization
Human Health Hazard Assessment
Data Integration
Asbestos is a known carcinogen and the hazard
was accepted a priori. The Health Evaluation
focused on foe Dose-Response Analysis.
Hazard ID
Confirm potential
hazards identified
during
scoping/problem
formulation
Dose-Response
Analysis
Dose-
Response;
Selection of
PODs;
IUR Derivation
Summaries for
Endpoints
IUR c
(Sections 3.2.4)
: B)
Cancer risk
considerations
(Sections 4.2 and
4.3)
Study Quality
Summary Table
(Data quality
ratings and
numerical scores)
(Appendix B)
Output of the
Systematic
Review
Stage
Figure 3-1. EPA Approach to Hazard Identification, Data Integration, and Dose-Response
Analysis for Chrysotile Asbestos
In the PF document, it was stated that the asbestos Risk Evaluation would focus on epidemiological
inhalation data on lung cancer and mesothelioma for all TSCA Title II fiber types, just as stated in the
1988 EPA IRIS Assessment on Asbestos ( IS. EPA. 1988b). This was based on the large database on
the health effects associated with asbestos exposure which has been cited in numerous U.S. and
international data sources. These data sources included, but were not limited to, EPA IRIS Assessment
IRIS Assessment on Asbestos (1988b). IRIS Assessment on Libbv Amphibole Asbestos (2014c).
National Toxicology Program (N I P) Report on Carcinogens. Fourteenth Edition (2016). NIOSH
Page 148 of 352

-------
Asbestos Fibers and Other Elongate Mineral Particles: State of the Science and Roadmap for Research
CaIU!)), ATSDR Toxicologic•.A Pnซ(tte for Asbestos (2001a). IARC Monographs on the Evaluation of
Carcinogenic Risks to Human nic. Metals. Fibres, and Dusts. Asbestos (Chrvsotile. Amosite.
Crocidolite. Tremolite. Actinolite. and Anthophyllite) (2012b). and World Health. Organization. (WHO)
Chrvsotile Asbestos (2014).
EPA conducted comprehensive searches for reasonably available information on health hazards of
asbestos, as described in Strategy for Conducting Literature Searches for Asbestos: Supplemental File
for the TSCA Scope Document (EPA-HQ-OPPT-2016-0736). The relevant studies were evaluated using
the data quality criteria in the Application of Systemic Review in TSCA Risk Evaluations document (U.S.
EPA. 2018a). The process and results of this systematic review are available in a supplemental
document (see Systematic Review Supplemental File: Data Quality Evaluation and Data Extraction of
Human Health Hazard Studies).
This EPA human health hazard assessment consists of hazard identification and dose-response
assessment as described in EPA's Framework for Human Health Risk Assessment to Inform Decision
Making (U.S. EPA. 2014a). Hazards were identified from consensus documents. EPA integrated
epidemiological studies of asbestos with other readily available information to select the data to use for
dose-response assessment. Dose-response modeling was performed for the hazard endpoints with
adequate study quality and acceptable data sets.
After publication of the PF document, EPA determined that only chrysotile asbestos is still imported into
the U.S. either in raw form or in products; the other five forms of asbestos have neither known, intended,
nor reasonably foreseen manufacture, import, processing, or distribution. EPA will consider other legacy
uses and associated disposals of asbestos in a separate and forthcoming Part 2 of the risk evaluation for
asbestos (as noted in the Preamble). Therefore, for this document, in order to inform the estimation of an
exposure-response function allowing for the derivation of a chrysotile asbestos IUR, EPA identified
epidemiological studies on mesothelioma and lung cancer in cohorts of workers using chrysotile
asbestos in commerce. To identify studies with the potential to be used to derive an IUR, EPA also
screened and evaluated new studies that were published since the EPA IRIS assessment conducted in
1988.
The new literature was screened against inclusion criteria in the PECO statement, and the literature was
further screened to identify only hazard studies with inhalation exposure to chrysotile asbestos. Cohort
data deemed as "key" was entered directly into the data evaluation step based on its relevance to the risk
evaluation. The relevant (e.g., useful for dose-response for the derivation of the IUR) study cohorts were
further evaluated using the data quality criteria for human studies. Only epidemiological hazard studies
by inhalation and only chrysotile asbestos exposures were included.
EPA developed unique data quality criteria for epidemiological studies on asbestos exposure and
mesothelioma and lung cancer for the study domains of exposure, outcome, study participation,
potential confounding, and analysis which were tailored to the specific needs of evaluating asbestos
studies for their potential to provide information on the exposure-response relationship between asbestos
exposure and risk of lung cancer and from mesothelioma, (see Section 3.2.4 and Systematic Review
Supplemental File: Data Quality Evaluation and Data Extraction of Human Health Hazard Studies
(U.S. EPA. 2020h)). EPA considered studies of low, medium, or high confidence for dose-response
analysis for the derivation of the IUR. Information that was rated unacceptable was not included in the
risk evaluation (U.S. EPA. 2018a). The Supplemental File presents the data quality information on
human health hazard endpoints (cancer) for all acceptable studies (with low, medium, or high scores).
See section 3.2.4.
Page 149 of 352

-------
Following the data quality evaluation, EPA extracted a summary of data from each relevant cohort. In
the last step, the strengths and limitations of the data among the cohorts of acceptable quality were
evaluated for each cancer endpoint and a weight-of-the-scientific evidence narrative was developed.
Data for either mesothelioma or lung cancer was modeled to determine the dose-response relationship.
Finally, the results were summarized, and the uncertainties were presented. The process is described in
Section 3.2.4.
Section 3.2.4.3 describes the epidemiological studies chosen for the derivation of the IUR for chrysotile
asbestos.
3,2,2 Hazard Identification from Inhalation of Chrysotile Asbestos
Asbestos has an existing EPA IRIS Assessment, an ATSDR Toxicological Profile, and many other U.S.
and international assessments (see Section 1.3); hence, many of the hazards of asbestos have been
previously compiled and reviewed. Most of the information in these assessments is based on inhalation
exposures to human populations. Only inhalation exposures in humans are evaluated in this Part 1 of the
risk evaluation of asbestos. EPA identified key and supporting studies from previous peer reviewed
assessments and new studies published since 1988 and evaluated them against the data quality criteria
developed for all types of asbestos - including chrysotile asbestos. The evaluation criteria were tailored
to meet the specific needs of asbestos studies and to determine the studies' potential to provide
information on the exposure-response relationship between asbestos exposure and risk of lung cancer
and from mesothelioma.
During scoping and PF, EPA reviewed the existing EPA IRIS health assessments to ascertain the
established health hazards and any known toxicity values. EPA had previously, in the IRIS assessment
on asbestos (	88b). identified asbestos as a carcinogen causing both lung cancer and
mesothelioma from inhalation exposures and derived an IUR to address both cancers. No toxicity values
or IURs have yet been estimated for other cancers that have been identified by the International Agency
for Research on Cancer (IARC) and other government agencies. Given the well-established
carcinogenicity of asbestos for lung cancer and mesothelioma, EPA, in its PF document, had decided to
limit the scope of its systematic review to these two specific cancers and to inhalation exposures with
the goal of updating, or reaffirming, the existing unit risk. As explained in Section 1.4, the only COUs of
asbestos or asbestos containing products assessed in this risk evaluation are for chrysotile asbestos.
Thus, an IUR value for chrysotile asbestos only was developed.
3.2_.2.1_ Non-Cancer Hazards from Inhalation of Chrysotile Asbestos
Asbestos exposure is known to cause various non-cancer health outcomes including respiratory and
cardiovascular effects. Respiratory effects of asbestos are well-documented and include asbestosis, non-
malignant respiratory disease (NMRD), deficits in pulmonary function, diffuse pleural thickening
(DPT), and pleural plaques. Various immunological and lymphoreticular effects are suggested, but not
well-established (ATSDR. 2001a: U.S. EPA. 1988b)
These non-cancer effects are adverse. Asbestosis and NMRD have been observed to be a cause of
mortality in many asbestos exposed cohorts. DPT and pleural plaques decrease pulmonary function
(' r U \	), Pulmonary deficits are considered to be adverse for an asbestos-exposed population,
because a shift in distribution of pulmonary function in an exposed population results in a considerable
increase in the proportion of individuals with a significant degree of pulmonary impairment below a
clinically adverse level.
There is no RfC for general asbestos {i.e., actinolite, amosite, anthophyllite, chrysotile, crocidolite,
Page 150 of 352

-------
tremolite), derived by EPA or any of the consensus organizations; only Libby Amphibole asbestos has a
RfC which is based on pleural plaques S * j_\\ 20] 4c,).
3.2.2.2	Cancer Hazards from Inhalation of Chrysotile Asbestos
Many authorities have established that there are causal associations between asbestos exposures and
lung cancer and mesotheliomas in humans based on epidemiologic studies CNTP. 2016; I ARC. 2.012b;
AT SDR. 2001a; U.S. EPA. 1988b; I ARC. 1987; U.S. EPA. 1986; IARC. 1977V EPA also noted in the
scope that there is a causal association between exposure to asbestos and cancer of the larynx and cancer
of the ovary (IARC. 2012b). and that there is also suggestive evidence of a positive association between
asbestos and cancer of the pharynx (IARC. 2012b; NRC. 2006). stomach (IARC. 2012b; AT SDR.
2.001a) and colorectum CNTP. : . 1' \
1980). In addition, the scope document reported increases in lung cancer mortality in both workers and
residents exposed to various asbestos fiber types, including chrysotile asbestos, as well as fiber mixtures
(IARC. 2012b). Mesotheliomas, tumors arising from the thin membranes that line the chest (thoracic)
and abdominal cavities and surround internal organs, are relatively rare in the general population, but are
observed at much higher frequencies in populations of asbestos workers. All types of asbestos fibers
have been reported to cause mesothelioma - including chrysotile asbestos (IARC. 2012b;
1988b. 1986).
During PF, EPA reviewed the existing EPA IRIS health assessments (U.S. EPA. 2014c. 1988b) to
ascertain the established health hazards and any known toxicity values. EPA had previously (U.S. EPA.
1988b. 1986) identified asbestos as a carcinogen causing both lung cancer and mesothelioma and
derived a unit risk based on epidemiologic studies to address both cancers. The U.S. Institute of
Medicine (IOM. 2006) and the International Agency for Research on Cancer (IARC. 2012b) have
evaluated the evidence for causation of cancers of the pharynx, larynx, esophagus, stomach, colon, and
rectum, and IARC has evaluated the evidence for cancer of the ovary. Both the U.S. Institute of
Medicine and IARC concluded that asbestos causes laryngeal cancer and IARC concluded that asbestos
causes ovarian cancer. No toxicity values or IURs have yet been estimated for either laryngeal or
ovarian cancers.
3.2.2.3	Mode of Action (MOA) Considerations for Chrysotile Asbestos
EPA evaluated the evidence supporting plausible modes of action (MOA) of chrysotile asbestos
carcinogenicity for specific tumor locations using the modified Hill criteria for MO A analysis described
in EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005). EPA considered available
evidence from animal cancer bioassays, genotoxicity studies, specific MO As proposed in the literature,
and the analysis previously presented in the IRIS Toxicological Review of Libby Amphibole Asbestos
2014c) and the International Agency for Research on Cancer (IARC) proposed a mechanism for the
carcinogenicity of asbestos fibers [see Figure 4-2 in IARC (2012b)l.
EPA specifically considered MO As for lung carcinogenicity and mesothelioma of chrysotile asbestos.
There is insufficient chemical-specific information about larynx, ovary, pharynx, stomach and
colorectum tumors to support MOA analysis for these tumor types.
Potential Modes of Action for Chrysotile Asbestos Lung Carcinogenicity and Mesothelioma
Physicochemical properties of chrysotile fibers
Chrysotile asbestos falls into the serpentine asbestos mineral group. The chrysotile crystal structure
results from the association of a tetrahedral silicate sheet with an octahedral brucite-like sheet. These
Page 151 of 352

-------
two sheets form a silicate layer with a slight misfit that causes curling to form concentric cylinders with
the silicate layer on the inside and brucite-like layer on the outside of the cylinder. The fibrils are held
together by Van der Waals interparticle forces so that when the chrysotile fiber is broken up large
numbers of smaller fibers are generated (Fubini and Arean. 1999).
It has been proposed that the pathogenic potential of asbestos fibers depends on their aspect ratio and
fiber size (IARC. 2012b). While some shorter asbestos fibers have been shown to be cleared by the
system more efficiently, evidence from in vitro genotoxicity studies in Chinese hamster lung cells
suggests that short and intermediate chrysotile fibers may induce micronuclei formation and sister
chromatid exchange (Lu et at... 1994). There is some evidence that aspect ratio and size may play
differing roles in the onset and progression of lung cancer and mesothelioma. NIOSH (2008) reported an
association of lung cancer with fibers longer than 10 |im and thicker than 0.15 |im, while mesothelioma
was more closely associated with shorter, thinner fibers (~5 |im long and 0.1 |im thin). However, this
evidence is equivocal as multiple epidemiologic studies [summarized in the I ARC monograph, IARC
(2012b)] have reported the presence of short fibers (<5 |im, typically associated with fibrosis) in the
lung and pleural tissue of malignant mesothelioma patients.
Fiber aerodynamic diameter is a key determinant of the extent of deposition and penetration to different
parts of the respiratory tract (TOM 2006). Fibers with an aerodynamic diameter less than 3 (.im, which
includes chrysotile asbestos, are capable of penetration into the deep pulmonary region (NIOSH. 201 la).
Generation of reactive oxygen and reactive nitrogen species
In addition to aspect ratio and size, the surface of asbestos fibers has reactivity that may generate
reactive oxygen species (ROS), reactive nitrogen species (RNS) and lead to iron mobilization and or
biodepositon as reviewed in Miller et al. (2.014). The surface reactivity of asbestos fibers, including
chrysotile asbestos, has been implicated in the pathogenesis of lung cancer and mesothelioma. The
ability of chrysotile, and other asbestos fibers, to produce ROS and RNS depends on the presence of iron
ions on the surface of the fibers (Gazzano et al.. 2005). While chrysotile asbestos is a low iron
containing asbestos fiber, it has been shown to produce ROS (Wang et al.. 2019; Miller et al.. 2014;
Kopnin et al.. 2004). It has been postulated that lung macrophages encounter inhaled asbestos fibers and
proceed to undergo phagocytosis but are unable to complete the process leading to "frustrated
phagocytosis24" resulting in oxidative stress. Chrysotile fiber, and other asbestos fiber, derived-oxidative
stress has been shown to damage cellular macromolecules, such as proteins, lipids and nucleic acids
(Miller et al.. 2014; Gulumian. 1999; Ghio et al.. 1998) and apoptosis (Upadhyav and Karop. 2003;
Simeonova and Luster. 1995) which may then contribute and play key roles in the onset and progression
of asbestos related diseases, such as lung cancer and mesothelioma.
Overall MOA conclusions
Evidence from both in vitro and in vivo studies strongly suggest that the physicochemical properties of
chrysotile asbestos fiber along with the reactive oxidants generated by these are key in the pathogenesis
of asbestos related diseases such as lung cancer and mesothelioma. However, there is currently
insufficient information to determine the MOA for either chrysotile lung carcinogenicity or
mesothelioma. Chrysotile asbestos mesothelioma and lung carcinogenicity may be mediated by different
underlying complex mechanisms that have yet to be fully elucidated. In the absence of other information
about MOA, EPA often takes the health-protective approach of assuming a linear no-threshold risk
model consistent with a mutagenic MOA (	305).
24 Frustrated phagocytosis: When a phagocyte fails to engulf its target, in this case the asbestos fiber, and the toxic agent
(asbestos) results in the target being released or spread into the environment (Mularski et al. 20.1.8).
Page 152 of 352

-------
3,2,3 Derivation of a Chrysotile Asbestos Inhalation Unit Risk
3.2.3.1	Considerations in Derivation of a Chrysotile Asbestos Inhalation Unit Risk
EPA did not have a previous, recent risk assessment of asbestos on which to build; therefore, the
literature was reviewed to determine whether a new IUR needed to be developed. As the RE process
progressed, several decisions were made that refined and narrowed the scope of the RE. It was
determined during PF that the RE would focus on epidemiologic data on mesothelioma and lung cancer
by the inhalation route. The existing EPA IUR for asbestos was developed in 1988 was based on 14
epidemiologic studies that included occupational exposure to chrysotile, amosite, or mixed-mineral
asbestos exposures (chrysotile, amosite, crocidolite). However, EPA's research to identify COUs
indicated that only chrysotile asbestos is currently being imported in the raw form or imported in
products. The other five forms of asbestos identified for this risk evaluation are no longer manufactured,
imported, processed, or distributed in the United States. This commercial chrysotile asbestos is therefore
the substance of concern for this quantitative assessment and thus EPA sought to derive an IUR specific
to chrysotile asbestos. The epidemiologic studies available for risk assessment all include populations
exposed to commercial chrysotile asbestos, which may contain small, but variable amounts of
amphibole asbestos. Because chrysotile asbestos is the only form of asbestos in the United States with
COUs in this document, studies of populations exposed only to chrysotile asbestos provide the most
informative data for the purpose of developing the TSCA risk estimates for the COUs for chrysotile
asbestos. EPA will consider legacy uses and associated disposals of asbestos in a separate and
forthcoming Part 2 of the risk evaluation for asbestos.
3.2.3.2	Rationale for Asbestos-Specific Data Evaluation Criteria
For the first 10 TSCA REs, a general set of study evaluation criteria was developed. These data
evaluation criteria were not tailored to any specific exposure or outcome. In the PF step of the asbestos
assessment, it was accepted that exposure to asbestos was a known cause of lung cancer and
mesothelioma, and that the purpose of the systematic review would be the identification of studies which
could inform the estimation of an exposure-response function allowing for the derivation of an asbestos
inhalation unit risk for lung cancer and mesothelioma combined. The study domains of exposure,
outcome, study participation, potential confounding, and analysis were further tailored to the specific
needs of evaluating asbestos studies for their potential to provide information on the exposure-response
relationship between chrysotile asbestos exposure and risk of lung cancer and mesothelioma (U.S. EPA.
2020hY
In terms of evaluating exposure information, asbestos is unique among these first 10 TSCA chemicals as
it is a fiber and has a long history of different exposure assessment methodologies. For mesothelioma,
this assessment is also unique with respect to the impact of the timing of exposure relative to the cancer
outcome as the time since first exposure plays a dominant role in modeling risk. The most relevant
exposures for understanding mesothelioma risk were those that occurred decades prior to the onset of
cancer. Asbestos measurement methodologies have changed over those decades, from early
measurement of total dust particles measured in units of million particles per cubic foot of air (mppcf)
by samplers called midget impingers to fibers per milliliter (f/ml), or the equivalent fibers per cubic
centimeter (f/cc), where fiber samples were collected on membrane filters and the fiber count per
volume of air was measured by analyzing the filters using phase contrast microscopy (PCM). In several
studies encompassing several decades of asbestos exposures, matched samples from midget impingers
and membrane filters were compared to derive job- (or location-) specific factors allowing for the
conversion of earlier midget impinger measurements to estimate PCM measurement of asbestos air
concentrations. While some studies were able to provide these factors for specific locations and jobs,
Page 153 of 352

-------
other studies were only able to derive one factor for all jobs and locations. The use of such data has
allowed asbestos researchers to investigate the risk of asbestos and successfully model lung cancer and
mesothelioma mortality over several decades of evaluation (U.S. EPA. 2014c. 1988b. 1986). Thus, the
general exposure evaluation criteria were adjusted to be specific to exposure assessment methodologies
such as midget impingers and PCM with attention to the use of job-exposure-matrices (JEMs) to
reconstruct workers' exposure histories and the reporting of key metrics needed to derive exposure-
response functions for lung cancer and mesothelioma.
In terms of evaluating the quality of outcome information, lung cancer is relatively straightforward to
evaluate as an outcome. Specific International Classification of Disease (ICD) codes for lung cancer
have existed for the entire time period of the studies evaluated here making it possible to identify cases
from mortality databases. On the other hand, there was no diagnostic code for mesothelioma in the
International Classification of Diseases prior to the introduction of the 10th revision (ICD-10) which was
not implemented in the United States until 1999. Before ICD-10, individual researchers employed
different strategies (e.g., generally searched original death certificates for mention of mesothelioma,
considered certain ICD codes known to be substitutes for mesothelioma coding in the absence of a
specific ICD code). Thus, the general outcome evaluation criteria were adjusted to be specific to
mesothelioma and outcome ascertainment strategies.
Mesothelioma is a very rare cancer. As noted by U.S. EPA (2014c). the "Centers for Disease Control
and Prevention estimated the death rate from mesothelioma, using 1999 to 2005 data, as approximately
23.2 per million per year in males and 5.1 per million per year in females (CDC. 2009)." While very
rare, the overwhelmingly dominant cause of mesothelioma is asbestos exposure (Tossavainen. 1997).
making the observance of mesothelioma in a population a very specific indicator for asbestos exposure.
The prevailing risk model for mesothelioma is an absolute risk model, which assumes there is no risk at
zero exposure (' _ \ * Peto et a! . 1982; Peto. 1978). This use of an absolute risk model differs
from the standard use of a relative risk model for lung and other cancers. For the relative risk model, the
risk of lung cancer in an asbestos exposed population multiplies a background risk in an unexposed
population. Thus, an important consideration of study quality is the evaluation of that comparison
population. However, for mesothelioma, no comparison population is needed to estimate the absolute
risk among people exposed to asbestos, and therefore the criteria in the study participation domain (that
includes comparison population) were adjusted for mesothelioma.
In terms of evaluating potential confounding, the generic potential confounding section was adapted to
recognize that there are both direct and indirect methods for controlling for some confounders.
Specifically, methodologies that involve only internal comparisons within a working population may
indirectly control for smoking and other factors assuming these factors do not vary with asbestos
exposure concentrations in the workplace. In contrast, mesothelioma is much simpler to evaluate for
potential confounding as diagnostic X-ray contrast medium "Thorotrast" and external beam
radiotherapy are the only other known non-fibrous risk factors for mesothelioma, and these are unlikely
to be confounders because these rare procedures are not routinely done on healthy workers. Screening
programs typically X-ray all workers - regardless of their cumulative asbestos exposure. There are other
fibrous risk factors for mesothelioma such as fluoroedenite (Grosse et at. 2014) and erionite (IARC.
2012a). but exposures to these materials which are not used in conjunction with chrysotile asbestos in
COUs.
In terms of evaluating analysis, the evaluation criteria were adapted for both mesothelioma and lung
cancer. For mesothelioma, the Peto model (Peto et at.. 1982; Peto. 1978) has traditionally been used for
summary data published in the literature (U.S. EPA. 1986) but also has been used with individual-level
data (e.g..Berman and Crimp (2.008a)). so studies were considered acceptable if the authors reported
Page 154 of 352

-------
sufficient information on the results of using the Peto model or presented sufficient information to fit the
Peto model post hoc. For lung cancer, a wider selection of statistical models was acceptable, with the
preference generally given to modeling that used individual data in the analysis.
3.2.3.3 Additional considerations for final selection of studies for exposure-response
As shown in Figure 1-8, EPA's literature search identified more than 24,000 studies, but for the final
data evaluation 26 papers covering seven cohorts were identified, and these cohorts are listed in Table
3-2.
In reviewing these available studies, EPA distinguished between studies of exposure settings where only
commercial chrysotile asbestos was used or where workers exposed only to commercial chrysotile
asbestos could be identified, and situations where chrysotile asbestos was used in combinations with
amphibole asbestos forms and the available information does not allow exposures to chrysotile and
amphibole asbestos forms to be separated. Studies in the latter group were judged to be uninformative
with respect to the cancer risks from exposure to commercial chrysotile and were excluded from further
consideration (e.g., Slovenia cohort: Dodic et al., 2007; 2003).
All the studies determined to be informative for lung cancer and mesothelioma analysis were based on
historical occupational cohorts. Some cohorts have been the subject of multiple publications; in these
cases, only data from the publication with the longest follow-up for each cohort or the most relevant
exposure-response data were used unless otherwise specified.
Studies were deemed informative for lung cancer risk assessment if either the relative risk of lung cancer
per unit of cumulative chrysotile asbestos exposure in fibers per cc-year (f/cc-yrs) were available from
fitting log-linear or additive relative risk models or the data needed to fit such models as described
below. The group of Balangero, Italy cohort studies including Pira et al. (2009) was excluded for lack of
results from models using a continuous measure of exposure. Studies that presented lung cancer risks
only in relation to impinger total dust exposure were excluded from consideration unless they provided
at least a data-based, study-specific factor for converting concentrations from mppcf to f/cc.
EPA identified studies of five independent occupational cohorts exposed only to commercial chrysotile
asbestos that provided adequate data for assessment of lung cancer risks: chrysotile asbestos textile
manufacturing workers in North Carolina and South Carolina, USA (Loomis et al.. 2009: Hein et at.
2007) and Chongqing, China (Dene et al.. 2012) and chrysotile asbestos miners in Quebec, Canada
(Liddell et al.. 1997). and Qinghai, China (2014: Wane et al.. 2 ). A pooled analysis of the two U.S.
studies (NC and SC) chrysotile asbestos textile cohorts (Elliott et al.. 2012) also provides informative
data about analysis of pooled as well as individual data from both cohorts. In addition, Berman and
Crump (2008b) provide informative risk estimates for the Quebec miner cohort based on modeling dose-
response data that were not available in the original study.
Studies were considered informative for mesothelioma risk assessment if risk estimates from fitting the
EPA mesothelioma model to individual-level data or data needed to fit the model as described below
were available. None of the original publications reported risk estimates from fitting the Peto model.
However, Berman & Crump (2008b) provide risk estimates for the Quebec miners and South Carolina
workers from analyses of original, individual-level data Liddell et al. (1997) and Hein et al. (2007).
respectively. Comparable risk estimates were generated for North Carolina textile workers (Loomis et
al.. 2009) using tabulated mesothelioma data (Loomis et al.. 2019). Data needed to fit Peto
mesothelioma model have not been published for any other cohort exposed to chrysotile asbestos only.
Page 155 of 352

-------
Table 3-2. Study Cohort, Individual studies and Study Quality of Commercial Chrysotile Asbestos
Study Cohort
Author, Year
Study Quality**

Herman and Cmmo (2008b)


vn et al. (1994)


Cole et )


Dement et al. (1983b)


Dement et al. (1994)
Lung Cancer

Dement and Brown (1994)
1.6 High
South Carolina, US
Edwards et al. (2014)


Elliott et al. (2.012)
Mesothelioma

Hein et al. (2007)
1.7 Medium

Loom is et al. (2012)


SRC (2019c)


Stavner et al. (1997)


Stavner et al. (2008)


Wane et al. (2012)
Lung Cancer
1.6 High
Qinghai, China - miners
Wane et al. (2013b)

Wane et al. (2014)

Piolatto et al. (1990)

Balangero, Italy*
a et al. (2009)
NA
Pira et al. (2017)

Rubinoetal. (1979)


Dement et al. (2008)


Elliott et al. (2012)
Lung Cancer

Loomis et al. (2009)
1.7 Medium
North Carolina, US
Loomis et al. (2010)


Loomis et al. (2012)
Mesothelioma

Loomis et al. (2019)
1.5 High

SRC (2019a)

Salonit Anhovo, Slovenia*
Dodic Fikfak (2003)
NA
Dodic Fikfak et al. (2007)

Berman and Crump (2.008b)


Gibbs and Lachano ,)
Lung Cancer***
Low* * *

Liddell et al. (1997)

Liddell et al. (1998)
Quebec, Canada
Liddell an strone (2002)
Mesothelioma
Medium***

)

Mcdonald 3! ปr^Vซh)

SRC (2019b)


Vacek (1998)


Courtice et al. (2016)

Chongqing, China -
Dene et al. (2012)
Lung Cancer
1.4 High
asbestos products factory
Wang et al. (2014)
including textiles
Wane et al (2013a)

10 et al. (2001)

Page 156 of 352

-------
* Cohorts from Italy and Slovenia are not considered further (see text above the table)
** Detailed information on Study quality is in Systematic Review Supplemental File: Data Quality Evaluation and Data
Extraction of Human Health Hazard Studies
*** Study quality was downgraded for this cohort during conflict resolution between primary review and QA/QC review.
Downgrading was due to lack of PCM or TEM-equivalent exposure estimates and potentially significant co-exposure to
tremolite or other amphiboles.
3.2.3.4 Statistical Methodology
The first step in deriving a cancer unit risk for risk estimation is to identify potency factors for lung
cancer and mesothelioma. Cancer potency values are either extracted from published epidemiology
studies or derived from the data within those studies. Once the cancer potency values have been
obtained, they are adjusted for differences in air volumes between workers and other populations
(Section 3.3.3.4.3). Those adjusted values can be applied to the U.S. population as a whole in the
standard EPA life-table analyses. These life-table analyses allow for the estimation of an exposure
concentration associated with a specific extra risk of cancer incidence caused by asbestos. The unit risks
for lung cancer and mesothelioma are estimated separately and then combined to yield the cancer
inhalation unit risk.
3.2.3.4.1 Cancer Risk Models for Asbestos Exposures
A cancer risk model predicts the probability of cancer in an individual with a specified history of
exposure to a cancer-causing agent. In the case of inhalation exposure to chrysotile asbestos, the cancer
effects of chief concern are lung cancer and mesothelioma, and exposure history is the product of the
level and timing of the asbestos exposure. The most common model forms are described below.
Lung Cancer
For lung cancer, the risk from epidemiologic studies from exposure to asbestos is usually quantified
using a linear relative risk model of the following form (Berman and Crump. 2008b; U.S. EPA. 1988b.
1986):
RR = a (1 + CE Kl)
where:
RR	= Relative risk of lung cancer
CE	= Cumulative exposure to asbestos (f/cc-yrs), equals the product of exposure
concentration (f/cc) and the duration of exposure (years). In many publications, exposure estimates are
"lagged" to exclude recent exposures, since lung cancer effects usually take at least 10 years to become
apparent. In this case, cumulative exposure is indicated as CE10 to represent the 10-year lag period.
Kl	= Lung cancer potency factor (f/cc-yrs)"1.
a	= The ratio of baseline (unexposed) risk in the study population compared to the
reference population. If the reference population is well-matched to the study population, a is usually
assumed to be constant=l and is not treated as a fitting parameter. If the general population is used as
the reference population, then a may be different from 1 and is treated as a fitting parameter.
A re-parametrization with a = exp (J3o) is called the linear relative rate model. For epidemiologic studies
where, individual data analysis was conducted, other models have been used for modeling lung cancer.
Page 157 of 352

-------
These include both linear relative rate model (e.g., Hein et ai. 2007). the Cox proportional hazard model
(e.g., U.S. EPA. (2014c); Wang et al. (2014)) and other log-linear relative rate models (e.g., Elliott et al.
(2012); Loom is et al. (2009)). Results from all these model types were considered to be informative in
estimating the lung cancer potency factor (Kl) and were carried forward for further consideration.
Mesothelioma
For mesothelioma, the risk model is usually an absolute risk model that gives the risk of mesothelioma
in an individual following exposure to asbestos that is a function of the concentration and length of time
since first exposure. The model form below was originally proposed by Peto	J) and Peto
|) and was subsequently used by U.S. EPA (1986). Berman and Crump (2008b) adapted this model
for variable exposure, which is used in this Part 1 of the risk evaluation for asbestos.
Im = C Km Q
where:
Im	=	Rate of mesothelioma (cases per person year)
C	=	Concentration of asbestos (f/cc)
Km	=	Mesothelioma potency factor (f/cc-yrs3)"1
Q	=	A cubic function of the time since first exposure (TSFE) and the duration (d) of
exposure, as follows:
•	for TSFE <10	Q = 0
•	for 10d+10	Q = (TSFE - 10)3 - (TSFE - 10 - d)3
3.2.3.4.2 Derivation of Potency Factors
Values for the cancer potency factors (Kl and Km in the equations above) are derived by fitting a risk
model to available exposure-response data from epidemiological studies of workers exposed to asbestos.
Fitting is performed using the method of Maximum Likelihood Estimation (MLE), assuming that the
observed number of cases in a group is a random variable described by the Poisson distribution.
In general, the preferred model for fitting utilizes individual-level observations. This allows for the
exposure metric to be treated as a continuous variable, and also allows for the inclusion of categorical
covariates of potential interest such as gender, calendar interval, race, and birth cohort. When the
individual data are not available, then the data for individuals may be summarized into groups according
to a key exposure metric (CE10 for lung cancer, TSFE for mesothelioma), and the mid-point of the
range for each exposure metric is usually used in the fitting, unless means/medians of exposure metric
were available. In cases where the upper bound of the highest exposure category was not reported in the
publication, the value for the upper bound of the highest exposure category was assumed to be the
maximum exposure reported in the publication. Background parameter a in lung cancer model was both
assumed fixed at 1 and fitted. Results with lower AIC are shown in the tables below. Full modeling
results for both cases are shown in Appendix J.
Page 158 of 352

-------
In cases where study authors reported a potency factor derived using an appropriate model, that value
was retained for consideration. In cases where the authors did not report a potency factor derived by an
appropriate method, EPA estimated the potency factor by fitting a model to data summarized into
groups, if they were reported. EPA fitting was performed using SAS. Appendix G provides the SAS
codes that were employed. As a quality check, calculations were also performed using Microsoft Excel.
Both methods yielded the same results to three or more significant figures.
When the potency factors were estimated by the study authors, EPA relied upon the confidence bounds
reported by the authors. These were generally Wald-type bounds. The inhalation unit risk (see below) is
derived from the one-sided 95% upper bound (which is equivalent to the upper bound of the two-sided
90% upper bound). In the literature, authors typically report two-sided 95% confidence intervals (i.e.,
from the 2.5% to the 97.5% bounds). In these instances, EPA computed the standard error of the effect
estimate from the published results and used that value to estimate the 5% and 95% confidence bounds,
assuming a normal distribution.
When EPA performed the fitting, 90% two-sided confidence bounds around the potency factors were
derived using the profile likelihood method. In this method, the 100(1-a) confidence interval is
computed by finding the two values of the potency factor that yield a log-likelihood result that is equal
to the maximum log-likelihood minus 0,5-%2( l -a, 1), i.e., central chi-square distribution with one degree
of freedom and confidence level 1-a. For a 90% confidence interval, this is equal to the maximum log-
likelihood minus 1.353.
3.3.3.4.3 Extrapolation from Workers to the General Population to
Derive an Inhalation Unit Risk
Because EPA defines the cancer inhalation unit risk for asbestos as an estimate of the increased cancer
risk from inhalation exposure to a concentration of 1 f/cc for a lifetime25, and the cancer potency factors
are derived by fitting risk models to exposure-response data based on workers, it is necessary to adjust
the worker-based potency factors to derive values that are applicable to an individual with a different
exposure pattern (e.g., a bystander to consumer/DIY COU). The extrapolation is based on the
assumption that the ratio of the risk of cancer in one population compared to another (both exposed to
the same level of asbestos in air) is related to the ratio of the amount of asbestos-contaminated air that is
inhaled per unit time (e.g., per year).
For workers, EPA assumes a breathing rate of 10 m3 of air per 8-hour work day (	2009). If
workplace exposure is assumed to occur 240 workdays/year, the volume of air inhaled in a year is
calculated as follows:
Volume Inhaled (worker) =10 m3/workday • 240 workdays/yr = 2,400 m3/yr
For a resident, EPA usually assumes a breathing rate of 20 nrVday (	>09). If exposure is
assumed to be continuous (24 hours per day, 365 days per year), the volume inhaled in a year is
calculated as follows:
Volume Inhaled (resident) = 20 m3/day • 365 days/yr = 7,300 m3/yr
In this case, the extrapolation factor from worker to resident is:
25 Note that the lifetime inhalation unit risk is then applied to specific environmental exposure scenarios applicable to current
chrysotile asbestos uses; for specific worker exposure scenarios, the extrapolation factor described may not be applied.
Page 159 of 352

-------
Extrapolation factor = 7,300 / 2,400 = 3.042
In the tables below (Section 3.2.4), the potencies are shown as calculated from epidemiological studies,
and the worker to other populations extrapolation factor is applied in the life-table analyses so that the
unit risks and IUR incorporate that extrapolation factor.
3.2.3.4.4 Life-Table Analysis and Derivation of Inhalation Unit Risk
Potency factors are not analogous to lifetime unit risks or cancer slope factors, and do not directly
predict the excess risk of lung cancer or mesothelioma in an exposed individual. Rather, the potency
factors are used in lifetable analyses for lung cancer and mesothelioma to predict the risk of cancer as a
result of the exposure in a specified year of life. However, it is important to recognize that cancer risk in
a particular year of life is conditional on the assumption that the individual is alive at the start of the
year. Consequently, the risk of a chrysotile asbestos-related cancer within a specified year of life is
calculated as a function of the probability of being alive at the start of the year, the background
probability of getting cancer, and the increased risk of getting cancer from chrysotile asbestos exposure
within the specified year. The lifetime risk is then the sum of all the yearly risks. This procedure is
performed to calculate the lifetime risk both for an unexposed individual (Ro) and for an individual with
exposure to chrysotile asbestos (Rx).
"Extra risk" for cancer is a calculation of risk which adjusts for background incidence rates of the same
type of cancer, by estimating risk at a specified exposure level and is calculated as follows (U.S. EPA.
2012V
Extra Risk = (Rx - Ro) / (1 - Ro)
For mesothelioma, because background risk (Ro) is assumed to be zero, extra risk is the same as absolute
risk (Rx).
The unit risk is risk of incident cancer per unit asbestos concentration (fiber/cc or f/cc) in inhaled air.
The unit risk is calculated by using life table analysis to find the exposure concentration (EC) that yields
a 1% (0.01) extra risk of cancer. The 1% value is referred to as the Benchmark Response (BMR). This
value is used because it represents a cancer response level that is near the low end of the observable
range (U.S. EPA. 2012.1
As described in Section 3.2.2.3, because MOA for chrysotile asbestos is uncertain, following the
recommendations of the Guidelines for Carcinogen Risk Assessment (	35) a linear
extrapolation to low doses was used. Given the EC at 1% extra risk (ECoi), the unit risk is the slope of a
linear exposure-response line from the origin through the ECoi:
Unit risk = 0.01 / ECoi
A unit risk value may be calculated based on both the best estimate and the 95% upper confidence
bound (UB) on the potency factor. The value based on the upper 95% confidence bound is normally
used for decision-making, since it corresponds to a lower 5% confidence bound (LB) on the exposure
level yielding 1% extra risk (LECoi). Inhalation unit risk is derived by statistically combining risks of
lung cancer and mesothelioma. This procedure is described below in the section on combining unit risks.
Life table calculations require as input the all-cause mortality rates and cause-specific cancer incidence
rate for the general population in each year of life. The all-cause mortality data were obtained from the
Page 160 of 352

-------
National Vital Statistics Report Vol 68 No 7 Table 1 ("20171 which provides data from the U.S.
population in 2017. Lung-cancer incidence rates were obtained by downloading 2017 data for malignant
neoplasms of bronchus and lung (ICD-10 C33-C34) from CDC Wonder (http://wonder.cdc.gov/ucd-
Til). Because cause-specific rates were given for 5-year intervals, the cause-specific rate for each
5-year interval was applied to each age within the interval. For mesothelioma, the incidence rate in the
absence of asbestos exposure was assumed to be zero.
The detailed equations for calculating lifetime excess cancer risk for a specified exposure concentration
in the presence of competing risks are based on the approach used by NRC ( |) for evaluating lung
cancer risks from radon. The equations are detailed in Appendix H. The SAS code for lung cancer life
table analysis was provided to EPA by NIOSH26 and was adapted for use by a) entering the data noted
above, b) adding an equation to compute extra risk, and c) adding a macro to solve for the EC. The SAS
code for mesothelioma was created by inserting user-defined equations for the mesothelioma risk model
into the NIOSH code. The SAS codes for performing the mesothelioma and lung cancer life table
calculations are provided in Appendix I. As a quality check, life table calculations were also performed
using Microsoft Excel. Both methods yielded the same results to three or more significant figures.
3.2.3.5 Study Descriptions and Model Fitting Results
The asbestos exposure data and exposure assessment methods in studies of the Charleston, South
Carolina textile plant (Elliott et at.. 2012; Hein et at.. 2007) are exceptionally detailed compared to most
asbestos studies. The methods used were innovative at the time, a large number of exposure
measurements cover the relevant study period, and detailed process and work history information were
available and utilized in estimating exposures. The exposure data used in studies of North Carolina
plants (Loomis et at.. 2.019; Elliott et al. ) are also high quality. The methods were similar to those
developed for the studies of the South Carolina plant. However, relative to the South Carolina study, the
number of exposure measurements is smaller, and the historical process and work-history data are less
detailed. Nevertheless, the exposure data are of higher quality than those utilized in other studies of
occupational cohorts exposed to chrysotile asbestos. For both U.S. textile cohorts, the exposure
assessment methods and results have been published in full detail (Dement et at.. 1983b; Dement et al.,
2009).
Studies of the asbestos products factory in Chongqing, China (Courtice et al.. 2016; Wang et al.. 2013 a;
Deng et al.. 2012.; Yano et al.. 2001) provide informative data on a cohort that has not been included in
previous risk assessments. The methods used to estimate worker exposures for exposure-response
analyses appear to have emulated those used in the U.S. textile-industry studies. Nevertheless,
confidence in the exposure data is lower because exposure measurements were made only in later years
in the study period, the number of measurements is small, and the methodology is not reported in detail.
Information about the assessment of exposures for the Quebec asbestos mining and milling cohorts is
presented in several papers (LiddelLa_nti„ \j_>nstrong. 2002; 1998; Vacei	Liddell et al.. 1997;
1993a; 1980a; McDonald et al.. 1980b). but the reports are lacking important details and are sometimes
in conflict. Nevertheless, it is evident that exposure measurements do not cover the entire study period.
The number of measurements is not consistently reported but appears to be smaller than for either of the
U.S. textile cohorts, while the number of distinct jobs was larger. Moreover, all the reported
measurements were of total dust, rather than fibers. Some reports have suggested or used a conversion
factor, but the use of single factor for all operations is likely to introduce substantial exposure
26 Beta Version. SAS 30NOV18, provided by Randall Smith, National Institute for Occupational Safety & Health.
Page 161 of 352

-------
misclassification since the relationship between total dust and fiber counts has been shown to vary
considerably by process.
Fewer details are available about the assessment of exposures for studies of chrysotile asbestos miners in
China (2014; 2013b; Wane et at.. ). Although workshop- and job title-specific fiber concentrations
were estimated in the study in China, these estimates were based on a small number of paired samples
and important details of the exposure assessment are not available. The quality of the exposure data is
therefore difficult to judge.
Cohorts are listed in order of the quality of exposure assessment with the highest quality cohorts first.
The cohorts from SC and NC were judged to have the highest quality exposure assessment and only
those results were carried forward for consideration on the cancer-specific unit risks and the overall
IUR. For the rest of the cohorts, results of modeling are reported, but not carried forward.
3.2.3.5.1 Highest quality cohorts with results carried forward for IUR
derivation
South Carolina asbestos textile plant
Mortality in a cohort of workers at an asbestos textile plant in Charleston, South Carolina, USA has been
reported in several papers (Elliott et at.. 2012; 2008; Hein et at.. 2007; Stavner et at.. 1997; Brown et at...
1994; 19*' I. ! ^roent et at. 1983a). Workers employed for at least one month between 1940 and 1965
were included; the cohort originally included only white men but was later expanded to include non-
whites and women.
The Charleston plant produced asbestos textiles from raw chrysotile asbestos fibers imported from
Canada (Quebec and British Columbia) and Rhodesia (now Zimbabwe). Purchased crocidolite yarns
were also woven in a small separate operation for about 25 years, but crocidolite was never carded or
spun on site (Dement et at.. 1994). The total amount of crocidolite handled was 0.03% of the amount of
asbestos processed annually (Dement et at.. 1994).
Methods and results of exposure assessment for this cohort were published in detail by Dement et al.,
(1983b) and summarized in subsequent publications (e.g., Hein et at.. 2007). Engineering controls for
dust levels were introduced in the plant beginning in the 1930s and the facility was believed to represent
the best practice in the industry at the time (Dement et at.. 1983b). Estimates of individual exposure
were based on 5952 industrial hygiene air samples between 1930 and 1975. All samples before 1965
were obtained by midget impinger; both impinger and membrane filter samplers were used from 1965
until 1971, and afterward only membrane filter samplers were used. Phase-contrast microscopy (PCM)
was used in conjunction with membrane filter sampling to estimate concentrations of fibers >5|im in
length. Further details of historical fiber counting rules are not reported, but fibers <0.25 |im in diameter
cannot be visualized by PCM and are normally not counted. Paired and concurrent samples by both
methods were used to estimate job and operation-specific conversion factors from mppcf to f/cc. One
hundred and twenty paired samples were collected in 1965 and 986 concurrent samples were collected
during 1968-1971. Statistical analysis of the data indicated no significant trends in fiber/dust ratios over
time and no significant differences among operations, except for preparation. Consequently, conversion
factors of 8 PCM f/cc per mppcf for preparation and 3 PCM f/cc per mppcf for all other operations were
adopted for further analysis. Fiber concentrations were estimated for 9 departments and 4 job categories
by linear regression, accounting for time-related changes in process and dust control. Individual
cumulative exposures were estimated by linking this job-exposure-matrix to detailed occupational
histories for each worker.
Page 162 of 352

-------
The most up to date data for lung cancer and mesothelioma in the cohort were reported by Hein et al.
(2007) based on follow-up of 3072 workers through 2001; 198 deaths from lung cancer and 3 deaths
from mesothelioma were observed. Quantitative exposure-response relationships for lung cancer were
estimated by Poisson regression modeling using a linear relative rate form. Cumulative chrysotile
asbestos exposure in f/cc-yrs was lagged by 10 years and entered as a continuous variable with sex, race
and age as covariates. Elliott et al. (2012) performed a similar analysis, except some members of the
cohort were excluded to improve comparability with a cohort of textile workers from North Carolina
(see below).
Hein et al. (2.007) did not report exposure-response analysis or detailed data for mesothelioma in the
Charleston cohort. All death certificates for deaths before ICD-10 in 1999 were investigated (Hein,
personal communication) for mention of mesothelioma (3 deaths), no mesothelioma deaths after 1999
were observed. Berman & Crump (2008b) estimated Km for the cohort from analyses of the original data
obtained from the study investigators (see Table 3-3). They did not reject the hypothesis of linearity in
the variable exposure Peto model, so results of the linear model are shown.
Table 3-3. Model Fitting Results for the South Carolina Cohort
Liulpoinl
Source
la hie in
original
publication
I'olcnc)
ki o
mi.i:
I'aclor
r Km
ซ>5% I li
K\p<
C'oncen
associal
ISM K (1
Risk)
IX ,,|
mi.i:
>sure
(ration
ctl with
Zn Lxlra
(l'/cc)
i.i:c mi
5% I.B
1. i I'el i mo
(per
mi.i:
I nit Risk
l'/cc)
95% I li
Lung Cancer
Hein et al. (2007)
linear
Table 5
1.98E-2
2.80E-2
4.67E-2
3.30E-2
2.14E-1
3.03E-1
EPA modeling of
Hein et al. (2007)
grouped data
linear
NIOSH
1.76E-02
2.64E-02
5.25E-02
3.50E-02
1.90E-01
2.86E-01
Elliott et al.
(2012) linear
Table 2
2.35E-2
3.54E-2
3.93E-2
2.61E-2
2.54E-1
3.83E-1
Elliott et al.
(2012)
exponential
Table 2
5.10E-3
6.36E-3
1.67E-1
1.34E-1
5.99E-2
7.47E-2
Mesothelioma
Berman and
Crump (2008b)
using the data
from Hein et al.
(2007) variable
exposure Peto
Table 4
1.5E-9
3.3E-9
3.9E-1
1.8E-1
2.6E-2
5.7E-2
1) Data summarized by NIOSH from the data of Hein et al. (2007) as well as details for the modeling for lung cancer
are provided in Appendix J, Section 1. Details for the modeling of mesothelioma are provided in Berman and Crump
(2008b")
2)	In EPA modeling of Hein et al. (2007) grouped data, alpha= 1.35 and upper bound on the highest exposure interval
was assumed 699.8 f/cc (the maximum exposure reported in the publication).
3)	In calculations involving Elliott et al. (2012), the 95% upper bound on potency factor was calculated from the
reported 97.5% upper bound as described above.
4)	Berman and Crump (2.008b) reported mesothelioma potency number (KM) with 2 significant digits.
Page 163 of 352

-------
Selection of the results from the South Carolina cohort
As discussed above, for lung cancer, the modeling of individual data is preferred so results from Hein et
al. (200?) as well as two results of Elliott et al. ( ) were carried forward for further consideration.
For mesothelioma, only the results of modeling of the South Carolina cohort data by Berman and Crump
(2008b) are available, and those are carried forward for the unit risk derivation.
North Carolina asbestos textile plants
Loom is et al. (2019; 2009) reported on mortality in a cohort of workers in four North Carolina asbestos
textile mills that had not been studied previously. Three of the plants were operationally similar to the
South Carolina plant, but did not have equivalent exposure controls. They produced yarns and woven
goods from raw chrysotile asbestos fibers, mostly imported from Canada. A fourth, smaller plant
produced several asbestos products using only purchased yarns. The latter plant lacked adequate
exposure data and was included in comparisons of cohort mortality to the general population, but not in
exposure-response analyses for lung cancer or mesothelioma. One of the three larger plants also carded,
twisted and wove amosite fibers in a separate facility for 13 years (Loomis et al.. 2009). Quantitative
data on the amounts of amosite used are not available. However, the operation was isolated from general
production and no amosite fibers were found in TEM analysis of archived samples from that plant or
any other (Elliott et al.. 2.012.).
Workers employed at least 1 day between 1950 and 1973 were enumerated from company records: 5770
workers (3975 men and 1795 women) and files of state and national health agencies were included and
followed for vital status through 2003. Causes of death were coded to the ICD revision in force at the
time of death. All conditions mentioned on the death certificate, including intermediate causes and other
significant conditions were coded. Death certificate data were examined for any mention of
mesothelioma and for ICD codes often applied to mesothelioma before a specific code for mesothelioma
was introduced in 1999.
Exposure assessment methods and results are described by Dement et al. (2009). The approach was
similar to that used in South Carolina (Dement et al.. 1983b) with updated statistical methods. Asbestos
fiber concentrations were estimated from 3420 air samples taken from 1935 to 1986. Sampling until
1964 was by impinger; membrane filter sampling was introduced in 1964 and both methods were used
until 1971, with only membrane filter sampling thereafter. Fibers longer than 5 |im captured on
membrane filters were counted by PCM to estimate concentrations; further details of historical fiber
counting rules are not available. Paired and concurrent samples by both methods were used to estimate
plant-, operation- and period-specific factors for converting dust to PCM-equivalent fiber
concentrations. Fiber/dust ratios did not change significantly over time, so plant- and operation-specific
conversion factors (range 1.6 (95% CI 0.4-2-8) fibers/mppcf to 8.0 (95% CI 7.4-8.7) fibers/mppcf) were
used for further analysis. Fiber concentration data were analyzed using multivariable mixed models to
estimate average concentrations by plant, department, job and time period. The operation and job
categories of the job-exposure matrix were similar to those developed for South Carolina (2009; Dement
et al.. 1983a). These estimates were linked to individual work history records to estimate average and
cumulative exposure to asbestos fibers for each worker. Detailed job titles within departments were
missing for 27% of workers, mostly short-term; in these cases, exposure was estimated using the plant,
period and department average (Loomis et al.. 2009). For years prior to 1935, when no exposure
measurements and few work history records were available, exposures were assumed to have been equal
to those in 1935, before dust controls were implemented.
Page 164 of 352

-------
In total, 277 deaths from lung cancer occurred during follow-up. Exposure-response analyses for lung
cancer included 3803 workers in production jobs in 3 of the 4 study plants and 181 lung cancer deaths.
Data were analyzed using conventional log-linear Poisson regression models adjusted for age, sex, race,
decade of follow-up and birth cohort. Results were reported as relative rates per 100 f/cc-yrs with
exposure lags of 0 to 30 years (Loomis et at.. 2009).
Elliott et al. (2012) also evaluated exposure-response relationships for lung cancer in the North Carolina
cohort using Poisson regression with both log-linear and additive relative rate model forms. Models
were adjusted for age, sex, race, calendar period and birth cohort. Results were reported per 100 f/cc-yrs
of cumulative fiber exposure with lags of 0, 10 or 20 years. Results of modeling with lag of 10 years are
shown in Table 3-4.
During the follow-up of the North Carolina cohort, four deaths were coded to mesothelioma according
to the ICD-10, and, prior to the implementation of ICD-10 in 1999, four deaths coded as cancer of the
pleura and one death coded as cancer of the peritoneum were observed (2.019; Loomis et al.. 2.009).
Because Loomis et al. ( ) reported only pleural cancers before ICD-10, EPA used variable exposure
Peto model for the post-1999 subcohort reported in that publication (see Table 3-4).
Table 3-4. Model Fitting Results for the North Carolina Cohort
Endpoint
Source
Table in
Original
Publication
Potency Factor
Kl or Km
Exposure
Concentration
associated with
BMR (1% Extra
Risk) (f/cc)
Lifetime Unit Risk
(per f/cc)


MLE
95%
ECoi
LECoi
MLE
95%



UB
MLE
5% LB
UB

Elliott et al.
(2012) linear
Table 2
1.20E-3
2.71E-3
7.70E-1
3.41E-1
1.30E-2
2.93E-2

Elliott et al.








(2.012)
Table 2
9.20E-4
1.40E-3
9.25E-1
6.08E-1
1.08E-2
1.64E-2

exponential








Loomis et al.







Lung Cancer
(2009)
exponential
Table 6
1.01E-3
1.47E-3
8.43E-1
5.79E-1
1.19E-2
1.73E-2

EPA








modeling of
Loomis et al.
(2009)
Table 5
5.15E-4
1.02E-3
1.79
9.06E-1
5.57E-3
1.10E-2

grouped data
linear








Loomis et al.







Mesothelioma
(2019),
variable
Text, page
475
2.96E-9
5.87E-9
1.97E-1
9.92E-2
5.08E-2
1.0 IE-1

exposure
Peto






1)	Details for the modeling are provided in Appendix J, Section 2.
2)	In EPA modeling of the Loomis et al. (2009') lung cancer grouped data, fitted alpha= 1.18 and the upper bound
on the highest exposure interval was assumed 2,194 f/cc (the maximum exposure reported in the publication).
3) In calculations involving Loomis et al. (20091 and Elliott et al. (2012) lung cancer modeling, the 95% upper
bound on potency factor was calculated from the reported 97.5% upper bound as described above.
Page 165 of 352

-------
Selection of the results from the North Carolina Cohort
As discussed above, for lung cancer, the modeling of individual data is preferred so results from Loomis
et al. (2009) as well as two results of Elliott et al. (2012) are carried forward for further consideration.
The mesothelioma results from the Loomis et al. (2.019) sub-cohort of workers that were evaluated with
ICD-10 are carried forward for unit risk derivation.
3.2.3.5.2 Other cohorts with results not carried forward for IUR derivation
Chongqing, China, Asbestos Products Factory
An initial report on mortality among workers at a plant in Chongqing, China, that produced a variety of
asbestos products was published by Yano et al. (2001). A fixed cohort of 515 men employed at least one
year and active as of 1 January 1972 was established and followed for mortality using plant records.
Women were not included in the original cohort as none were hired before 1970. Further analyses based
on extended follow-up were reported in subsequent papers (Courtice et al.. 2016; Wang et al.. 2013a;
Dene et al..: ). The 2008 follow-up of the cohort added 279 women employed between 1970 and
1972 (Wang et al. 2013a).
The Chongqing plant opened in 1939 and expanded in the 1950s; a range of asbestos products, including
textiles, friction materials, rubber-impregnated goods and cement were produced (Yano et al.. 2001).
The plant is reported to have used chrysotile asbestos from two mines in Sichuan Province; amphibole
contamination in bulk samples from these mines assessed by transmission electron microscopy (TEM)
was found to be below the limit of detection (LOD <0.001%, Courtice et al. ( ), Yano et al. (2.001)).
An independent study of commercial chrysotile asbestos extracted from six mines in China reported
tremolite content of 0.002 to 0.312% by weight (Tossavainen et al.. 2001). but it is not clear whether
these mines supplied chrysotile asbestos to the Chongqing factory.
Deng et al. ( ) reported on the methods of exposure assessment. Fiber concentrations for four
operations (raw materials processing, textile carding and spinning, textile weaving and maintenance, and
rubber and cement production) were estimated from 556 area measurements taken every 4 years from
1970 to 2006. Only total dust was measured before 1999, while paired measurements of dust and fibers
were taken subsequently. A total of 223 measurements of fiber concentration by PCM were available.
Paired dust and fiber samples from 1999-2006 were used to estimate dust to PCM fiber-equivalent
concentrations for the 1970-1994 using an approach similar to that of Dement et al. (2009) and the
estimated and measured concentrations were combined for analysis; however, no details were reported
on what operations and jobs these estimates represent. Individual cumulative fiber exposures were
estimated from the concentration data and the duration of employment in each area of the plant. Work
histories were reported to have been stable with few job changes (Deng et al.. 2012).
Exposure-response data for lung cancer in the Chongqing cohort have been reported in several papers.
Deng et al. ( ) analyzed data for 586 men and women followed to 2006 and reported quantitative
risk estimates for cumulative chrysotile asbestos exposure obtained by fitting log-linear and additive
relative rate models with adjustment for age, smoking and calendar period. Wang et al. (2014) published
additional analyses of the same study population but truncated the follow-up period from 1981 to 2006
to make it more comparable with a study of Chinese asbestos miners (described below). The vital status
of this cohort was updated to 2008 and an analysis including follow-up from 1972 to 2008 was
published by Courtice et al. (2016). The latter papers provide quantitative risk estimates from internal
analyses with log-linear relative rate models. Papers on the Chongqing cohort provide informative
exposure-response information in units of f/cc-years from Cox or Poisson regression analyses. However,
Page 166 of 352

-------
there is potential for misclassification of exposures due to the relatively small number of exposure
measurements, the lack of fiber measurements before 1999 and use of area rather than personal sampling
(Deng et at.. 2012). Modeling results from Deng et al. ( ) are provided in Table 3-5. Result of
modeling with lag of 10 years are shown.
Table 3-5 Model Fitting Results for the Chongqing China Cohort
Knclpoinl
Source
Table in
Original
Publication
I'olcnc)
k
mi.i:
I'actor
*>5%
I li
K\p<
Conccii
associal
IIMR (1
Risk)
IX ,,|
mi.i:
>SUIT
t rat ion
oil with
Zn Kxlra
(l/cc)
i.i-:c„i
5% I.B
Li lot in
Risk (p
mi.i:
ic I nit
or l'/cc)
*>5%
I li
Lung
Cancer
IX-iiiJ cl al ( )
exponential
Tabic 3
:.USL-3
3.U2L-3
4.UWL-1
:.s:l-i
2.44L-:
3.55L-2
Dene et al. (2012)
Linear
Table 3
4.21E-3
4.56E-3
2.19E-1
2.03E-1
4.56E-2
4.94E-2
Details for the modeling are provided in Deng et al. (2012)
Data for mesothelioma were reported for follow-up through 2008 of the expanded cohort including
women (Wane et al.. 2013a). Three deaths coded as mesothelioma according to the ICD-10 (2 among
men and 1 among women) were recognized and only SMRs were reported separately for men and
women (Wane et al.. 2013a). Data on the exposure levels of the mesothelioma cases are not available,
however, so model fitting was not possible. No other analyses of mesothelioma have been reported for
the Chongqing cohort.
Quebec, Canada Asbestos Mines and Mills
Data from studies of miners, millers and asbestos products factory workers at several facilities in
Quebec, Canada are reported in multiple publications (Liddell and Armstrone. 2002; 1998; Vacek. 1998;
Liddell et al.. 1997; 1993a; 1980a; McDonald et al.. 1980b). The earliest publication, McDonald et al.
( 3b), included 11,379 miners and millers from Quebec, Canada who were born between 1891 and
1920 and had worked for at least a month in the mines and mills and were followed to 1975. Additional
findings based on follow-up of the cohort to 1988 were reported by McDonald et al. (1993 a). and further
extended to 1992 by Liddell et al. (1997). Trace amounts of tremolite have been reported in samples
from the Canadian mines (IARC. 2012b). with the amounts varying between mines (Liddell et al..
1997).
The most detailed description of exposure assessment methods used in the Quebec studies is given by
Gibbs and Lachance (1972). Additional details and updates are given in later publications (e.g., Liddell
et al. (1997); McDonald eta I l iฐ80b)). Total dust concentrations (in mppcf) were estimated using
midget impinger measurements taken from 1948 to 1966 (Gibbs and Lachanc 2). Several different
figures are reported for the total number of dust measurements used to estimate exposures: Gibbs and
Lachance (1972) reported 3096; McDonald et al. (1980b) reported "well over 4000," and McDonald et
al. (1980a) reported 10,205. Annual dust concentrations for 5783 unique jobs were assigned according a
13-point scale with categories of 0.5, 2, 7, 12, 17, 22, 27, 32, 37, 42, 47, 70 and 140 mppcf. The authors
describe the categories as "approximating to the mean," but the methods of analyzing the exposure
measurements and developing the categories are not reported. Different approaches were used to
estimate exposures in earlier and later years when dust data were judged to be inadequate; exposures in
years before 1948 were reportedly estimated by expert assessment based on interviews with workers and
Page 167 of 352

-------
company personnel, while those after 1966 were estimated by extrapolation from the previously
measured levels (Liddett et at... 1997). Cumulative dust exposure (in mppcf-years) for each worker was
estimated from the assigned dust concentrations and individual work histories; estimated exposures in
years before 1938 were multiplied by 1.65 to account for longer work weeks at that time (liddett et at..
1997). Fibers reportedly accounted for 8-15% of total dust (Gibbs and Lachance. 1972). Most exposure-
response analyses for the cohort were reported relative to cumulative dust exposure in mppcf. However,
in a case-control study of lung cancer, McDonald et al. (1980a) adopted an overall conversion factor of
3.14 f/cc per mppcf, citing 11,819 fiber measurements (methods of measurement and analysis not
described), "unfortunately with little overlap" with the dust data. In another publication, McDonald et al.
( 3b) suggested fiber concentrations per cc would be between 1 and 7 per mppcf. Liddell et al. (1984)
subsequently reported conversion factors ranging from 3.44 to 3.67 f/cc per mppcf. Gibbs (1994)
reported a 95% confidence interval of 0.58(D)0 68 to 55.7(D)0 68, where D is the dust concentration
measured by impinger, for the ratio of fibers to dust (units not specified). Gibbs and Lachance (1972).
reported that the correlation between midget impinger and membrane filter counts (0.32) was poor and
suggested that "no single conversion factor was justified." Berman (2010) performed an analysis of dust
samples from the Quebec mines and found that one third of the PCM structures samples in the dust were
not asbestos, and that about one third of structures counted by PCM were also counted by TEM. These
findings along with the uncertainties concerning what is an appropriate conversion factor raise
significant concerns about the accuracy of the f/cc estimates of exposure from the Quebec studies.
Most analyses of the Quebec cohort compared workers' mortality to the general population using SMRs
(e.g., Liddett et al. (1997; 1993a); McDonald et at. (1980b)). Liddell et al. (1998) conducted a nested
case-control study of lung cancer in a subset of workers at the mines and mills that were included in the
previous cohort studies and workers from an asbestos products factory. Subsequent publications by
Vacek et al. (1998). and Liddell and Armstrong (2002) presented more detailed analyses on a subset of
the cohort to examine the role of intensity and timing of exposure, and of potential effect modification
by cigarette smoking. All exposure-response analyses of lung cancer in the Quebec studies utilized total
dust exposure expressed in mppcf. Estimates of Kl or analogous additive relative risk measures have not
been reported for these studies.
Berman and Crump (2008b) estimated Kl for the Quebec cohort using summarized data in Liddell et al.
(1997). A single conversion factor for all operations of 3.14 f/cc per mppcf was assumed in this analysis
(and mesothelioma analysis below). Results of lung cancer modeling with lag of 10 years are presented
in Table 3-6.
Liddell et al. (1997) reported 38 cases of mesothelioma in the last follow-up through 1992. There is a
considerable uncertainty about potency (Km) estimates in this cohort. Berman and Crump (2008b)
conducted testing of linearity in the Quebec cohort (35 cases of mesothelioma were used in their
analysis) using the variable exposure Peto model and statistically rejected linearity (p < 0.00001)
resulting in sublinearity and thus estimated values of Km from the non-linear model that were one and a
half orders of magnitude higher than in linear model (0.02E-8 vs 0.72E-8). Because no confidence
interval was reported for the non-linear model, only the linear model result for the "Mines and mills at
Asbestos" (based on eight cases) is shown in the Table 3-6 because it was specific to mining and
linearity was not rejected.
Page 168 of 352

-------
Table 3-6. Model Fitting Results for the Quebec, Canada Cohort
Kndpoinl
Source
Table in
Original
publication
Polenc\
\s
Mil.
l-'aclor
.i
95%
IB
Kxp<
Coiicen
associal
liMH
Kxlra
	(17
l.( in
Mil.
>sure
(ration
ed willi
(1 %
Kisk)
1.1. Cm
5% I.I!
Li lei in
Kisk (p
Mil.
le I nil
er f/cc)
95%
11!
Lung Cancer
Berman and Crump
(2008b) modeling
of grouped data
linear
Table B1
2.9E-4
4.1E-4
3.2
2.3
3.1E-3
4.4E-3
Mesothelioma
Berman and Crump
(2008b) variable
exposure Peto
model
Table 1
1.2E-10
2.1E-10
4.9
2.8
2.1E-3
3.6E-3
1.	Details for the modeling are provided in Berman and Crump (2008b).
2.	In Berman and Crump (2008b) modeling of the lung cancer, alpha=1.15 was fitted.
3.	Berman and Crump (2008b) reported lung cancer and mesothelioma potency numbers (KL and KM) with 2
significant digits.
Qinghai, China Asbestos Mine
Wang et al. (2014;	reported findings from exposure-response analyses of a cohort of 1539
workers at a chrysotile asbestos mine in Qinghai Province, China who were on the registry January 1,
1981 and had been employed for at least one year. The cohort was followed for vital status from 1981 to
2006.
The mine opened in 1958 (no closing date reported) and produced commercial chrysotile asbestos with
no detectable tremolite asbestos content (LOD 0.1%, Wane et al. ( )). Total dust concentrations in
the mine were measured periodically between 1984 and 1995 by area sampling in fixed locations (Wane
et al.. 2012). Sampling was performed according to Chinese national standards. The number of
measurements during this period is not reported. An additional 28 measurements were taken in 2006 in 8
different workshops. Dust concentrations in mg/m3 were converted to f/cc using a linear regression
model based on 35 paired measurements taken in 1991. Fiber concentrations were estimated by
workshop and job title for the period 1984-2006, apparently using a single conversion factor. The
estimation methods are not described in detail in English-language publications, but further details may
be available in Chinese-language publications referenced by Wang et al. (2013b; 2012). but not
reviewed here. As recognized by the authors Wane et al. (2 ), there is potential for exposure
measurement error due to the conversion from mppcf to f/cc-yrs which was based on 35 paired samples
that were collected in only one year, for an unspecified number of operations.
Wang et al. (2013b) report estimates of SMRs and standardized rate ratios (SRRs) for lung cancer by
categorical levels of f/cc-yrs, stratified by smoking status. EPA used these combined data for smokers
and non-smokers to estimate a value and confidence interval for KL based on the linear relative risk
model.
Wang et al. (2014) presented rate ratios for categorical and continuous exposure variables using log-
linear Cox proportional hazards models adjusted for age and smoking. The findings from the Cox model
Page 169 of 352

-------
are useful for risk assessment in that asbestos exposure is modeled as a continuous variable using
individual level data, which generally provides a more statistically powerful examination of exposure-
response relationships than a grouped analysis. Furthermore, the Cox PH analyses by Wang et al. (2014)
adjusted for smoking, whereas the earlier SMR and SRR analyses (Wane et al.. ^ ) did not.
Modeling results with lag of 10 years are shown in Table 3-7.
No data on mesothelioma have been reported for the Qinghai mining cohort.
Table 3-7. Mode Fitting Results for the Qinghai, China Cohort
Lnilpoint
Source
Tabic in
Original
Publication
I'olcnc)
h
mil
factor
95%
11;
L\p<
(onceii
associal
IIMR (1
Risk)
IX M|
mil
>SUIT
t rat ion
oil with
Zn Lxtra
(f/cc)
LL( ,,|
5% I.B
Li lot in
Risk (p
mi.i:
ic I nit
or f/cc)
95%
I li
Lung
Cancer
EPA modeling of
Wang etal. (2013b)
grouped data linear
Tables 5
and 6
2.72E-2
3.51E-2
3.40E-2
2.63E-2
2.94E-1
3.80E-1
Wang et al. (2014)
exponential
Table 3
1.82E-3
2.63E-3
4.68E-1
3.24E-1
2.14E-2
3.09E-2
1)
2)
3)
Details for the modeling are provided in Appendix J, Section 3.
In EPA modeling of the Wang et al. (200b) grouped data, alpha was fixed and the upper bound on the highest
exposure interval was assumed 1097 f/cc (the maximum exposure reported in Wang et al. (2014) for this
cohort). The data in Tables 5 and 6 were combined in modeling.
In calculations involving Wang et al. (2014) results of lung cancer modeling, the reported hazard ratio at
exposure level of 100 f/cc-yrs was 1.2 and it was used to calculate the potency factor as follows: potency factor
= In (1.2)/100.
3.2.3.6 Lung Cancer and Mesothelioma Potencies Ranges by Industry
Historically, it has been proposed in the asbestos literature, that lung cancer and mesothelioma potencies
may differ by industry (e.g., U.S. EPA (1986). Berman and Crump (2008b) and references therein). The
estimated potencies of lung cancer (Kl) are available from both North and South Carolina cohorts and
from two other cohorts (Quebec, Canada; Qinghai, China). Regarding mesothelioma, estimated potency
estimates (Km) are available from both North Carolina and South Carolina cohorts, and Quebec, Canada
cohort. These results allow comparison of lung cancer and mesothelioma potencies by industry (textile
vs. mining); one remaining cohort included multiple industries and was not included in the comparison
(Chongqing, China). Because there are at most two cohorts in each industry category, only a rough
comparison is possible by looking at range of Kl and Km for each industry. Results are in Table 3-8
below. It is clear that the range of potencies in each cell is very wide; however, this limited data
indicates that among these cohorts exposed only to chrysotile asbestos, there is no evidence that the
potencies of lung cancer and mesothelioma are different between textile and mining industries.
Page 170 of 352

-------
Table 3-8. Comparison of Cancer Potencies (Kl and Km) by Industry
1 nd list rv
Cancer Outcome
Cancer Potenci
MM.
(Ki or Km)
95% I li
Textiles
Lung Cancer (Kl)
5.15E-4 - 2.95E-2
1.02E-3 - 3.54E-2
Mining
2.9E-4 - 2.72E-2
4.1E-4 - 3.51E-2
Textiles
Mesothelioma (Km)
1.95E-9 - 2.96E-9
3.3E-9 - 5.87E-9
Mining
1.2E-10-7.2E-9
2.1E-10 - NA
Textiles cohorts (Loo mis et aL 2009; He in et aL 2007); Mining cohorts (Quebec, Canada (95% UB for non-linear model Km
is N/A); Qinghai, China (only for lung cancer)). The cohort from Chongqing, China (lung cancer only) was not included
here, but its lung cancer potencies are intermediate and would not change the lung cancer ranges provided in the Table.
3.2.3.7 Summary of Results of North and South Carolina Cohorts
As discussed above, the cohorts from NC and SC, and the models based on individual-level data are
listed in the Table 3-9 below.
Table 3-9. Cohorts and Preferred Statistical Models for SC and NC Cohorts
Cohort
Endpoint
Source
Potency Factor
Kl or Km
Exposure Concentration
associated with BMR
(1% Extra Risk) (f/cc)
Lifetime Unit
Risk (per f/cc)



MLE
95% UB
ECoi
MLE
LECoi
5% LB
MLE
95% UB
South
Carolina
Lung Cancer
Hein et al. (2007) linear
1.98E-2
2.80E-2
4.67E-2
3.30E-2
2.14E-1
3.03E-1

Elliott et al. (2012.)
linear
2.35E-2
3.54E-2
3.93E-2
2.61E-2
2.54E-1
3.83E-1


Elliott et al. (2012)
exponential
5.10E-3
6.36E-3
3.67E-1
1.34E-1
5.99E-2
7.47E-2

Mesothelioma
Berman and Crump
(2008b) usine the data
from Hein et al. (2007)
variable exposure Peto
1.5E-9
3.3E-9
3.9E-1
1.8E-1
2.6E-2
5.7E-2
North
Carolina
Lung Cancer
Elliott et al. (2012)
linear
1.20E-3
2.71E-3
7.70E-1
3.41E-1
1.30E-2
2.93E-2


Elliott et al. (2012)
exponential
9.20E-4
1.40E-3
9.25E-1
6.08E-1
1.08E-2
1.64E-2


Loomis et al. (2009)
exponential
1.01E-3
1.47E-3
8.43E-1
5.79E-1
1.19E-2
1.73E-2

Mesothelioma
Loomis et al. (2.019)
variable exposure Peto
2.96E-9
5.87E-9
1.97E-1
9.92E-2
5.08E-2
1.0 IE-1
3.2.3.8 Derivation of Inhalation Unit Risk of Cancer Incidence: Issues to Consider
3.2.3.8.1 Biases in the Cancer Risk Values
Addressing underascertainment of mesothelioma
Unlike for lung cancer, where the relative risk model is used, the model used for mesothelioma is an
absolute risk model. For mesothelioma, the undercounting of cases (underascertainment) is a particular
Page 171 of 352

-------
concern given the limitations of the ICD classification systems used prior to 1999. In practical terms,
this means that some true occurrences of mortality due to mesothelioma are missed on death certificates
and in almost all administrative databases such as the National Death Index. Even after the introduction
of a special ICD code for mesothelioma with the introduction of ICD-10 in 1999, detection rates were
still imperfect (Camidge et al.. 2.006; Pinheiro et al. 2004). and the reported numbers of cases typically
reflect an undercount of the true number (note that the North Carolina cohort was updated in 2003, soon
after the introduction of ICD-10). The undercounts are explained by the diagnostic difficulty of
mesothelioma, both because of its rarity, variety of clinical presentations, and complexity of cytological
confirmation. For example, primary diagnosis of pleural mesothelioma is by chest exam and pleural
effusion, but the latter is absent in 10-30% of pleural mesothelioma cases (e.g., Ismail-Khan et al..
2006).
There is no single or set of morphological criteria that are entirely specific for mesothelioma (Whitaker.
2000). Peritoneal mesothelioma diagnosis is challenging, because mesothelioma and ovarian or
peritoneal serous carcinoma have a common histogenesis, and may be difficult to differentiate
morphologically (Davidson. 2011). To account for various sources of underascertainment of
mesothelioma deaths, U.S. EPA Q ), following Kopylev et al. (2011). developed a multiplier of risk
for mesothelioma deaths before and after introduction of ICD-10. Although this procedure was
developed based on the Libby Worker cohort, the problematic diagnostic issues described above are
agnostic to the fiber type exposure. The developed multiplier ,HV\.	is 1.39 with confidence
interval (0.80, 2.17). Table 3-10 shows the mesothelioma unit risks adjusted for underascertainment.
Table 3-10. Addressing Underascertainment of Mesothelioma
Cohort
Source
Mesothelioma
Mi l Unit
risk
(per f/cc)
Mesothelioma
UB unit risk
(per f/cc)
Adjusted
Mesothelioma
MLE Unit
Risk
(per f/cc)
Adjusted
Mesothelioma
UB risk
(per f/cc)
South
Carolina
Berman and Crump
(2008b) using the
data from Hein et al.
(2007) variable
exposure Peto
2.6E-2
5.7E-2
3.6E-2
7.9E-2
North
Carolina
Loom is et al. (2019)
variable exposure
Peto
5.08E-2
1.01E-1
7.06E-2
1.40E-1
Addressing inhalation cancer risks corresponding to other cancer endpoints
There is evidence that other cancer endpoints may also be associated with exposure to the commercial
forms of asbestos. IARC concluded that there was sufficient evidence in humans that commercial
asbestos (chrysotile, crocidolite, amosite, tremolite, actinolite, and anthophyllite) was causally
associated with lung cancer and mesothelioma, as well as cancer of the larynx and the ovary (Straif et
al.. 2009). EPA lacked quantitative estimates of the risks of cancers of the larynx and the ovary from
chrysotile asbestos. Failing to account for those risks in the IUR necessarily underestimates the total
cancer risk associated with chrysotile asbestos.
An adjustment factor for these other cancers has been developed by comparing the excess deaths from
lung cancer with the number of excess deaths from other cancers.
Page 172 of 352

-------
Adjustment factor = (excess lung cancer + excess other cancer)/(excess lung cancer)
This approach has been applied to estimate adjustment factors for laryngeal and ovarian cancers using
data from studies of chrysotile asbestos exposed workers that reported findings for these sites (see
Appendix M). The adjustment factor for laryngeal cancer is 1.02 and the adjustment factor for ovarian
cancer is 1.04. The combined adjustment factor for lung cancer to address other cancers is 1.06. Table 3-
11 shows the lung cancer unit risks adjusted for other cancers.
Table 3-11. Addressing Risk of Other Cancers
Cohort
Source
l.ung Cancer
Mi l. I nil
Uisk (per
I7cc)
Lung
Cancer I'U
I nil Uisk
(per l'/cc)
Adjusted
Lung
Cancer
Mi l. I nil
Uisk (per
r/cc)
Adjusted
1 .ting
Cancer I'U
I nil Uisk
(per l'/cc)
South
Carolina
Hein et al. (2007)
linear
2.14E-1
3.03E-1
2.27E-1
3.21E-1
Elliott etal. (2012)
linear
2.54E-1
3.83E-1
2.69E-1
4.06E-1
Elliott etal. (2012)
exponential
5.99E-2
7.47E-2
6.35E-2
7.92E-2
North
Carolina
Elliott et al. (2012)
linear
1.30E-2
2.93E-2
1.38E-2
3.11E-2
Elliott etal. (2012)
exponential
1.08E-2
1.64E-2
1.14E-2
1.74E-2
Loomis et al. (2009)
exponential
1.19E-2
1.73E-2
1.26E-2
1.84E-2
3.2.3.8.2 Combining Lung Cancer Unit Risk and Mesothelioma Unit
Risk
Once the cancer-specific lifetime unit risks are obtained, the two are then combined. It is important to
note that this estimate of overall risk describes the risk of cancer at either of the considered sites and is
not just the risk of an individual developing both cancers concurrently. Because each of the unit risks is
itself an upper bound estimate, summing such upper bound estimates across mesothelioma and lung
cancer is likely to overpredict the upper bound on combined risk. Therefore, following the
recommendations of the Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005). a statistically
appropriate upper bound on combined risk was derived as described below.
Because the estimated risks for mesothelioma and lung cancer were derived using maximum likelihood
estimation, it follows from statistical theory that each of these estimates of risk is approximately
normally distributed. For independent normal random variables, a standard deviation for a sum is easily
derived from individual standard deviations, which are estimated from confidence intervals: standard
deviation = (upper bound - central estimate) ^ Z0.95, where Z0.95 is a standard normal quantile equal
to 1.645. For normal random variables, the standard deviation of a sum is the square root of the sum of
the squares of individual standard deviations. It is important to mention here that assumption of
independence above is a theoretical assumption, but U.S. EPA (2014c) conducted an empirical
evaluation and found that the assumption of independence in this case does not introduce substantial
error.
Page 173 of 352

-------
In order to combine the unit risks, first obtain an estimate of the standard deviation (SD) of the sum of
the individual unit risks as:
SD = V [ [(UB LC - CE LC) - 1,645]2 + [(UB M - CE M) - 1.645 ]2]
Where,
UB - upper bound unit risk; CE - central estimate of unit risk; LC - lung cancer
M - mesothelioma
Then, the combined central estimate of risk (CCE) of either mesothelioma or lung cancer is CCE = (CE
LC + CE M) per f/cc, and the combined IUR is CCE + SD x 1.645 per f/cc.
3.2.3.9 Derivation of Inhalation Unit Risk of Cancer Incidence
To illustrate the range of estimates in the estimates of the cancer incidence IUR, central risks and upper
bounds for the combined IUR for North and South Carolina cohorts are presented in Table 3-12.
Table 3-12. Range of Estimates of Estimated Central Unit Risks and IURs for North and South
Carolina Cohorts
Lung
Cancer
Source
Central
1 nit
Uisk
Luii"
Cancer
I pper
Hound
In it
Uisk
l.ung
Cancer
.Mesothelioma
Source
Ccnlral
In it Uisk
Meso
I pper
Hound
In it
Uisk
Meso
Com hined
Central
1 nit Uisk
(l.llllg
Cancer +
Meso)
Lifetime
Cancer
11 U
(per
f/cc)
North Carolina Cohort
Elliott et al.
( ) Linear
1.38E-2
3.11E-2
Loomis et al. (2019)
variable exposure Peto
7.06E-2
1.40E-1
0.084
0.16
Elliott et al.
C )
Exponential
1.14E-2
1.74E-2
Loomis et al. (2.019)
variable exposure Peto
7.06E-2
1.40E-1
0.082
0.15
Loomis et al.
(2009")
Exponential
1.26E-2
1.84E-2
Loomis et al. (2019)
variable exposure Peto
7.06E-2
1.40E-1
0.083
0.15
South Carolina Cohort
Hein et al.
(2007) Linear
2.3E-1
3.2E-1
Berman and Crump
(2008b) using the data
from Hein et al. (2007)
variable exposure Peto
3.6E-2
7.9E-2
0.26
0.37
Elliott et al.
(: ) Linear
2.7E-1
4.1E-1
Berman and Crump
(2008b) using the data
from Hein et al. (2007)
variable exposure Peto
3.6E-2
7.9E-2
0.31
0.45
Elliott et al.
(2012)
Exponential
6.35E-2
7.92E-2
Berman and Crump
(2008b) using the data
from Hein et al. (2007)
variable exposure Peto
3.6E-2
7.9E-2
0.10
0.15
Page 174 of 352

-------
The values of the estimated IURs range from 0.15 per f/cc to 0.45 per f/cc (Table 3-12). There is a three-
fold difference between lowest and highest IUR estimates - a very low range of model uncertainty in
risk assessment. Because of low model uncertainty, EPA selected a median IUR value. Because there
are six IUR values, the median is 0.155 per f/cc, which is between values 0.15 per f/cc and 0.16 per f/cc.
Rounding to two significant digits, EPA selected 0.16 per f/cc based on modeling of North Carolina
cohort (linear model for lung cancer and variable exposure Peto model for mesothelioma) as the
chrysotile asbestos lifetime cancer incidence IUR, shown in Table 3-13.
Table 3-13. Estimates of Selected Central Risk and IUR of Cancer Incidence for Chrysotile
Asbestos
Lung Cancer
Source
Central
I nil
Uisk
Lung
Cancer
I pper
Bound
1 nil
Uisk
I-" ii g
Cancer
Mesothelioma
Source
Central
I nil Uisk
Meso
I pper
lion nd
I nil
Uisk
Meso
Coin billed
Ccnlral
Incidence
I nil Uisk
(Lung
Cancer +
Meso)
Lifetime
Cancer
Incidence
11 U
(per I'/cc)
NC Elliott et al.
(2012) Linear
1.38E-2
3.11E-2
NC Loomis et al.
(2019) variable
exposure Peto
7.06E-2
1.40E-1
0.08
0.16
The definition of the IUR is for a lifetime of exposure. For the estimation of lifetime risks for each condition of use, the
partial lifetime (or less than lifetime) IUR has been calculated using the lifetable approach and values for different
combination of age of first exposure and duration of exposures are presented in Appendix K.
Uncertainties in the cancer risk values are presented in Section 4.3.5 and 4.3.6.
3.2,4 Potentially Exposed or Susceptible Subpopulations
TSCA requires that a risk evaluation "determine whether a chemical substance presents an unreasonable
risk of injury to health or the environment, without consideration of cost or other non-risk factors,
including an unreasonable risk to a potentially exposed or susceptible subpopulation identified as
relevant to the risk evaluation by the Administrator, under the conditions of use." TSCA ง 3(12) states
that "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, pregnant women, workers, or the elderly."
During problem formulation (	2018d), EPA identified potentially exposed and susceptible
subpopulations for further analysis during the development and refinement of the life cycle, conceptual
models, exposure scenarios, and analysis plan. In this section, EPA addresses the potentially exposed or
susceptible subpopulations identified as relevant based on greater susceptibility. EPA addresses the
subpopulations identified as relevant based on greater exposure in Section 2.3.3.
Factors affecting susceptibility examined in the available studies on asbestos include lifestage, gender,
genetic polymorphisms and lifestyle factors. Additional susceptible subpopulations may include
pregnant workers and children exposed prenatally. There is some evidence of genetic predisposition for
mesothelioma related to having a germ line mutation in BAP1 (Testa et at.. 2011). Cigarette smoking in
an important risk factor for lung cancer in the general population. In addition, lifestage is important
relative to when the first exposure occurs. The long-term retention of asbestos fibers in the lung and the
long latency period for the onset of asbestos-related respiratory diseases suggest that individuals
Page 175 of 352

-------
exposed earlier in life may be at greater risk to the eventual development of respiratory problems than
those exposed later in life (ATSDR. 2001a). Appendix K of this RE illustrates this point in the IUR
values for less than lifetime COUs. For example, the IUR for a two-year old child first exposed to
chrysotile asbestos for 40 years is 1.30 E-l while the IUR for a 20-year old first exposed to asbestos for
40 years is 4.86 E-2.
Page 176 of 352

-------
4 RISK CHARACTERIZATION
4.1 Environmental Risk
EPA made refinements to the conceptual models during the PF that resulted in the elimination of the
terrestrial exposure, including biosolids, pathways. Thus, environmental hazard data sources on
terrestrial organisms were determined to be out of scope and excluded from data quality evaluation and
further consideration in the risk evaluation process.
In the PF, EPA identified the need to better determine whether there were releases to surface water and
sediments from the COUs in this risk evaluation and whether risk estimates for aquatic (including
sediment-dwelling) organisms should be included in the risk evaluation. Thus, reasonably available
environmental hazard data/information on aquatic toxicity was carried through the systematic review
process (data evaluation, data extraction and data integration).
EPA reviewed reasonably available information on environmental hazards posed by chrysotile asbestos.
A total of four on-topic and in scope environmental hazard studies were identified for chrysotile
asbestos and were determined to have acceptable data quality with overall high data quality (Appendix
E). In addition, the Systematic Review Supplemental File: Asbestos Data Quality Evaluation of
Environmental Hazard Studies presents details of the data evaluations for each study, including scores
for each metric and the overall study score. These laboratory studies indicated reproductive,
development, and sublethal effects at a concentration range of 104-108 fibers/L, which is equivalent to
0.01 to 100 MFL, to aquatic environmental receptors following chronic exposure to chrysotile asbestos.
On the exposure side of the equation, Table 2-1 presents asbestos monitoring results from the last two
six-year Office of Water sampling programs (encompassing 1998 through 2011). Results of the next six-
year review cycle is anticipated to be completed in 2023. The data show a low number of samples
(approximately 3.5% of over 14,000 samples over a 12-year period) above the reported minimum
reporting limit (MRL) of 0.2 MFL. This exposure value is within the range of hazard values reported to
have effects on aquatic organisms (0.01 to 100 MFL). EPA believes there is low or no potential for
environmental risk to aquatic or sediment-dwelling receptors from the COUs included in this Part 1 of
the risk evaluation for asbestos because water releases associated with the COUs are not expected and
were not identified.
Also, after the PF was released, EPA was still in the process of identifying potential asbestos water
releases for the COUs to determine the need to evaluate risk to aquatic and sediment-dwelling
organisms. EPA continued to search EPA databases as well as the literature and engaged in a dialogue
with industries to shed light on potential releases to water. The available information indicated that there
were surface water releases of asbestos; however, it is unclear of the source of the asbestos and the fiber
type present. In the draft Risk Evaluation, EPA concluded that, based on the reasonably available
information in the published literature, provided by industries using asbestos, and reported in EPA
databases, there were minimal or no releases of asbestos to surface water associated with the COUs that
EPA is evaluating (see Appendix D). Therefore, in the draft Risk Evaluation, EPA concluded there was
no unreasonable risk to aquatic or sediment-dwelling environmental organisms.
EPA has considered peer review and public comments on this conclusion and has decided to retain the
finding made in the draft Risk Evaluation {i.e., that there were minimal or no releases of asbestos to
surface water associated with the COUs that EPA is evaluating in this Part 1 of the risk evaluation for
Page 177 of 352

-------
asbestos). EPA is confident that the minimal water release data available and reported more fully in the
PF - and now presented again in Appendix D - cannot be attributed to chrysotile asbestos from the
COUs in this Part 1 of the risk evaluation for asbestos. Assessing possible risk to aquatic organisms
from the exposures described would not be reasonably attributed to the COUs. However, based on the
decision to develop a scope and risk evaluation for legacy uses and associated disposals of asbestos (Part
2 of the final Risk Evaluation for asbestos), EPA expects to address the issue of releases to surface water
based on those other uses.
Therefore, EPA concludes there is low or no risk to aquatic or sediment-dwelling organisms from
exposure to chrysotile asbestos. In addition, terrestrial pathways, including biosolids, were excluded
from analysis at the PF stage.
4.2 Human Health Risk
4,2,1 Risk Estimation Approach
EPA usually estimates extra cancer risks for repeated exposures to a chemical using an equation where
Risk = Human Exposure (e.g., LADC) x IUR. Then estimates of extra cancer risks would be interpreted as
the incremental probability of an individual developing cancer over a lifetime as a result of exposure to
the potential carcinogen (i.e., incremental or extra individual lifetime cancer risk).
However, as discussed in Section 3.2, this assessment is unique with respect to the impact of the timing
of exposure relative to the cancer outcome as the time since first exposure plays a dominant role in
modeling risk. The most relevant exposures for understanding mesothelioma risk were those that
occurred decades prior to the onset of cancer and subsequent cancer progression. For this reason, EPA
has used a less than lifetime exposure calculation.
The general equation for estimating cancer risks for less than lifetime exposure from inhalation of
asbestos, from the Office of Land and Emergency Management Framework for Investigating Asbestos-
contaminated SuperfundSites (U.S. EPA. 2008). is:
ELCR = EPC • TWF • IURltl
where:
ELCR = Excess Lifetime Cancer Risk, the risk of developing cancer as a
consequence of the site-related exposure
EPC = Exposure Point Concentration, the concentration of asbestos fibers in air
(f/cc) for the specific activity being assessed
IURltl = Less than lifetime Inhalation Unit Risk per f/cc
[For example: the notation for the less than lifetime IUR could start at age 16 with 40
years duration IUR(i6,40). Values for different combination of starting age and duration
can be found in Table Apx K-l in Appendix K.
TWF = Time Weighting Factor, this factor accounts for less-than-continuous
Page 178 of 352

-------
exposure during a one-year exposure27, and is given by:
TWF = [ฆ
Exposure time (hours per day)
Exposure frequency (days per year)
24 hours
365 days
The general equation above can be extended for more complex exposure scenarios by computing the
time-weighted-average exposure of multiple exposures (e.g., for 30-minute task samples within a full 8-
hour shift). Similarly, when multiple exposures may each have different risks, those may be added
together (e.g., for episodic exposures during and between DIY brake work).
There are three points to emphasize in the application of the general equation:
1.	The EPC must be expressed in the same units as the IUR for chrysotile asbestos. The units of
concentration employed in this risk evaluation are f/cc as measured by phase contrast microscopy28.
2.	The concentration-response functions on which the chrysotile asbestos IUR is based varies as a
function of time since first exposure. Consequently, estimates of cancer risk depend not only on
exposure concentration, frequency and duration, but also on age at first exposure. Therefore, it is
essential to use an IUR value that matches the exposure period of interest (specifically the age of first
exposure and the duration of exposure).
3.	When exposures of full-shift occupational workers are to be evaluated, the TWF should be adjusted to
account for differences in inhalation volumes between workers and non-workers. As noted in Appendix
G, EPA assumes workers breath 10 m3 air during an 8-hour shift and non-workers breath 20 m3 in 24
hours. The hourly ratio of those breathing volumes is the volumetric adjustment factor for workers
(V(worker)) [(10/8) / (20/24) = 1.5], Thus, for workers, the formula, ELCR = EPC • TWF • IURltl, is
extended as ELCR = EPC • TWF • V • IURltl.
If the worker began work at age 16 years and worked for 40 years, the appropriate unit risk
factor for cancer risk of chrysotile asbestos (taken from Table Apx K-l (Less Than Lifetime (or
Partial lifetime) IUR) in Appendix K) would be:
IUR(i6,40) = 0.0612 per f/cc
Based on these two factors, the excess lifetime cancer risk would be computed as:
27	See U.S. EPA (.1.994) and Part F update to RAGS inhalation guidance U.S. EPA (2009
28	PCM-equivalent (PCMe) concentrations measured using TEM could also be used.
Page 179 of 352
TWF(worker) = (8 hours / 24 hours) • (240 days / 365 days) = 0.2192, and
V (worker) 1 • 5

-------
ELCR = EPC in f/cc • 0.2192 • 1.5 • (0.0612 per f/cc)
BOX 4-1
IUR values for other combinations of age at first exposure and duration of exposure can be
found in Table Apx K-l: Less Than Lifetime (or Partial Lifetime) IUR and in 0: Sensitivity
Analysis of Exposures for DIY/Bystander Scenarios
For example:
•	First exposure at age
•	First exposure at age
•	First exposure at age
•	First exposure at age
•	First exposure at age
0 with 78 years exposure:
16 with 62 years exposure:
16 with 40 years exposure:
16 with 20 years exposure:
16 with 10 years exposure:
IUR(o,78) =0.16 per f/cc
IUR(i6,62) = 0.0641 per f/cc
IUR(iMo) = 0.0612 per f/cc
IUR(i6,20) = 0.0468 per f/cc
IUR(i6,io) = 0.0292 per f/cc
The use scenarios and populations of interest for cancer risk estimation for partial lifetime chronic
exposures are presented in Table 4-1.
EPA provided occupational exposure results representative of central tendency conditions and high-end
conditions. A central tendency was assumed to be representative of occupational exposures in the center
of the distribution for a given condition of use. EPA used the 50th percentile (median), mean (arithmetic
or geometric), mode, or midpoint values of a distribution as representative of the central tendency
scenario. EPA's preference was to provide the 50th percentile of the distribution. However, if the full
distribution was not known, EPA assumed that the mean, mode, or midpoint of the distribution
represented the central tendency depending on the statistics available for the distribution. EPA provided
high-end results at the 95th percentile. If the 95th percentile was not available, or if the full distribution
was not known and the preferred statistics were not available, EPA estimated a maximum estimate in
lieu of the high-end. Refer to Table 2-24. and Table 2-25 for occupational and consumer exposures,
respectively.
EPA received occupational monitoring data for some of the COUs (chlor-alkali and sheet gaskets) and
those data were used to estimate risks. For the other COUs, EPA used monitoring information from the
reasonably available information. Risks for both workers and ONUs were estimated when data were
reasonably available. Cancer risk was calculated for the central and high-end exposure estimates. Excess
cancer risks were expressed as number of cancer cases per 10,000 (or 1 x 10"4).
It was assumed that the exposure frequency (i.e., the amount of days per year for workers or occupational
non-users exposed to asbestos) was 240 days per year and the occupational exposure started at age 16
years with a duration of 40 years. EPA typically uses a benchmark cancer risk level of lxlO"4 for
workers/ONUs and lxlO"6 for consumers/bystanders for determining the acceptability of the cancer risk
in a population. For consumers (DIY and bystanders; see Section 4.2.3.1), the exposure frequency
assumed was 62 years, assuming exposure starting at 16 years old and continuing through their lifetime
(78 years). Exposure frequency was also based on data from the EPA Exposure Factors Handbook (U.S.
EPA. 2011) for exposure to chrysotile asbestos resulting from the COUs. As noted in Box 4-1, other
age/duration assumptions may be made.
Page 180 of 352

-------
Table 4-1. Use Scenarios and Populations of Interest for Cancer Endpoints for Assessing
	Occupational Risks Following Inhalation Exposures to Chrvsotile Asbestos	
Populations and Toxicological
Approach
Occupational I se Scenarios of Asbestos
Population of Interest and
Exposure Scenario: Workers
Adult and adolescent workers (>16 years old) exposed to chrysotile
asbestos 8-hours/day for 240 days/year for working 40 years
Population of Interest and
Exposure Scenario:
Occupational Non-Users
(ONUs)
Adults and adolescents of both sexes (>16 years old) indirectly
exposed to chrysotile asbestos while being in the same building
during product use.
Health Effects of Concern,
Concentration and Time
Duration
Cancer Health Effects: Cancer Incidence
Chrysotile Asbestos Cancer IUR (see Section 3.2.4)
•	Lifetime Inhalation Unit Risk per f/cc (from Table 3-13)
o Incidence of Cancer
o 0.16 per f/cc
•	Less than Lifetime Inhalation Unit Risk oer f/cc (IURi tm
o Uses values from life tables for different
combination of starting age of exposure and duration
(see Table APX-K-1)
Uses a Time Weighting Factor, this factor accounts for less-than-
continuous exposure during a one-year exposure
Notes:
Adult workers (>16 years old) include both healthy female and male workers.
Table 4-2. Use Scenarios and Populations of Interest for Cancer Endpoints for Assessing
Consumer Risks Following Inhalation Exposures to Chrysotile Asbestos
Populations and Toxicological
Approach
I se Scenarios of Asbestos
Population of Interest and
Exposure Scenario: Users (or
I)o-It- Yourselfers; DIY)
Consumer Users:
Adults and adolescents of both sexes (>16 years old) exposed to
chrysotile asbestos
Population of Interest and
Exposure Scenario: Bystanders
Individuals of any age indirectly exposed to chrysotile asbestos
while being in the same work area of the garage as the consumer
Health Effects of Concern,
Concentration and Time
Duration
Cancer Health Effects: Incidence of Cancer
Chrysotile Asbestos Cancer IUR (see Section 3.2.4)
•	Lifetime Inhalation Unit Risk per f/cc (from Table 3-13)
o Incidence of Cancer
o 0.16 per f/cc
•	Less than Lifetime Inhalation Unit Risk oer f/cc dURrrn
o Uses values from life tables for different
combination of starting age of exposure and duration
(see Table APX-K-1)
Uses a Time Weighting Factor, this factor accounts for less-than-
continuous exposure during a one-year exposure
29
Re-entrainment of asbestos can occur indoors in a garage. Both users and bystanders can be exposed.
29 Settled Asbestos Dust Sampling and Analysis 1st Edition Steve M. Hays, James R. Millette CRC Press 1994
Page 181 of 352

-------
Reported Respirator Use by COU
EPA evaluated inhalation exposure for workers and consumers using personal monitoring data either
from industry or journal articles. Respirators may be used when effective engineering controls are not
feasible as per OSHA's 29 CFR ง 1910.134(a). The knowledge of the range of respirator APFs is
intended to assist employers in selecting the appropriate type of respirator that could provide a level of
protection needed for a specific exposure scenario. EPA received information from industry on certain
COUs that specified the types of respirators currently being used. This information is summarized in
Table 4-3. The APF EPA applied for this risk calculation is provided in bold (based on the discussion in
Section 2.3.1.2). When no respirator usage was provided or it was deemed inadequate for the COU, EPA
provided a hypothetical APF. It is important to note that based on published evidence for asbestos (see
Section 2.3.1.2), the nominal APF may not be achieved for all respirator users.
Table 4-3. Reported Respirator Use by COU for Asbestos Occupational Exposures
Condition of
I se
Monitoring
Data?
Respirator I se Text
API lor Risk
Calculation
Chlor-alkali
Yes,
provided by
industry
(EPA-HQ-
OPPT-2016-
0736-0052,
Enclosure C)
W orkers engaged in the most hazardous activities
(e.g., those with the highest likelihood of
encountering airborne asbestos fibers) use
respiratory protection. Examples include workers
who: handle bags of asbestos; clean up spilled
material; operate glove boxes; and perform
hydroblasting of spent diaphragms. The types of
respirator used range from half-face air-purifying
respirators to supplied air respirator hoods,
depending on the nature of the work.
Half-face air-
purifying APF of 10
Supplied air
respirator hoods APF
of 25 for specific
tasks3
APF to use for the
risk calculation: 10
to 25
Sheet gasket
stamping
Yes,
provided by
industry
Workers wearN95 filtering facepiece masks. A
site-specific industrial hygiene evaluation
determined that asbestos exposures were not high
enough to require employee respirator use. (Note:
the EPA risk estimates indicate that these workers
should be wearing appropriate respirators, which
is not an N95 mask. See footnote 1).
Half mask with
N951
Hypothetical APF
to use for the risk
calculation: 10 to 25
Sheet gasket
use
(Chemical
Production)
Yes,
provided by
industry
When replacing or servicing asbestos-containing
sheet gaskets, workers in the titanium dioxide
industry wear respirators, either airline respirators
or cartridge respirators with P-100 HEP A filters.
Cartridge respirators
with P-100 HEP A
filters APF 10
Airline respirators:
APF 10
APF to use for the
risk calculation: 10
Oilfield
brake blocks
Yes, from the
literature
No information is reasonably available on
respirator use for this COU. A safety data sheet
obtained by EPA did not list respirator use (see
Section 2.3.1.6.1).
Hypothetical APF
to use for the risk
calculation: 10 to 25
Aftermarket
automotive
Yes,
provided in
An unknown amount of respirator use occurs
among these workers. OSHA's asbestos standard
Hypothetical APF
to use for the risk
Page 182 of 352

-------
brakes and
clutches
literature
requires establishments to use control methods to
ensure that exposures are below permissible
exposure limits. OSHA has also reported:
"Respiratory protection is not required during
brake and clutch jobs where the control methods
described below are used" (OSHA. 2006).
Nonetheless, some respirator use among workers
in this industry is expected.
calculation: 10 to 25
Other gasket
vehicle
friction
product
(UTV)
No2
No information is reasonably available on
respirator use for this COU, but worker activities
are expected to be similar to those for aftermarket
automotive brakes and clutches.
Hypothetical APF
to use for the risk
calculation: 10 to 25
1	OSHA Asbestos Standard 1910.1001 states that negative pressure and filtering masks should not be used for asbestos
exposure. The N95 is a negative pressure mask.
2	EPA is using worker exposure data from the sheet gasket replacement in the chemical manufacturing industry as a surrogate
for the exposures that may occur when workers service UTV friction products.
Source: OSHA (2006'). Asbestos-Automotive Brake and Clutch Repair Work: Safety and Health Information Bulletin. SHIB
07-26-06. Available online at: https://www.osha.gov/dts/shib/shib072606.htniL
1 See Table 2-7.
4.2.2 Risk Estimation for Workers: Cancer Effects Following Less than Lifetime
Inhalation Exposures by Conditions of Use
This section presents the risk estimates for workers and ONUs exposed to chrysotile asbestos for the
COUs included in this Part 1 of the risk evaluation for asbestos. EPA typically uses a benchmark cancer
risk level of 1x10" for workers/ONUs for determining the acceptability of the cancer risk in a worker
population. Risk estimates that exceed the benchmark (i.e., cancer risks greater than the cancer risk
benchmark) are shaded and in bold. Before presenting the estimates, discussion of how personal
protective equipment (PPE) is considered is warranted.
For all COUs that were quantitatively assessed (except the Super Guppy scenario), there were risks to
workers without respirators as PPE for both central and high-end exposure estimates; including those
scenarios for which short-term exposure concentrations were available to include in the analysis. When
PPE were applied (some known, some hypothetical), risks were not exceeded for some COUs (chlor-
alkali and oilfield brake blocks) but they were exceeded for others (sheet gasket stamping - central and
high-end, short-term exposure estimates; sheet gasket use - high-end exposure estimate; aftermarket
auto brakes and other vehicle friction products - high-end and high-end short-term exposure estimates;
and other gaskets [UTV] - high-end exposure estimates). Industry submissions indicated no use of
respirators (e.g., sheet gasket stampers using N95 respirators is not protective based on OSHA
regulations), or respirators with an APF of 10 or 25 (chlor-alkali) and an APF of 10 (gasket use). It is
important to note that based on published evidence for asbestos, nominal APF may not be achieved for
all respirator users (see Section 2.3.1.2).
ONUs were not assumed to use PPE, so APFs do not apply in estimated risks to ONUs. Results show
some COUs with cancer risk exceedances for ONUs for both central and high-end exposure estimates
(sheet gasket use and other gaskets [UTV]). For all other quantitatively assessed COUs (except the
Super Guppy scenario), at least one of the ONU scenarios exceeded the cancer risk benchmark. Thus,
exceedances were observed for ONUs in every quantitatively assessed COU (except the Super Guppy
scenario).
Page 183 of 352

-------
4.2.2.1 Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures
for Chlor-alkali Industry
Exposure data from the chlor-alkali industry were presented for two sampling durations (full shift and
short-term) in Table 4-4. and Table 4-5., respectively (taken from Table 2-8). Short term samples were
assumed to be approximately 30 minutes in duration. Data on exposure at central tendency (median) and
the high-end (95th percentile) are presented along with the Excess Lifetime Cancer Risk (ELCR) for
each exposure distribution.
Table 4-4. Excess Lifetime Cancer Risk for Chlor-alkali Industry Full Shift Workers and ONUs
	(Personal Samples) before consideration of PPE and any relevant APF	
Occupational
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age
16 years)
Exposure
Asbestos Worker
ONU30
Asbestos Worker
ONU
Scenario
Central
High-
Central
High-
Central
High-
Central
High-

Tendency
end
Tendency
end
Tendency
end
Tendency
end
Producing,
handling, and








disposing of
asbestos
0.0049
0.034
< 0.0025
<0.008
9.9 E-5
6.8 E-4
5.0 E-5
1.6 E-4
diaphragms: Full








shift exposure








Asbestos Workers: ELCR (Central Tendency) 0.0049 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Asbestos Workers: ELCR (High-end> = 0.034 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (Central Tendency) 0.0025 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (High-end) 0.008 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Table 4-4. presents the inhalation cancer risk estimates for chlor-alkali workers and ONUs exposed to
asbestos. The exposure values in Table 4-4. were based on monitoring data from 3 chlor-alkali
companies. For asbestos workers, the benchmark cancer risk estimate of lxlO"4 was exceeded for the
high-end exposure estimate; but not the central tendency exposure estimate. For ONUs, the cancer
benchmark was exceeded for the high-end exposure value. Estimates exceeding the benchmark are
bolded and shaded in pink.
OSHA Standard 29 CFR 1910.1001(c)(2) for asbestos describes the 30-minute excursion limit. "The
employer shall ensure that no employee is exposed to an airborne concentration of asbestos in excess of
1.0 fiber per cubic centimeter of air (1 f/cc) as averaged over a sampling period of thirty (30) minutes as
determined by the method prescribed in Appendix A to this section, or by an equivalent method." Table
2-4 reports 30-minute short-term personal exposures. As these exposures may not represent chronic
exposures, risk estimates were not calculated based on these sample values in isolation. However,
workers exposed to these short-term exposure concentrations are likely to be exposed to chrysotile
asbestos at other times during their full-shift period. As these short-term exposure concentrations exceed
the full shift exposure concentrations, averaging the 30-minute values into a full 8-hour shift would
311 Excel file "Chlor-Alkali - Summary of Area Sampling Data (7-5-2019).xlsx list 15 area samples from Olin. Eleven area
samples from one facility all have exposure concentrations of exactly 0.004 f/cc with no mention of detection limit; four area
samples from another facility have exposure concentration of exactly 0.008 f/cc and these four samples are labeled 'Detection
limit was 0.008f/cc'." For the purposes of estimating risks, the sampling values of 0.0025 f/cc are used as the measure of
central tendency of ONU exposure and the values of 0.008 f/cc at the detection limit are used to represent the high-end of
ONU exposure.
Page 184 of 352

-------
result in an increased 8-hour TWA exposure concentration with increased risks. Table 4-5 uses 30
minutes as the short-term exposure concentration averaged with 7.5 hours at the full shift exposure
concentration. The 30-minute values are provided for asbestos workers at the central tendency and at the
high-end, but risks are not calculated just for them. The revised 8-hour TWA for a full shift containing
one 30-minute exposure value per day is provided along with the risk associated with that revised full-
shift exposure value.
There are no short-term values for ONUs, presumably because the short-term sampling is specifically
limited to asbestos workers.
Table 4-5. Excess Lifetime Cancer Risk for Chlor-alkali Industry Workers (Short-Term Personal
Samples from Table 2-4, 8-hour full shift) before consideration of PPE and any relevant APF
Occupational
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at
age 16 years)
Exposure
Asbestos Worker
ONU
Asbestos Worker
ONU
Scenario
Central
High-
Central
High-
Central
High-
Central
High-

Tendency
end
Tendency
end
Tendency
end
Tendency
end
Producing,
handling, and
disposing of
asbestos
30 min







diaphragms:
Short-term
value:
0.024
0.512
N/A
N/A
—
—
—
—
exposures
(exactly 30-
minutes); and 30-
8-hr TWA:
0.0061*
0.0639**
N/A
N/A
1.2 E-4
1.3 E-3
—
—
minute short term








samples within a
full shift)*.








* This 8-hour TWA includes the 30-minute short-term exposure within an 8-hour Ml shift and is calculated as follows:
{[(0.5 hour) • (0.024 f/cc) + (7.5 hours) • (0.0049 f/cc from Table 4-2)]/8 hours}=0.0061 f/cc
** This 8-hour TWA includes the 30-minute short-term exposure within an 8-hour full shift and is calculated as follows:
{[(0.5 hour) • (0.512 f/cc) + (7.5 hours) • (0.034 f/cc from Table 4-2)]/8 hours}=0.0639 f/cc.
ELCR (Central Tendency) {[(0.5 hour) ปEPC(3ominutei + (7.5 hours) • EPC(fuiishifti] / 8 hours}. • 0.2192 • 1.5 • 0.0612.
ELCR (High-end) = {[(0.5 hour) • EPCoominutei + (7.5 hours) • EPC(fuii shift >] / 8 hours} • 0.2192 • 1.5 • 0.0612.
ELCR (Central Tendency) {[(0.5 hour) • 0.024 + (7.5 hours) • 0.0049] / 8 hours}. • 0.2192 • 1.5 • 0.0612.
ELCR (High-end) {[(0.5 hour) • 0.512 + (7.5 hours) • 0.034] / 8 hours} • 0.2192 • 1.5 '0.0612.
The results in Table 4-5 show that when a 30-minute high exposure short-term exposure concentration is
included as part of a full shift exposure estimation, the result is that workers are likely exposed at higher
concentrations than other full-shift workers who are not exposed to short-term exposures monitored for
OSHA compliance, thereby posing an even higher excess lifetime cancer risk. Note that this will be true
regardless of the frequency at which they may be exposed to those 30-minute short-term sample values
within the 8-hour TWA, as the inclusion of high 30-minute exposures will always be higher than the
standard full-shift TWA.
Applying APFs to Data from Both Full Shift Work and Short-Term Work
ELCRs for chlor-alkali workers that assumes that they will be wearing PPE with APFs of 10 and 25 for
8-hour TWAs and various combinations of 30 minutes and 7.5 hour exposures are presented in Table
4-6, Table 4-7, Table 4-8, Table 4-9 and Table 4-10.
Page 185 of 352

-------
Table 4-6. Excess Lifetime Cancer Risk for Chlor-alkali Industry Full Shift Workers and ONUs
(from Table 4-4) after consideration of PPE with APF=10 for all workers (excluding ONUs)
Occupational Exposure Scenario
Asbestos Worker
Central Tendency
High-end
Producing, handling, and disposing of asbestos diaphragms:
Full shift exposure
9.9 E-6
6.8 E-5
Table 4-7. Excess Lifetime Cancer Risk for Chlor-alkali Industry Full Shift Workers and ONUs
(from Table 4-4) after consideration of PPE with APF=25 for all workers (excluding ONUs)
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age
16 years)
Asbestos Worker
Central Tendency
High-end
Producing, handling, and disposing of asbestos diaphragms:
Full shift exposure
3.9 E-6
2.7 E-5
Table 4-6 and Table 4-7 show the risk estimates when an APF of 10 or 25 is applied to all full shift
worker exposures. In both scenarios, the risk estimates for the workers are below the benchmark of 10"4
(1 E-4). Since the assumption is that ONUs do not wear respirators, application of APFs do not apply
and so their risk estimates do not change (i.e., the benchmark cancer risk estimate of lxlO"4 was
exceeded for ONUs for high-end exposures). Table 4-3 indicated the respirators that ACC reported to
EPA are currently used by some chlor-alkali workers and both APF of 10 and 25 are used depending on
the activity being performed. It is not clear whether the workers monitored for either short-term or full
shift exposures were wearing respirators at the time of the collection of air samples.
Table 4-8. Excess Lifetime Cancer Risk for Chlor-alkali Industry Short-Term Personal Samples
(from Table 4-5) after consideration of PPE with APF=25 for short-term workers for 0.5 hours
	 (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central
Tendency
High-end
Producing, handling, and disposing
of asbestos diaphragms: Short-term
exposures (exactly 30-minutes); and
30-minute short term samples within
a full shift)
9.4 E-5
6.7 E-4
The central risks for 7.5 hours at 0.0049 f/cc with no APF were calculated and added to the 0.5 hour risk at 0.024 f/cc and
APF=25 and then the sum divided by 8 hours. The high-end risks for 7.5 hours at 0.005 f/cc were calculated and added to the
0.5 hour risk at 0.35 f/cc and APF=25 and then sum divided by 8 hours.
Central:	Risk for 7.5 hours = 0.0049 f/cc • 0.2192 • 1.5 • 0.0612
Risk for 0.5 hours = 0.024 f/cc • 0.2192 • 1.5 • 0.0612 / (APF of 25)
Risk for 8 hours = [7.5 • 1.2 E-4 + 0.5 • 2.4 E-5]/8
High-end: Risk for 7.5 hours = 0.034 f/cc • 0.2192 • 1.5 • 0.0612
Risk for 0.5 hours = 0.512 f/cc • 0.2192 • 1.5 • 0.0612 / (APF of 25)
Risk for 8 hours = [7.5 • 8.4 E-4 + 0.5 • 3.3 E-4]/8
Page 186 of 352

-------
Table 4-9. Excess Lifetime Cancer Risk for Chlor-alkali Industry Short-Term Personal Samples
(from Table 4-5) after consideration of PPE and with APF=10 for full-shift workers and with
Occupational l-.xposure Scenario
KI.CU (40 yr exposure starling at age 16 years)
Asbestos Worker
Central
Tendency
lligh-end
Producing, handling, and disposing
of asbestos diaphragms: Short-term
exposures (exactly 30-minutes); and
30-minute short term samples within
a full shift).
1.0 E-5
9.0 E-5
The central risks for 7.5 hours at 0.005 f/cc and APF=10 were calculated and added to the 0.5 risk at 0.024 f/cc and
APF=25 and then sum divided by 8 hours. The high-end risks for 7.5 hours at 0.005 f/cc and APF=10 were
calculated and added to the 0.5 risk at 0.024 f/cc and APF=25 and then sum divided by 8 hours.
Central :	Risk for 7.5 hours = 0.005 f/cc • 0.2192 • 1.5 • 0.0612 / (APF of 10)
Risk for 0.5 hours = 0.024 f/cc • 0.2192 • 1.5 • 0.0612 / (APF of 25)
Risk for 8 hours = [7.5 • 1.2 E-5 + 0.5 • 2.4 E-5]/8
High-end: Risk for 7.5 hours = 0.034 f/cc • 0.2192 • 1.5 • 0.0612 / (APF of 10)
Risk for 0.5 hours = 0.512 f/cc • 0.2192 • 1.5 • 0.0612 / (APF of 25)
Risk for 8 hours = [7.5 • 8.4 E-5 + 0.5 • 3.3 E-4]/8
Table 4-10. Excess Lifetime Cancer Risk for Chlor-alkali Industry Short-Term Personal Samples
(from Table 4-5) after consideration of PPE and with APF=25 for full-shift workers and with
	APF=25 for short-term workers (excluding ONUs)	
Occupational Kxposurc Scenario
KI.CU (40 yr exposure starting at age 16
years)
Asbestos Worker
Central
Tendency
lligh-end
Producing, handling, and disposing of asbestos
diaphragms: Short-term exposures (exactly 30-
minutes); and 30-minute short term samples within
a full shift).
4.9 E-6
5.1 E-5
Here the method is simply to divide the risks in Table 4-5 by 25:
Central	Risk from Table 4-5 = 1.5E-4/25
High	Risk from Table 4-5 = 1.3E-3/25
Table 4-8, Table 4-9, and Table 4-10 present the ELCR for short-term exposures for chlor-alkali
workers. The three scenarios represented are: (1) APF of 25 for short-term (30-minute exposure) and no
APF for 7.5 hours; (2) APF of 25 for short-term exposures and APF of 10 for the remaining 7.5 hours;
and (3) APF of 25 for both short-term and remaining 7.5 hours. The high-end risk estimates exceeded
the benchmark for workers in only the first of the three scenarios presented. None of the other
combinations of APFs exceeded the benchmark.
Page 187 of 352

-------
4.2.2.2 Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures
for Sheet Gasket Stamping
Table 4-11 presents the ELCRs for workers stamping gaskets from sheets, using exposure data from two
sampling durations (8-hour full shift; 30-minute short-term). The central tendency and high-end
exposure values are presented along with the ELCR for each exposure distribution in Table 4-11 and
Table 4-12. The exposure levels (personal samples) for full shift workers are from Table 2-10. The high-
end 8-hour TWA exposure value for workers (0.059 f/cc) is an estimate, and this full-shift exposure
level was not actually observed. This estimate assumes the highest measured short-term exposure of the
gasket stamping worker could persist for an entire day.
Table 4-11. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Full Shift Workers and
ONUs (from Table 2-10, Personal Samples) before consideration of PPE and any relevant APF
Occupational
Exposure
Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age
16 years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Sheet gasket
stamping: 8-hr
TWA exposure
0.014
0.059
0.0024
0.010
2.8 E-4
1.2 E-3
4.8 E-5
2.0 E-4
Asbestos Workers: ELCR (Central Tendency) 0.014 f/cc • 0.2192 • 1.5 • 0.0612 per f/cc
Asbestos Workers: ELCR (High-end> = 0.059 f/cc • 0.2192 • 1.5 • 0.0612 per f/cc
ONU: ELCR (Central Tendency) 0.0024 f/cc • 0.2192 • 1.5 • 0.0612 per f/cc
ONU: ELCR (High-end) 0.01 f/cc • 0.2192 • 1.5 • 0.0612 per f/cc
Table 4-11. presents the inhalation cancer risk estimates for workers stamping asbestos-containing sheet
gaskets and for ONUs exposed to asbestos. For asbestos workers, the benchmark cancer risk estimate of
lxlO"4 was exceeded for both central tendency and high-end exposure estimates. For ONUs, the cancer
benchmark was exceeded for the high-end exposure values. Estimates exceeding the benchmark are
shaded in pink and bolded.
Table 4-12 presents the inhalation cancer risk estimates for workers stamping sheet gaskets and for
ONUs exposed to asbestos, using an averaging of short-term exposures (assuming 30 minutes) and full
shift exposures (7.5 hours per day of the full shift TWA exposure) based on monitoring data. The central
tendency short-term exposure value for workers (0.024 f/cc) is the arithmetic mean of ten short-term
measurements reported in a study of one worker at a company that stamps sheet gaskets containing
asbestos. The high-end short-term exposure value for workers (0.059 f/cc) is the highest measured short-
term exposure value from the available monitoring data. This exposure value occurred during a 30-
minute sample.
Page 188 of 352

-------
Table 4-12. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Short-term Exposures within
an 8-hour Full Shift (from Table 2-10, Personal Samples) before consideration of PPE and any
relevant APF
Occupational
Exposure
Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16
years)
Asbestos Worker
ONU
Asbestos Worker
ONI
V
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Sheet gasket
stamping:
Short-term
exposures
(-30- minute;
and -30-
minute short
term samples
within a full
shift)*.
30 min
value:
0.024
8-hr TWA:
0.0146*
0.059
0.059*
0.0042
0.0025*
0.010
0.010*
2.9 E-4
1.2 E-3
4.8 E-5
2.0 E-4
* Short-term exposures are assumed to be 30 minutes in duration. For the purposes of risk estimation, short term exposures
are averaged with full shift exposure by assuming 30 minutes per day of short-term exposure with an additional 7.5 hours per
day of the full shift TWA exposure.
ELCR = {[(0.5 hour)*EPC(30mmute, + (7.5 hours)* EPC(Fuii shift,] / 8 hours} • 0.2192 • 1.5 • 0.0612.
Asbestos Worker: ELCR (Central Tendency) {[(0.5 hour)*0.024 + (7.5 hours)* 0.014] / 8 hours} • 0.2192 • 1.5 • 0.0612.
Asbestos Worker: ELCR,High-end, = {[(0.5 hour)*0.059 + (7.5 hours)* 0.059] / 8 hours} • 0.2192 • 1.5 • 0.0612.
For asbestos workers, the benchmark cancer risk estimate of lxlO"4 was exceeded for both central
tendency and high-end exposure estimates. For ONUs, the cancer benchmark was exceeded for the high-
end exposure values. Estimates exceeding the benchmark are shaded in pink and bolded.
Applying APFs to Data from Both Full Shift Work and Short-Term Work
ELCRs for workers who stamp sheet gaskets using PPE with hypothetical APFs of 10 and 25 applied for
8-hour TWAs and various combinations of 30 minutes and 7.5 hour exposures are presented in Table
4-13, Table 4-14., Table 4-15, and Table 4-16.
Table 4-13. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Full Shift Workers and
ONUs (from Table 4-11) after consideral
tion of PPE using an APF=10 (excluding ONUs)
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Sheet gasket stamping: 8-hr TWA exposure
2.8 E-5
1.2 E-4
Page 189 of 352

-------
Table 4-14. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Full Shift Workers and
ONUs (from Table 4-11) after consideration of PPE using an APF=25 (excluding ONUs)
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Sheet gasket stamping: 8-hr TWA
exposure
1.1 E-5
4.7 E-5
For full shift worker scenarios, the benchmark cancer risk estimate of lxlO"4 was exceeded for workers
with high-end exposures when a hypothetical APF of 10 was applied; all other worker scenarios were
below the benchmark (central tendency for hypothetical APFs of 10 and 25 and high-end exposures with
an APF of 25). Since the assumption is that ONUs do not wear respirators, application of APFs do not
apply and so their risk estimates do not change (i.e., the benchmark cancer risk estimate of lxlO"4 was
exceeded for ONUs for high-end exposures).
Table 4-15. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Short-term Exposures within
an 8-hour Full Shift (from Table 4-12) after consideration of PPE using an APF=10 for both full-
	shift and short-term exposures (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Sheet gasket stamping: Short-
term exposures
2.9 E-5
1.2 E-4
Table 4-16. Excess Lifetime Cancer Risk for Sheet Gasket Stamping Short-term Exposures within
an 8-hour Full Shift (from Table 4-12) after consideration of PPE using an APF=25 for both full-
	shift and short-term exposures (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Sheet gasket stamping: Short-term
exposures
1.2 E-5
4.7 E-5
Tables 4-15 and 4-16 present the ELCR for short-term exposures for sheet gasket stamping workers. The
two scenarios represented are (all hypothetical applications of an APF): (1) APF of 10 for short-term
(30-minute exposure) and an APF of 10 for 7.5 hours; and (2) APF of 25 for both short-term and
remaining 7.5 hours. The high-end risk estimates exceeded the benchmark for workers in only the first
scenario presented. None of the other combinations of hypothetical APFs exceeded the benchmark.
4.2.2.3 Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures
for Sheet Gasket Use in Chemical Production
Exposure data from sheet gasket use (replacing gaskets) - using titanium dioxide production as an
example - were presented for 8-hour full shift exposures in Table 2-11. These data are based on reports
from ACC for gasket removal/replacement at titanium dioxide facilities. The 8-hour TWA exposures
assume that the workers removed gaskets throughout the day during maintenance. Data on the exposure
at the central and high-end estimates are presented along with the ELCR for each exposure distribution
in Table 4-6. The high-end value for 8-hr TWA worker exposure (0.094) is based on the highest
Page 190 of 352

-------
exposure measurement (see Section 2.3.1.4.5). No data are available for evaluating worker short-term
exposures for this COU (see 2.3.1.4.5).
Table 4-17. Excess Lifetime Cancer Risk for Sheet Gasket Use in Chemical Production (using data
from titanium dioxide production), 8-hour TWA (from Table 2-11., Personal Samples) before
		consideration of PPE and any relevant APF	
Occupational
Exposure
Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-end
Central
Tendency
High-end
Sheet gasket use:
8-hr TWA
exposure
0.026
0.094
0.005
0.016
5.2 E-4
1.9 E-3
1.0 E-4
3.2 E-4
Asbestos Workers: ELCR (Central Tendency) 0.026 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Asbestos Workers: ELCR (High-end> = 0.094 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (Central Tendency) 0.005 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (High-end) 0.016 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Table 4-17. presents the inhalation cancer risk estimates based on data for workers replacing sheet
gaskets in titanium dioxide production and for ONUs exposed to asbestos. For asbestos workers, the
benchmark cancer risk estimate of lxlO"4 was exceeded for both central tendency and high-end exposure
estimates. For ONUs, the cancer benchmark was also exceeded for both the central tendency and the
high-end exposure values. Estimates exceeding the benchmark are shaded in pink and bolded.
Applying APFs
ELCRs for workers who repair/replace sheet gaskets and ONUs exposed to asbestos using PPE with
hypothetical APFs of 10 and 25 applied for 8-hour TWAs are presented in Table 4-18. and Table 4-19.
Based on data received from ACC, the current APF used for these activities is 10.
Table 4-18. Excess Lifetime Cancer Risk for Sheet Gasket Use in Chemical Production, 8-hour
TWA (from Table 4-6) after consideration of PPE using the APF=10 reflecting the current use of
		respirators (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Sheet gasket use: 8-hr
TWA exposure
5.2 E-5
1.9 E-4
Table 4-19. Excess Lifetime Cancer Risk for Sheet Gasket Use in Chemical Production, 8-hour
TWA (from Table 4-6) after consideration of PPE using an APF=25 (excluding ONUs)
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central
Tendency
High-end
Sheet gasket use: 8-hr TWA exposure
2.1 E-5
7.6 E-5
In both scenarios, the risk estimates for the workers are below the benchmark of lxlO"4 for the central
tendency risk estimate. The benchmark is exceeded when a hypothetical APF of 10 is used for the high-
Page 191 of 352

-------
end scenario; but not when the APF of 25 is applied to the high-end scenario. As shown in Table 4-3,
ACC reported that titanium dioxide sheet gasket workers use respirators with an APF of 10. Estimates
exceeding the benchmark are shaded in pink and bolded.
4.2.2.4 Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures
for Oilfield Brake Blocks
Qualitatively, the information available to EPA confirms that some brake blocks used in domestic
oilfields contain asbestos, as demonstrated by a safety data sheet provided by a supplier. It is reasonable
to assume that wear of the brake blocks over time will release some asbestos fibers to the air. However,
the magnitude of these releases and resulting worker exposure levels are not known. Only one study on
brake blocks was located and used to estimate exposures. In an effort to provide a risk estimate for this
activity, estimated exposures from Table 2-13 were used to represent the central tendencies of exposures
for workers and ONUs; there is no estimate for high-end exposures. More information on the limitations
of these data is provided in Section 2.3.1.5.3.
Table 4-20. Excess Lifetime Cancer Risk for Oil Field Brake Block Use, 8-hour TWA (from Table
	2-13 before consideration of PPE and any relevant APF	
Occupational Exposure
Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Brake Block use: 8-hr
TWA exposure
0.03
...
0.02
...
6.0 E-4
...
4.0 E-4
...
Asbestos Workers: ELCR (Central Tendency) 0.03 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (Central Tendency) 0.02 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Table 4-20. presents the inhalation cancer risk estimates for workers around brake block use and for
ONUs exposed to asbestos. For workers and ONUs, the benchmark cancer risk estimate of lxlO"4 was
exceeded for central tendency. No high-end exposures were available for this activity. Estimates
exceeding the benchmark are shaded in pink and bolded.
Applying APFs
ELCRs for workers who work near oil field brake blocks exposed to asbestos using PPE with
hypothetical APFs of 10 and 25 applied for 8-hour TWAs are presented in Table 4-21. and Table 4-22.
Table 4-21. Excess Lifetime Cancer Risk for Oil Field Brake Block Use, 8-hour TWA (from Table
	4-20) after consideration of PPE using an APF=10 (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Brake Block use: 8-hr TWA
exposure
6.0 E-5
—
Page 192 of 352

-------
Table 4-22. Excess Lifetime Cancer Risk for Oil Field Brake Block Use, 8-hour TWA (from Table
	4-20) after consideration of PPE using an APF=25 (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central
Tendency
High-end
Brake Block use: 8-hr TWA
exposure
2.4 E-5
—
In both scenarios, the risk estimates for the workers using either the hypothetical APF of 10 or 25 are
below the benchmark of 1 E-4.
4.2,2.5 Risk Estimation for Cancer Effects Following Chronic Inhalation Exposures
for Aftermarket Auto Brakes and Clutches
Exposure data from aftermarket auto brakes and clutches were presented for two sampling durations (8-
hour TWA and short-term) in Table 2-15. The exposure levels are based on an 8-hour TWA from Table
2-15., which are based on 7 studies found in the literature. ELCRs for short-term data from Table 2-15.
are also presented.
Table 4-23. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes
and Clutches in an Occupational Setting, 8-hour TWA Exposure (from Table 2-15.) before
		consideration of PPE and any relevant APF	
Occupational
Exposure Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16
years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Repairing or
replacing brakes with
asbestos-containing
aftermarket
automotive parts: 8-
hour TWA exposure
0.006
0.094
0.001
0.002
1.2 E-4
1.9 E-3
2.0 E-5
4.0 E-5
Asbestos Workers: ELCR (Central Tendency) 0.006 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Asbestos Workers: ELCR (High-end> = 0.094 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (Central Tendency) 0.001 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (High-end) 0.002 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Table 4-23. presents the inhalation cancer risk estimates for workers repairing and replacing auto brakes
and clutches and for ONUs exposed to asbestos. For workers, the benchmark cancer risk estimate of
lxlO"4 was exceeded for central tendency and high-end. For ONUs, the benchmark of 1 E-4 was not
exceeded. Estimates exceeding the benchmark are shaded in pink and bolded.
Table 4-24. presents the inhalation cancer risk estimates for workers repairing or replacing aftermarket
auto brakes and clutches and for ONUs exposed to asbestos, using an averaging of short-term exposures
(assuming 30 minutes per day) and full shift exposures (7.5 hours per day of the full shift TWA
exposure) based on 7 studies located in the literature. For asbestos workers, the benchmark cancer risk
estimate of lxlO"4 was exceeded for both central tendency and high-end exposure estimates. For ONUs,
the cancer benchmark was not exceeded. Estimates exceeding the benchmark are shaded in pink and
bolded.
Page 193 of 352

-------
Table 4-24. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes
and Clutches in an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from
	Table 2-15.) before consideration of PPE and any relevant APF	
Occupational
Exposure Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16
years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Repairing or
replacing brakes with
asbestos-containing
aftermarket
automotive parts:
short-term exposure
(~30- minute; and
~30-minute short
tenn samples within a
full shift)*.
30 min value: 0.006
8-hr TWA: 0.006*
0.836
0.140*
0.001
0.001*
0.002
0.002*
1.2 E-4
2.8 E-3
2.0 E-5
4.0 E-5
* Short-term exposures are assumed to be 30 minutes in duration. For the purposes of risk estimation, short term exposures
are averaged with full shift exposure by assuming 30 minutes per day of short-tenn exposure with an additional 7.5 hours per
day of the full shift TWA exposure.
ELCR = {[(0.5 hour)*EPC(30mmute, + (7.5 hours)* EPC,Fuii shift,] / 8 hours}. • 0.2192 • 1.5 • 0.0612.
Asbestos Worker: ELCR (Central Tendency) {[(0.5 hour)*EPC,30mmute, + (7.5 hours)* EPC(Fuii shift,] / 8 hours} • 0.2192 • 1.5 •
0.0612.
Asbestos Worker: ELCR (High-end, = {[(0.5 hour)*EPC(30minute, + (7.5 hours)* EPC(fuiishift,] / 8 hours} • 0.2192 • 1.5 • 0.0612.
Asbestos Worker: ELCR (Central Tendency) = {[(0.5 hour)*0.006 + (7.5 hours)* 0.006] / 8 hours} • 0.2192 • 1.5 • 0.0612.
Asbestos Worker: ELCR (High-end) = {[(0.5 hour)*0.836 + (7.5 hours)* 0.094] / 8 hours} • 0.2192 • 1.5 • 0.0612.
ONU: ELCR (Central Tendency) = {[(0.5 hour)*0.001 + (7.5 hours)* 0.001] / 8 hours} • 0.2192 • 1.5 • 0.0612.
ONU: ELCR (High-end) = {[(0.5 hour)*0.002 + (7.5 hours)* 0.002] / 8 hours} • 0.2192 • 1.5 • 0.0612.
Applying APFs to Data from Both Full Shift Work and Short-Term Work
ELCRs for workers who repair/replace auto brakes and clutches exposed to asbestos using PPE with
hypothetical APFs of 10 and 25 applied for 8-hour TWAs and various combinations of 30 minutes and
7.5 hour exposures are presented in: Table 4-25, Table 4-26, Table 4-27, and Table 4-28.
Table 4-25. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes
and Clutches in an Occupational Setting, 8-hour TWA Exposure (from Table 4-23) after
	consideration of PPE with APF=10 (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Repairing or replacing brakes
with asbestos-containing
aftermarket automotive parts: 8-
hour TWA exposure
1.2 E-5
1.9 E-4
Page 194 of 352

-------
Table 4-26. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes
and Clutches in an Occupational Setting, 8-hour TWA Exposure (from Table 4-24.) after
	consideration of PPE with APF=25 (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Repairing or replacing brakes
with asbestos-containing
aftermarket automotive parts: 8-
hour TWA exposure
4.8 E-6
7.6 E-5
For asbestos workers wearing a hypothetical respirator at APF 10, the benchmark cancer risk estimate of
lxlO"4 was exceeded for high-end exposure estimates; all other scenarios (hypothetical APF of 10 for
central tendency and hypothetical APF of 25 for both central and high-end exposures) had risk estimates
below the benchmark. Estimates exceeding the benchmark are shaded in pink and bolded.
Table 4-27. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes
and Clutches in an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from
	Table 4-24) after consideration of PPE with APF=10 (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Repairing or replacing brakes with
asbestos-containing aftermarket
automotive parts: short-term
exposure
1.2 E-5
2.8 E-4
Table 4-28. Excess Lifetime Cancer Risk for Repairing or Replacing Aftermarket Auto Brakes
and Clutches in an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from
	Table 4-24) after consideration of PPE with APF=25 (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Repairing or replacing brakes with asbestos-
containing aftermarket automotive parts:
short-term exposure
4.8 E-6
1.1 E-4
Table 4-27. and Table 4-28. display the ELCRs for short-term exposures for workers repairing or
replacing auto brakes and using hypothetical APFs of 10 and 25. For asbestos workers exposed to
asbestos, the benchmark cancer risk estimate of lxlO"4 was exceeded for high-end exposures, but not
central tendency exposures, after consideration of both hypothetical APF 10 and APF 25. Estimates
exceeding the benchmark are shaded in pink and bolded.
Page 195 of 352

-------
4.2.2,6 Risk Estimation for Cancer Effects Following Chronic Exposures for Other
Vehicle Friction Products
As discussed in Section 2.3.1.8, EPA is using the exposure estimates for aftermarket auto brakes and
clutches for the other vehicle friction products COU. Therefore, the risk estimates will mimic those for
the aftermarket auto brakes scenarios. Exposure data from aftermarket auto brakes and clutches were
presented for two sampling durations (8-hour TWA and short-term) in Table 2-15. The exposure levels
are based on an 8-hour TWA from Table 2-15., which are based on 7 studies found in the literature.
ELCRs for short-term data from Table 2-15. are also presented.
In addition, as noted in Section 2.3.1.8.2, there is a limited use of asbestos-containing brakes for a
special, large transport plane (the "Super-Guppy") by the National Aeronautics and Space
Administration (NASA).
Table 4-29. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in
an Occupational Setting, 8-hour TWA Exposure (from Table 2-15.) before consideration of PPE
		and any relevant APF	
Occupational
Exposure Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16
years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Installing brakes with
asbestos-containing
automotive parts in
exported cars: 8-hour
TWA exposure
0.006
0.094
0.001
0.002
1.2 E-4
1.9 E-3
2.0 E-5
4.0 E-5
Asbestos Workers: ELCR (Central Tendency) 0.006 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Asbestos Workers: ELCR (High-end> = 0.094 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (Central Tendency) 0.001 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (High-end) 0.002 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Table 4-29 presents the inhalation cancer risk estimates for workers repairing and replacing auto brakes
and clutches in exported cars and for ONUs exposed to asbestos. For workers, the benchmark cancer
risk estimate of lxlO"4 was exceeded for central tendency and high-end. For ONUs, the benchmark of
lxl0"4 was not exceeded. Estimates exceeding the benchmark are shaded in pink and bolded.
Page 196 of 352

-------
Table 4-30. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in
an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from Table 2-15.)
	 before consideration of PPE and any relevant APF	
Occupational
Exposure Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16
years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Repairing or
replacing brakes with
asbestos-containing
aftermarket
automotive parts in
exported cars: short-
term exposure (~30-
minute; and ~30-
minute short tenn
samples within a full
shift)*.
30 min value: 0.006
8-hr TWA: 0.006*
0.836
0.140*
0.001
0.001*
0.002
0.002*
1.2 E-4
2.8 E-4
2.0 E-5
4.0 E-5
*	Short-term exposures are assumed to be 30 minutes in duration. For the purposes of risk estimation, short term exposures
are averaged with full shift exposure by assuming 30 minutes per day of short-tenn exposure with an additional 7.5 hours per
day of the full shift TWA exposure. ELCR = {[(0.5 hour)*EPC(30mmute, + (7.5 hours)* EPC(fuii shift,] / 8 hours}. • 0.2192 • 1.5
•	0.0673.
Asbestos Worker: ELCR (Central Tendency) {[(0.5 hour)*EPC(30mmute, + (7.5 hours)* EPC(Fuii shift,] / 8 hours} • 0.2192 • 1.5 •
0.0612.
Asbestos Worker: ELCR (High-end, = {[(0.5 hour)*EPC(3ominute, + (7.5 hours)* EPC(fuii shift,] / 8 hours} • 0.2192 • 1.5 • 0.0612
Asbestos Worker: ELCR (Central Tendency) {[(0.5 hour)*0.006 + (7.5 hours)*0.006] / 8 hours} • 0.2192 • 1.5 • 0.0612
Asbestos Worker: ELCR,High-end, = {[(0.5 hour)*0.836 + (7.5 hours)*0.094 / 8 hours} • 0.2192 • 1.5 • 0.0612
ONU: ELCR (Central Tendency) = {[(0.5 hour)*0.001 + (7.5 hours)* 0.001] / 8 hours} • 0.2192 • 1.5 • 0.0612
ONU: ELCR (High-end) = {[(0.5 hour)*0.002 + (7.5 hours)* 0.002] / 8 hours} • 0.2192 • 1.5 • 0.0612
Table 4-30 presents the inhalation cancer risk estimates for workers repairing or replacing aftermarket
auto brakes and clutches in exported cars and for ONUs exposed to asbestos, using an averaging of
short-term exposures (assuming 30 minutes per day) and full shift exposures (7.5 hours per day of the
full shift TWA exposure) based on 7 studies located in the literature. For asbestos workers, the
benchmark cancer risk estimate of lxlO"4 was exceeded for both central tendency and high-end exposure
estimates. For ONUs, the benchmark of lxlO"4 was not exceeded. Estimates exceeding the benchmark
are shaded in pink and bolded.
Applying APFs to Data from Both Full Shift Work and Short-Term Work
ELCRs for workers who repair/replace auto brakes and clutches in exported cars exposed to asbestos
using PPE with hypothetical APFs of 10 and 25 applied for 8-hour TWAs and various combinations of
30 minutes and 7.5 hour exposures are presented in Table 4-31, Table 4-32, Table 4-33 and Table 4-34.
Page 197 of 352

-------
Table 4-31. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in
an Occupational Setting, 8-hour TWA Exposure (from Table 4-29) after consideration of PPE
	with APF=10 (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Installing brakes with asbestos-
containing automotive parts in
exported cars: 8-hour TWA
exposure
1.2 E-5
1.9 E-4
Table 4-32. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in
an Occupational Setting, 8-hour TWA Exposure (from Table 4-24.) after consideration of PPE
	with APF=25 (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Installing brakes with asbestos-
containing aftermarket automotive
parts in exported cars: 8-hour TWA
exposure
4.8 E-6
7.6 E-5
For asbestos workers wearing a hypothetical respirator at APF 10, the benchmark cancer risk estimate of
lxlO"4 was exceeded for high-end exposure estimates; all other scenarios (hypothetical APF of 10 for
central tendency and hypothetical APF of 25 for both central and high-end exposures) had risk estimates
below the benchmark. Since the assumption is that ONUs do not wear respirators, application of APFs
do not apply and so their risk estimates do not change. Estimates exceeding the benchmark are shaded in
pink and bolded.
Table 4-33. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in
an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from Table 4-30)
	after consideration of PPE with APF=10 (excluding ONUs)	
Occupational Exposure
Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Installing brakes with asbestos-
containing aftermarket
automotive parts in exported
cars: short-term exposure
1.2 E-5
2.8 E-4
Page 198 of 352

-------
Table 4-34. Excess Lifetime Cancer Risk for Installing Brakes and Clutches in Exported Cars in
an Occupational Setting, Short-term Exposures Within an 8-hour Full Shift (from Table 4-30)
	after consideration of PPE with APF=25 (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
Installing brakes with asbestos-
containing aftermarket automotive
parts in exported cars: short-term
exposure
4.8 E-6
1.1 E-4
Table 4-33 and Table 4-34 display the ELCRs for short-term exposures for workers repairing or
replacing auto brakes in exported cars and using hypothetical APFs of 10 and 25. For asbestos workers
exposed to asbestos, the benchmark cancer risk estimate of lxlO"4 was exceeded for high-end exposures,
but not central tendency exposures, after consideration of both hypothetical APF 10 and APF 25.
Estimates exceeding the benchmark are shaded in pink and bolded.
4.2.2.7 Risk Estimation for Cancer Effects Following Chronic Exposures for
Replacing Brakes on the NASA Large Transport Plane (i.e., Super Guppy)
Table 4-35. Excess Lifetime Cancer Risk for Replacing Brakes on the NASA Large Transport
Plane (i.e., Super Guppy) in an Occupational Setting, 8-hour TWA Exposure (from Table 2-17)
	 before consideration of PPE and any relevant APF	
Occupational
Exposure Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16
years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Central
Tendency
High-
end
Installing brakes with
asbestos-containing
aircraft parts in the
NASA Large
Transport Plane: 8-
hour TWA exposure
0.003
0.0089
N/A
N/A
3.7 E-7
1.1 E-6
N/A
N/A
TWFuSER Brakes (3.3-hours on 3.6 days every year) ~ (3.3 llOUrS / 24 llOUrS) • (3.6 dayS / 365 days) = 0.001356
User: ELCR (Central Tendency) 0.003 f/cc • 0.001356 • 1.5 • 0.0612 perf/cc
User: ELCR (High-end) 0.0089 f/cc • 0.001356 • 1.5 • 0.0612 perf/cc
Page 199 of 352

-------
Table 4-36. Excess Lifetime Cancer Risk for Replacing Brakes on the NASA Large Transport
Plane (ie., Super Guppy) in an Occupational Setting, Short-term Exposures Within an 8-hour
Occupiilioiiiil
l-lxposurc Scciiiirio
I'Aposlll'C I.C\cls I f/cc)
IH.( K (40 \ r exposure sliirlin^ ill iiปc l(>
\ciirs)
Ashcslos Worker
OM
Ashcslos Worker
OM
( cnlnil TcihIciio
llilili-
end
(cnlnil
TcihIciio
lliปh-
end
( en 1 nil
TcihIciio
llilili-
end
( en 1 ml
1 cihIciio
lliปh-
end
Repairing or
replacing brakes with
asbestos-containing
aircarft parts: short-
term exposure (-30-
minute; and -30-
minute short term
samples within a full
shift)*.
30 min value: 0.022
8-hr TWA: 0.0059*
0.045
0.014*
N/A
N/A
7.3 E-7
1.7 E-6
N/A
N/A
Central Tendency Exposure includes the 30-minute short-term exposure within each 3.3 hour brake change as follows:
{[(0.5 hour) • (0.022 f/cc) + (2.8 hours) • (0.003 f/cc)]/3.3 hours}=0.0059 f/cc
High End Exposure includes the 30-minute short-term exposure within each 3.3 hour brake change as follows:
{[(0.5 hour) • (0.045 f/cc) + (2.8 hours) • (0.0089 f/cc)]/3.3 hours}=0.014 f/cc
TWFuserBrakes = (3.3 hours / 24 hours) • (3.6 days / 365 days) = 0.001356
Worker: ELCR (Central Tendency) 0.0059 f/cc • 0.001356 • 1.5 • 0.0612 per f/cc
Worker: ELCR (High-end) 0.014 f/cc • 0.001356 • 1.5 • 0.0612 per f/cc
These risk estimates fall below the benchmark for both the central tendency and high-end. Respirator
usage is also not required by NASA because measured exposures were below applicable occupational
exposure limits (NASA. 2020a) and the work is performed in a special, ventilated walk-in booth
specifically built for this activity (see Section 2.3.1.8.2). Because the risk estimates already do not show
exceedances, there is no reason to consider or incorporate hypothetical PPE and an APF. Despite
respiratory protection not being required, NASA informed EPA that some certified technicians choose to
use half mask air-purifying respirator with P-100 particulate filters when replacing the brake pads.
4.2.2.8 Risk Estimation for Cancer Effects Following Inhalation Exposures for
Gasket Installation/Servicing in UTVs
Multiple publications (see Section 2.3.2.2) report on occupational exposures associated with installing
and servicing gaskets in automobiles. The exposure data used for this COU are presented in Table 2-23.
Data on the exposure at the central and high-end estimates are presented along with the ELCR for each
exposure distribution in Table 4-35.
Page 200 of 352

-------
Table 4-37. Excess Lifetime Cancer Risk for UTV Gasket Installation/Servicing in an Occupational
Setting, 8-hour TWA Exposure (from Table 2-23.) before consideration of PPE and any relevant
APF
Occupational Exposure
Scenario
Exposure Levels (f/cc)
ELCR (40 yr exposure starting at age 16
years)
Asbestos Worker
ONU
Asbestos Worker
ONU
Central
High-
Central
High-
Central
High-
Central
High-

Tendency
end
Tendency
end
Tendency
end
Tendency
end
UTV (based on gasket
repair/replacement in
vehicles: 8-hr TWA
0.024
0.066
0.005
0.015
4.8 E-4
1.3 E-3
1.0 E-4
3.0 E-4
exposure)








Asbestos Workers: ELCR (Central Tendency) 0.024 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
Asbestos Workers: ELCR (High-end> = 0.066 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (Central Tendency) 0.005 f/cc • 0.2192 • 1.5 • 0.0612 perf/cc
ONU: ELCR (High-end) 0.015 f/ccป 0.2192* 1.5 '0.0612 perf/cc
Table 4-35. presents the inhalation cancer risk estimates for workers installing and/or servicing gaskets
in utility vehicles and for ONUs exposed to asbestos. For both workers and ONUs, the benchmark
cancer risk estimate of lxlO"4 was exceeded for both central tendency and high-end exposures. Estimates
exceeding the benchmark are shaded in pink and bolded.
Applying APFs
ELCRs for workers who install/service gaskets in UTVs exposed to asbestos using PPE with
hypothetical APFs of 10 and 25 applied for 8-hour TWAs are presented in Table 4-36. and Table 4-37.
Table 4-38. Excess Lifetime Cancer Risk for UTV Gasket Installation/Servicing in an
Occupational Setting, 8-hour TWA Exposure (from Table 4-35) after consideration of PPE with
	APF=10 (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
UTV (based on gasket
repair/replacement in vehicles: 8-hr
TWA exposure)
4.8 E-5
1.3 E-4
Table 4-39. Excess Lifetime Cancer Risk for UTV Gasket Installation/Servicing in an
Occupational Setting, 8-hour TWA Exposure (from Table 4-35) after consideration of PPE with
	APF=25 (excluding ONUs)	
Occupational Exposure Scenario
ELCR (40 yr exposure starting at age 16 years)
Asbestos Worker
Central Tendency
High-end
UTV (based on gasket repair/replacement
in vehicles: 8-hr TWA exposure)
1.9 E-5
5.3 E-5
For asbestos workers using respirators with a hypothetical APF of 10, the benchmark cancer risk
estimate of lxlO"4 was exceeded for the high-end exposure estimate; all other scenarios (hypothetical
APF of 10 for central tendency and hypothetical APF of 25 for both central and high-end exposures) had
risk estimates below the benchmark. Estimates exceeding the benchmark are shaded in pink and bolded.
Page 201 of 352

-------
4.2.2.8. Summary of Risk Estimates for Cancer Effects for Occupational Inhalation
Exposure Scenarios for All COUs
Table 4-38 summarizes the risk estimates for inhalation exposures for all occupational exposure
scenarios for asbestos evaluated in this RE. EPA typically uses a benchmark cancer risk level of lxlO"4
for workers/ONUs for determining the acceptability of the cancer risk in a worker population. Risk
estimates that exceed the benchmark (i.e., cancer risks greater than the cancer risk benchmark) are
shaded and in bold.
Table 4-40. Summary of Risk Estimates for Inhalation Exposures to Workers and ONUs by CPU
cou
Population
Exposure Duration and
Level
Cancer Risk
Estimates (before
applying PPE)
Cancer Risk
Estimates (with
APF=10C)
Cancer Risk
Estimates (with
APF=25C)
Diaphragms for
chlor-alkali
industry
Section 4.2.2.1.
Worker
Central Tendency (8-hr)
9.9 E-5
9.9 E-6
3.9 E-6
High-end (8-hr)
6.8 E-4
6.8 E-5
2.7 E-5
Central Tendency short term
1.2 E-4
9.4 E-5a
1.0 E-5d
4.9 E-6b
High-end short term
1.3 E-3
6.7 E-4a
9.0 E-5d
5.1 E-5b
ONU
Central Tendency (8-hr)
5.0 E-5
N/A
N/A
High-end (8-hr)
1.6 E-4
N/A
N/A
Asbestos Sheets -
Gasket Stamping
Section 4.2.2.2
Worker
Central Tendency (8-hr)
2.8 E-4
2.8 E-5
1.1 E-5
High-end (8-hr)
1.2 E-3
1.2 E-4
4.7 E-5
Central Tendency short term
2.9 E-4
2.9 E-5e
1.2 E-5f
High-end short term
1.2 E-3
1.2 E-4e
4.7 E-5f
ONU
Central Tendency (8-hr)
4.8 E-5
N/A
N/A
High-end (8-hr)
2.0 E-4
N/A
N/A
Central Tendency short term
5.1 E-5
N/A
N/A
High-end short term
2.0 E-4
N/A
N/A
Asbestos Sheet
Gaskets - use
(based on repair/
replacement data
from TiO: industry)
Section 4.2.2.3
Worker
Central Tendency (8-hr)
5.2 E-4
5.2 E-5
2.1 E-5
High-end (8-hr)
1.9 E-3
1.9 E-4
7.6 E-5
ONU
Central Tendency (8-hr)
1.0 E-4
N/A
N/A
High-end (8-hr)
3.2 E-4
N/A
N/A
Oil Field Brake
Blocks
Section 4.2.2.4
Worker
Central Tendency (8-hr)
6.0 E-4
6.0 E-5
2.4 E-5
ONU
Central Tendency (8-hr)
4.0 E-4
N/A
N/A
Aftennarket Auto
Brakes
Section 4.2.2.5
Worker
Central Tendency (8-hr)
1.2 E-4
1.2 E-5
4.8 E-6
High-end (8-hr)
1.9 E-3
1.9 E-4
7.6 E-5
Central Tendency short-term
1.2 E-4
1.2 E-5e
4.8 E-6f
High-end short-term
2.8 E-3
2.8 E-4e
1.1 E-4f
ONU
Central Tendency (8-hr)
2.0 E-5
N/A
N/A
Page 202 of 352

-------


High-end (8-hr)
4.0 E-5
N/A
N/A
Central Tendency short-term
2.0 E-5
N/A
N/A
High-end short-term
4.0 E-5
N/A
N/A
Other Vehicle
Friction Products
Section 4.2.2.6
Worker
Central Tendency (8-hr)
1.2 E-4
1.2 E-5
4.8 E-6
High-end (8-hr)
1.9 E-3
1.9 E-4
7.6 E-5
Central Tendency short term
1.2 E-4
1.2 E-5e
4.8 E-6f
High-end w short term
2.8 E-3
2.8 E-4e
1.1 E-4f
ONU
Central Tendency (8-hr)
2.0 E-5
N/A
N/A
High-end (8-hr)
4.0 E-5
N/A
N/A
Central Tendency short-term
2.0 E-5
N/A
N/A
High-end short-term
4.0 E-5
N/A
N/A
Other Vehicle
Friction Products:
Super Guppy
Section 4.2.2.6
Worker
Central Tendency (8-hr)
3.7 E-7
N/A
N/A
High-end (8-hr)
1.1 E-6
N/A
N/A
Central Tendency (short-term)
7.3 E-7
N/A
N/A
High-end (short-term)
1.7 E-7
N/A
N/A
Other Gaskets -
Utility Vehicles
Section 4.2.2.7
Worker
Central Tendency (8-hr)
4.8 E-4
4.8 E-5
1.9 E-5
High-end (8-hr)
1.3 E-3
1.3 E-4
5.3 E-5
ONU
Central Tendency (8-hr)
1.0 E-4
N/A
N/A
High-end (8-hr)
3.0 E-4
N/A
N/A
N/A: Not Assessed; ONUs are not assumed to wear respirators
aNo APF applied for 7.5 hours, APF of 25 applied for 30 minutes.
bAPF 25 applied for both 30 mins and 7.5 hours
0 As shown in Table 4-3, EPA has information suggesting use of respirators for two COUs (chlor-alkali: APF of 10 or 25; and
sheet gasket use: APF of 10 only). Application of all other APFs is hypothetical.
d APF 25 for 30 minutes, APF 10 for 7.5 hours
e APF 10 for 30 minutes, APF 10 for 7.5 hours
f APF 25 for 30 minutes, APF 25 for 7.5 hours
For workers, with the exceptions of the central tendancy, full shift chlor-alkali worker and all scenarios
assessed for brake pad replacements for the NASA Super Guppy, cancer risks were indicated for all
quantitatively assessed conditions of use under high-end and central tendency exposure scenarios when
PPE was not used. With the use of PPE at APF of 10, most risks were reduced but still persisted for
chlor-alkali (for high-end estimates when short-term exposures were considered), sheet gasket stamping
(high-end only), sheet gasket use (high-end only), auto brake replacement (high-end only for 8-hour and
central and high-end estimates when short-term exposures are considered), and UTV gasket replacement
(high-end only). When an APF of 25 was applied, risk was still indicated for the auto brakes and other
vehicle friction products high-end short-term exposure scenarios.
For ONUs, the benchmark for risk is exceeded for most high-end estimates and most central tendency
estimates. The exceptions for central tendency exceedances are for the following COUs: chlor-alkali (8-
hour), sheet gasket stamping (8-hour), and auto brake replacement (8-hour and short-term exposure
scenarios). The exceptions for high-end exceedances are for the aftermarket auto brakes and other
vehicle friction products scenarios.
Page 203 of 352

-------
4,2,3 Risk Estimation for Consumers: Cancer Effects by Conditions of Use
4.2.3.1 Risk Estimation for Cancer Effects Following Episodic Inhalation Exposures
for DIY Brake Repair/Replacement
EPA assessed chronic chrysotile exposures for the DIY (consumer) and bystander brake repair/
replacement scenario based on repeated exposures resulting from recurring episodic exposures from
active use of chrysotile asbestos related to DIY brake-related activities. These activities include
concomitant exposure to chrysotile asbestos fibers which are reasonably anticipated to remain within
indoor and outdoor use facilities. It is well-understood that asbestos fibers in air will settle out in dust
and become re-entrained in air during any changes in air currents or activity within the indoor and
outdoor use facilities. On the other hand, in occupational settings, regular air sampling would capture
both new and old fibers and have industrial hygiene practices in place to reduce exposures.
EPA used the following data on exposure frequency and duration, making assumptions when needed:
•	Exposure frequency of active use of chrysotile asbestos related to DIY brake repair and
replacement of 3 hours on 1 day every 3 years or 0.33 days per year. This is based on the
information that brakes are replaced every 35,000 miles, and an average number of miles driven
per year per driver in the U.S. of 13,476 miles/year (U.S. DOT. 2018).
•	An estimate assuming a single brake change at age 16 years old is presented.
•	Estimates for exposure duration of 62 years and assuming exposure for a DIY mechanic starting
at 16 years old and continuing through their lifetime (78 years) is presented. EPA also did a
sensitivity analyses with different ages at first exposure and different exposure durations (see
Appendix L and the uncertainties Section 4.3.7).
•	Exposure frequency of concomitant exposure to chrysotile asbestos resulting from COUs was
based on data in the EPA Exposure Factor Handbook Q v \ 201 I). 'Doers' are the
respondents who engage or participated in the activity.31 According to Table 16-16 of the
Handbook, the median time 'Doers' spent in garages is approximately one hour per day. The 95th
percentile of time 'Doers' spent in garages is approximately 8 hours. According to Table 16-57
of the Handbook, the median time spent near outdoor locations is 5 minutes, and the 95th
percentile of time is 30 minutes.
•	Over the interval of time between the recurring episodic exposures of active COUs, the fraction
of the exposure concentrations from active use of chrysotile asbestos is unknown, however some
dispersion of fibers can reasonably be expected to occur over time. For example, if 50% of fibers
were removed from garages each year, the concentration at the end of the first year would be
50%, at the end of the second year would be 25%, and at the end of the third year would be 13%.
In this example, the mean exposure over the 3-year interval would be approximately 30% of the
active COUs. In order to estimate the chrysotile asbestos concentration over the interval of time
between the recurring episodic exposures of active COUs in the garages, EPA simply assumed
approximate concentrations of 30% of the active COUs over the 3-year interval. In order to
estimate the chrysotile asbestos concentration over the interval of time between the recurring
episodic exposures of active COUs in outdoor driveways, EPA simply assumed approximate
concentrations of 2% of the active COUs over the 3-year interval based on 95% reduction of
fibers each year.
31
This Part 1 of the risk evaluation for asbestos uses the term "consumer" or Do-It-Yourselfer (DIY) or DIY mechanic to
refer to the "doer" referenced in the Exposure Factor Handbook.
Page 204 of 352

-------
• Exposure frequency of bystander exposures are similar to those of active user (i.e., Doers) and
may occur at any age and exposure durations are assumed to continue for a lifetime; with an
upper-bound estimate of 78 years of exposure (i.e., ages 0-78) No reduction factor was applied
for indoor DIY brake work inside residential garages. A reduction factor of 10 was applied for
outdoor DIY brake work32. A sensitivity analysis is presented in Appendix L which includes a
lower-bound estimate for a bystander of 20 years (ages 0-20) (see the uncertainties Section
4.3.7).
Excess lifetime cancer risk for people engaging in DIY brake repair (consumers) and
replacement
ELCRdiy Brakes	EPCdiy Brakes • TWFdIY Brakes * IURlTL(DIY Brakes) +
EPCconcomitant Exposures * TWFConcomitant Exposures * IURLTL(Concomitant Exposures)
TWFdiy Brakes (3-hours on 1 day every 3 years) (3/24)*(l/3)*(l/365) = 0.0001142
IURltl(diy Brakes) = IUR(16,62) = 0.0641 per f/cc
TWFconcomitant Exposures (1-hour per day every day) — (l/24)*(365/365) = 0.04167
IURlTL(Concomitant Exposures) — IUR(16,62) — 0.0641 per f/cC
Excess lifetime cancer risk for bystanders to DIY brake repair and replacement
ELCR-Bystander =	EPCBystander to DIY brake work * TWFBystander to DIY brake work * IUR-Lifetime +
EPCBystander to Concomitant Exposures * TWFBystander to Concomitant Exposures * lUR-Lifetime
TWFBystander to DIY brakes work (3-hours on 1 day every 3 years) = (3/24)*(l/3)*(l/365) = 0.0001142
IURLifetime =0.16 per f/cc
TWFBystander to Concomitant Exposures (1-hour per day every day) = (l/24)*(365/365) = 0.04167
Exposure values from Table 2-31 were used to represent indoor brake work (with compressed air) and
are the basis for the exposure levels used in Tables 4-39 through 4-42, EPA then assumed that the
concentration of chrysotile asbestos in the interval between brake work (every 3 years) is 30% of that
during measured active use.
Consumers and bystanders were assumed to spend one hour per day in their garages based on the 50th
percentile estimate in the EPA Exposure Factors Handbook. Based on these assumptions, the consumer
risk estimate was exceeded for central and high-end exposures based on replacing brakes every 3 years
(Table 4-39). Estimates exceeding the benchmark are shaded in pink and bolded.
32 As explained in Section 2.3.1.2, EPA evaluated consumer bystander exposure for the DIY brake outdoor scenario by
applying a reduction factor of 10 to the PBZ value measured outdoors for the consumer user. The reduction factor of 10 was
chosen based on a comparison between the PBZ and the < 3meter from automobile values measured indoors across all
activities identified in the study data utilized from Blake (a ratio of 6.5). The ratio of 6.5 was rounded up to 10, to account for
an additional reduction in concentration to which a bystander may be exposed in the outdoor space based on the high air
exchange rates and volume in the outdoors.
Page 205 of 352

-------
Tables 4-40 and 4-41 used the alternative assumptions for age at first exposure (16 years old) and
exposure duration (40 years) for the DIY user; and the assumptions for the exposure duration of the
bystander (lifetime). Table 4-41 presents another alternative estimate for both the DIY user (performing
work from ages 16-36, and a bystander being present from ages 0-20) for the one-hour/day scenario (
Table 4-40). The risk estimates note that the benchmark is exceeded for both these alternative estimates.
Table 4-41. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with
Compressed Air Use for Consumers and Bystanders (exposures from Table 2-31 without a
reduction factor) with Exposures at 30% of 3-hour User Concentrations between Brake/Repair
Consumer
Exposure Scenario
Exposure Levels (f/cc)
ELCR (62 yr exposure
starting at age 16
years)
ELCR (Lifetime
exposure)
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Aftennarket
automotive parts -
brakes (3-hour
TWA indoors every
3 years with
compressed air)
0.0445
0.4368
0.0130
0.0296
3.6 E-5
3.5 E-4
2.6 E-5
6.0 E-5
T WF concomitant Exposures (1 hour per day every day) (l/24)*(365/365) = 0.04167
IUR( 16,621=0.0641; IUR(Lifetime 1=0.16
DIY User: ELCR (Central Tendency) 0.0445 f/cc • 0.0001142 • 0.0641 per f/cc + 0.0445 • 0.3 • 0.04167 • 0.0641
DIY User: ELCR,High-endi = 0.4368 f/cc • 0.0001142 • 0.0641 per f/cc + 0.4368 • 0.3 • 0.04167 • 0.0641
DIY Bystander: ELCR (Central Tendency) 0.013 f/cc • 0.0001142 • 0.16 per f/cc + 0.013 • 0.3 • 0.04167 • 0.16
DIY Bystander: ELCR,High-endi = 0.0296 f/cc • 0.0001142 • 0.16 per f/cc + 0.0296 • 0.3 • 0.04167 • 0.16
Table 4-42. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with
Compressed Air Use for Consumers for 20-year duration (exposures from Table 2-31 without a
reduction factor) with Exposures at 30% of 3-hour User Concentrations between Brake/Repair
Consumer
Exposure Scenario
Exposure Levels (f/cc)
ELCR (20 yr exposure
starting at age 16
years)
ELCR ((20 yr
exposure starting at
age 0 years))
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Aftennarket
automotive parts -
brakes (3-hour
TWA indoors every
3 years with
compressed air)
0.0445
0.4368
0.0130
0.0296
2.6 E-5
2.6 E-4
1.7 E-5
3.9 E-5
TWFConcomitant Exposures (1 hour per day every day) (l/24)*(365/365) = 0.04167
IURi6,20)=0.0468; IURo,2o1=OX057 '
DIY User: ELCR (Central Tendency) 0.0445 f/cc • 0.0001142 • 0.0468 per f/cc + 0.0445 • 0.3 • 0.04167 • 0.0468
DIY User: ELCR,High-end> = 0.4368 f/cc • 0.0001142 • 0.0468 per f/cc + 0.4368 • 0.3 • 0.04167 • 0.0468
DIY Bystander: ELCR (Central Tendency) 0.013 f/cc • 0.0001142 • 0.1057 per f/cc + 0.013 • 0.3 • 0.04167 • 0.1057
DIY Bystander: ELCR,High-endi = 0.0296 f/cc • 0.0001142 • 0.1057 per f/cc + 0.0296 • 0.3 • 0.04167 • 0.1057
For Table 4-41, users were assumed to spend eight hours per day in their garages based on the 95th
percentile estimate in the EPA Exposure Factors Handbook (Table 16-16 in the Handbook). Bystanders
Page 206 of 352

-------
were assumed to spend one hour per day in their garages. Based on these assumptions, both the
consumer and the bystander risk estimates were exceeded for central tendency and high-end exposures
during use of compressed air. Estimates exceeding the benchmark are shaded in pink and bolded.
Table 4-43. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with
Compressed Air Use for Consumers and Bystanders (exposures from Table 2-31 without a
reduction factor) with Exposures at 30% of 3-hour User Concentrations between Brake/Repair
Consumer
Exposure
Scenario
Exposure Levels (f/cc)
ELCR (62 yr exposure
starting at age 16 years)
ELCR (Lifetime
exposure)
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-
end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-
end
Aftennarket
automotive parts -
brakes (3-hour
TWA indoors with
compressed air)
0.0445
0.4368
0.0130
0.0296
2.9 E-4
2.8 E-3
2.6 E-5
6.0 E-5
TWF Concomitant Exposures (8-hours per day every day) (8/24)*(365/365) = 0.3333
IUR( 16,621=0.0641; IUR(Lifetime 1=0.16
DIY User: ELCR (Central Tendency) 0.0445 f/cc • 0.0001142 • 0.0641 per f/cc + 0.0445 • 0.3 • 0.3333 • 0.0641
DIY User: ELCR,High-endi = 0.4368 f/cc • 0.0001142 • 0.0641 per f/cc + 0.4368 • 0.3 • 0.3333 • 0.0641
DIY Bystander: ELCR (Central Tendency) 0.013 f/cc • 0.0001142 • 0.16 per f/cc + 0.013 • 0.3 • 0.04167 • 0.16
DIY Bystander: ELCR,High-endi = 0.0296 f/cc • 0.0001142 • 0.16 per f/cc + 0.0296 • 0.3 • 0.04167 • 0.16
In Table 4-42 the assumption is that DIY brake/repair replacement with compressed air is limited to a
single brake change at age 16 years. EPA then assumed that the concentration of chrysotile asbestos
following this COU decreases 50% each year as was assumed in all the indoor exposure scenarios. EPA
then assumed that both the DIYer and the bystander would remain in the house for 10 years. Risks were
determined for the 10-year period by calculating the risk with the appropriate partial lifetime IUR and
re-entrainment exposure over 10 years, averaging 10% of the brake/repair concentrations each year
(total 10-year cumulative exposure is 50% in first year plus 25% in second year is for all practical
purposes equal to a limit of one year at the 3-hour concentration divided by 10 years).
Table 4-44. Risk Estimate using one brake change at age 16 years with 10 years further exposure.
Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement with Compressed Air
Use for Consumers and Bystanders (exposures from Table 2-31 without a reduction factor)
(Consumers 1 hour/day spent in garage; Bystanders 1 hour/day
Consumer
Exposure Scenario
Exposure Levels (f/cc)
ELCR (62 yr exposure
starting at age 16
years)
ELCR (Lifetime
exposure)
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Aftennarket
automotive parts -
brakes (3-hour
TWA indoors once
at 16 yrs old; with
compressed air)
0.0445
0.4368
0.0130
0.0296
5.4 E-6
5.3 E-5
3.4 E-6
7.8 E-6
TWF Concomitant Exposures (1 hour per day every day) (l/24)*(365/365) = 0.04167
IURi6,io)=0.0292; IUR(o,ior0.0634
DIY User: ELCR (Central Tendency) 0.0445 f/cc • 0.000005524 • 0.0292 per f/cc + 0.0445 • 0.1 • 0.04167 • 0.0292
Page 207 of 352

-------
DIY User: ELCR (High-end) = 0.4368 f/cc • 0.000005524 • 0.0292 per f/cc + 0.4368 • 0.1 • 0.04167 • 0.0292
DIY Bystander: ELCR (Central Tendency) 0.013 f/cc • 0.000005524 • 0.0634 per f/cc + 0.013 • 0.1 • 0.04167 • 0.0634
DIY Bystander: ELCR (High-end) = 0.0296 f/cc • 0.000005524 • 0.0634 per f/cc + 0.0296 • 0.1 • 0.04167 • 0.0634
Exposure Levels in Table 4-43 are from Table 2-31 and the assumption is used that the concentration of
chrysotile asbestos in the interval between brake works is 2% of that during measured active use. Users
and bystanders were assumed to spend 5 minutes per day in the driveway each day based on the 50th
percentile estimate in the EPA Exposure Factors Handbook (in Table 16-57 in the Handbook). The
reduction factor is 10 for bystanders33. Neither of the risk estimates for consumers or bystanders in
Table 4-43 exceeded the risk benchmark for either the central tendency or high-end estimates.
Table 4-45. Excess Lifetime Cancer Risk for Outdoor DIY Brake/repair Replacement for
Consumers and Bystanders (5 minutes per day in driveway) (from Table 2-31 with a reduction
		factor of 10) 		
CoilMlllHT
r.xposurc Scenario
l'l\|)OMIIV I.C'\C'ls (I'/ccI
l-'.I.CK ((>2 j r exposure
sdirlinii ill iiซe l(>
\ oil rs)
I I.( K (lifetime
exposure)
DIYIsit
DIY IS^sliindor
DIY I sit
DIY KtsliindiT
( en I ml
Teink'iio
lli^h-cnd
(cilllill
Tendeno
lli^h-cnd
( Cllll'ill
Tendeno
lliiili-cud
( on I nil
Tendeno
Ili'Ji-ond
Aftermarket
automotive parts -
brakes (3-hour TWA
outdoors)
0.007
0.0376
0.0007
0.0038
8.2 E-8
4.4 E-7
2.1 E-8
1.1 E-7
TWFConcomitant Exposures (0.0833 hours per day every day) (0.08333/24)*(365/365) = 0.003472
IUR(16,62)=0.0641; IUR(Lifetime)=0.16
DIY User: ELCR (Central Tendency) 0.007 f/cc • 0.0001142 • 0.0641 per f/cc + 0.007 • 0.02 • 0.003472 • 0.0641
DIY User: ELCR (High-end) = 0.0376 f/cc • 0.0001142 • 0.0641 per f/cc + 0.0376 • 0.02 • 0.003472 • 0.0641
DIY Bystander: ELCR (Central Tendency) 0.0007f/cc • 0.0001142 • 0.16 per f/cc + 0.0007 • 0.02 • 0.003472 • 0.16
DIY Bystander: ELCR (High-end) = 0.0038 f/cc • 0.0001142 • 0.16 per f/cc + 0.0038 • 0.02 • 0.003472 • 0.16
33 As explained in Section 2.3.1.2, EPA evaluated consumer bystander exposure for the DIY brake outdoor scenario by
applying a reduction factor of 10 to the PBZ value measured outdoors for the consumer user. The reduction factor of 10 was
chosen based on a comparison between the PBZ and the < 3 meter from automobile values measured indoors across all
activities identified in the study data utilized from Blake (a ratio of 6.5). The ratio of 6.5 was rounded up to 10, to account for
an additional reduction in concentration to which a bystander may be exposed in the outdoor space based on the high air
exchange rates and volume in the outdoors.
Page 208 of 352

-------
Table 4-46. Excess Lifetime Cancer Risk for Outdoor DIY Brake/Repair Replacement for
Consumers and Bystanders (30 minutes per day in driveway) (from Table 2-31 with a reduction
		factor of 10) 		
Occupational
Exposure Scenario
Exposure Levels (f/cc)
ELCR (62 yr exposure
starting at age 16
years)
ELCR (Lifetime
exposure)
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Aftennarket
automotive parts -
brakes (3-hour TWA
outdoors)
0.007
0.0376
0.0007
0.0038
2.4 E-7
1.3 E-6
5.9 E-8
3.2 E-7
TWF Concomitant Exposures (0.5 hours per day every day) (0.5/24)*(365/365) = 0.02083
IUR( 16,621=0.0641; IUR(Lifetime 1=0.16
DIY User: ELCR (Central Tendency) 0.007 f/cc • 0.0001142 • 0.0641 perf/cc + 0.007 • 0.02 • 0.02083 • 0.0641
DIY User: ELCR, High-end i = 0.0376 f/cc • 0.0001142 • 0.0641 perf/cc + 0.0376 • 0.02 • 0.02083 • 0.0641
DIY Bystander: ELCR (Central Tendency) 0.0007 f/cc • 0.0001142 • 0.16 per f/cc + 0.0007 • 0.02 • 0.02083 • 0.16
DIY Bystander: ELCR, High-end i = 0.0038 f/cc • 0.0001142 • 0.16 perf/cc + 0.0038 • 0.02 • 0.02083 • 0.16
Exposure Levels from Table 2-31 are used in Table 4-44. The assumption that the concentration of
chrysotile asbestos in the interval between brake works is 2% of that during measured active use. Users
and bystanders were assumed to spend 30 minutes per day in the driveway each day based on the 95th
percentile estimate in the EPA Exposure Factors Handbook (in Table 16-57 in the Handbook). The
reduction factor is 10 for bystanders. The risk estimates for the DIY consumer exceeded the risk
benchmark for the high-end exposure only, whereas the risk estimates were not exceeded for either
scenario for the bystanders.
4.2.3.2 Risk Estimation for Cancer Effects following Episodic Inhalation Exposures
for UTV Gasket Repair/replacement
EPA assessed chrysotile exposures for the DIY (consumer) and bystander UTV gasket
repair/replacement scenario based on aggregated exposures resulting from recurring episodic exposures
from active use of chrysotile asbestos related to DIY brake-related activities. These activities include
concomitant exposure to chrysotile asbestos fibers which are reasonably anticipated to remain within
indoor use facilities. It is well-understood that asbestos fibers in air will settle out in dust and become re-
entrained in air during any changes in air currents or activity indoors. On the other hand, in occupational
settings, regular air sampling would capture both new and old fibers and have industrial hygiene
practices in place to reduce exposures.
For the risk estimations for the UTV gasket COU, EPA used the same data/assumptions identified in
Section 4.2.3.1 for brakes for exposure frequency and duration; with the exception that there is no
outdoor exposure scenario. A sensitivity analysis is presented which includes a lower-bound estimate for
a bystander of 20 years (ages 0-20) (see Appendix L and the uncertainties Section 4.3.7).
The assumption is that DIY UTV gasket replacement is limited to a single gasket change at age 16
years. EPA then assumed that the concentration of chrysotile asbestos in following this COU decreases
50% each year as was assumed in all the indoor exposure scenarios. EPA then assumed that both the
DIYer and the bystander would remain in the house for 10 years. Risks were determined for the 10-year
period by calculating the risk with the appropriate partial lifetime IUR.
Page 209 of 352

-------
Based on these assumptions, the consumer and bystander risk estimates were exceeded for both central
and high-end exposures based on a single UTV gasket change and remaining in the house for 10 years.
Estimates exceeding the benchmark are shaded in pink and bolded.
Table 4-47. Risk Estimate using one UTV gasket change at age 16 years with 10 years further
exposure. Excess Lifetime Cancer Risk for Indoor DIY UTV gasket change for Consumers and
Bystanders (exposures from Table 2-31 without a reduction factor) (Consumers 1 hour/day spent
in garage; Bystanders 1 hour/day)
Consumer
Exposure Scenario
Exposure Levels (f/cc)
ELCR (62 yr exposure
starting at age 16
years)
ELCR (Lifetime
exposure)
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Aftennarket
automotive parts -
brakes (3-hour
TWA once,
indoors)
0.024
0.066
0.012
0.03
2.9 E-6
8.0 E-6
3.2 E-6
7.9 E-6
TWF Concomitant Exposures (1 hour per day every day) (l/24)*(365/365) = 0.04167
IUR(i6,io)=0.0292; IUR(o,ior0.0634
DIY User: ELCR (Central Tendency) 0.024 f/cc • 0.000005524 • 0.0292 perf/cc + 0.024 • 0.1 • 0.04167 • 0.0292
DIY User: ELCR, High-end i = 0.066 f/cc • 0.000005524 • 0.0292 perf/cc + 0.066 • 0.1 • 0.04167 • 0.0292
DIY Bystander: ELCR (Central Tendency) 0.012 f/cc • 0.000005524 • 0.0634 perf/cc + 0.012 • 0.1 • 0.04167 • 0.0634
DIY Bystander: ELCR,High-endi = 0.03 f/cc • 0.000005524 • 0.0634 per f/cc + 0.03 • 0.1 • 0.04167 • 0.0634
Table 4-48. Excess Lifetime Cancer Risk for Indoor DIY UTV Gasket /Repair Replacement for
Consumers and Bystanders (exposures from Table 2-31) (Users 1 hour/day spent in garage;
		Bystanders 1 hour/day) 		
Consumer
Exposure Scenario
Exposure Levels (f/cc)
ELCR (62 yr exposure
starting at age 16
years)
ELCR (Lifetime
exposure)
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-end
Aftennarket UTV
parts - gaskets
(indoors every 3
years)
0.024
0.066
0.012
0.030
1.9 E-5
5.3 E-5
2.4 E-5
6.1 E-5
TWF Concomitant Exposures (1 hour per day every day) (l/24)*(365/365) = 0.04167
IUR( 16,621=0.0641; IUR(Lifetime 1=0.16
DIY User: ELCR (Central Tendency) 0.024 f/cc • 0.0001142 • 0.0641 perf/cc + 0.024 • 0.3 • 0.04167 • 0.0641
DIY User: ELCR,High-end> = 0.066 f/cc • 0.0001142 • 0.0641 per f/cc + 0.066 • 0.3 • 0.04167 • 0.0641
DIY Bystander: ELCR (Central Tendency) 0.012 f/cc • 0.0001142 • 0.16 perf/cc + 0.012 • 0.3 • 0.04167 • 0.16
DIY Bystander: ELCR, High-end i = 0.030 f/cc • 0.0001142 • 0.16 perf/cc + 0.030 • 0.3 • 0.04167 • 0.16
The exposure values from Table 2-31 were used to estimate ELCRs in Table 4-46 for indoor DIY gasket
repair/replacement (one-hour/day assumption). The assumption is that the concentration of chrysotile
asbestos in the interval between gasket work (every 3 years) is 30% of that during measured active use.
Consumers and bystanders were assumed to spend one hour per day in their garages based on the 50th
percentile estimate in the EPA Exposure Factors Handbook (in Table 16-16 in the Handbook). Based on
these assumptions, both the consumer and the bystander risk estimates were exceeded for central
tendency and high-end exposures. Estimates exceeding the benchmark are shaded in pink and bolded.
Page 210 of 352

-------
Table 4-49. Excess Lifetime Cancer Risk for Indoor DIY Gasket/Repair Replacement for
Consumers and Bystanders (exposures from Table 2-31) (Consumers 8-hours/day spent in garage;
		Bystanders 1 hour/day)		
Consumer
Exposure
Scenario
Exposure Levels (f/cc)
ELCR (62 yr exposure
starting at age 16 years)
ELCR (Lifetime
exposure)
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-
end
Central
Tendency
High-end
Central
Tendency
High-end
Central
Tendency
High-
end
Aftennarket
automotive parts -
brakes (indoors
every three years)
0.024
0.066
0.012
0.030
1.5 E-4
4.2 E-4
2.4 E-5
6.1 E-5
TWF Concomitant Exposures (8-hours per day every day) (8/24)*(365/365) = 0.3333
IUR( 16,621=0.0641; IUR(Lifetime 1=0.16
DIY User: ELCR (Central Tendency) 0.024 f/cc • 0.0001142 • 0.0641 per f/cc + 0.024 • 0.3 • 0.3333 • 0.0641
DIY User: ELCR,High-endi = 0.066 f/cc • 0.0001142 • 0.0641 per f/cc + 0.066 • 0.3 • 0.3333 • 0.0641
DIY Bystander: ELCR (Central Tendency) 0.012 f/cc • 0.0001142 • 0.16 per f/cc + 0.012 • 0.3 • 0.04167 • 0.16
DIY Bystander: ELCR,High-endi = 0.030 f/cc • 0.0001142 • 0.16 per f/cc + 0.030 • 0.3 • 0.04167 • 0.16
The exposure values from Table 2-31 were used to estimate ELCRs in Table 4-47 for indoor DIY gasket
repair/replacement (eight hours/day assumption). The assumption is that the concentration of chrysotile
asbestos in the interval between replacement is 30% of that during measured active use. Users were
assumed to spend eight hours per day in their garages based on the 95th percentile estimate in the EPA
Exposure Factors Handbook (U.S. EPA. 2011). Bystanders were assumed to spend one hour per day in
their garages. Based on these assumptions, both the consumer and the bystander risk estimates were
exceeded for central tendency and high-end exposures. Estimates exceeding the benchmark are shaded
in pink and bolded.
4.2.3.3 Summary of Consumer and Bystander Risk Estimates by COU for Cancer
Effects Following Inhalation Exposures
Table 4-48 summarizes the risk estimates for inhalation exposures for all consumer exposure scenarios.
Risk estimates that exceed the benchmark (i.e., cancer risks greater than the cancer risk benchmark) are
shaded and in bold.
Ranging from using an estimate for a single brake job at 16 years of age, and estimates for age at first
exposure (16 years old for DIY users and 0 years for bystanders) and exposure duration (62 years for
DIY users and 78 years for bystanders), for all COUs that were quantiatively assessed, there were risks
to consumers (DIY) and bystanders for all high-end and central tendency exposures from brake
repair/replacement and UTV gasket repair/replacement scenarios except outdoor brake scenarios
(outdoor scenario was not evaluated for gasket replacement). One outdoor brake scenario showed risks
to the DIY consumer for the high-end exposure scenario (30 minutes/day in the driveway).
To evaluate sensitivity to the age at first exposure and exposure duration assumptions, EPA conducted
multiple sensitivity analyses assuming that exposure of DIY users was limited to a single brake change
at age 16 years as well as durations of exposure as short as 20 years with different ages of first exposure.
Section 4.3.7 provides a summary of the detailed analyses in Appendix L. These sensitivity analyses
show that in four of the five scenario pairings different durations and age of first exposure, only one of
24 possible scenarios changed from exceeding the benchmark cancer risk level of lxlO"6 to no
exceedance (DIY user, brake repair outdoors, 30 minutes/ day, high-end only). In the fifth scenario
Page 211 of 352

-------
(Sensitivity Analysis 2), there was no change in any of the 24 scenarios exceeding risk benchmarks. All
analyses are in Appendix L.
Table 4-50. Summary of Risk Estimates for Inhalation Exposures to Consumers and Bystanders
		by CPU (Cancer benchmark is 10-6)			
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer
Risk
Estimates

Brakes
Repair/replacement

DIY
Central
Tendency
3.6 E-5

Indoor, compressed air, once
every 3 years for 62 years
starting at 16 years, exposures
at 30% of active used between
Section

High-end
3.5 E-4

4.2.3.1
Bystander
Central
Tendency
2.6 E-5

uses, 1 hour/d in garage


High-end
6.0 E-5

Brakes Repair/ replacement
Indoor, compressed air, once
every 3 years for 62 years

DIY
Central
Tendency
2.6 E-4

Section

High-end
2.6 E-3

starting at 16 years, exposures
at 30% of active used between
uses, 8 hours/d in garage
4.2.3.1
Bystander
Central
Tendency
1.7 E-5



High-end
3.9 E-5

Brakes
Repair/replacement

DIY
Central
Tendency
5.4 E-6
Imported
asbestos
products
Indoor, compressed air, once at
16 years, staying in residence
for 10 years, exposures at 10%
of active use, 1 hour/d in
Section

High End
5.3 E-5
4.2.3.1
Bystander
Central
Tendency
3.4 E-6

garage


High-end
7.8 E-6

Brakes Repair/ replacement
Outdoor, once every 3 years
for 62 years starting at 16

DIY
Central
Tendency
8.2 E-8

Section

High-end
4.4 E-7

years, exposures at 2% of
active used between uses, 5
min/d in driveway
4.2.3.1
Bystander
Central
Tendency
2.1 E-8



High-end
1.1 E-7

Brakes Repair/ replacement
Outdoor, once every 3 years

DIY
Central
Tendency
2.4 E-7

for 62 years starting at 16
years, exposures at 2% of
active used between uses, 30
min/d in driveway
Section
4.2.3.1

High-end
1.3 E-6

Bystander
Central
Tendency
5.9 E-8


High-end
3.2 E-7
Page 212 of 352

-------
Imported
Asbestos
Products
Gaskets Repair/ replacement in
UTVs
Indoor, 1 hour/d, once every 3
years for 62 years starting at
16 years
exposures at 30% of active
used between uses, 1 hour/d in
garage
Section
4.2.3.2
DIY
Central
Tendency
1.9 E-5
High-end
5.3 E-5
Bystander
Central
Tendency
2.4 E-5
High-end
6.1 E-5
Gaskets Repair/ replacement in
UTVs
Indoor, 1 hour/d, once every 3
years for 62 years starting at
16 years
exposures at 30% of active
used between uses, 8 hour/d in
garage
Section
4.2.3.2
DIY
Central
Tendency
1.5 E-4
High-end
4.2 E-4
Bystander
Central
Tendency
2.4 E-5
High-end
6.1 E-5
Gasket Repair
Repair/replacement
Indoor, once at 16 years,
staying in residence for 10
years, exposures at 10% of
active use, 1 hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
2.9 E-6
High end
8.0 E-6
Bystander
Central
Tendency
3.2 E-6
High-end
7.9 E-6
4.3 Assumptions and Key Sources of Uncertainty
4,3.1 Key Assumptions and Uncertainties in the Uses of Asbestos in the U.S.
EPA researched sources of information to identify the intended, known, or reasonably foreseen asbestos
uses in the U.S. Beginning with the February, 2017 request for information (see 2017 Public Meeting)
on uses of asbestos and followed by both the Scope document (June 2017d) and Problem Formulation
(June 2018d), EPA has refined its understanding of the current conditions of use of asbestos in the U.S.
This has resulted in identifying chrysotile asbestos as the only fiber type manufactured, imported,
processed, or distributed in commerce at this time and under six COU categories. EPA received
voluntary acknowledgement of asbestos import and use from a handful of industries that fall under these
COU categories. Some of the COUs are very specialized, and with the exception of the chlor-alkali
industry, there are many uncertainties with respect to the extent of use, the number of workers and
consumers involved and the exposures that might occur from each activity. For example, the number of
consumers who might change out their brakes on their cars with asbestos-containing brakes ordered on
the Internet or the number of consumers who might change out the asbestos gaskets in the exhaust
system of their UTVs is unknown.
On April 25, 2019, EPA finalized an Asbestos Significant New Use Rule (SNUR) under TSCA section 5
that prohibits any manufacturing (including import) or processing for discontinued uses of asbestos from
restarting without EPA having an opportunity to evaluate each intended use for risks to health and the
environment and to take any necessary regulatory action, which may include a prohibition. By finalizing
Page 213 of 352

-------
the asbestos SNUR to include manufacturing (including import) or processing discontinued uses not
already banned under TSCA, EPA is highly certain that manufacturing (including import), processing,
or distribution of asbestos is not intended, known or reasonably foreseen beyond the 6 product
categories identified herein.
EPA will consider legacy uses and associated disposal and other fiber types of asbestos in Part 2 of the
final Risk Evaluation on asbestos (see Preamble).
4.3.2 Key Assumptions and Uncertainties in the Environmental (Aquatic) Assessment
While the EPA has identified reasonably available aquatic toxicity data to characterize the overall
environmental hazards of chrysotile asbestos, there are uncertainties and data limitations regarding the
analysis of environmental hazards of chrysotile asbestos in the aquatic compartment. Limited data are
available to characterize effects caused by acute exposures of chrysotile asbestos to aquatic organisms.
Only one short-term aquatic invertebrate study was identified (Belanger et al.. 1986b). In addition, the
reasonably available data characterizes the effects of chronic exposure to waterborne chrysotile asbestos
in fish and clams. While these species are assumed to be representative for aquatic species, without
additional data to characterize the effects of asbestos to a broader variety of taxa, the broader ecosystem-
level effects of asbestos are uncertain. The range of endpoints reported in the studies across different life
stages meant that a single definitive, representative endpoint could not be determined, and the endpoints
needed to be discussed accordingly. Several of the effects reported by Belanger et al. (e.g., gill tissue
altered, fiber accumulation, and siphoning activity) are not directly related to endpoints like mortality or
reproductive effects and therefore the biological relevance is unclear. Lastly, the effect concentrations
reported in these studies may differ from the actual effect concentrations due to the inconsistent
methodologies for determining aquatic exposure concentrations of asbestos measured in different
laboratories.
During development of the PF, EPA was still in the process of identifying potential asbestos water
releases for the COUs. After the PF was released, EPA continued to search EPA databases as well as the
literature and engaged in a dialogue with industries to shed light on potential releases to water. In
addition to the Belanger et al. studies, EPA evaluated the following lines of evidence that suggested
there is minimal or no releases of chrysotile asbestos to water: (1) 96% of-14,000 samples from
drinking water sources are below the minimum reporting level of 0.2 MFL and less than 0.2% are above
the MCL of 7 MFL for humans; (2) the source of the asbestos fibers is not known to be from a TSCA
condition of use described in the draft Risk Evaluation; and (3) TRI data have not shown releases of
asbestos to water (Section 2.2.1).
The available information indicated that there were surface water releases of asbestos; however, it is
unclear of the source of the asbestos and the fiber type present. In the draft Risk Evaluation, EPA
concluded that, based on the reasonably available information in the published literature, provided by
industries using asbestos, and reported in EPA databases, there were minimal or no releases of asbestos
to surface water associated with the COUs that EPA is evaluating (see Appendix D). Therefore, EPA
concluded there is no unreasonable risk to aquatic or sediment-dwelling environmental organisms.
EPA has considered peer review and public comments on this conclusion and has decided to keep the
finding made in the draft Risk Evaluation (i.e., that there were minimal or no releases of asbestos to
surface water associated with the COUs that EPA is evaluating in this Part 1 of the risk evaluation for
asbestos). This is because EPA is confident that the minimal water release data available and reported
more fully in the PF - and now presented again in Appendix D - cannot be attributed to chrysotile
asbestos from the COUs in this Part 1 of the risk evaluation for asbestos. Assessing possible risk to
Page 214 of 352

-------
aquatic organisms from the exposures described would not be reasonably attributed to the COUs.
However, based on the decision to develop a scope and risk evaluation for legacy uses and associated
disposals of asbestos (Part 2 of the final Risk Evaluation for asbestos), EPA expects to address the issue
of releases to surface water based on those other uses.
While this does introduce some uncertainty, EPA views it as low and has confidence in making a
determination of no exposure regarding potential releases to water for the COUs in this Part 1 of the risk
evaluation for asbestos. This conclusion is also based on the information in Section 2.3 in which, for the
major COUs {i.e., chlor-alkali, sheet gasket stamping and sheet gasket use), there is documentation of
collecting asbestos waste for disposal via landfill. Finally, there are no reported releases of asbestos to
water from TRI.
4,3.3 Key Assumptions and Uncertainties in the Occupational Exposure Assessment
The method of identifying asbestos in this Part 1 of the risk evaluation for asbestos is based on fiber
counts made by phase contrast microscopy (PCM). PCM measurements made in occupational
environments were used both in the exposure studies and in the studies used to support the derivation of
the chrysotile asbestos IUR. PCM detects only fibers longer than 5 |im and >0.4 |im in diameter, while
transmission electron microscopy (TEM), often found in environmental monitoring measurements, can
detect much smaller fibers. Most of the studies used in this Part 1 of the risk evaluation for asbestos have
reported asbestos concentrations using PCM.
In general, when enough data were reasonably available, the 95th and 50th percentile exposure
concentrations were calculated using reasonably available data {i.e., the chlor-alkali worker monitoring
data). In other instances, EPA had very little monitoring data available on occupational exposures for
certain COUs {e.g., sheet gasket stamping and brake blocks) or limited exposure monitoring data in the
published literature as well. Where there are few data points available, it is unlikely the results will be
representative of worker exposure across the industry depending on the sample collection location (PBZ
or source zone) and timing of the monitoring.
EPA acknowledges that the reported inhalation exposure concentrations for the industrial scenario uses
may not be representative for the exposures in all companies within that industry. For example, there are
only three chlor-alkali companies who own a total of 15 facilities in the U.S. that use chrysotile asbestos
diaphragms, but their operations are different, where some of them hydroblast and reuse their chrysotile
asbestos-containing diaphragms and others replace them. The exposures to workers related to these two
different activities are different.
EPA also received data from one company that fabricates sheet gaskets and one company that uses sheet
gaskets. These data were used, even though there are limitations, such as the representativeness of
practices in their respective industries.
All the raw chrysotile asbestos imported into the U.S. is used by the chlor-alkali industry for use in
asbestos diaphragms. The number of chlor-alkali plants in the U.S. is known and therefore the number
of workers potentially exposed is reasonably certain. In addition, estimates of workers employed in this
industry were provided by the chlor-alkali facilities. However, the number of workers potentially
exposed during other COUs is very limited. Only two workers were identified for stamping sheet
gaskets, and two titanium dioxide manufacturing facilities were identified in the U.S. who use asbestos-
containing gaskets. However, EPA is not certain if asbestos-containing sheet gaskets are used in other
industries and to what extent. For the other COUs, no estimates of the number of potentially exposed
workers were submitted to EPA by industry or its representatives, so estimates were used. Therefore,
Page 215 of 352

-------
numbers of workers potentially exposed were estimated; and these estimates could equally be an over-
estimate or an under-estimate.
Finally, there is uncertainty in how EPA categorized the exposure data. Each PBZ and area data point
was classified as either "worker" or "occupational non-user." The categorizations are based on
descriptions of worker job activity as provided in worker monitoring data, in the literature and EPA's
judgment. In general, PBZ samples were categorized as "worker" and area samples were categorized as
"occupational non-user." Exposure data for ONUs were not available for most scenarios. EPA assumes
that these exposures are expected to be lower than worker exposures, since ONUs do not typically
directly handle asbestos nor are in the immediate proximity of asbestos.
4.3.4 Key Assumptions and Uncertainties in the Consumer Exposure Assessment
Due to lack of specific information on DIY consumer exposures, the consumer assessment relies on
available occupational data obtained under certain environmental conditions expected to be more
representative of a DIY consumer scenario (no engineering controls, no PPE, residential garage).
However, the studies utilized still have uncertainties associated with the environment where the work
was done. In Blake et al. (2003). worker exposures were measured at a former automobile repair facility
which had an industrial sized and filtered exhaust fan unit to ventilate the building during testing while
all doors were closed. A residential garage is not expected to have a filtered exhaust fan installed and
operating during DIY consumer brake repair/replacement activities.
The volume of a former automobile repair facility is considerably larger than a typical residential garage
and will have different air exchange rates. While this could raise some uncertainties related to the
applicability of the measured data to a DIY consumer environment, the locations of the measurements
utilized for this evaluation minimize that uncertainty.
There is some uncertainty associated with the length of time EPA assumes the brake repair/replacement
work takes. The EPA assumed it takes a DIY consumer user about three hours to complete brake
repair/replacement work. This is two times as long as a professional mechanic. While it is expected to
take a DIY consumer longer, it is also expected DIY consumers who do their own brake
repair/replacement work would, over time, develop some expertise in completing the work as they
continue to do it every three years.
There is also some uncertainty associated with the assumption that a bystander would remain within
three meters from the automobile on which the brake repair/replacement work is being conducted for the
entire three-hour period EPA assumes it takes the consumer user to complete the work. However,
considering a residential garage with the door closed is relatively close quarters for car repair work, it is
likely anyone observing (or learning) the brake repair/replacement work would not be able to stay much
further away from the car than three meters. Remaining within the garage for the entire three hours also
has some uncertainty, although it is expected anyone observing (or learning) the brake
repair/replacement work would remain for the entire duration of the work or would not be able to
observe (or learn) the task.
While industry practices have drifted away from the use of compressed air to clean brake drums/pads,
no information was found in the literature indicating consumers have discontinued such work practices.
To consider potential consumer exposure to asbestos resulting from brake repair/replacement activities,
EPA uses data which included use of compressed air. However, EPA recognizes this may be a more
conservative estimate because use of compressed air typically could cause considerable dust/fibers to
become airborne if it is the only method used.
Page 216 of 352

-------
There were no data identified through systematic review providing consumer specific monitoring for
UTV exhaust system gasket repair/replacement activities. Therefore, this evaluation utilized published
monitoring data obtained in an occupational setting, performed by professional mechanics, on
automobile exhaust systems as a surrogate for estimating consumer exposures associated with UTV
gasket removal/replacement activities. There is some uncertainty associated with the use of data from an
occupational setting for a consumer environment due to differences in building volumes, air exchange
rates, available engineering controls, and the potential use of PPE. As part of the literature review, EPA
considered these differences and utilized reasonably available information which was representative of
the expected consumer environment.
There is uncertainty associated with the use of an automobile exhaust system gasket repair/replacement
activity as a surrogate for UTV exhaust system gasket repair/replacement activity due to expected
differences in the gasket size, shape, and location. UTV engines and exhaust systems are expected to be
smaller than a full automobile engine and exhaust system, therefore the use of an automobile exhaust
system gasket repair may slightly overestimate exposure to the consumer. At the same time, the smaller
engine and exhaust system of a UTV could make it more difficult to access the gaskets and clean the
surfaces where the gaskets adhere therefore increasing the time needed to clean and time of exposure
resulting from cleaning the surfaces which could underestimate consumer exposure.
There is uncertainty associated with the assumption that UTV exhaust system gasket repair/replacement
activities would take a consumer a full three hours to complete. While there was no published
information found providing consumer specific lengths of time to complete a full repair/replacement
activity. The time needed for a DIY consumer to complete a full UTV exhaust system gasket
repair/replacement activity can vary depending on several factors including location of gaskets, number
of gaskets, size of gasket, and adherence once the system is opened up and the gasket removed. Without
published information, EPA assumes this work takes about three hours and therefore three-hour time
frames to estimate risks for this evaluation.
Finally, EPA has made some assumptions regarding both age at start of exposure and duration of
exposure for both the DIY users and bystanders for both the brake and UTV gasket scenarios. Realizing
there is uncertainty around these assumptions, specifically that they may over-estimate exposures, EPA
developed a sensitivity analysis approach specifically for the consumer exposure/risk analysis (see
appropriate part of Section 4.3.8 below) and also performed a sensitivity analysis using five different
scenarios (Appendix L), including a single brake change (or UTV gasket) repair/replacement activity at
16 years of age.
4,3,5 Key Assumptions and Uncertainties in the Human Health IUR Derivation
The analytical method used to measure exposures in the epidemiology studies is important in
understanding and interpreting the results as they were used to develop the IUR. As provided in more
detail in Section 3, the IUR for "current use" asbestos {i.e., chrysotile) is based solely on studies of PCM
measurement as TEM-based risk data are limited in the literature and the available TEM results for
chrysotile asbestos lack modeling results for mesothelioma. In TEM studies of NC and SC (Loomis et
at. 2010; Stavner et at.. 20081 models that fit PCM vs. TEM were generally equivalent (about 2 AIC
units), indicating that fit of PCM is similar to the fit of TEM (for these two cohorts), providing
confidence in those PCM measurements for SC and NC. Given that confidence in the PCM data and the
large number of analytical measurements, exposure uncertainty is considered low in the cohorts used for
IUR derivation.
Page 217 of 352

-------
The endpoint for both mesothelioma and lung cancer was mortality, not incidence. Incidence data are
not available for any of the cohorts. However, for lung cancer, EPA was able to use background rates of
incidence data in lifetables (Appendix H) as a way to address this bias of using mortality data for risk
estimates instead of incidence data. For mesothelioma, this adjustment in lifetable methodology was not
possible, because the mesothelioma model uses an absolute risk. Thus, using mortality data for
mesothelioma remains a downward bias in the selected IUR value. However, for mesothelioma, the
median length of survival with mesothelioma is less than 1 year for males, with less than 20% surviving
after 2-years and less than 6% surviving after 5-years.
By definition, the IUR only characterizes cancer risk. It does not include any risks that may be
associated with non-cancer health effects. Pleural and pulmonary effects from asbestos exposure (e.g.,
asbestosis and pleural thickening) are well-documented (U.S. EPA. 1988b). although there is no
reference concentration (RfC) for these non-cancer health effects specifically for chrysotile asbestos.
The IUR for chrysotile asbestos is 0.16 per f/cc (Section 3.2.4). Based on this IUR, the chrysotile
asbestos exposure concentration, over a lifetime, that would be expected to cause 1 cancer per 1,000,000
people (1E-6) in the general population is 6E-6 f/cc. The IRIS assessment of Libby amphibole asbestos
(	2014b) derived a RfC for non-cancer health effects, and at that concentration (9 E-5 f/cc),
the risk of cancer for chrysotile asbestos was 1 E-5 [IUR*RfC = (0.16 per f/cc)*(9 E-5 f/cc)]. Thus, at a
target risk of 1 cancer per 1,000,000 people (1E-6) and exposures at or below 6E-6 f/cc to meet that
target risk, the chrysotile asbestos cancer toxicity value appeared to be the clear risk driver as meeting
that target risk for the general population (including consumers, DIY and bystanders) would result in
lower non-cancer risks than at the Libby Amphibole asbestos RfC (i.e., 6E-6 f/cc < 5E-5 f/cc).
The POD associated with the only non-cancer toxicity value is for Libby amphibole asbestos - 0.026 f/cc
(U.S. EPA. 2014b). Although the non-cancer toxicity of chrysotile asbestos may be different from Libby
amphibole asbestos, there is uncertainty that the IUR for chrysotile asbestos may not fully encompasses
the health risks associated with occupational exposure to chrysotile asbestos. Several of the COU-related
exposures evaluated for human health risks in Section 4.2 are at or greater than the POD for non-cancer
effects associated with exposure to Libby amphibole asbestos.
4.3.6 Key Assumptions and Uncertainties in the Cancer Risk Values
Although direct comparison of cancer slopes for PCM and TEM fibers is impossible because different
counting rules for these methods result in qualitatively and quantitatively different estimates of asbestos
exposure, comparing the fit of models based on different analytical methods is possible. In TEM studies
of NC and SC (Loomis et ai. 2010; Stavner et at.. 2008). models that fit PCM vs. TEM were generally
equivalent (about 2 AIC34 units), indicating that fit of PCM is similar to the fit of TEM (for these two
cohorts), providing confidence in those PCM measurements for SC and NC, whose data is the basis for
chrysotile asbestos IUR.
Another source of uncertainty in the exposure assessment is that early measurements of asbestos fiber
concentrations were based on an exposure assessment method (midget impinger) that estimated the
combined mass of fibers and dust, rather than on counting asbestos fibers. The best available
methodology for conversion of mass measurements to fiber counts is to use paired and concurrent
sampling by both methods to develop factors to convert the mass measurements to estimated fiber
counts for specific operations. There is uncertainty in these conversion factors, but it is minimized in the
34 AIC stands for Akaike Information Criterion; a measure of relative quality used for evaluation of statistical models.
Page 218 of 352

-------
studies of SC and NC chrysotile asbestos textile workers due to the availability of an extensive database
of paired and concurrent samples and the ability to develop operation-specific conversion factors.
Uncertainty in the estimation of these conversion factors and their application to estimate chrysotile
asbestos exposures will not be differential with respect to disease because this uncertainty is independent
of cancer status.
As noted in Section 3.2.3.3 the exposure data from the SC and NC cohorts are of higher quality than
those utilized in other studies of occupational cohorts exposed to chrysotile asbestos. Given the
confidence in the PCM data and the large number of analytical measurements, exposure uncertainty is
overall low in the SC and NC cohorts, as high-quality exposure estimates are available for both cohorts.
Statistical error in estimating exposure levels is random and not differential with respect to disease {i.e.,
independent). Therefore, to the extent that such error exists, it is likely to produce either no bias or bias
toward the null under most circumstances {e.g., Kim et	); Armstrong (1998)).
Epidemiologic studies are observational and as such are potentially subject to confounding and selection
biases. Most of the studies of asbestos exposed workers did not have information to control for cigarette
smoking, which is an important risk factor for lung cancer in the general population. In particular, the
NC and SC studies of textile workers, which were chosen as the most informative studies, did not have
this information.
However, the bias related to this inability to control for smoking is believed to be small (Blair et at..
2007); Siemiatvcki et a 0) because the exposure-response analyses for lung cancer were based on
internal comparisons and for both studies the regression models included birth cohort, thus introducing
some control for the changing smoking rates over time. It is unlikely that smoking rates among workers
in these facilities differed substantially enough with respect to their cumulative chrysotile exposures to
induce important confounding in risk estimates for lung cancer. Mesothelioma is not related to smoking
and thus smoking could not be a confounder for mesothelioma.
For the purpose of combining risks, it is assumed that the unit risks of mesothelioma and lung cancer are
normally distributed. Because risks were derived from a large epidemiological cohort, this is a
reasonable assumption supported by the statistical theory and the independence assumption has been
investigated and found a reasonable assumption (	'14c).
4,3,7 Confidence in the Human Health Risk Estimations
Workers/Occupational Non-Users
Depending on the variations in the exposure profile of the workers/occupational non-users, risks could
be under- or over-estimated for all COUs. The estimates for extra cancer risk were based on the EPA-
derived IUR for chrysotile asbestos. The occupational exposure assessment made standard assumptions
of 240 days per year, 8 hours per day over 40 years starting at age 16 years35 . This assumes the workers
and occupational non-users are regularly exposed until age 56. If a worker changes jobs during their
career and are no longer exposed to asbestos, this may overestimate exposures. However, if the worker
stays employed after age 56, it would underestimate exposures.
The concentration-response functions on which the chrysotile asbestos IUR is based varies as a function
of time since first exposure. Consequently, estimates of cancer risk depend not only on exposure
concentration, frequency and duration, but also on age at first exposure. To approximate the impact of
35 The Fair Labor Standards Act of 1938 allows adolescents to work an unrestricted number of hours at age 16 years.
Page 219 of 352

-------
different assumptions for occupational exposures, Table 4-49 can be used to understand what percentage
of the risk in the baseline occupational exposure scenario remains for different ages at first exposure and
different durations of exposure.
Table 4-51. Ratios of risks for alternative exposure scenarios using scenario-specific partial
lifetime IURs from Appendix K by age at first exposure and duration of exposure compared to
baseline occupational exposure scenarios (baseline scenario: first exposure at 16 years for 40 years
	duration)	

Duration of exposure (years)
Age at first exposure (years)
20
40
16
0.0468/0.0612 = 0.76
0.0612/0.0612= 1
20
0.0374/0.0612 = 0.61
0.0486/0.0612 = 0.79
30
0.0209/0.0612 = 0.34
0.0269/0.0612 = 0.44
Other occupational exposure scenario can be evaluated by selecting different values for the age at first
exposure and the duration of exposure from the table of partial lifetime IUR values in Appendix K.
Exposures for ONUs can vary substantially. Most data sources do not sufficiently describe the proximity
of these employees to the exposure source. As such, exposure levels for the ONU category will vary
depending on the work activity. It is unknown whether these uncertainties overestimate or underestimate
exposures.
Cancer risks were indicated for all of the worker COUs and most of the consumer/bystander COUs. If
additional factors were not considered in the RE, such as exposures from other sources (e.g., legacy
asbestos sources), the risks could be underestimated. Legacy uses and associated disposals of asbestos
will be considered in Part 2 of the risk evaluation for asbestos (see Preamble).
In addition, several subpopulations (e.g., smokers, genetically predisposed individuals, COU workers
who change their own asbestos-containing brakes, etc.) may be more susceptible than others to health
effects resulting from exposure to asbestos. These conditions are discussed in more detail for potentially
exposed or susceptible subpopulations and aggregate exposures in Section 4.4 and Section 4.5.
Consumer DIY/Bystanders
Similarly, risks for consumers/bystanders could be under- or over-estimated for their COU. Unlike
occupational scenarios, there are no standard assumptions for consumers and bystanders, EPA
conducted sensitivity analyses to evaluate some alternative scenarios for consumers/bystanders as
described below.
For consumers (see Table 4-48) EPA considered age at first exposure of 16 years with duration of
exposure 62 years and for bystanders EPA considered age at first exposure of 0 years with lifetime
duration (78 years). To evaluate sensitivity to these assumptions, EPA conducted multiple sensitivity
analyses assuming that duration of exposure as short as 10 years with different ages of first
exposure. Tables 4-50 and 4-51 below show the different scenarios covered in the sensitivity analysis
and the associated adjustment factor that may be used to calculate a different risk number. In Table 4-50,
DIY exposures with different ages at start of exposure (16, 20 or 30 years old) are paired with different
durations of exposure (20, 40 or 62) and Table 4-51 shows the same for bystanders (age at start is
always zero but the three exposure durations are 20, 40 and 78). All analyses are presented in Appendix
Page 220 of 352

-------
L and show that using the ratios in both Tables 4-49 and 4-50 do not change the overall risk picture in
almost all scenarios. Table 4-52 provides a summary of the detailed analyses in Appendix L.
Table 4-52. Ratios of risks for alternative exposure scenarios using scenario-specific partial
lifetime IURs from Appendix K by age at first exposure and duration of exposure compared to
baseline consumer DIY exposure scenarios (baseline scenario: first exposure at 16 years for 62
	years duration)	

Duration of exposure (years)
Age at first exposure
(years)
20
40
62
16
0.0468/0.0641 = 0.73
0.0612/0.0641 =0.95
0.0641/0.0641 = 1
20
0.0374/0.0641 = 0.58
0.0486/0.0641 =0.76
-
30
0.0209/0.0641 = 0.33
0.0269/0.0641 =0.42
-
Table 4-53. Ratios of risks for alternative exposure scenarios using scenario-specific partial
lifetime IURs from Appendix K by age at first exposure and duration of exposure compared to
baseline consumer bystander exposure scenarios (baseline scenario: first exposure at 0 years for
	78 years duration)	

Duration of exposure (years)
Age at first exposure
(years)
20
40
78
0
0.106/0.16 = 0.66
0.144/0.16=0.90
0.16/0.16 = 1
Table 4-52 provides a summary of the detailed analyses in Appendix L. These sensitivity analyses show
that in four of the five scenario pairings, only one of 24 possible scenarios changed from exceeding the
benchmark cancer risk level of lxlO"6 to no exceedance (DIY user, brake repair outdoors, 30
minutes/day, high-end only). In the fifth scenario (Sensitivity Analysis 2), there was no change in any of
the 24 scenarios. All analyses are in Appendix L.
Page 221 of 352

-------
Table 4-54. Results of Sensitivity Analysis of Exposure Assumptions for Consumer DIY/Bystander
	Episodic Exposure Scenarios	
Sensitivity
Analysis'
DIY (age at start and
age at end of duration)
livstander (age at
start and age at end
of duration)
Change in Kisk
from Kxceedance
to No Kxceedance
Scenario Affected
Baseline
16-78
0-78
None
17/24 Exceed
Benchmarks
1
16-36
0-20
1/24
DIY user, Brake
repair, 30 min/day,
high-end
2
20-60
0-40
0/24
None
3
20-40
0-40
1/24
DIY user, Brake
repair, 30 min/day,
high-end
4
30-70
0-40
1/24
DIY user, Brake
repair, 30 min/day,
high-end
5
30-50
0-20
1/24
DIY user, Brake
repair, 30 min/day,
high-end
1 Includes all brake repair/replacement and gasket repair replacement scenarios - a total of 24. See Table 4-45
Assumptions About Bystanders
The EPA Exposure Factors Handbook (2011) provides the risk assessment community with data-derived
values to represent human activities in a variety of settings. For the purposes of this Part 1 of the risk
evaluation for asbestos, understanding the amount of time consumers spend in a garage is important to
develop an exposure scenario for DIYers/mechanics who change their own brakes or gaskets and
bystanders to those activities. Table 16-16 in the Handbook, entitled Time Spent (minutes/day) in
Various Rooms at Home and in All Rooms Combined, Doers Only, has a section on time spent in a
garage.
The total number of respondents to the survey question on time spent in the garage was 193 and the
minimum and maximum reported times were one minute and 790 minutes (-13 hours). Again, these
respondents are "doers," defined as people who reported being in that location (i.e., the garage). In this
analysis, it was assumed that the 50th percentile would represent a central tendency estimate for being
present in the garage (one hour/day) and the 95th percentile would represent a high-end estimate for
being present in the garage (8-hours).
EPA understands that a bystander in this exposure situation (DIY automotive and UTV repair) is most
likely to be a family member (minor or adult relative) with repeated access to the garage used to repair
vehicles. As a familial bystander, and not a neighbor or someone visiting, EPA considered that these
bystanders would have similar exposures to the garage, and thus to any chrysotile asbestos fibers in the
same garage environment as the DIY user. EPA used the same median time of one hour per day as the
bystander's estimated central tendency and the same estimate of high-end exposures. EPA noted that the
younger doers appear to spend somewhat more time in the garage (EFH Table 16-16). In the same table
of time spent per day in the garage, some data on doers is shown for ages 1-17 years which can be
Page 222 of 352

-------
aggregated to find the mean time spent in a garage. The mean for these children is 77 minutes per day
based on 22 young doers, which is similar to the one-hour median based on all 193 doers. EPA also
noted that male doers had a median of 94 minutes compared to female doers who had a median of 30
minutes per day in the garage. It is possible that familial bystanders are unlike the DIY users and spend
little time in the garage. If this were true, then with little or no time spent in the garage, their risks would
be limited.
Finally, as part of the sensitivity analysis, understanding that a bystander in a doer family may spend
somewhat less time in the garage than the 50th percentile time of one hour (60 minutes/day), Table 4-53
below shows the data available in the Exposure Factors Handbook that present other percentiles broken
down by age and gender. In its original analysis, EPA used 60 minutes/day. If 10 minutes/day were used
for the bystander and in keeping with deriving a risk estimate following a single brake or gasket change
and a time-in-residence of only 10 years, the calculated risk values would be:
At 10 minutes/day in the garage following a single brake change and the next 10 years in the
house, the by-stander risks would be 6.9 E-8 for the central tendency and 1.6 E-7 for the high-
end estimates.
At 10 minutes/day in the garage following a single UTV gasket change and the next 10 years in
the house, the by-stander risks would be 6.4 E-8 for the central tendency and 1.6 E-7 for the
high-end estimates.
Table 4-55. Time Spent (minutes/day) in Garage, Doers Only (Taken from Table 16-16 in EFH, 2011
(lender mid Age
Percentiles in 1 lie Distribution o
'Survev Respondents
Usmge
5"'
25th
501 It
75th
95 th
All ages
5
20
60
150
480
Men
10
30
94
183
518
Women
5
15
30
120
240
1-4 yrs old
15
52
100
115
120
5 to 11 yrs old
10
25
30
120
165
12-17 yrs old
10
20
51
148
240
Potential Number of Impacted Individuals
Table 4-54 provides an estimate of the number of impacted individuals for both occupational and
consumer exposure scenarios. Some of the estimates have a higher level of confidence than others. For
example, EPA is reasonably certain about the number of chlor-alkali workers given the information
submitted by industry. For some of the other COUs, the estimated numbers presented in the draft risk
evaluation were modified based on peer review and public comments.
The following text accompanies the estimates presented in Table 4-54:
Chlor-Alkali Workers and ONUs
There is a total of 3,050 employees at the 15 chlor-alkali plants we have identified as using diaphragms;
with approximately 75-148 potentially exposed to asbestos during various activities associated with
constructing, using and deconstructing asbestos diaphragms. Subtracting the 75 to 148 workers
potentially exposed to asbestos results in approximately 2,900 to 3,000 other employees who work at the
same or adjoining plant. This is an upper bound estimate of the number of ONUs and only an unknown
Page 223 of 352

-------
subset of these workers may be ONUs. EPA has low certainty in this number because some of these
sites are very large and make different products in different parts of the facility (one site is 1,100 acres
and has 1,300 employees). Thus, this approach may overestimate the number of ONUs for asbestos
diaphragms.
Sheet Gaskets - Stamping (Workers and ONUs)
EPA found only two gasket sampling sites handling asbestos containing sheet gasket; one worker and
two ONUs per site. However, there may be more gasket stamping sites processing asbestos containing
sheet gasket in US. Thus, the uncertainty in this number of impacted individuals is high.
Sheet Gaskets - Use (Workers and ONUs)
The Bureau of Labor Statistics 2016 data for the NAICS code 325180 (Other Basic Inorganic Chemical
Manufacturing) indicates an industry-wide aggregate average of 25 directly exposed workers per facility
and 13 ONUs per facility. The total number of use sites is unknown.
Oilfield Brake Blocks (Workers and ONUs)
According to 2016 Occupational Employment Statistics data from the Bureau of Labor Statistics (BLS)
and 2015 data from the U.S. Census' Statistics of U.S. Businesses. EPA used BLS and Census data for
three NAICS codes: 211111, Crude Petroleum and Natural Gas Extraction; 213111, Drilling Oil and Gas
Wells; and 213112, Support Activities for Oil and Gas Operations, there are up to 61,695 workers and
66,108 ONU. See Table 2-12 for the breakdown by each category. It is not known how many of these
workers are exposed to asbestos.
Aftermarket Automobile Brakes/Linings/Clutches (Workers and ONUs)
As discussed in Sections 2.3.1.7.2, EPA has changed its estimates of the number of workers/ONUs who
may be exposed to chrysotile asbestos from replacing aftermarket automobile brakes/linings/clutches.
The Draft Risk Evaluation estimated that there are 749,900 automotive service technicians and
mechanics that may be exposed as workers to aftermarket automotive brakes/linings, and clutches, and
another 749,900 that may be exposed as ONUs.
Both peer review and public commenters questioned this estimate; which was based on assuming that
100% of workers in automotive repair could potentially be exposed to a much lower percentage of
asbestos-containing brakes on the market. One comm.enter suggested that, at worst case, asbestos brakes
represent 0.002% of the market, and that it is not rational to conclude that 100% of workers in
automotive repair could potentially be exposed to 0.002% of brakes on the market. While EPA
disagreed with some of the specifics in the calculation, EPA does agree such an adjustment is warranted.
Government data on imports are characterized by Harmonized Tariff Schedule (HTS) code. Most
friction materials are reported under HTS code 68 1 3.36 Products containing asbestos should be reported
under HTS 6813.20. Within HTS 6813.20, the 8-digit codes 6813.20.10 and 6813.20.15 represent brakes
linings and pads for use in civil aircraft and in other vehicles (e.g., cars), respectively. Although some of
the shipments coded with an HTS code for asbestos may be misclassified and may not contain asbestos,
36 HTS 6813 is defined as "Friction material and articles thereof (for example, sheets, rolls, strips, segments, discs, washers,
pads), not mounted, for brakes, for clutches or the like, with a basis of asbestos, of other mineral substances or of cellulose,
whether or not combined with textile or other materials."
Page 224 of 352

-------
it is also possible that some of the shipments coded with a non-asbestos HTS code are also misclassified
and actually do contain asbestos. The figures for imports of brakes in HTS code 6813.20.15 seem to be
the best option available for estimating the potential market share of asbestos brakes.
According to U.S. Census Bureau data, the average annual value of imports in HTS 6813.20.15 during
the period from 2010 to 2019 was $1,949,006.37 According to the web page of a market research group,
the demand for aftermarket automotive brake is approximately $4.3 billion per year for all of North
America.38 Based on this data, asbestos brakes may represent approximately 0.05% of aftermarket
automotive brakes.
Assuming that the number of potentially exposed individuals is equal to the apparent market share of
asbestos brakes and applying a 0.05% adjustment factor to the estimates of 749,900 yields a value of
375 for both workers and ONUs (see Table 4-54).
While there would still be uncertainty in these newer estimates, the level of confidence is higher than in
the 100%) estimates in the draft risk evaluation.
1J TV Sheet Gaskets (Workers and ONUs)
Based on Bureau of Labor Statistics and several assumptions detailed in section 2.3.1.9, EPA estimate
1,500 workers for UTV service technicians and mechanics. It is not known how many of them service
and/or repair UTV with asbestos containing gasket.
Aftermarket Automobile Brakes/Linings/Clutches (Consumers/DIY/Bystanders)
In the draft Risk Evaluation, EPA calculated the number of consumers that could be purchasing and
performing DIY brake jobs based on housing units and number of vehicles per household, and the
results of a survey suggesting that 50% of all U.S. households have at least one automotive DIYer (see
Section 4.3.7 in the Draft Risk Evaluation for Asbestos). The estimate was that 31,857,106 automotive
DIYers replace brake pads.
However, as noted above, both peer review and public commenters questioned this estimate; which was
based on assuming that 100% of the estimated DIYers purchased and used asbestos-containing brakes.
Using the same logic previously described {i.e., a market share estimate of 0.05%), and applying a
0.05%) adjustment factor to the estimates of 31,857,106 yields a value of 15,929 for DIYers (see Table
4-54).
While there would still be uncertainty in these newer estimates, the level of confidence is higher than in
the 100%) estimates in the draft risk evaluation.
CO Us for Which No Estimates May be Made
EPA could not develop a reasonable estimate of potentially impacted individuals for two COUs: other
vehicle friction products (workers/ONUs) and UTV gasket replacement/repair (DIY/bystanders).
37	Source: U.S. International Trade Commission DataWeb (USITC DataWeb), using data retrieved from the U.S. Bureau of
the Census (accessed August 12, 2020).
38	Since this figure is not limited to the United States, it may underestimate the fraction of U.S. brakes that may contain
asbestos (i.e., the denominator is too large).
Page 225 of 352

-------
Table 4-56. Summary of Estimated Number of Exposed Workers and DIY Consumers3.
Condition of I se
Industrial am
Workers
1 Commercial
OM
1)
Consumer
IV
IKstandcrs
Asbestos diaphragms - chlor-
alkali
75-148
<2900-3000
-
-
Sheet gaskets - stamping
>2
>4
-
-
Sheet gaskets - use
25/facility (no. of
facilities Unknown)
13/facility (no. of
facilities Unknown
-
-
Oilfield brake blocks
<61,695 (total;
number exposed to
asbestos unknown)
<66,108 (total;
number in vicinity
of asbestos
unknown)
-
-
Aftermarket automotive
brakes/linings, clutches
375
375
15,929
Unknown
Other Vehicle Friction Products
(brakes installed in exported cars)
Unknown
Unknown
-
-
Other gaskets - UTVs
-1500 (total;
number exposed to
asbestos unknown
Unknown
Unknown
Unknown
a See Text for details.
4.4 Other Risk-Related Considerations
4,4,1 Potentially Exposed or Susceptible Subpopulations
EPA identified workers, ONUs, consumers, and bystanders as potentially exposed populations. EPA
provided risk estimates for workers and ONUs at both central tendency and high-end exposure levels for
most COUs. EPA determined that bystanders may include lifestages of any age.
For inhalation exposures, risk estimates did not differ between genders or across lifestages because both
exposures and inhalation hazard values are expressed as an air concentration. EPA expects that
variability in human physiological factors (e.g., breathing rate, body weight, tidal volume) could affect
the internal delivered concentration or dose of asbestos.
Workers exposed to asbestos in workplace air, especially if they work directly with asbestos, are most
susceptible to the health effects associated with asbestos due to higher exposures. Some workers not
associated with the COU in this Part 1 of the risk evaluation for asbestos may experience higher
exposures to asbestos, such as, but not limited to, asbestos removal workers, firefighters, demolition
workers and construction workers (Lamdrigam et ai. 2004); and these populations will be considered
when EPA evaluates legacy uses and associated disposals of asbestos in Part 2 of the risk evaluation for
asbestos. Although it is clear that the health risks from asbestos exposure increase with greater exposure
and longer exposure time, investigators have found asbestos-related diseases in individuals with only
brief exposures. Generally, those who develop asbestos-related diseases show no signs of illness for a
long time after exposure (ATSDR. 2.001a).
A source of variability in susceptibility between people is smoking history or the degree of exposure to
other risk factors with which asbestos interacts. In addition, the long-term retention of asbestos fibers in
Page 226 of 352

-------
the lung and the long latency period for the onset of asbestos-related respiratory diseases suggest that
individuals exposed earlier in life may be at greater risk to the eventual development of respiratory
problems than those exposed later in life (ATSDR. 2001a). Appendix K of this Part 1 of the risk
evaluation for asbestos illustrates this point in the IUR values for less than lifetime COUs. For example,
the IUR for a one-year old child first exposed to chrysotile asbestos for 40 years is 1.31 E-l while the
IUR for a 20-year old first exposed to asbestos for 40 years is 5.4 E-2. Using the central tendency
bystander exposure value of 0.032 f/cc, the resulting risk estimates are 1.7 x E-4 and 7.2 x E-5,
respectively. There is also some evidence of genetic predisposition for mesothelioma related to having a
germline mutation in BAP1 (Testa et at.. ).
4.4.2 Aggregate and Sentinel Exposures
Section 2605(b)(4)(F)(ii) of TSCA requires the EPA, as a part of the risk evaluation, to describe whether
aggregate or sentinel exposures under the conditions of use were considered and the basis for their
consideration. The EPA has defined aggregate exposure as "the combined exposures to an individual
from a single chemical substance across multiple routes and across multiple pathways (40 CFR ง
702.33)."
Aggregate exposures for chrysotile asbestos were not assessed by routes of exposure in this Part 1 of the
risk evaluation for asbestos since only inhalation exposure was evaluated. Although there is the
possibility of dermal exposures occuring for the chrysotile asbestos COUs, it is unlikely that they would
contribute to mesothelioma and lung cancer. As discussed in the scope and PF documents, the only
known hazard associated with dermal exposure to asbestos is the formation of warts. But perhaps most
importantly, with risk estimations already exceeding benchmarks from simply inhalation exposures,
adding other, different exposures/hazards does not seem pragmatic.
Pathways of exposure were not combined in this Part 1 of the risk evaluation for asbestos. Although it is
possible that workers exposed to asbestos might also be exposed as consumers (e.g., by changing brakes
at home), the number of workers/users is potentially small. The individual risk estimates already indicate
risk; aggregating the pathways would increase the risk.
In addition, the potential for exposure to other uses/fiber types of asbestos (besides chrysotile), legacy
asbestos for any populations or subpopulation, due to activities such as home or building renovations,
as well as occupational or consumer exposures identified in this Part 1 of the risk evaluation for
asbestos, is possible. EPA will consider legacy uses and associated disposals of asbestos in Part 2 of the
Risk Evaluation on asbestos (see Preamble).
The EPA defines sentinel exposure as "the exposure to 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 ง 702.33)." In terms of this Part 1 of the risk evaluation for asbestos, the
EPA considered sentinel exposure the highest exposure given the details of the conditions of use and the
potential exposure scenarios. EPA considered sentinel exposure for chrysotile asbestos in the form of a
high-end level scenario for occupational, ONU, consumer DIY, and bystander exposures resulting from
inhalation exposures for each COU.
Page 227 of 352

-------
4.5 Risk Conclusions
4,5.1 Environmental Risk Conclusions
Based on the reasonably available information in the published literature, provided by industries using
asbestos, and reported in EPA databases, there is minimal or no releases of asbestos to surface water and
sediments associated with the COUs in this risk evaluation. Therefore, EPA concludes there is no
unreasonable risk to aquatic or sediment-dwelling environmental organisms. In addition, terrestrial
pathways, including biosolids, were excluded from analysis at the PF stage.
4,5.2 Human Health Risk Conclusions to Workers
Table 4-38 provides a summary of risk estimates for workers and ONUs. For workers in all of
quantitatively assessed COU categories identified in this Part 1 of the risk evaluation for asbestos cancer
risks were exceeded for all high-end exposures (except for other vehicle friction products: Super Guppy
and high-end exposures were not assessed for oilfield brake blocks) and for most at the central tendency
(except for chlor-alkali diaphragms and other vehicle friction products: Super Guppy). In addition, for
ONUs, cancer risks were exceeded for both central tendency and high-end exposures for sheet gasket
use and UTV gasket replacement. Cancer risks for ONUs were indicated for high-end exposures only for
chlor-alkali, sheet gasket stamping. With the assumed use of respirators as PPE at APF of 10, most risks
would be reduced but still persisted for high-end exposures for sheet gasket stamping, sheet gasket use,
aftermarket auto brake replacement, other vehicle friction products and UTV gasket replacement. When
respirators with an APF of 25 was assumed, risk was still indicated for the aftermarket auto brakes and
the other vehicle friction products for high-end short-term exposure scenario. It is important to note that
based on published evidence for asbestos (see Section 2.3.1.2), nominal APF may not be achieved for all
respirator users. ONUs are not assumed to be using PPE to reduce exposures to asbestos.
4,5.3 Human Health Risk Conclusions to Consumers
Table 4-48 provides a summary of risk estimates for consumers and bystanders. Cancer risks were
exceeded for all consumer and bystander exposure scenarios except for some of the outdoor brake
repair/ replacement exposure scenarios. Cancer risk, however, was still exceeded for the
consumer/DIYer 30 minutes/day outdoor brake repair/replacement exposure scenario for high-end
exposures.
Page 228 of 352

-------
5 Risk Determination
5.1 Unreasonable Risk
5.1.1 Overview
In each risk evaluation under TSCA ง 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. In making this determination, 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); the severity of hazard (including the nature of the hazard, the irreversibility of the
hazard); and uncertainties. EPA takes into consideration the Agency's confidence in the data used in the
risk estimate. This includes an evaluation of the strengths, limitations and uncertainties associated with
the information used to inform the risk estimate and the risk characterization. This approach is in
keeping with the Agency's final rule, Procedures for Chemical Risk Evaluation Under the Amended
Toxic Substances Control Act (82 FR 33726).
Under TSCA, conditions of use are defined as the circumstances, as determined by the Administrator,
under which the substance is intended, known, or reasonably foreseen to be manufactured, processed,
distributed in commerce, used, or disposed of (TSCA ง3(4)).
An unreasonable risk may be indicated when health risks under the conditions of use are identified by
comparing the estimated risks with the risk benchmarks and where the risks affect the general
population or certain potentially exposed or susceptible subpopulations (PESS), such as consumers. For
other PESS, such as workers, an unreasonable risk may be indicated when risks are not adequately
addressed through expected use of workplace practices and exposure controls, including engineering
controls or use of personal protective equipment (PPE). This Part 1 of the risk evaluation for asbestos
evaluated the cancer risk to workers and occupational non-users and consumers and bystanders from
inhalation exposures only, and in this risk determination of chrysotile asbestos, respirator PPE (where
present) and its effect on mitigating inhalation exposure was considered.
For cancer endpoints, EPA uses the term "greater than risk benchmark" as one indication for the
potential of a chemical substance to present unreasonable risk; this occurs, for example, if the lifetime
cancer risk value is 5xl0"2, which is greater than the benchmarks of lxlO"4 to lxlO"6. Conversely, EPA
uses the term "does not indicate unreasonable risk" when EPA does not have a concern for the potential
of the chemical substance to present unreasonable risk. More details are described below.
The degree of uncertainty surrounding cancer risk is a factor in determining whether or not unreasonable
risk is present. Where uncertainty is low and EPA has high confidence in the hazard and exposure
characterizations (for example, the basis for the characterizations is measured or monitoring data or a
robust model and the hazards identified for risk estimation are relevant for conditions of use), the
Agency has a higher degree of confidence in its risk determination. EPA may also consider other risk
factors, such as severity of endpoint, reversibility of effect, or exposure-related considerations such as
magnitude or number of exposures, in determining that the risks are unreasonable under the conditions
of use. Where EPA has made assumptions in the scientific evaluation and whether or not those
assumptions are protective, will also be a consideration. Additionally, EPA considers the central
tendency and high-end scenarios when determining unreasonable risk. High-end risk estimates (e.g.,
Page 229 of 352

-------
95th percentile) are generally intended to cover individuals or subpopulations with greater sensitivity or
exposure, and central tendency risk estimates are generally estimates of average or typical exposure.
Conversely, EPA may make a no unreasonable risk determination for conditions of use where the
substance's hazard and exposure potential, or where the risk-related factors described previously, lead
EPA to determine that the risks are not unreasonable.
5.1.2 Risks to Human Health
EPA estimates cancer risks by estimating the incremental increase in probability of an individual in an
exposed population developing cancer over a lifetime (excess lifetime cancer risk (ELCR)) following
exposure to the chemical under specified use scenarios. However, for asbestos, EPA used a less than
lifetime exposure calculation because the time of first exposure impacts the cancer outcome (see Section
4.2.1). Standard cancer benchmarks used by EPA and other regulatory agencies are an increased cancer
risk above benchmarks ranging from 1 in 1,000,000 to 1 in 10,000 {i.e., lxlO"6 to lxlO"4 or also denoted
as 1 E-6 to 1 E-4) depending on the subpopulation exposed. Generally, EPA considers benchmarks
ranging from lxlO"6 to lxlO"4 as appropriate for the general population, consumer users, and non-
occupational PESS.39
For the purposes of this risk determination, EPA uses lxlO"6 as the benchmark for consumers (e.g., do-
it-yourself mechanics) and bystanders. In addition, consistent with the 2017 NIOSH guidance,40 EPA
uses lxlO"4 as the benchmark for individuals in industrial and commercial work environments subject to
Occupational Safety and Health Act (OSHA) requirements. It is important to note that lxlO"4 is not a
bright line, and EPA has discretion to make risk determinations based on other benchmarks and
considerations as appropriate. It is also important to note that exposure-related considerations (e.g.,
duration, magnitude, population exposed) can affect EPA's estimates of the ELCR. When making an
unreasonable risk determination based on injury to health of workers, EPA also makes assumptions
regarding workplace practices and the implementation of the required hierarchy of controls from OSHA.
EPA assumes that feasible exposure controls, including engineering controls, administrative controls, or
use of personal protective equipment (PPE) are implemented in the workplace. EPA's decisions for
unreasonable risk to workers are based on high-end exposure estimates, in order to capture not only
exposures for PESS but also to account for the uncertainties related to whether or not workers are using
PPE
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 1.4.4.
39	As an example, when EPA's Office of Water in 2017 updated the Human Health Benchmarks for Pesticides, the
benchmark for a "theoretical upper-bound excess lifetime cancer risk" from pesticides in drinking water was identified as 1 in
1,000,000 to 1 in 10,000 over a lifetime of exposure (EPA. Human Health Benchmarks for Pesticides: Updated 2017
Technical Document. January 2017. https://www.epa.gov/sites/prodncfion/files/2015-10/docnments/Mi-benchmarks-
techdoc.pdf). Similarly, EPA's approach under the Clean Air Act to evaluate residual risk and to develop standards is a two-
step approach that includes a "presumptive limit on maximum individual lifetime [cancer] risk (MIR) of approximately 1 in
10 thousand" and consideration of whether emissions standards provide an ample margin of safety to protect public health "in
consideration of all health information, including the number of persons at risk levels higher than approximately 1 in 1
million, as well as other relevant factors" (54 FR 38044, 38045, September 14, 1989).
40	NIOSH 20.1.6). Current intelligence bulletin 68: NIOSH chemical carcinogen policy, available at
https ://www. cdc .gov/niosh/docs/2017- 100/pdf/2017-100.pdf.
Page 230 of 352

-------
5,1,3 Determining Environmental Risk
In the draft Risk Evaluation, EPA concluded that, based on the reasonably available information in the
published literature, provided by industries using asbestos, and reported in EPA databases, there were
minimal or no releases of asbestos to surface water associated with the COUs that EPA is evaluating
(see Appendix D). Therefore, EPA concluded there was no unreasonable risk to aquatic or sediment-
dwelling environmental organisms.
EPA has considered peer review and public comments on this conclusion and has decided to retain the
finding made in the draft Risk Evaluation {i.e., that there were minimal or no releases of asbestos to
surface water associated with the COUs that EPA is evaluating in this Part 1 of the risk evaluation for
asbestos). EPA is confident that the minimal water release data available and reported more fully in the
PF - and now presented again in Appendix D - cannot be attributed to chrysotile asbestos from the
COUs in this Part 1 of the risk evaluation for asbestos. Assessing possible risk to aquatic organisms
from the exposures described would not be reasonably attributed to the COUs. However, based on the
decision to develop a scope and risk evaluation for legacy uses and associated disposals of asbestos (Part
2 of the final Risk Evaluation for asbestos), EPA expects to address the issue of releases to surface water
based on those other uses. EPA did not evaluate risks to terrestrial organisms from any conditions of use
and the unreasonable risk determinations do not account for any risks to terrestrial organisms.
Additional details regarding terrestrial organism exposures are in Section 1.4.4.
5.2 Risk Determination for Chrysotile Asbestos
EPA's determination of unreasonable risk for the conditions of use of chrysotile asbestos in this Part 1
of the risk evaluation for asbestos is based on health risks to workers, occupational non-users (exposed
to asbestos indirectly by being in the same work area), consumers, and bystanders (exposed indirectly by
being in the same vicinity where consumer uses are carried out).
As described in Section 4, significant risk of cancer incidence was identified. Section 26 of TSCA
requires that EPA make decisions consistent with the "best available science" and based on "weight of
the scientific evidence." As described in EPA's framework rule for risk evaluation [82 FR 33726]
weight of the scientific evidence includes consideration of the "strengths, limitations and relevance of
the information." EPA believes that public health is best served when EPA relies upon the highest
quality information for which EPA has the greatest confidence.
The only fiber type of asbestos that EPA identified as manufactured (including imported), processed, or
distributed under the conditions of use is chrysotile asbestos, the serpentine variety. Chrysotile asbestos
is the prevailing form of asbestos currently mined worldwide. Therefore, it is reasonable to assume that
commercially available products fabricated overseas are made with chrysotile asbestos. Any asbestos
being imported into the U.S. in articles for the conditions of use EPA has identified in this document is
believed to be chrysotile asbestos. Based on EPA's determination that chrysotile asbestos is the only
form of asbestos imported into the U.S. as both raw form and as contained in articles, EPA performed a
quantitative assessment for chrysotile asbestos. The other five forms of asbestos are no longer
manufactured, imported, or processed in the United States and are now subject to a significant new use
rule (SNUR) that requires notification (via a Significant New Use Notice (SNUN)) of and review by the
Agency should any person wish to pursue manufacturing, importing, or processing crocidolite
(riebeckite), amosite (cummingtonite-grunerite), anthophyllite, tremolite or actinolite (either in raw form
or as part of articles) for any use (40 CFR 721.11095). Under the final asbestos SNUR, EPA will be
made aware of manufacturing, importing, or processing for any intended use of the other forms of
asbestos. If EPA finds upon review of a SNUN that the significant new use presents or may present an
Page 231 of 352

-------
unreasonable risk (or if there is insufficient information to permit a reasoned evaluation of the health and
environmental effects of the significant new use), then EPA would take action under TSCA section 5(e)
or (f) to the extent necessary to protect against unreasonable risk. In this Part 1 of the risk evaluation for
asbestos, EPA evaluated the following categories of conditions of use of chrysotile asbestos:
manufacturing; processing; distribution in commerce; occupational and consumer uses; and disposal.
EPA will also develop a scope document that will include other asbestos conditions of use, including
"legacy" uses (e.g., in situ building materials) and evaluation of additional fiber types (Part 2 of the final
risk evaluation for asbestos, see Preamble). As explained in the problem formulation document and
Section 1.4 of this Part 1 of the risk evaluation for asbestos, EPA did not evaluate the following:
emission pathways to ambient air from commercial and industrial stationary sources or associated
inhalation exposure of the general population or terrestrial species; the drinking water exposure pathway
for asbestos; the human health exposure pathway for asbestos in ambient water; emissions to ambient air
from municipal and industrial waste incineration and energy recovery units; on-site releases to land that
go to underground injection; or on-site releases to land that go to asbestos National Emission Standards
for Hazardous Air Pollutants (NESHAP) (40 CFR part 61, subpart M) compliant landfills or exposures
of the general population (including susceptible populations) or terrestrial species from such releases.
This Part 1 of the risk evaluation on asbestos describes the physical-chemical characteristics that are
unique to chrysotile asbestos, such as insolubility in water, suspension and duration in air,
transportability, the friable nature of asbestos-containing products, which attribute to the potential for
asbestos fibers to be released, settled, and to again become airborne under the conditions of use (re-
entrainment41). Also unique to asbestos is the impact of the timing of exposure relative to the cancer
outcome; the most relevant exposures for understanding cancer risk were those that occurred decades
prior to the onset of cancer. In addition to the cancer benchmark, the physical-chemical properties and
exposure considerations are important factors in considering risk of injury to health. To account for the
exposures for occupational non-users and, in certain cases bystanders, EPA derived a distribution of
exposure values for calculating the risk for cancer by using area monitoring data (i.e., fixed location air
monitoring results) where available for certain conditions of use and when appropriate applied exposure
reduction factors when monitoring data was not available, using data from published literature.
The risk determination for each COU in this Part 1 of the risk evaluation for asbestos considers both
central tendency and high-end risk estimates for workers, ONUs, consumers and bystanders. Where
relevant, EPA considered PPE for workers. For many of the COUs both the central tendency and high-
end risk estimates exceed the risk benchmark while some only exceeded the risk benchmark at the high-
end for each of the exposed populations evaluated. However, the risk benchmarks do not serve as a
bright line for making risk determinations and other relevant risk-related factors and EPA's confidence
in the underlying data were considered. In particular, risks associated with previous asbestos exposures
are compounded when airborne asbestos fibers settle out and again become airborne where they can
cause additional exposures and additional risks. Thus, EPA focused on the high-end risk estimates rather
than central tendency risk estimates to be most protective of workers, ONUs, consumers, and
bystanders. Additionally, as discussed in Section 4.5.3, for workers and ONUs exposed in a workplace,
EPA considered extra risks of 1 cancer per 10,000 people. At this risk level (1E-4), if the non-cancer
effects (e.g., asbestosis and pleural thickening) of chrysotile asbestos are similar to Libby amphibole
asbestos, the non-cancer effects of chrysotile asbestos are likely to contribute additional risk to the
overall health risk of chrysotile asbestos beyond the risk of cancer. Several of the COU-related
exposures evaluated for human health risks in Section 4.2 are at or greater than the point of departure
(POD) for non-cancer effects associated with exposure to Libby amphibole asbestos. Thus, the overall
health risks of chrysotile asbestos are underestimated based on cancer risk alone and support the
41 Settled Asbestos Dust Sampling and Analysis 1st Edition Steve M. Hays, James R. Millette CRC Press 1994
Page 232 of 352

-------
Agency's focus on using the high-end risk estimates rather than central tendency risk to be protective of
workers and ONUs.
The limited conditions of use of asbestos in conjunction with the extensive regulations safeguarding
against exposures to asbestos helped to focus the scope of this Part 1 of the risk evaluation for asbestos
on occupational and consumer scenarios where chrysotile asbestos in certain uses and products is
known, intended, or reasonably foreseen. EPA did not quantitatively assess each life cycle stage and
related exposure pathways as part of this document. Existing EPA regulations and standards address
exposure pathways to the general population and terrestrial species as well as exposures to chlor-alkali
industry occupational populations {i.e., workers and ONUs) for the asbestos waste pathway {e.g., the
asbestos NESHAP, particularly 40 CFR งง 61.144(a)(9), 61.150). As such, the Agency did not evaluate
these pathways.
The risk determinations are organized by conditions of use and displayed in a table format. Presented
first are those life cycle stages where EPA assumes the absence of asbestos exposure, and the conditions
of use that do not present an unreasonable risk are summarized in a table. EPA then presents the risk
determination for the chrysotile asbestos-containing brakes conditions of use for the NASA "Super
Guppy." Those conditions of use were determined not to present an unreasonable risk. The risk
determinations for the conditions of use that present an unreasonable risk are depicted in section
5.2.1 (Occupational Processing and Use of Chrysotile Asbestos) and section 5.2.2 (Consumer Uses of
Chrysotile Asbestos). For each of the conditions of use assessed in this Part 1 of the risk evaluation on
asbestos, a risk determination table is presented based on relevant criteria pertaining to each exposed
population {i.e., health only for either workers, occupational non-users, consumers, or bystanders as
indicated in table headings) is provided and explained below.
Import, Distribution in Commerce and Disposal of Chrysotile Asbestos
EPA assumed the absence of exposure to asbestos at certain life cycle stages. Raw chrysotile asbestos
and asbestos-containing products are imported into the U.S. in a manner where exposure to asbestos is
not anticipated to occur. According to information reasonably available to EPA, raw chrysotile asbestos
is imported in bags wrapped in plastic where they are contained in securely locked shipping containers.
These shipping containers remain locked until they reach the chlor-alkali plants (Enclosure B: Asbestos
Controls in the Chlor-Alkali Manufacturing Process httm://www.regulations.gov/document?I) ==EPA-
HQ-OPPT-2 736-0052). Asbestos articles (or asbestos-containing products) are assumed to be
imported and distributed in commerce in a non-friable state, enclosed in sealed boxes, where fibers are
not expected to be released.
EPA also assumes the absence of asbestos exposure during the occupational disposal of asbestos sheet
gaskets scraps during gasket stamping and the disposal of spent asbestos gaskets used in chemical
manufacturing plants. This assumption is based on the work practices followed and discussed in Section
2.3.1 that prevent the release of asbestos fibers.
Considering these exposure assumptions, EPA finds no unreasonable risk to health or the environment
for the life cycle stages of import and distribution in commerce of chrysotile asbestos for all the
conditions of use evaluated in this Part 1 of the risk evaluation on asbestos. EPA also finds no
unreasonable risk to health or the environment for occupational populations for the disposal of asbestos
sheet gaskets scraps during gasket stamping and the disposal of spent asbestos gaskets used in chemical
manufacturing plants.
Page 233 of 352

-------
In addition, there is a limited use of asbestos-containing brakes (categorized under other vehicle friction
products) for a special, large NASA transport plane (the "Super-Guppy"). (See sections 2.3.1.8.2 and
4.2.2.6). EPA calculated risk estimates using occupational exposure monitoring data provided by
NASA. EPA assumes 12 hours of brake changes occur every year starting at age 26 years with 20 years
exposure.
The Excess Lifetime Cancer Risk for Super Guppy Brake/Repair Replacement for Workers is:
Full Shift: Central Tendency - 3.7 E-7
Full Shift: High-End -1.1 E-6
Short-Term: Central Tendency - 7.3 E-7
Short-Term: High-End - 1.7 E-6
Because these risk estimates fall below the benchmark for both the central tendency and high-end and
after considering the engineering controls and work practices in place discussed in section 2.3.1.8.2,
EPA finds these COUs (import/manufacture, distribution, use and disposal) do not present an
unreasonable risk of injury to health.
Conditions of I so that Do Not Present sin I nre;is<>ii;ihie Kisk to Health or Knvironnient
•	Import of chrysotile asbestos and chrysotile asbestos-containing products
•	Distribution of chrysotile asbestos-containing products
•	Use of chrysotile asbestos brakes for a specialized, large NASA transport plane
•	Disposal of chrysotile asbestos-containing sheet gaskets processed and/or used in the industrial
setting and chrysotile asbestos-containing brakes for a specialized, large NASA transport plane
5.2.1 Occupational Processing and Use of Chrysotile Asbestos
EPA identified the following conditions of use where chrysotile asbestos is processed and/or used in
occupational settings: asbestos diaphragms in chlor-alkali industry, processed asbestos-containing sheet
gaskets, asbestos-containing sheet gaskets in chemical production, asbestos-containing brake blocks in
the oil industry, aftermarket automotive asbestos-containing brakes/ linings and other vehicle friction
products and other asbestos-containing gaskets. 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. Assigned protection factors (APFs) are provided in Table 1 under ง
1910.134(d)(3)(i)(A) (see Table 2-3) and refer to the level of respiratory protection that a respirator or
class of respirators is expected to provide to employees when the employer implements a continuing,
effective respiratory protection program. Where applicable, in the following tables, EPA provides risk
estimates with PPE using APFs derived from information provided by industry. However, there is some
uncertainty in taking this approach as based on published evidence for asbestos (see Section 2.3.1.2),
nominal APFs may not be achieved for all respirator users.
Occupational non-users (ONUs) are not expected to wear PPE since they do not directly handle the
chemical substance or articles thereof. Additionally, because ONUs are expected to be physically farther
away from the chemical substance than the workers who handle it, EPA calculated an exposure
Page 234 of 352

-------
reduction factor for ONUs based on the monitoring data {i.e., fixed location air monitoring results)
provided by industry and the information available in the published literature (see section 2.3.1.3).
As explained in Section 5.2, EPA considers the high-end risk estimates for risk to workers, occupational
non-users, consumers, and bystanders for this risk determination of chrysotile asbestos.
Table 5-1. Risk Determination for Chrysotile Asbestos: Processing and Industrial Use of Asbestos
Diaphragms in Chlor-alkali Industry (refer to section 4.2.2.1 for the risk characterization)
Criteria for Risk
Determination
Exposed Population
Workers
Occupational Non-Users
Life cycle
Stage
Processing and Industrial Use
Processing and Industrial Use
TSCA Section 6(b)(4)(A)
Unreasonable Risk
Determination
Presents an unreasonable risk of injury to health
(workers and occupational non-users).
Unreasonable Risk Driver
Cancer resulting from chronic
inhalation exposure
Cancer resulting from chronic
inhalation exposure
Benchmark (Cancer)
10"4 excess cancer risks
10"4 excess cancer risks
Risk Estimates without PPE
8-hour TWA
9.9 E-5 Central Tendency
6.8 E-4 High-end
Short Term
1.2	E-4 Central Tendency
9.4 E-5 Central Tendency3
1.3	E-3 High-end
6.7 E-4 High-enda
8-hour TWA
5.0 E-5 Central Tendency
1.6 E-4 High-end
Short Term
Not available
Risk Estimates with PPE
APF=10
8-hour TWA
9.9 E-6 Central Tendency
6.8	E-5 High-end
Short Term
1.0 E-5 Central Tendency
9.0	E-5 High-end
APF=25
8-hour TWA
3.9	E-6 Central Tendency
2.7 E-5 High-end
Short Term
4.9 E-6 Central Tendency
5.1	E-5 High-end
Not Assessed; ONUs are not assumed
to wear respirators
Risk Considerations
EPA calculated risk estimates using
occupational exposure monitoring
data provided by industry (Section
2.3.1.3). Without respiratory PPE the
high-end risk estimates exceed the
EPA calculated risk estimates using
area monitoring data (i.e., fixed
location air monitoring results)
provided by industry (Section
2.3.1.3), which supports EPA's
Page 235 of 352

-------
10x"4 risk benchmark; however,
when expected use of respiratory PPE
is considered for some worker tasks
(APF=10 and APF=25), the risk
estimates do not exceed the risk
benchmark (at the central tendency
and high-end). As depicted in Table
2-7 and documented by industry13, of
the eight asbestos-related worker
tasks, workers wear respiratory PPE
during three tasks (Asbestos
Unloading/Transport, Glovebox
Weighing and Asbestos Handling,
and Hydroblasting), but do not wear
respiratory PPE during five of the
tasks (Asbestos Slurry, Depositing,
Cell Assembly, Cell Disassembly,
and Filter Press). Although the use of
respiratory PPE during three of the
worker tasks reduces asbestos
exposure and overall risk to workers,
respiratory PPE is not worn
throughout an entire 8-hour shift. The
industry data depicted in Table 2-7
indicates workers without respiratory
PPE are exposed to asbestos fibers
where the maximum short-term PBZ
samples for three tasks (cell
assembly, cell disassembly and filter
press) are in the range of some tasks,
and higher than one task (Asbestos
unloading/Transport), where
respiratory PPE is used. Considering
that respiratory PPE is not worn for
all worker tasks where occupational
exposure monitoring data indicates
the presence of airborne asbestos
fibers, the physical-chemical
properties of asbestos and
considering the severe and the
irreversible effects associated with
asbestos inhalation exposures, these
conditions of use (for processing and
use) present unreasonable risk to
workers.
Finally, as discussed in Section 4.5.3,
for workers and ONUs exposed in a
workplace, EPA considered extra
risks of 1 cancer per 10,000 people.
expectation that ONU inhalation
exposures are lower than inhalation
exposures for workers directly
handling asbestos materials (Table 2-
8). There is some uncertainty in the
ONU exposure estimate because
much of the reported area monitoring
data were reported as "less than"
values, which may represent non-
detects. One facility did not clearly
distinguish whether measurements
were area samples or personal
breathing zone samples. EPA
considered both the high-end and
central tendency risk estimates in its
determination, and although the high-
end exceeds the cancer risk
benchmark of lxlO"4' both risk
estimates are fairly similar. Based on
the benchmarks exceedances and
considering the physical-chemical
properties of asbestos, the expected
absence of respiratory PPE, and the
severe and irreversible health effects
associated with asbestos inhalation
exposures, these conditions of use
(for processing and use) present
unreasonable risk to ONUs.
Page 236 of 352

-------
At this risk level (1E-4), if the non-
cancer effects (e.g., asbestosis and
pleural thickening) of chrysotile
asbestos are similar to Libby
amphibole asbestos, the non-cancer
effects of chrysotile asbestos are
likely to contribute additional risk to
the overall health risk of chrysotile
asbestos beyond the risk of cancer.
Thus, the overall health risks of
chrysotile asbestos are
underestimated based on cancer risk
alone and support the Agency's focus
on using the high-end risk estimates
rather than central tendency risk to be
protective of workers and ONUs.
aNo APF applied for 7.5 hours, APF of 25 applied for 30 minutes.
industry provided descriptions of the PPE used in Enclosure C: Overview of Monitoring Data and PPE Requirements
https://www.regulations.gov/document?D=EPA-HO-OPPT-2016-0736-0Q52
Table 5-2. Risk Determination for Chrysotile Asbestos: Processing Asbestos-Containing Sheet
	Gaskets (refer to section 4.2.2.2 for the risk characterization)	
Criteria for Risk
Determination
Exposed Population
Workers
Occupational Non-Users
Life cycle
Stage
Processing
Processing
TSCA Section 6(b)(4)(A)
Unreasonable Risk
Determination
Presents an unreasonable risk of injury to health
(workers and occupational non-users)
Unreasonable Risk Driver
Cancer resulting from chronic
inhalation exposure
Cancer resulting from chronic inhalation
exposure
Benchmark (Cancer)
10"4 excess cancer risks
10"4 excess cancer risks
Risk Estimates without PPE
8-hour TWA
2.8	E-4 Central Tendency
1.2 E-3 High-end
Short Term
2.9	E-4 Central Tendency
1.2 E-3 High-end
8-hour TWA
4.8 E-5 Central Tendency
2.0	E-4 High-end
Short Term
5.1	E-5 Central Tendency
2.0 E-4 High-end
Risk Estimates with PPEb
APF = 1
An APF of 1 was assigned to the
respiratory PPE provided to
workers based on industry
information b
8-hour TWA
2.8 E-4 Central Tendency
1.2 E-3 High-end
Not Assessed; ONUs are not assumed to
wear respirators
Page 237 of 352

-------

Short Term
2.9 E-4 Central Tendency
1.2 E-3 High-end

Risk Considerations
EPA calculated risk estimates
using occupational exposure
monitoring data provided by
industry and in the published
literature (Section 2.3.1.4). The
use of N95 respirators was
reported by industry51 to be worn
by a worker cutting gaskets.
However, the OSHA Asbestos
Standard 29 CFR 1910.1001 states
that such respirators should not be
used to mitigate asbestos exposure.
However, using N95 respirators
are not protective as OSHA
Asbestos Standard 29 CFR
1910.1001 prohibit the use of
filtering facepiece respirators for
protection against asbestos fibers.
Thus, the N95 respirator has an
assigned APF=1 due to
ineffectiveness as respiratory PPE
for mitigating asbestos exposure.
Absent effective respiratory PPEb
risk estimates for both central
tendency and high-end exceeds the
benchmark of lxl0"4. Based on the
benchmark exceedances and
considering the physical-chemical
properties of asbestos and the
severe and irreversible health
effects associated with asbestos
inhalation exposures, this
condition of use presents
unreasonable risk to workers.
EPA calculated risk estimates using
monitoring data provided by industry and
in the published literature. ONU
inhalation exposures are expected to be
lower than inhalation exposures for
workers directly handling asbestos
materials and based on exposure
measurements in the published literature
comparing workers to non-workers, EPA
estimated a reduction factor of 5.75 for
ONUs which was applied to the exposure
estimate for workers (Section 2.3.1.3)..
High-end risk estimates exceed the cancer
risk benchmark of lxlO"4. Based on the
benchmark exceedances and considering
the physical-chemical properties of
asbestos and the severe and irreversible
health effects associated with asbestos
inhalation exposures, this condition of use
presents unreasonable risk to ONUs.
'Industry provided description of PPE A.CC (2017a').
bRisk to workers was calculated using hypothetical respirator PPE of APF=10 and APF=25 in the risk evaluation. However,
the risk estimates based on the hypothetical APF were not used in the risk determination based on industry description of
current respiratory PPE.
Page 238 of 352

-------
Table 5-3. Risk Determination for Chrysotile Asbestos: Industrial Use of Asbestos-Containing
Sheet Gaskets in Chemical Production
(Titanium Dioxide Example is Representative of this COU; refer to section 4.2.2.3 for the risk
	characterization)	
Criteria for Risk
Exposed Population
Determination
Workers
Occupational Non-Users
Life Cycle Stage
Industrial Use
Industrial Use
TSCA Section
6(b)(4)(A)
Unreasonable Risk
Determination
Presents an unreasonable risk of injury to health
(workers and occupational non-users)
Unreasonable Risk
Driver
Cancer resulting from chronic
inhalation exposure
Cancer resulting from chronic inhalation
exposure
Benchmark (Cancer)
10"4 excess cancer risks
10"4 excess cancer risks
Risk Estimates
without PPE
8-hour TWA
5.2 E-4 Central Tendency
1.9 E-3 High-end
8-hour TWA
1.0 E-4 Central Tendency
3.2 E-4 High-end
Risk Estimates with
current PPEa
APF=10
8-hour TWA
5.2 E-5 Central Tendency
1.9 E-4 High-end
Not Assessed; ONUs are not assumed to
wear respirators
Risk Considerations
EPA calculated risk estimates using
occupational exposure monitoring
data provided by industry and in the
published literature (Section
2.3.1.5). Based on respiratory PPE
used according to industry3 EPA also
calculated the risk estimates using an
APF of 10; however, even with PPE
and considering the physical-
chemical properties of asbestos and
the severe and irreversible health
effects associated with asbestos
inhalation exposures, high-end risk
estimates for this condition of use
exceed the benchmark of lxlO"4 and
presents unreasonable risk to
workers.
EPA calculated risk estimates using
monitoring data provided by industry
and in the published literature. Based on
exposure measurements in the published
literature, EPA estimated a reduction
factor of 5.75 for ONUs (Section
2.3.1.4.). Because asbestos fibers
released during the worker activities
described in Section 2.3.1.5. EPA
considered it appropriate to use the
high-end estimate when determining
ONU risk; however, both central
tendency and high-end estimates
showed risk. Based on the central
tendency and high-end risk estimates
exceeding the benchmark of lxlO"4, the
expected absence of respiratory PPE,
the physical-chemical properties of
asbestos and the severe and irreversible
effects associated with asbestos
inhalation exposures, this condition of
use presents unreasonable risk to ONUs.
"ฆIndustry provided description of PPE ACC (2017a).
Page 239 of 352

-------
Table 5-4. Risk Determination for Chrysotile Asbestos: Industrial Use and Disposal of Asbestos-
Containing Brake Blocks in Oil Industry (refer to section 4.2.2.4 for the risk characterization)
Criteria for Risk
Determination
Exposed Population
Workers
Occupational Non-Users
Life cycle
Stage
Industrial Use and Disposal
Industrial Use and Disposal
TSCA Section
6(b)(4)(A)
Unreasonable Risk
Determination
Presents an unreasonable risk of injury to health
(workers and occupational non-users)
Unreasonable Risk
Driver
Cancer resulting from chronic
inhalation exposure
Cancer resulting from chronic
inhalation exposure
Benchmark (Cancer)
10"4 excess cancer risks
10"4 excess cancer risks
Risk Estimates without
PPE
8-hour TWA
6.0 E-4
8-hour-TWA
4.0 E-4
Risk Estimates with
PPE
APF=1
Workers are not assumed to wear
respirators
Not Assessed; ONUs are not
assumed to wear respirators
Risk Considerations
(applies to both
workers and ONUs)
The estimated exposure scenario used in the risk evaluation is based on
one 1988 study of Norway's offshore petroleum industry and relevance
to today's use of oil field brake blocks in the United States is uncertain.
EPA is aware that brake blocks are imported, distributed, and used in the
U.S. although the full extent of use could not be determined. According
to industry3, Drawworks machineries are always used and serviced
outdoors, close to oil wells. Information on processes and worker
activities are insufficient to determine the proximity of ONUs to
workers. ONU inhalation exposures are expected to be lower than
inhalation exposures for workers directly handling asbestos materials.
Although EPA has calculated a single conservative risk estimate for
workers and for ONUs, EPA does not expect routine use of respiratory
PPE. Considering the cancer risk benchmark of 1x10-4 is exceeded, the
physical-chemical properties of asbestos and the severe and irreversible
effects associated with asbestos inhalation exposures, these conditions of
use present unreasonable risk for both workers and ONUs.
a Industry provided data Popik (2018)
Page 240 of 352

-------
Table 5-5. Risk Determination for Chrysotile Asbestos: Commercial Use and Disposal of
Aftermarket Automotive Asbestos-Containing Brakes/Linings and Other Vehicle Friction
Products (Commercial Mechanic Brake Repair/Replacement is Representative for both COUs;
	refer to section 4.2.2.5 and 4.2.2.6 for the risk characterization)	
Criteria for Risk
Determination
Exposed Population
Workers
Occupational Non-Users
Life Cycle Stage
Commercial Use
Commercial Use
TSCA Section
6(b)(4)(A)
Unreasonable Risk
Determination
Presents an unreasonable risk of injury
to health
Does not present an unreasonable risk of
injury to health
Unreasonable Risk
Driver
Cancer resulting from chronic
inhalation exposure
Cancer resulting from chronic inhalation
exposure
Benchmark (Cancer)
10"4 excess cancer risks
10"4 excess cancer risks
Risk Estimates
without PPE
8-hour TWA
1.2 E-4 Central Tendency
1.9 E-3 High-end
Short Term
1.2 E-4 Central Tendency
2.8 E-3 High-end
8-hour TWA
2.0 E-5 Central Tendency
4.0 E-5 High-end
Short Term
2.0 E-5 Central Tendency
4.0 E-5 High-end
Risk Estimates with
PPE
APF = 1
Workers are not assumed to wear
respirators; Respirators only required
by OSHA if PEL exceeded.
8-hour TWA
1.2 E-4 Central Tendency
1.9 E-3 High-end
Short Term
1.2 E-4 Central Tendency
2.8 E-3 High-end
Not Assessed; ONUs are not assumed to
wear respirators
Risk Considerations
EPA calculated risk estimates based
on data provided in the published
literature and OSHA monitoring data
(Table 2-14). Although OSHA
standards require certain work
practices and engineering controls to
minimize dust, respiratory PPE is not
required unless the permissible
exposure limit (PEL) is exceeded.
With the expected absence of PPE, the
cancer benchmark is exceeded (for
both central tendency and high-end).
Based on the exceedance of the
benchmark of lxlO"4 and
consideration of the physical-chemical
EPA calculated risk estimates data
provided in the published literature.
ONU inhalation exposures are expected
to be lower than inhalation exposures for
workers. EPA estimated a reduction
factor of 8.4 (Section 2.3.1.7) for ONUs.
Because asbestos fibers released during
the worker activities described in
Section 2.3.1.7.2 can settle and again
become airborne where they can be
inhaled by ONUs, EPA considered it
appropriate to use the high-end estimate
when determining ONU risk. Based on
no exceedance of the benchmark of
lxlO"4, even with the expected absence
Page 241 of 352

-------
properties of asbestos and the severe
and irreversible effects associated with
asbestos inhalation exposures, these
conditions of use present unreasonable
risk to workers.
of respiratory PPE, these conditions of
use do not present unreasonable risk to
ONUs.
Table 5-6. Risk Determination for Chrysotile Asbestos: Commercial Use and Disposal of Other
Asbestos-Containing Gaskets
(Commercial Mechanic Gasket Repair/Replacement is Representative for this COU; refer to
	section 4.2.2.7 for the risk characterization)	
Criteria for Risk
Determination
Exposed Population
Workers
Occupational Non-Users
Life cycle
Stage
Commercial Use
Commercial Use
TSCA Section 6(b)(4)(A)
Unreasonable Risk
Determination
Presents an unreasonable risk of injury to health
(workers and occupational non-users)
Unreasonable Risk Driver
Cancer resulting from chronic
inhalation exposure
Cancer resulting from chronic
inhalation exposure
Benchmark (Cancer)
10"4 excess cancer risks
10"4 excess cancer risks
Risk Estimates without
PPE
8-hour TWA
4.8 E-4 Central Tendency
1.3 E-3 High-end
8-hour TWA
1.0 E-4 Central Tendency
3.0 E-4 High-end
Risk Estimates with PPE
APF=1
Workers are not assumed to wear
respirators
8-hour TWA
4.8 E-4 Central Tendency
1.3 E-3 High-end
Not Assessed; ONUs are not assumed
to wear respiratory PPE.
Risk Considerations
EPA calculated risk estimates using
exposure scenarios based on
occupational monitoring data
(breathing zone of workers) for
asbestos-containing gasket
replacement in vehicles. Although,
risk to workers was calculated using
hypothetical respirator PPE of
APF=10 and APF=25, workers are
not expected to wear respiratory PPE
during gasket repair and replacement
in a commercial setting. Based on the
expected absence of PPE and the
benchmark of lxlO"4 is exceeded (for
both central tendency and high-end),
EPA calculated risk estimates using
exposure scenarios based on
occupational monitoring data (work
area samples in the vicinity of the
workers) for asbestos-containing
gasket replacement in vehicles. EPA
estimated a reduction factor of 5.75
(Section 2.3.1.9) for ONUs. Due to
the severe and irreversible effects
associated with asbestos inhalation
exposures and that asbestos fibers
released during the worker activities
described in Section 2.3.1.9 can settle
and again become airborne where
they can be inhaled by ONUs, EPA
Page 242 of 352

-------
the physical-chemical properties of
asbestos and the severe and
irreversible effects associated with
asbestos inhalation exposures, these
conditions of use present
unreasonable risk to workers.
considered it appropriate to use the
high-end estimate when determining
ONU risk; however both central
tendency and high-end risk estimates
presented risk. Based on the
exceedance of the benchmark of
lxlO"4 (for both central tendency and
high-end), and the expected absence
of respirators, the physical-chemical
properties of asbestos and the
potential severity of effect associated
with inhalation exposures to asbestos,
these conditions of use present
unreasonable risk to ONUs.
5.2.2 Consumer Uses of Chrysotile Asbestos
The consumer uses of asbestos include aftermarket automotive asbestos-containing brakes/linings, and
other asbestos-containing gaskets. Consumers and bystanders are not assumed to wear respiratory PPE.
Therefore, EPA did not assess risk estimates with PPE for the conditions of use for these exposed
populations.
Table 5-7. Risk Determination for Chrysotile Asbestos: Consumer Use and Disposal of
Aftermarket Automotive Asbestos-Containing Brakes/Linings
(Do-it-Yourself Consumer Brake Repair/Replacement is Representative for both COUs; refer to
	section 4.2.3.1 for the risk characterization)	
Criteria for Risk
Determination
Exposed Population
Do-it-Yourself Mechanic
Bystander
Life cycle
Stage
Consumer Use and Disposal
Consumer Use and Disposal
TSCA Section
6(b)(4)(A)
Unreasonable Risk
Determination
Presents an unreasonable risk of injury to health
(consumers and bystanders)
Unreasonable Risk
Driver
Cancer resulting from chronic
inhalation exposure
Cancer resulting from chronic inhalation
exposure
Benchmark (Cancer)
10"6 excess cancer risks
10"6 excess cancer risks
Risk Estimates
without PPE
Indoor, compressed air
Indoor, compressed air
1 hour/day; once every 3 years for
62 years (starting age 16)
Exposures at 30% of active used
between uses, 1 hour/d in garage
3.6 E-5 Central Tendency
3.5 E-4 High-end
1 hour/day; once every 3 years for 62
years (starting age 16)
Exposures at 30% of active used
between uses, 1 hour/d in garage
2.6 E-5 Central Tendency
6.0 E-5 High-end
Page 243 of 352

-------

Indoor, compressed air
Indoor, compressed air
8-hour/day; once every 3 years for
62 years (starting age 16)
Exposures at 30% of active used
between uses, 8 hours/d in garage
2.6 E-4 Central Tendency
2.6 E-3 High-end
8-hour/day; once every 3 years for 62
years (starting age 16)
Exposures at 30% of active used
between uses, 8 hours/d in garage
1.7 E-5 Central Tendency
3.9 E-5 High-end
Indoor, compressed air
Indoor, compressed air
Indoor, compressed air, once at 16
years, staying in residence for 10
years, exposures at 10% of active
used between uses, 1 hour/d in
garage
5.4 E-6 Central Tendency
5.3 E-5 High-end
Indoor, compressed air, once at 16
years, staying in residence for 10
years, exposures at 10% of active used
between uses, 1 hour/d in garage
3.4 E-6 Central Tendency
7.8 E-6 High-end
Outdoor
Once every 3 years for 62 years
(starting age 16)
Exposures at 2% of active used
between uses, 5 min/d in driveway
8.2 E-8 Central Tendency
4.4 E-7 High-end
Outdoor
Once every 3 years for 62 years
(starting age 16)
Exposures at 2% of active used between
uses, 5 min/d in driveway
2.1 E-8 Central Tendency
1.1 E-7 High-end
Outdoor
Once every 3 years for 62 years
(starting age 16)
Exposures at 2% of active used
between uses, 30 min/d in driveway
2.4 E-7 Central Tendency
1.3 E-6 High-end
Outdoor
Once every 3 years for 62 years
(starting age 16)
Exposures at 2% of active used between
uses, 30 min/d in driveway
5.9 E-8 Central Tendency
3.2. E-7 High-end
Risk Estimates with
PPE
Not Assessed; Consumers are not
assumed to wear respiratory PPE
Not Assessed; Bystanders are not
assumed to wear respiratory PPE
Risk Considerations
EPA calculated risk estimates are
based on data provided in the
published literature and surrogate
monitoring data from occupational
brake repair studies. EPA considered
4 different exposure scenarios with
different assumptions on the duration
of exposure, whether indoors in a
garage using compressed air or
outside without compressed air.
Although DIY brake and clutch work
is more likely to occur outdoors, it
may also occur inside a garage.
Additionally, considering that many
DIY mechanics have access to air
EPA calculated risk estimates are based
on data provided in the published
literature and surrogate monitoring data
from occupational brake repair studies.
No reduction factor was applied for
indoor DIY brake work inside
residential garages due to the expected
close proximity of bystanders inside a
garage. In the absence of data to
estimate a reduction factor for outdoor
brake work, EPA assumed a reduction
factor of 10 (Section 2.3.2.1). Because
asbestos fibers released during the DIY
(consumer) activities described in
Section 2.3.2.1 can settle and again
Page 244 of 352

-------
compressors, EPA expects that at
least some DIY mechanics may use
compressed air to clean dust from
brakes or clutches and can spend up
to a full day (8-hours) in their garage
and working three hours specifically
on brakes and clutches. Because
asbestos fibers released during the
DIY (consumer) activities described
in Section 2.3.2.1 can settle and again
become airborne where they can be
inhaled by bystanders, EPA
considered it appropriate to use the
high-end estimate when determining
consumer risk. EPA chose a
conservative and protective brake and
clutch repair/replacement exposure
scenario of 3 hours/day once every 3
years inside a garage using
compressed air to account for the
possibility that some DIY mechanics
may fit this exposure scenario. EPA
also used a less conservative brake
and clutch repair/replacement
exposure scenario of once in a
lifetime, 1 hour per day, while inside
a garage, using compressed air. As
part of the analysis, EPA made some
assumptions regarding both age at the
start of exposure and the duration of
exposure. Realizing there is
uncertainty around these assumptions,
EPA developed a sensitivity analysis
approach specifically for the
consumer exposure/risk analysis (see
Section 4.3.7 and Appendix L.) Under
the chosen indoor exposure scenarios,
the cancer benchmark is exceeded
(both central tendency and high-end).
Based on the benchmark exceedances
and considering the physical-chemical
properties of asbestos and the severe
and irreversible health effects
associated with asbestos inhalation
exposures, this condition of use
presents unreasonable risk
unreasonable risk to consumers.
become airborne where they can be
inhaled by bystanders, EPA considered
it appropriate to use the high-end
estimate when determining bystander
risk. EPA also chose a conservative and
protective brake repair/replacement
exposure scenario of 3 hours/day while
inside a garage up to 8-hours once every
3 years, using compressed air to account
for the possibility that some bystanders
(e.g., children watching parents) may fit
this exposure scenario. EPA also used a
less conservative brake and clutch
repair/replacement exposure scenario of
once in a lifetime, 1 hour per day, while
inside a garage, using compressed air.
As part of the analysis, EPA made some
assumptions regarding both age at the
start of exposure and the duration of
exposure. Realizing there is uncertainty
around these assumptions, EPA
developed a sensitivity analysis
approach specifically for the bystander
exposure/risk analysis (see Section 4.3.7
and Appendix L.) Based on the
exceedance (both central tendency and
high-end) of the benchmark of lxlO"6 for
the chosen indoor exposure scenarios,
the expected absence of respiratory PPE,
the physical-chemical properties of
asbestos and the potential severity of
effects associated with inhalation
exposures to asbestos, these conditions
of use present unreasonable risk to
bystanders.
Page 245 of 352

-------
Table 5-8. Risk Determination for Chrysotile Asbestos: Consumer Use and Disposal of Other
Asbestos-Containing Gaskets
(Do-it-Yourself Consumer Gasket Repair/Replacement is Representative for this COU; refer to
section 4.2.3.2 for the risk characterization)	
Criteria for Risk
Exposed Population
Determination
Do-it-Yourself Mechanic
Bystander
Life cycle
Stage
Consumer Use and Disposal
Consumer Use and Disposal
TSCA Section
6(b)(4)(A)
Unreasonable Risk
Determination
Presents unreasonable risk of injury to health
(consumers and bystanders)
Unreasonable Risk
Driver
Cancer resulting from chronic inhalation
exposure
Cancer resulting from chronic inhalation
exposure
Benchmark (Cancer)
10"6 excess cancer risks
10"6 excess cancer risks

Indoor
1 hour/day; once every 3 years for 62
years (starting age 16)
Exposures at 30% of active used
between uses, 1 hour/d in garage
1.9 E-5 Central Tendency
5.3 E-5 High-end
Indoor
1 hour/day; once every 3 years for 62
years (starting age 16)
Exposures at 30% of active used between
uses, 1 hour/d in garage
2.4 E-5 Central Tendency
6.1 E-5 High-end
Risk Estimates
without PPE
Indoor
8-hour/day; once every 3 years for 62
years (starting age 16)
Exposures at 30% of active used
between uses, 8 hours/d in garage
1.5 E-4 Central Tendency
4.2 E-4 High-end
Indoor
8-hour/day; once every 3 years for 62
years (starting age 16)
Exposures at 30% of active used between
uses, 8 hours/d in garage
2.4 E-5 Central Tendency
6.1 E-5 High-end

Indoor
Indoor, compressed air, once at 16
years, staying in residence for 10
years, exposures at 10% of active used
between uses, 1 hour/d in garage
2.9 E-6 Central Tendency
8.0 E-6 High-end
Indoor
Indoor, compressed air, once at 16
years, staying in residence for 10 years,
exposures at 10% of active used between
uses, 1 hour/d in garage
3.2 E-6 Central Tendency
7.9 E-6 High-end
Risk Estimates with
PPE
Not Assessed; Consumers are not
assumed to wear respiratory PPE
Not Assessed; Bystanders are not assumed
to wear respiratory PPE
Risk Considerations
EPA assumed that the duration of gasket
repair activity was 3 hours a day and that
a DIY mechanic is likely to perform one
gasket repair once every 3 years and can
spend up to a full day (8-hours) in their
garage. This scenario assumes all the
EPA assumed that the duration of
bystander exposure was 1 hour a day once
every 3 years. EPA also presents a less
conservative gasket repair/replacement
exposure scenario of 1 hour a day, once in
a lifetime gasket repair/replacement at age
Page 246 of 352

-------
work is conducted indoors (within a
garage) and both the consumer and
bystander remain in the garage for the
entirety of the work. EPA presents this
conservative and protective gasket
repair/replacement exposure scenario
approach to account for the possibility
that some DIY mechanics may fit this
exposure scenario. EPA also presents a
less conservative gasket
repair/replacement exposure scenario of
1 hour a day, once in a lifetime gasket
repair/replacement at age 16. EPA made
some assumptions regarding both age at
the start of exposure and the duration of
exposure. Realizing there is uncertainty
around these assumptions, EPA
developed a sensitivity analysis
approach specifically for the consumer
exposure/risk analysis (see Section 4.3.7
and Appendix L.) Because asbestos
fibers released during the DIY activities
described in Section 2.3.2.2, can settle
and again become airborne where they
can be inhaled EPA considered it
appropriate to use the high-end estimates
when determining consumer risk. Based
on the exceedance of the benchmark of
lxlO"6, at both the central tendency and
high-end estimates, the expected absence
of respiratory PPE, the physical-
chemical properties of asbestos and the
severe and irreversible effects associated
with asbestos inhalation exposures, these
conditions of use present unreasonable
risk to consumers.
16. EPA made some assumptions
regarding both age at the start of exposure
and the duration of exposure. Realizing
there is uncertainty around these
assumptions, EPA developed a sensitivity
analysis approach specifically for the
consumer exposure/risk analysis (see
Section 4.3.7 and Appendix L.) Because
asbestos fibers released during the DIY
activities described in Section 2.3.2.2 can
settle and again become airborne where
they can be inhaled by bystanders, EPA
considered it appropriate to use the high-
end estimate when determining bystander
risk. Based on the exceedance of the
benchmark of lxlO"6, at both the central
tendency and high-end estimates, the
expected absence of respiratory PPE, the
physical-chemical properties of asbestos
and the severe and irreversible effects
associated with asbestos inhalation
exposures, these conditions of use present
unreasonable risk to bystanders.
5.3 Unreasonable Risk Determination Conclusion
5.3.1 No Unreasonable Risk Determinations
TSCA Section 6(b)(4) requires EPA to conduct risk evaluations to determine whether chemical
substances present unreasonable risk under their conditions of use. In conducting risk evaluations, "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, either in a single
decision document or in multiple decision documents." " 40 CFR 702.47. Pursuant to TSCA section
6(i)(l), a determination of "no unreasonable risk" shall be issued by order and considered to be final
agency action. Under EPA's implementing regulations, "[a] determination by EPA that the chemical
Page 247 of 352

-------
substance, under one or more of the conditions of use within the scope of the risk evaluations, does not
present an unreasonable risk of injury to health or the environment will be issued by order and
considered to be a final Agency action, effective on the date of issuance of the order." 40 CFR
702.49(d).
EPA has determined that the following conditions of use of chrysotile asbestos do not present an
unreasonable risk of injury to health or the environment:
•	Import of chrysotile asbestos and chrysotile asbestos-containing products (Section5.2, Section 4,
Section 3, Section 2)
•	Distribution of chrysotile asbestos-containing products (Section 5.2, Section 4, Section 3,
Section 2)
•	Use of chrysotile asbestos brakes for a specialized, large NASA transport plane (Section 5.2,
Section 4.2.2.6, Section 3, Section 2.3.1.8.2)
•	Disposal of chrysotile asbestos-containing sheet gaskets processed and/or used in the industrial
setting and asbestos-containing brakes for a specialized, large NASA transport plane (Section
5.2, Section 4.2.2.6, Section 3, Section 2.3.1.8.2)
This subsection of the Part 1 of the risk evaluation for asbestos therefore constitutes the order required
under TSCA Section 6(i)(l), and the "no unreasonable risk" determinations in this subsection are
considered to be final agency action effective on the date of issuance of this order. All assumptions that
went into reaching the determinations of no unreasonable risk for these conditions of use, including any
considerations excluded for these conditions of use, are incorporated into this order.
The support for each determination of "no unreasonable risk" is set forth in Section 5.2, "Risk
Determination for Chrysotile Asbestos." This subsection also constitutes the statement of basis and
purpose required by TSCA Section 26(f).
5.3.2 Unreasonable Risk Determinations
EPA has determined that the following conditions of use of chrysotile asbestos present an unreasonable
risk of injury:
•	Processing and Industrial use of Chrysotile Asbestos Diaphragms in the Chlor-alkali Industry
•	Processing and Industrial Use of Chrysotile Asbestos-Containing Sheet Gaskets in Chemical
Production
•	Industrial Use and Disposal of Chrysotile Asbestos-Containing Brake Blocks in Oil Industry
•	Commercial and Consumer Use and Disposal of Aftermarket Automotive Chrysotile Asbestos-
Containing Brakes/Linings
•	Commercial Use and Disposal of Other Chrysotile Asbestos-Containing Vehicle Friction
Products
•	Commercial and Consumer Use and Disposal of Other Chrysotile Asbestos-Containing Gaskets
EPA will initiate TSCA Section 6(a) risk management actions on these conditions of use as required
under TSCA Section 6(c)(1). Pursuant to TSCA Section 6(i)(2), the "unreasonable risk" determinations
for these conditions of use are not considered final agency action.
Page 248 of 352

-------
5.4 Risk Determination for Five other Asbestiform Varieties
For the asbestos risk evaluation, EPA adopted the TSCA Title II definition of asbestos which includes
the varieties of six fiber types - chrysotile (serpentine), crocidolite (riebeckite), amosite
(cummingtonite-grunerite), anthophyllite, tremolite or actinolite. In this Part 1 of the risk evaluation for
asbestos, EPA only assessed conditions of use for chrysotile asbestos. The Agency will evaluate legacy
uses and associated disposals and other fiber types of asbestos in Part 2 of the risk evaluation for
asbestos. Part 2 will begin with a draft scope document (see Figure P-l in the Preamble). Those legacy
uses could include the other five asbestiform varieties included in the TSCA Title II definition. As such,
risk determinations for conditions of use that include those asbestiform varieties would be made in a
subsequent document.
Page 249 of 352

-------
6 REFERENCES
Abundo. ML; Almaguer. D; Driscoll. R. (1994). Health Hazard Evaluation Report HETA 93-1133-2425,
Electrode Corporation, Chardon, Ohio (pp. 34). (NTIS/02994339_a). ACC. (2017b). Use of
Non-Friable Asbestos Containing Gaskets in Titanium Dioxide Manufacturing. Information
submitted by email to EPA. October 30, 2017. ACC (American Chemistry Council).
ACC. (2018). Responses to EPA Questions Regarding Asbestos-Containing Gaskets. Information
submitted by email to EPA. September 27, 2018. Washington, DC: ACC (American Chemistry
Council).
Acheson. ED; Gardner. MI; Pippg	Ime. LP. (1982). Mortality of two groups of women who
manufactured gas masks from chrysotile and crocidolite asbestos: a 40-year follow-up. Occup
Environ Med 39: 344-348.
Addison. WE; Neal. GH; Whi	(1966). Amphi-boles—Part IV—Surface properties of amosite and
crocido-lite. J Am Chem Soc A: 79-81.
Antao. VC; Pinheir	ssell. IT. (2009). Asbestosis mortality in the USA: facts and predictions.
Occup Environ Med 66: 335-338. http://dx.doi.c	5/oem.2008.039172.
Arias. E; Heron. M; Xu. J. (2017). United States life tables, 2013. National Vital Statistics Reports; vol
66 no 3. Hyattsville, MD: National Center for Health Statistics.
strong. BG. (1998). Effect of measurement error on epidemiological studies of environmental and
occupational exposures. Occup Environ Med 55: 651-656.
http://dx.doi.oo	)em.55.10.651.
PR. (2001a). Toxicological profile for asbestos (update). (CIS/03/00067). Atlanta, Georgia: U.S.
Department Of Health And Human Services, Public Health Service, Agency for Toxic
Substances and Disease Registry, http://www.atsdr.cdc.gov/toxprofiles/index.asp.
DR. (2001b). Toxicological Profile for Asbestos (Update, September 2001). 440.
AWWA. (2019). RE: Asbestos; Significant New Use Rule, Docket No. EPA-HQ-OPPT-2018-0159.
Washington, DC: American Water Works Association
Axiall-Westlake. (2017). Responses to EPA Questions on Asbestos Use in Chlor-alkali Plants. Axiall-
Westlake (Westlake Chemical).
Badollet. MS. (1951). Asbestos, a mineral of unparalleled properties. 54: 151-160.
Bales. RC; Morgan. JJ. (1985). SURFACE-CHARGE AND ADSORPTION PROPERTIES OF
CHRYSOTILE ASBESTOS IN NATURAL-WATERS. Environ Sci Technol 19: 1213-1219.
Bel anger. SE; Cher	ims. J. (1990). Functional and pathological impairment of japanese
medaka (oryzias-latipes) by long-term asbestos exposure. Aquat Toxicol 17: 133-154.
Belanger. SE; Chem--. DS; Cairns J. J. R. (1986a). Seasonal behavioral and growth changes of juvenile
Corbicula-fluminea exposed to chrysotile asbestos. Water Res 20: 1243-1250.
Belanger. SE; Cher	ims J. J. R. (1986b). Uptake of chrysotile asbestos fibers alters growth and
reproduction of Asiatic clams. Can J Fish Aquat Sci 43: 43-52. http://dx.doi.<	9/f86-
006.
Belanger. SE; Cher	ims. J; Mcguire. MI. (1987). Using Asiatic clams as a biomonitor for
chrysotile asbestos in public water supplies. J Am Water Works Assoc 79: 69-74.
http://dx.doi.org/10.1002/i. 1551-8833.19^ ' tl^2817.x.
Belanger. SE; Schurr. K; Attei iohara. AF. (1986c). Effects of chrysotile asbestos on coho salmon
and green sunfish: evidence of behavioral and pathological stress. Environ Res 39: 74-85.
Berman. DW. (2010). Comparing milled fiber, Quebec ore, and textile factory dust: has another piece of
the asbestos puzzle fallen into place [Review], CritRev Toxicol 40: 151-188.
http://dx.doi.org/] 0.3109/104084409'
Berman. DW; Crump. KS. (2008a). A meta-analysis of asbestos-related cancer risk that addresses fiber
size and mineral type. Crit Rev Toxicol 38: 49-73.
http://dx.doi.on 80/10408440802273156.
Page 250 of 352

-------
Berman. DW; Crump. KS. (2008b). Update of potency factors for asbestos-related lung cancer and
mesothelioma. CritRev Toxicol 38: 1-47. http://dx.doi.org. 10 l08Q/1040844080J.!7-> lo"/.
Blair. A; Stewart. P; Lubin. JH; Forastiere. F. (2007). Methodological issues regarding confounding and
exposure misclassification in epidemiological studies of occupational exposures [Review], Am J
Ind Med 50: 199-207. http://dx.doi.org/10.1002/aiim..2.02.81.
Blake. CL; Dotson. GS: Harbison. (2006). Assessment of airborne asbestos exposure during the
servicing and handling of automobile asbestos-containing gaskets. Regul Toxicol Pharmacol 45:
214-222. http://dx.doij	h.2006.04.007.
Blake. CL; Dotson. GS; Harbison. RD. (2008). Evaluation of asbestos exposure within the automotive
repair industry: a study involving removal of asbestos-containing body sealants and drive clutch
replacement. Regul Toxicol Pharmacol 52: 324-331.
http://dx.doi.on	.yrtph.2.008.09.001.
Blake. CL; Van Orden. PR; Banasik. M; Harbison. RD. (2003). Airborne asbestos concentration from
brake changing does not exceed permissible exposure limit. Regul Toxicol Pharmacol 38: 58-70.
http://dx.doi.on	50273-2300(03)00069-2.
Boelt'H i Simmons. C; Heweu 3> (2011). Exposure data from multi-application, multi-industry
maintenance of surfaces and joints sealed with asbestos-containing gaskets and packing. J Occup
Environ Hyg 8: 194-209. http://dx.doi.org/10.1080/15459624.2011.557936.
Boell ; Crawford. G:.N; Podraza. DM. (2002). Airborne fiber exposure assessment of dry asbestos-
containing gaskets and packings found in intact industrial and maritime fittings. AIHA J 63: 732-
740.
Branham. (2018). Responses to Questions for Branham Site Visit. Information submitted to EPA in
advance of the Branham site visit. Available online at (accessed
Brown. DP; Dement. JM; Okun. A. (1994). Mortality patterns among female and male chrysotile
asbestos textile workers. J Occup Med 36: 882-888.
CalEPA. (2003). Public health goal for Asbestos in drinking water. Sacramento, CA: Office of
Environmental Health Hazard Assessment, California Environmental Protection Agency.
https://oeh.ha.ca. gov/media/down.loads/water/chemicals/ph.g/ph4asbestos926( If.
Camargo. MC; Stavner I * Muitf. K; Reina U U Mptu 1 I K'men 1' 1 andrigan. PI. (2011).
Occupational exposure to asbestos and ovarian cancer: a meta-analysis. Environ Health Perspect
119: 1211-1217. http://dx.doi.orgf h' I _\9/eln> * < ^ L
Camidge. PR; Stockton. PL; Bain. M. (2006). Factors affecting the mesothelioma detection rate within
national and international epidemiological studies: Insights from Scottish linked cancer registry-
mortality data. Br J Cancer 95: 649-652. http://dx.doi.org/10.1038/si.bic.66Q3293.
CDC. (2009). Malignant mesothelioma mortality: United States, 1999—2005. MMWR Morb Mortal
Wkly Rep 58: 393-396.
Choi. I; Smith. RW. (1972). Kinetic study of dissolution of asbestos fibers in water. J Colloid Interface
Sci 40. http://dx.doi.on	0021-9797(72)90014-8.
Cohen. HI; Van Orden. DR. (2008). Asbestos exposures of mechanics performing clutch service on
motor vehicles. J Occup Environ Hyg 5: 148-156.
http://dx.doi.on 80/15459620701853227.
Cole. SR; Richardson. DB; Chu. H; Naimi \l_ (2013). Analysis of occupational asbestos exposure and
lung cancer mortality using the g formula. Am J Epidemiol 177: 989-996.
http://dx.doi.on 93/aie/kws343.
Coot	eehy. JW; O'Brien. DM; Mcglothlin. JP; Todd. WF. (1987). In-Pepth Survey Report:
Evaluation of Brake Prum Service Controls at United States Postal Service Vehicle Maintenance
Facility, Louisville, Kentucky, Report No. CT-152-11B (pp. 152-111). (NIOSH/OO177547).
Cooper. TC; Sheehy. JW; O'Brien. DM; McGlothlin. JP; Todd. WF. (1988). In-depth survey report:
Evaluation of brake drum service controls at Cincinnati Gas and Electric Garages, Cincinnati,
Page 251 of 352

-------
Evanston, and Monroe, Ohio and Covington, Kentucky. (ECTB 152-22b). Cincinnati, OH:
NIOSH.
Courtice. MN: Wang. X; Lin. S; Yu. IT; Berman. DW; Yano. E. (2016). Exposure-response estimate for
lung cancer and asbestosis in a predominantly chrysotile-exposed Chinese factory cohort. Am J
Ind Med 59: 369-378. http://dx.doi.org/10.1002/aiim..2.2579.
Cowan. DM; Cheng. TJ; Ground. M; Sahmel. J; Varughese. A; Madl. AK. (2015). Analysis of
workplace compliance measurements of asbestos by the U.S. Occupational Safety and Health
Administration (1984-2011). Regul Toxicol Pharmacol 72: 615-629.
http://dx.doi.ofg/~i 0.1016/i.yrtph.201 ^ 0 002.
Davidson. B. (2011). The diagnostic and molecular characteristics of malignant mesothelioma and
ovarian/peritoneal serous carcinoma [Review], Cytopathology 22: 5-21.
http://dx.doi.on	003.2010.00829.x.
Dement. JM; Browi (1994). Lung cancer mortality among asbestos textile workers: a review and
update. Ann Occup Hyg 38: 525-532, 412.
Dement. JM; Brown. DP; Okun. A. (1994). Follow-up study of chrysotile asbestos textile workers:
cohort mortality and case-control analyses. Am J Ind Med 26: 431-447.
Dement. JM; Harris. RL. Jr; Symons. MJ; Shy. CM. (1983a). Exposures and mortality among chrysotile
asbestos workers: Part 1: Exposure estimates. Am J Ind Med 4: 399-419.
http://dx.doi.oo 02/aiim.4700040303.
Dement. JM; Harris. RL. Jr; Symons. MJ; Shy. CM. (1983b). Exposures and mortality among chrysotile
asbestos workers: Part II: Mortality. Am J Ind Med 4: 421-433.
http://dx.doi.oo 02/aiim.4700040304.
Dement. JM; Kuempel. ED; Zumwalde. RD; Smith. ' v ^aynci 1 1 j__< >omis. D (2008). Development
of a fibre size-specific job-exposure matrix for airborne asbestos fibres. J Occup Environ Med
65: 605-612. http://dx.doi.org/ tO I j;6/oem..I>ป ^
Dement. JM; Myers. D; Loomis. D; Richardson. D; Wolf. S. (2009). Estimates of historical exposures
by phase contrast and transmission electron microscopy in North Carolina USA asbestos textile
plants. Occup Environ Med 66: 574-583. http://dx.doi.cn^ U' 11 'o/oem.2008 0 10 I tO.
Deng. O; Wang. X; Wang. M; Lan. Y. (2012). Exposure-response relationship between chrysotile
exposure and mortality from lung cancer and asbestosis. Occup Environ Med 69: 81-86.
http://dx.doi.oo	)em.2011.064899.
Dodic Fikfak. M. (2003). The amphibole hypothesis—a nested case-control study of lung cancer and
exposure to chrysotile and amphiboles. Arh Hig Rada Toksikol 54: 169-176.
Dodic Fikfak. M; Kriebet. D; Ouinn. MM; Eisen. EA; Wegman. PH. (2007). A case control study of
lung cancer and exposure to chrysotile and amphibole at a Slovenian asbestos-cement plant. Ann
Occup Hyg 51: 261-268. http://dx.doi.ore j3/an.nhyg/m.em003.
Donovan. EP; Donovan. BL; Sahmel. J; Scott. PK; Paustenbac (2011). Evaluation of bystander
exposures to asbestos in occupational settings: a review of the literature and application of a
simple eddy diffusion model [Review], CritRev Toxicol 41: 52-74.
http://dx.doi.Org/10.3109/10408444.2 ;6639.
EC. (2014). S1NPHON1E: Schools Indoor Pollution and Health Observatory Network in Europe.
(JRC91160). Luxembourg: European Union, http://dx.doi.org/10.2788/99220.
ECHA. (2014). Committee for Risk Assessment (RAC) opinion on an Annex XV dossier proposing
restrictions on chrysotile. Helsinki, Finland. https://echa.europa.eu/documents/l 0162/70F519ad~
84dc Sett ฐe09 ป "H. b6dd3.
Edwards. JK; Cole. SR; Chu. H; Olshan.	chardson. DB. (2014). Accounting for outcome
misclassification in estimates of the effect of occupational asbestos exposure on lung cancer
death. Am J Epidemiol 179: 641-647.
Page 252 of 352

-------
EHM (Environmental Health Management). (2013). Report for OSHA Compliance Monitoring at
Branham Corporation. Prepared by EHM. Prepared for Branham Corporation. January 4, 2013.
Louisville, KY: Branham Corporation.
Elliott. L; Loomis. D; Dement. J; Hein. MI; Richardson. D; Stayner. L. (2012). Lung cancer mortality in
North Carolina and South Carolina chrysotile asbestos textile workers. Occup Environ Med 69:
385-390. http://dx.dou >	ซ6/oemet!-201 I 100229.
Favero-Longo. SE; Turci. F; Tomatis. M; Castelli. D; Bonfante. P; Hochella. MF; Piervittori. R; Fubini.
(2005). Chrysotile asbestos is progressively converted into a non-fibrous amorphous material
by the chelating action of lichen metabolites. J Environ Monit 7: 764-766.
http://dx.doi.oo	)507569f.
Finlev BI , Richter. RO; Mowat FS: Mlynarek. S: Paustenbach. DJ; Warmerdam. JM; Sheehan
(2007). Cumulative asbestos exposure for US automobile mechanics involved in brake repair
(circa 1950s-2000). J Expo Sci Environ Epidemiol 17: 644-655.
http://dx.doi.on	si.ies.7500553.
Fowler. DP. (2000). Exposures to asbestos arising from bandsawing gasket material. Appl Occup
Environ Hyg 15: 404-408. http://dx.doi.org/10.1080/1047322C
Fubit	n. CO. (1999). Chemical aspects of the toxicity of inhaled mineral dusts. Chem Soc Rev
28: 373-381. http://dx.doi.org/10.1039/a805639k.
Gardner. Ml: Powell. CA. (1986). Mortality of asbestos cement workers using almost exclusively
chrysotile fibre. J Soc Occup Med 36: 124-126.
Gazzano. E; Foresti. E; Leset. IG: Tomatis. M; Riganti. C; Fubini. B; Ro\	igo. D. (2005).
Different cellular responses evoked by natural and stoichiometric synthetic chrysotile asbestos.
Toxicol Appl Pharmacol 206: 356-364.
Germani. D; Belli. S: Bruno. C; Grignoli. M; Nesti. M; Pirastu. R; Comba. P. (1999). Cohort mortality
study of women compensated for asbestosis in Italy. Am J Ind Med 36: 129-134.
G	adiiska. MB; Xiang. OH; Mason. RP. (1998). In vivo evidence of free radical formation
after asbestos instillation: an ESR spin trapping investigation. Free Radic Biol Med 24: 11-17.
Gibbs. GW. (1994). The assessment of exposure in terms of fibres. Ann Occup Hyg 38: 477-487, 409-
410.
Gibbs. GW; Lachance. M. (1972). Dust exposure in chrysotile asbestos mines and mills of Quebec. Arch
Environ Health 24: 189-197. http://dx.doi.org/10.1080/00Q398 2.10666068.
Godbev. FW; Cooper. TC; Sheehy. JW; O'Brien. DM; Van Wagenen. HP; Mcglothlin. ID; Todd. WF.
(1987). In-Depth Survey Report: Evaluation of Brake Drum Service Controls at United States
Postal Service, Vehicle Maintenance Facility, Nashville, Tennessee, Report No. CT-152-20B
(pp. 152-120). (NIOSH/00175415).
Gronow. JR. (1987). The dissolution of asbestos fibres in water. Clay Miner 22: 21-35.
http://dx.doi.on 80/clavmin. 1987.02.2.1.03.
Grosse. Y; Loomis. D; Guvton. KZ; Lauby-Secretai _ Ghissassi. F; Bouvard. V; Benbrahim-Tallaa.
L; Guha. N; Scoccianti. C; Mattock. H; Strail K; International Agency for Research on Cancer
Monograph Workit (2014). Carcinogenicity of fluoro-edenite, silicon carbide fibres and
whiskers, and carbon nanotubes. Lancet Oncol 15: 1427-1428.
Gulumian. M. (1999). The role of oxidative stress in diseases caused by mineral dusts and fibres: current
status and future of prophylaxis and treatment [Review], Mol Cell Biochem 196: 69-77.
http://dx.doi.Org/http://dx.doi.org/10.1023/A: 1006918212866.
Hein. MJ; Stayr	hman. E; Dement. JM. (2007). Follow-up study of chrysotile textile workers:
Cohort mortality and exposure-response. Occup Environ Med 64: 616-625.
http://dx.doi.on	3em.2006.031005.
Hodgson. JT; Darnton. A. (2000). The quantitative risks of mesothelioma and lung cancer in relation to
asbestos exposure. Ann Occup Hyg 44: 565-601.
Page 253 of 352

-------
Hwang. €Y. (1983). Size and shape of airborne asbestos fibres in mines and mills. Br J Ind Med 40:
273-279.
I ARC. (1977). I ARC monographs on the evaluation of carcinogenic risk of chemicals to man: Asbestos.
Lyon, France: World Health Organization. http://monoeraphs.iarc.fr/ENG/Monoeraphs/vol 1 -
42/m.onol4.pdf.
I ARC. (1987). Asbestos and certain asbestos compounds [I ARC Monograph], In Overall evaluations of
carcinogenicity: An updating of IARC monographs volumes 1 to 42 (pp. 106-116). Lyon,
France. http://monographs.iarc.fr/ENG/Monoeraphs/suppl7/index.php.
IARC. (2009). Monographs on the evaluation of carcinogenic risks to humans [Website],
http://monoeraphs.iarc.fr/ENG/Monoeraphs/PDFs/index.php.
IARC. (2012a). Arsenic, metals, fibres, and dusts. In Arsenic, Metals, Fibres and Dusts. Lyon, France.
IARC. (2012b). A review of human carcinogens. Part C: Arsenic, metals, fibres, and dusts [IARC
Monograph], Lyon, France: World Health Organization.
http://monoeraphs.iarc.fr/ENG/Monoeraphs/vol 1 OOC/monol OOC.pdf.
IQM. (2006). Asbestos: Selected cancers. Washington, DC: National Academies Press.
http://dx.doi.on 2.26/11665.
Ismail-Khan. R; Robinson. LA; Williams. CC. Jr.: Garrett. CR; Bepler. G; Simon. GR. (2006).
Malignant pleural mesothelioma: a comprehensive review. 13: 255-263.
http://dx.doi.on	480601300402.
Jaenicke. R (1980). Natural aerosols. In TJ Kneip; PJ Lioy (Eds.), Aerosols: anthropogenic and natural,
sources and transport: Ann. N. Y. Acad. Sci.
Jiane. GC; Madt. AK; Inemundson. KJ; Murbach. DM: Fehline. KA; Paustenbach. DJ; Finlev. BL.
(2008). A study of airborne chrysotile concentrations associated with handling, unpacking, and
repacking boxes of automobile clutch discs. Regul Toxicol Pharmacol 51: 87-97.
http://dx.doi.on	.vrtph.2008.02.009.
Kakooei. H; Hormozy. M; Marioryad. H. (2011). Evaluation of asbestos exposure during brake repair
and replacement. Ind Health 49: 374-380.
Kim. HN lardson. D; Loomis. D: Van Toneeren. M; Burstyne. I. (2011). Bias in the estimation of
exposure effects with individual- or group-based exposure assessment. J Expo Sci Environ
Epidemiol 21: 212-221. http://dx.doi.ore. >8/ies.2009.T4.
Kopnin. PB; Kravchenl	3v. ฅA; Pylev. LN: Kopnin. BP. (2004). Cell type-specific effects
of asbestos on intracellular ROS levels, DNA oxidation and G1 cell cycle checkpoint. Oncogene
23: 8834-8840.
Kopylev. L; Sullivan P_\ _ \ mikoi n_ H _ ป rteson. TF. (201 1). Monte Carlo analysis of impact of
underascertainment of Mesothelioma cases on underestimation of risk. Open Epidemiol J 4: 45-
53.
Landrigan PJ, L iov\ Pi, I'hurston G, Ucrkowitz ( hen. LC; Chillmd. SN: Gavett. SH;
Georgopoulos. PG: Geyh. AS: Levin. S: Perera. F: Rappaport. SM; Small. C. (2004). Health and
environmental consequences of the world trade center disaster [Review], Environ Health
Perspect 1 12: 73 1-739. http://dx.doi.org/10.1289/ehp.6702.
Liddell. FD: Armstrone. BG. (2002). The combination of effects on lung cancer of cigarette smoking
and exposure in quebec chrysotile miners and millers. Ann Occup Hyg 46: 5-13.
http://dx.doi.oo 93/annhyg/mef008.
Liddt'U i (t U xlonafti \D Vtcdonah! K (1997). The 1891 -1920 birth cohort of Quebec chrysotile
miners and millers: development from 1904 and mortality to 1992. Ann Occup Hyg 41: 13-36.
http://dx.doi.on	50003-4878(96)00044-0.
Liddell. FD: Mcdonald. AD: Mcdona (1998). Dust exposure and lung cancer in Quebec chrysotile
miners and millers. Ann Occup Hyg 42: 7-20.
Page 254 of 352

-------
Liddelt I'homas. DC; Gibbs. GW; Mcdonald., JC. (1984). Fibre exposure and mortality from
pneumoconiosis, respiratory and abdominal malignancies in chrysotile production in Quebec,
1926-75. Ann Acad Med Singapore 13: 340-344.
Liukonen. LR; Weir. FW. (2005). Asbestos exposure from gaskets during disassembly of a medium duty
diesel engine. Regul Toxicol Pharmacol 41: 113-121.
http://dx.doi.ori, 10 j_0| ฆ i.vrtph.2QQ4 10 003.
Loomis. D; Dement J; Richardson. D; Wolf. S. (2010). Asbestos fibre dimensions and lung cancer
mortality among workers exposed to chrysotile. Occup Environ Med 67: 580-584.
http://dx.doi.orp s s ' oem.2009.0 J*->
Loomis. D; Dement JM; Elliott. L; Richardson. D; Kuempel. ED; Stavner. L. (2012). Increased lung
cancer mortality among chrysotile asbestos textile workers is more strongly associated with
exposure to long thin fibres. Occup Environ Med 69: 564-568. http://dx.doi.(	5/oemed-
2 0676.
Loomis. D; Dement JM; Wolf. SH; Richardson. DB. (2009). Lung cancer mortality and fibre exposures
among North Carolina asbestos textile workers. Occup Environ Med 66: 535-542.
http://dx.doi.on	)em.2.008.044362.
Loomis. D; Richardson. DB; Elliott. L. (2019). Quantitative relationships of exposure to chrysotile
asbestos and mesothelioma mortality. Am J Ind Med 62: 471-477.
Lit. J; Keane^STOna MiiwEj(l994). IN-VITRO GENOTOXICITY STUDIES OF
CHRYSOTILE ASBESTOS FIBERS DISPERSED IN SIMULATED PULMONARY
SURFACTANT. Mutat Res 320: 253-259.
Madl. A.K; Hollins. DM; Devlin. KP; Donovan. EP; Dopan < > Scott. PK; Pere; \J_ (2014). Airborne
asbestos exposures associated with gasket and packing replacement: a simulation study and
meta-analysis. Regul Toxicol Pharmacol 69: 304-319.
http://dx.doi.oo	.vrtph.2014.04.007.
Madl. A.K; Scott. LL; Murbach. DM; Fehling. KA; Finl	ustenbach. DJ. (2008). Exposure to
chrysotile asbestos associated with unpacking and repacking boxes of automobile brake pads and
shoes. Ann Occup Hyg 52: 463-479. http://dx.doi.ore/10.1093/ann.hve/men028.
Maneold t_ I lark. K; Madl. A; Paustenbach. I ? (2006). An exposure study of bystanders and workers
during the installation and removal of asbestos gaskets and packing. J Occup Environ Hyg 3: 87-
98. http://dx.doi.or; 80/1545962.0500498067.
Mcdonald. JC; Gibbs. GW; Liddell. FD. (1980a). Chrysotile fibre concentration and lung cancer
mortality: a preliminary report. IARC Sci Publ 30: 811-817.
Mcdonald. JC; Liddell. FD; Dufresne. A; Mcdor	(1993a). The 1891-1920 birth cohort of
Quebec chrysotile miners and millers: mortality 1976-88. Br J Ind Med 50: 1073-1081.
Mcdonald. JC; Liddell. FD; Gibbs. GW; Eyssen. GE; Mcdonal (1993b). Dust exposure and
mortality in chrysotile mining, 1910-75. 1980. Br J Ind Med 50: 1058-1072.
McDonald. JC; Liddell. FDK; Gibbs. GW; Eyssen. GE; McDo	. (1980b). Dust exposure and
mortality in chrysotile mining, 1910-75. Occup Environ Med 37: 11-24.
MEMA. (Motor Equipment Manufacturers Association). (2016). RE: Request to Designate Asbestos as a
High-Priority Substance Under TSCA ง 6 Due to Use in Brake Friction Materials [Personal
Communication].
Miller. J; Thompsoi ukla. A. (2014). Asbestos-induced oxidative stress in lung pathogenesis. In I
Laher (Ed.), (pp. 1587-1610). Berlin, Germany: Springer-Verlag. http://dx.doi.ore/10.1007/978-
3-642-3 0018-9 201.
Mineis. S. (2018). Email correspondence from Steve Mingis (Branham Corporation) and EPA. June 12,
2018. Available online.
Page 255 of 352

-------
Mularski. A; Marie-Anafs. F; Mazzolini. J; Niedergang. F. (2018). Observing frustrated phagocytosis
and phagosome formation and closure using total internal reflection fluorescence microscopy
(TIRFM). In GRousselet (Ed.), (pp. 165-175). New York, NY: Humana Press.
http://dx.doi.orc 10 100 978-1-4939-78 * i
NASA. (2020a). NASA Operational Uses of Asbestos - Super Guppy Turbine Aircraft.
NASA. (2020b). NASA's Super Guppy Aircraft - Details on Use of Asbestos in Brake System.
NASA Occupational Health. (2020). Asbestos Training - Super Guppy Brake System.
Newhouse. ML; Sullivan. KR. (1989). A mortality study of workers manufacturing friction materials:
1941-86. Br J Ind Med 46: 176-179.
NIOSH. (2008). Asbestos fibers and other elongated mineral particles: State of the science and roadmap
for research. Revised draft. Department of Health and Human Services, Centers for Disease
Control and Prevention, National Institute for Occupational Safety and Health.
http://www.cdc.gOv/n.iosh/docket/archive/pdfs/.NIOSH.-099A/0099A-060608-CIB.pdf.
NIOSH. (201 la). Asbestos fibers and other elongate mineral particles: State of the science and roadmap
for research. In Current Intelligence Bulletin 62. (2011-159). Atlanta, GA: National Institute for
Occupational Safety and Health, Centers for Disease Control and Prevention.
NIOSH. (201 lb). Current intelligence bulletin 62: Asbestos fibers and other elongate mineral particles:
State of the science and roadmap for research [Revised April 2011] (Revised ed.). (DHHS
(NIOSH) Publication No. 2011-159). Atlanta, GA: National Institute for Occupational Safety
and Health, Centers for Disease Control and Prevention, https://www.cdc.gov/niosh/docs/2011-
159/pdfs/20	:.
NRC. (1983). Drinking water and health: Volume 5. Washington, DC: National Academies Press.
http://dx.doi.on 226/326.
NRC. (1988). Health Risks of Radon and Other Internally Deposited Alpha Emitters. BE1R IV.
Washington, D.C. http://books.nap.edu/openbook.php7recoixl id= 1026&page=Rl.
NRC. (2006). Asbestos: Selected cancers. Institute of Medicine (US) Committee on Asbestos: Selected
Health Effects. Washington, DC: The National Academies Press.
NTP. (2016). 14th Report on carcinogens. Research Triangle Park, NC: U.S. Department of Health and
Human Services, Public Health Service, https://ntp.niehs.nih.gov/pubhealth/roc/index-
1.html.https://hero.epa.gov/hero/index.cfm?action=search.view&reference id=5352390OSHA.
(2006). Asbestos-automotive brake and clutch repair work, Part 2. Washington, DC: U.S.
Department of Labor, Occupational Safety and Health Administration.
https://www.cdc. gov/niosh/docs/89-121 /pdfs/89-121 .pdf.
Paustenbach I > > Unle\ I'l 1^ J i Krorby. GP; Sheehan 1'ป (2004). Environmental and occupational
health hazards associated with the presence of asbestos in brake linings and pads (1900 to
present): a "state-of-the-art" review [Review], J Toxicol Environ Health B Crit Rev
7: 25-80. http://dx.doi.org/ 0/10937400490231494.
Paustenbach	. A.K; Donovan. E; Clark. K; Fehling. K; Lee. TC. (2006). Chrysotile asbestos
exposure associated with removal of automobile exhaust systems (ca. 1945-1975) by mechanics:
results of a simulation study. J Expo Sci Environ Epidemiol 16: 156-171.
http://dx.doi.on	;i jea.7500450.
Paustenbach. DJ; Richter. RO; Finlev. BL; Sheehan. PI. (2003). An evaluation of the historical
exposures of mechanics to asbestos in brake dust. Appl Occup Environ Hyg 18: 786-804.
http://dx.doi.orc 10 1080/104732203m I
Peng. WJ; Mi. J; Jiang. Y.H. (2015). Asbestos exposure and laryngeal cancer mortality. Laryngoscope
126: 1169-1174.
Peto. J. (1978). The hygiene standard for chrysotile asbestos. Lancet 1: 484-489.
http://dx.doi.on	30140-6736(78190145-9.
Page 256 of 352

-------
Peto. J; Seidman. H; Selikoff. II. (1982). Mesothelioma mortality in asbestos workers: implications for
models of carcinogenesis and risk assessment. Br J Cancer 45: 124-135.
Pinheiro. GA; Anta	ig. KM; Attfield. MP. (2004). Malignant mesothelioma surveillance: a
comparison of ICD 10 mortality data with SEER incidence data in nine areas of the United
States. Int J Occup Environ Health 10: 251-255.
Piolatto. G; Negri. E; La Vecchia. C; Pira. E; Decarli. A; Peto. J. (1990). An update of cancer mortality
among chrysotile asbestos miners in Balangero, northern Italy. Br J Ind Med 47: 810-814.
Pira. E; Pelucchi. C; Piolatto. PG; Negri. E; Bilei. T; La Vecchia. C. (2009). Mortality from cancer and
other causes in the Balangero cohort of chrysotile asbestos miners. Occup Environ Med 66: 805-
809. http://dx.doi.org/10.1136/oem.2008.044693.
Pira. E; Romano 1 ^ I 'onat. F. Pelucchi. C; La Vecchia ^ Uoffetui. P (2017). Mortality from cancer
and other causes among Italian chrysotile asbestos miners. Occup Environ Med 74: 558-563.
http://dx.doi.on !J ปoemed-20l> 10J"<73.
Popik. T (2018). "RE: [External] Oil field brake blocks." Email messages to Jay Jon, U.S.
Environmental Protection Agency. October 9, 2018; and October 25, 2018. Available online at
(accessed
kcid. A; de Klerk. N; Musk. AW. ((2011). Does Exposure to Asbestos Cause Ovarian Cancer? A
Systematic Literature Review and Meta-analysis [Review], Cancer Epidemiol Biomarkers Prev
20: 1287-1295. http://dx.doi.Org/l 71055-9965.EPI-10-1302.
Riata. RE; Riipinen. HM. (1998). Respirator and High Efficiency Particulate Air Filtration Unit
Performance in Asbestos Abatement. Appl Occup Environ Hyg 13: 32-40.
Richter. RO; Finlev. BL; Paustenbacl Williams. PR.D; Sheehan. PI. (2009). An evaluation of short-
term exposures of brake mechanics to asbestos during automotive and truck brake cleaning and
machining activities. J Expo Sci Environ Epidemiol 19: 458-474.
http://dx.doi.oo	es.2008.36.
Rohl. AN; Langer. AM; Wolff. MS; Weisman. I. (1976). Asbestos exposure during brake lining
maintenance and repair. Environ Res 12: 110-128.
Rubino.	)latto. G; Newhouse. ML; Scansetti. G; Aresini. GA; Murray. R. (1979). Mortality of
chrysotile asbestos workers at the Balangero Mine, northern Italy. Occup Environ Med 36: 187-
194.
Schlumberger. (2018). "Drawworks." Schlumberger Oilfield Glossary.
https ://www. glossary, oilfield, sib. com/T emu, s/d/drawworks. aspx.
Schreier. 11 UmuetT ป \ Ui, kutich. LM. (1987). WEATHERING PROCESSES OF ASBESTOS-RICH
SERPENTINITIC SEDIMENTS. Soil Sci Soc Am J 51: 993-999.
Sheehy. JW; Cooper. TC; O'Brien. DM; McgtotMin. JD; Froehlicl (1989). Control of asbestos
exposure during brake drum service. (CIS/90/00953).
Sheehy. JW; Godbey. FW; Cooper. TC; Lenihan. KL; Van Wagenen. HP; Mcglothlin. JD. (1987a). In-
Depth Survey Report: Control Technology for Brake Drum Service Operations at Ohio
Department of Transportation, Maintenance Facility, Lebanon, Ohio, CT-152-18b (pp. 152-118).
(NIOSH/00166990).
Sheehy. JW; To	ooper. TC; Van Wagenen. HP. (1987b). In-Depth Survey Report: Evaluation
of Brake Drum Service Controls at Cincinnati Bell Maintenance Facility, Fairfax, Ohio, Report
No. CT-152-21B (pp. 152-121). (NIOSH/00175414).
Siemiatvcki. J; Wacholder. '. * ^'war. R; Wald. L; Begin. D; Richardson. L; Rosenman. k t ^ >crin. M.
(1988). Smoking and degree of occupational exposure: Are internal analyses in cohort studies
likely to be confounded by smoking status? Am J Ind Med 13: 59-69.
Page 257 of 352

-------
Simeonova. P: Luster. M. (1995). IRON AND REACTIVE OXYGEN SPECIES IN THE ASBESTOS-
INDUCED TUMOR-NECROSIS-FACTOR-ALPHA RESPONSE FROM ALVEOLAR
MACROPHAGES. Am J Respir Cell Mol Biol 12: 676-683.
Speil. S; Leinewebt (1969). Asbestos minerals in modern technology. Environ Res 2: 166-208.
Spence. SK; Rocchi. PSJ. (1996). Exposure to asbestos fibres during gasket removal. Ann Occup Hyg
40: 583-588.
Spencer. JW: Plisko. Ml; Balzer. JL. (1999). Asbestos fiber release from the brake pads of overhead
industrial cranes. Appl Occup Environ Hyg 14: 397-402.
http://dx.doi.on 80/104732.299302800.
SRC. (2019a). A Summary on the North Carolina Cohort of Asbestos Textile Workers.
SRC. (2019b). A Summary on the Quebec Cohort of Asbestos Miners and Millers.
SRC. (2019c). A Summary on the South Carolina Cohort of Asbestos Textile Workers.
Stavner. L; Kuempel. E; Gilbert. S; Hein. M; Dement J. (2008). An epidemiological study of the role of
chrysotile asbestos fibre dimensions in determining respiratory disease risk in exposed workers.
Occup Environ Med 65: 613-619. http://dx.doi.org/ 5/oero.2007.035584.
Stavner. L; Smith. R; Bailet 1 *dbert. S; Steenland. K; Perriesn, ,<, \ 11ซปvป 11 Ij, 1_ ct
-------
parameters of Lemna gibba. Arch Environ Contam Toxicol 47: 281-289.
http://dx.doi.org/10.10Q7/ij
(2019). EPA query of Bureau of Labor Statistics online Occupational Employment Statistics
Query System [Website], Washington, DC: Bureau of Labor Statistics, https://data.bls.gov/oes/.
Census Bureau. (2016a). American FactFinder, 2012 Economic Census of the United States, Retail
Trade: Subject Series - Estab & Firm Size: Summary Statistics for Single Unit and Multiunit
Firms for the U.S.: 2012, EC1244SSSZ3. Retrieved from
https^/factfinder.census.gov/faces/tableservices/isf/pages/productview.xhtml?pid=ECN 20
S 44SSSZ3&prodType=table
Census Bureau. (2016b). American FactFinder, 2012 Economic Census of the United States, Retail
Trade: Subject Series - Product Lines: Product Lines Statistics by Industry for the U.S. and
States: 2012, EC1244SLLS1. Retrieved from
https://factfinder. census. gov/faces/tableservices/isf/pages/productview.xh.tml?pid=ECN 20
ISLLS1 &prodType=table
T (U.S. Department of Transportation,). (2018). Average annual miles per driver by age group.
Available online at https://www.fh.wa.dot.gov/ohim/onh00/bar8.htm (accessed
U.S. EPA. (1980). Ambient water quality criteria for asbestos [EPA Report], (EPA/440/5-80/022).
Washington, DC. http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=0000lLP6.txt.
U.S. EPA. (1985). Drinking water criteria document for asbestos. (600/X-84/199-1). Cincinnati, OH:
Environmental Criteria and Assessment Office, U.S. Environmental Protection Agency.
(1986). Airborne asbestos health assessment update. (EPA/600/8-84/003F). Washington DC:
U.S. Environmental Protection Agency, Environmental Criteria and Assessment.
EPA. (1988a). Asbestos Modeling Study. Final Report. Report from Versar to EPA. (560/3-
88/091). Washington, D.C.: Office of Toxic Substances.
K 1 T \ (1988b). IRIS summary for asbestos (CASRN 1332-21-4). Washington, DC: U.S.
Environmental Protection Agency, Integrated Risk Information System.
http ://www. epa. gov/iri s/ sub st/ tm.
(1994). Guidelines for Statistical Analysis of Occupational Exposure Data: Final. United
States Environmental Protection Agency :: U.S. EPA.
(1998). Guidelines for ecological risk assessment [EPA Report], (EPA/630/R-95/002F).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/risk/guidelines-ecological-risk-assessment.
(2005). Guidelines for carcinogen risk assessment [EPA Report] (pp. 1-166). (EPA/630/P-
03/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://www2.epa.gov/osa/guidelines-carcinogen-risk-assessment.
(2007). Current best practices for preventing asbestos exposure among brake and clutch
repair workers. (EPA-747-F-04-004). https://www.epa.gov/asbestos/current-best-practices-
preventing-asbestos-exposure-among-brake-and-clutch-repair-workers.
EPA. (2008). Framework for investigating asbestos-contaminated superfund sites (pp. 71).
(OSWER Directive #9200.0-68). Washington, DC: U.S. Environmental Protection Agency,
Office of Solid Waste and Emergency Response.
http://www.epa.gov/superfund/health/contaminants/asbestos/pdfs/framework asbestos guidance
.pdf.
(2009). Risk assessment guidance for superfund volume I: Human health evaluation manual
(Part F, supplemental guidance for inhalation risk assessment): Final [EPA Report], (EPA/540/-
R-070/002). Washington, DC. https://www.epa.gov/risk/risk-assessment-guidance-superfund-
rags-part-f.
(2011). Exposure factors handbook: 2011 edition (final) [EPA Report], (EPA/600/R-
090/052F). Washington, DC: U.S. Environmental Protection Agency, Office of Research and
Page 259 of 352

-------
Development, National Center for Environmental Assessment.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=236252. https://hero.epa.gov/hero/index.cf
m?action=search.view&reference id= 12	(2014a). Framework for human health
risk assessment to inform decision making. Final [EPA Report], (EPA/100/R-14/001).
Washington, DC: U.S. Environmental Protection, Risk Assessment Forum.
https://www.epa.gov/risk/framework-human-health-risk-assessment4nform-decision-making.
EPA. (2014b). Site-wide Baseline Ecological Risk Assessment Libby Asbestos Superfund Site.
https://www.epa.gOv/sites/production/files/2015-01/docum.en.ts/libbY-asbestos-site-wide-bera-l-
9-2015.pdf
L	(2014c). Toxicological review of libby amphibole asbestos: In support of summary
information on the Integrated Risk Information System (IRIS) [EPA Report], (EPA/635/R-
11/002F). Washington, DC: Integrated Risk Information System, National Center for
Environmental Assessment, Office of Research and Development.
https://cfpub.epa.gov/ncea/iris/iris documents/documents/toxreviews/1026tr.pdf.
EPA. (2015a). Memorandum of Understanding on Copper Mitigation in Watersheds and
Waterways between U.S. Environmental Protection Agency and various industry stakeholders.
https://www.epa.gov/sites/production/files/2015-l 1/docum. ents/copper brakepads mou.pdf.
(2015b). Peer review handbook [EPA Report] (4th ed.). (EPA/100/B-15/001). Washington,
DC: U.S. Environmental Protection Agency, Science Policy Council.
https://www.epa.gov/osa/peer-review-han.dbook-4th-edition-2.015.
(2017a). 1,4-dioxane (CASRN: 123-91-1) bibliography: Supplemental file for the TSCA
Scope Document [EPA Report], https://www.epa.gov/sites/production/files/^
06/docum ents/14dioxane comp bib.pdf.
(2017b). Asbestos (CASRN: 1332-21-4) bibliography: Supplemental file for the TSCA scope
document [EPA Report], Washington, DC. https://www.epa.gov/sites/production/files/2017-
06/documents/abestos comp bib.pdf.
EPA. (2017c). Federal Register: Procedures for chemical risk evaluation under the amended toxic
substances control act. Fed Reg 82: 33726-33753.
U.S. EPA. (2017d). Scope of the risk evaluation for Asbestos [EPA Report], (EPA-740-R1-7008).
Washington, DC: U.S. EPA, Office of Chemical Safety and Pollution Prevention (OCSPP),
Office of Pollution Prevention and Toxics (OPPT).
https://www.epa.gov/sites/production/files/2<' 1 <'•/documents/asbestos scoi J. 1 t>df.
EPA. (2017e). Toxics Release Inventory (TRI), reporting year 2015. Retrieved from
https://www.epa.gov/toxics-release-inventorv-tri-program/tri-data-and-tools
I EPA. (2018a). Application of systematic review in TSCA risk evaluations. (740-P1-8001).
Washington, DC: U.S. Environmental Protection Agency, Office of Chemical Safety and
Pollution Prevention, https://www.epa.gov/sites/production/files/2018-
06/docum ents/ final application of sr in tsซ	df.
(2018b). Meeting summary: Chemical Use Outreach Phone Call with Mexalit and EPA to
Discuss Asbestos in Products Exported to the United States. Washington, DC.
K 1 T \ (2018c). Meeting summary: Chemical Use Outreach Phone Call with Textiles Tecnicos and
EPA to Discuss Asbestos in Products Exported to the United States. Washington, DC.
EPA. (2018d). Problem formulation of the risk evaluation for asbestos. (EPA-740-R1 -7018).
Washington, DC: Office of Chemical Safety and Pollution Prevention, United States
Environmental Protection Agency.
!:_} iA (2020a). Final Risk Evaluation for Asbestos, Supplemental File: Consumer Exposure
Calculations.
(2020b). Final Risk Evaluation for Asbestos, Supplemental File: Occupational Exposure
Calculations (Chlor-Alkali).
Page 260 of 352

-------
L S J L.lv (2020c). Final Risk Evaluation for Asbestos, Systematic Review Supplemental File: Data
Quality Evaluation of Consumer Exposure.
EPA. (2020d). Final Risk Evaluation for Asbestos, Systematic Review Supplemental File: Data
Quality Evaluation of Ecological Hazard Studies.
(2020e). Final Risk Evaluation for Asbestos, Systematic Review Supplemental File: Data
Quality Evaluation of Environmental Fate and Transport Studies.
EPA. (2020f). Final Risk Evaluation for Asbestos, Systematic Review Supplemental File: Data
Quality Evaluation of Environmental Releases and Occupational Exposure.
K ^ 1 r \ (2020g). Final Risk Evaluation for Asbestos, Systematic Review Supplemental File: Data
Quality Evaluation of Environmental Releases and Occupational Exposure Data Common
Sources.
(2020h). Final Risk Evaluation for Asbestos, Systematic Review Supplemental File: Data
Quality Evaluation of Human Health Hazard Studies: Mesothelioma and Lung Cancer Studies.
LV. (2020i). Final Risk Evaluation for Asbestos, Systematic Review Supplemental File: Data
Quality Extraction Tables for Consumer Exposure.
(2020j). Final Risk Evaluation for Asbestos. Systematic Review Supplemental File: Data
Quality Evaluation of Physical-Chemical Properties studies. Washington, D.C.: U.S.
Environmental Protection Agency. Office of Chemical Safety and Pollution Prevention.
Upadhyav. D; Karop. DW. (2003). Asbestos-induced pulmonary toxicity: role of DNA damage and
apoptosis [Review], Exp Biol Med 228: 650-659.
LISGS. (2017). Mineral commodity summaries 2017. Washington, DC: U.S. Department of the Interior.
https://minerals.usgs.eov/minerals/piibs/mcs/2017/mcs! if.
LISGS. (2019). Mineral commodity summaries: Asbestos. USGS.
https://minerals.usgs.gov/minerals/pubs/commoditv/asbestos/mcs- sbes.pdf.
USGS. (2020). Asbestos: Mineral commodity summaries 2020 (pp. 26-27). (MCS2020). Reston, VA.
http://dx.doi.on	ncs2020.
Vacek. PM. (1998). Effects of the intensity and timing of asbestos exposure on lung cancer risk at two
mining areas in Quebec. J Occup Environ Med 40: 821-828.
Versar. (1987). Nonoccupational asbestos exposure. Revised Report. Washington, D C.: U.S.
Environmental Protection Agency.
Virta. R. (201 1). Asbestos. In Kirk-Othmer Encyclopedia of Chemical Technology, [online]: John Wiley
& Sons. http://dx.doi.org/10.10Q2 < M i _ ^ 61.0119020M01 >J J09.a01.pub3.
Virta. RL. (2002). Asbestos: U.S. Geological Survey open-file report no. 02-149 (pp. 35). (Report No.
02-149). Virta, RL. https://pubs.usgs.gov/of/2002/ofQ2~149/of	*.
Virta. RL. (2006). Asbestos. In Industrial Minerals & Rocks: Commodities, Markets, and Uses (7th ed.).
U.S.: SME.
Wang. X; Lin. S; Yano. E; Qiu. H; Yu. IT; Tse. L; Lan. Y; Wang. M. (2012). Mortality in a Chinese
chrysotile miner cohort. Int Arch Occup Environ Health 85: 405-412.
http://dx.doi.on	>00420-011-0685-9.
Wang. X; Lin. S; Yano. E; Yu. IT; Courtice. M; Lai iristiani. DC. (2014). Exposure-specific lung
cancer risks in Chinese chrysotile textile workers and mining workers. Lung Cancer 85: 119-124.
http://dx.doi.org IU J 0 k> i.lungcan.201 {0 j_ ON .
Wang. X; Lin. S; Yu. I; Qiu. H; Lan. Y; Yano. E. (2013a). Cause-specific mortality in a Chinese
chrysotile textile worker cohort. Cancer Sci 104: 245-249. http://dx.doi.org/l	1.12060.
Wang. X; Yano. E; Lin. S; Yu. IT; Lan. Y; Tse.	u. H; Christiani. DC. (2013b). Cancer mortality
in Chinese chrysotile asbestos miners: exposure-response relationships. PLoS ONE 8: e71899.
http://dx.doi.oo	ournal.pone.0071899.
Page 261 of 352

-------
Wang. Y; Hue I' h'ng. C: Zeng. Y; Yang. Ji; Zhan^ \u Otm^ I IViig. J. (2019). Chrysotile asbestos
induces apoptosis via activation of the p53-regulated mitochondrial pathway mediated by ROS in
A549 cells. Appl Clay Sci 105245. http://dx.doi.org/10 10ly i clay.201" 10>J l\
Washington State. (2010). RCW (Revised Code of Washington) Chapter 70.285: Brake Friction
Material. Washington State, http://app.leg.wa.gov/RCW/default.aspx?cite=70.2.85.030.
Weir _A I L.i < - Meraz. LB. (2001). Characterization of vehicular brake service personnel exposure
to airborne asbestos and particulate. Appl Occup Environ Hyg 16: 1139-1146.
http://dx.doi.org ijซ h '80/10473220]2 .1p;.
Whitakei (2000). The cytology of malignant mesothelioma [Review], Cytopathology 11: 139-151.
http://dx.doi.on	.1365-2303.2000.00247.x.
Whittaker. C; Rice. F; McKeman. L; Dankovic. D; Lentz. T; Macmahon. K; Kuempel. E; Zumwalde. R;
Schu (2016). Current Intelligence Bulletin 68: NIOSH Chemical Carcinogen Policy.
Whittaker, C; Rice, F; Mckernan, L; Dankovic, D; Lentz, T; Macmahon, K; Kuempel, E;
Zumwalde, R; Schulte, P.
WHO. (2014). Chrysotile asbestos. Geneva, Switzerland.
http://www.who.int/ipcs/assessm.ent/piiblic health/chrysotile asbestos summary.pdf.
Yano. E; Wang. ZM; Wang. XR; Wang. MZ; Le (2001). Cancer mortality among workers exposed
to amphibole-free chrysotile asbestos. Am J Epidemiol 154: 538-543.
Yeung. P; Patience. K; Apthorpe. L; Willcocks. D. (1999). An Australian study to evaluate worker
exposure to chrysotile in the automotive service industry. Appl Occup Environ Hyg 14: 448-457.
Page 262 of 352

-------
APPENDICES
Appendix A Regulatory History
A.l Federal Laws and Regulations
The federal laws and regulations applicable to asbestos are listed along with the regulating agencies
below. States also regulate asbestos through state laws and regulations, which are also listed within this
section.
Toxics Substances Control Act (TSCA), 1976
2601 et seq
The Toxic Substances Control Act of 1976 provides EPA with authority to require reporting, record-
keeping and testing requirements, and restrictions relating to chemical substances and/or mixtures.
Certain substances are generally excluded from TSCA, including, among others, food, drugs, cosmetics
and pesticides.
TSCA addresses the production, importation, use and disposal of specific chemicals including
polvchlorinated biphenyls (PCBsi asbestos, radon and lead-based paint. The Frank R. Lautenberg
Chemical Safety for the 21st Century Act updated TSCA in 2016 https://www.epa.gov/1 aws-
regulations/summarv-toxic-substances-control-act.
Asbestos Hazard Emergency Response Act (AHERA), 1986
TSCA Subchaptci U Vsbestos Hazard Emergency Respond tM \J._ 12641-2656
Defines asbestos as the asbestiform varieties of— chrysotile (serpentine), crocidolite (riebeckite),
amosite (cummingtonite-grunerite), anthophyllite, tremolite or actinolite.
Requires local education agencies (i.e., school districts) to inspect school buildings for asbestos and
submit asbestos management plans to appropriate state; management plans must be publicly available,
and inspectors must be trained and accredited.
Tasked EPA to develop an asbestos Model Accreditation Plan (MAP) for states to establish training
requirements for asbestos professionals who do work in school buildings and also public and
commercial buildings.
Asbestos-Containing Materials in Schools Rule (per AHERA), 1987
40 CFR Part 763. Subp
Requires local education agencies to use trained and accredited asbestos professionals to identify and
manage asbestos-containing building material and perform asbestos response actions (abatements) in
school buildings.
Page 263 of 352

-------
1989 Asbestos: Manufacture, Importation, Processing, and Distribution in Commerce
Prohibitions; Final Rule (also known as Asbestos Ban and Phase-out Rule (Remanded), 1989)

Dock n ซ•! rS-620481	M
EPA issued a final rule under Section 6 of Toxic Substances Control Act (TSCA) banning most
asbestos-containing products.
In 1991, this rule was vacated and remanded by the Fifth Circuit Court of Appeals. As a result, most of
the original ban on the manufacture, importation, processing or distribution in commerce for the
majority of the asbestos-containing products originally covered in the 1989 final rule was overturned.
The following products remain banned by rule under the Toxic Substances Control Act (TSCA):
o	Corrugated paper
o	Rollboard
o	Commercial paper
o	Specialty paper
o	Flooring felt
In addition, the regulation continues to ban the use of asbestos in products that have not historically
contained asbestos, otherwise referred to as "new uses" of asbestos (Defined by 40 CFR 763.163 as
"commercial uses of asbestos not identified in ง763.165 the manufacture, importation or processing of
which would be initiated for the first time after August 25, 1989.").
Restrictions on Discontinued Uses of Asbestos; Significant New Use Rule (SNUR), 2019
40 CFR Parts 9 and 721 - Restrictions on Discontinued Uses of Asbestos
Docket ID: EPA-HQ-OPPT-2018-0159; FRL 9991-33
This final rule strengthens the Agency's ability to rigorously review an expansive list of asbestos
products that are no longer on the market before they could be sold again in the United States. Persons
subject to the rule are required to notify EPA at least 90 days before commencing any manufacturing,
importing, or processing of asbestos or asbestos-containing products covered under the rule. These uses
are prohibited until EPA conducts a thorough review of the notice and puts in place any necessary
restrictions or prohibits use.
Other EPA Regulations:
Asbestos Worker Protection Rule, 2000
Extends OSHA standards to public employees in states that do not have an OSHA approved worker
protection plan (about half the country).
Page 264 of 352

-------
Asbestos Information Act, 1988
15 LI.S.C. ง2607ffi
Helped to provide transparency and identify the companies making certain types of asbestos-containing
products by requiring manufacturers to report production to the EPA.
Asbestos School Hazard Abatement Act (ASHAA), 1984 and Asbestos School Hazard Abatement
Reauthorization Act (ASHARA), 1990
20 LI.S.C. 4011 et sea, and Docket ID: OPTS-62Q48E; FRL-3269-8
Provided funding for and established an asbestos abatement loan and grant program for school districts
and ASHARA further tasked EPA to update the MAP asbestos worker training requirements.
Emergency Planning and Community Right-to-Know Act (EPCRA), 1986
42 LI.S.C. Chapter 116
Under Section 313, Toxics Release Inventory (TRI), requires reporting of environmental releases of
friable asbestos at a concentration level of 0.1%.
Friable asbestos is designated as a hazardous substance subject to an Emergency Release Notification at
40 CFR ง355.40 with a reportable quantity of 1 pound.
Clean Air Act, 1970
42 LI.S.C. ง7401 et sea.
Asbestos is identified as a Hazardous Air Pollutant.
Asbestos National Emission Standardfor Hazardous Air Pollutants (NESHAP), 1973
40 CFR Part 61. Subpart M of the Clean Air Act
Specifies demolition and renovation work practices involving asbestos in buildings and other facilities
(but excluding residences with 4 or fewer dwelling units single family homes).
Requires building owner/operator notify appropriate state agency of potential asbestos hazard prior to
demoliti on/renovati on.
Banned spray-applied surfacing asbestos-containing material for fireproofing/insulating purposes in
certain applications.
Requires that asbestos-containing waste material from regulated activities be sealed in a leak-tight
container while wet, labeled, and disposed of properly in a landfill qualified to receive asbestos waste.
Clean Water Act (CWA), 1972
33 LI.S.C. ง1251 et sea
Toxic pollutant subject to effluent limitations per Section 1317.
Page 265 of 352

-------
Safe Drinking Water Act (SDWA), 1974
42 U.S.C. ง300f et sea
Asbestos Maximum Contaminant Level Goals (MCLG) 7 million fibers/L (longer than lOum).
Resource Conservation and Recovery Act (RCRA), 1976
5901 et sea.
40 CFR 239-282
Asbestos is subject to solid waste regulation when discarded; NOT considered a hazardous waste.
Comprehensive Environmental Response, Compensation and Liability Act (CERCLA), 1980
3601 et sea.
40 CFR Part 302.4 - Designation of Hazardous Substances and Reportable Quantities
13 Superfund sites containing asbestos, nine of which are on the National Priorities List (NPL)
Reportable quantity of friable asbestos is one pound.
Other Federal Agencies:
Occupational Safety and Health Administration (OSHA):
Public L	Occupational Safety and Health Act, 1970
Employee permissible exposure limit (PEL) is 0.1 fibers per cubic centimeter (f/cc) as an 8-hour, time-
weighted average (TWA) and/or the excursion limit (1.0 f/cc as a 30-minute TWA).
Asbestos General Standard 29 CFR 1910
Asbestos Shipyard Standard 2
Asbestos Construction Standard 2.9 CFR 1926
Consumer Product Safety Commission (CPSC): Banned several consumer products. Federal Hazardous
Substances Act (FHSA) L*.0>. A>'0
Food and Drug Administration (FDA): Prohibits the use of asbestos-containing filters in pharmaceutical
manufacturing, processing and packing.
Mine Safety and Health Administration (MSHA): follows OSHA's safety standards.
Surface Mines 30 CFR part 56. subpart D
Underground Mines 30 CFR part 57. subpart D
Department of Transportation
Page 266 of 352

-------
Prescribes the requirements for shipping manifests and transport vehicle placarding applicable to
asbestos 40 CFR part 172.
Non-regulatory information of note:
NIOSH conducts related research and monitors asbestos exposure through workplace activities in an
effort to reduce illness and ensure worker health and safety.
A.2 State Laws and Regulations
Pursuant to AHERA, states have adopted through state regulation the EPA's Model Accreditation Plan
(MAP) for asbestos abatement professionals who do work in schools and public and commercial
buildings. Thirty-nine (39) states42 have EPA-approved MAP programs and twelve (12) states43 have
also applied to and received a waiver from EPA to oversee implementation of the Asbestos-Containing
Materials in Schools Rule pursuant to AHERA. States also implement regulations pursuant to the
Asbestos NESHAP regulations or further delegate those oversight responsibilities to local municipal
governments. While federal regulations set national asbestos safety standards, states have the authority
to impose stricter regulations. As an example, many states extend asbestos federal regulations - such as
asbestos remediation by trained and accredited professionals, demolition notification, and asbestos
disposal - to ensure safety in single-family homes. Thirty (30) states44 require firms hired to abate
asbestos in single family homes to be licensed by the state. Nine (9) states45 mandate a combination of
notifications to the state, asbestos inspections, or proper removal of asbestos in single family homes.
Some states have regulations completely independent of the federal regulations. For example, California
and Washington regulate products containing asbestos. Both prohibit use of more than 0.1% of asbestos
in brake pads and require laboratory testing and labeling.
Below is a list of state regulations that are independent of the federal AHERA and NESHAP
requirements that states implement. This may not be an exhaustive list.
California
Asbestos is listed on California's Candidate Chemical List as a carcinogen. Under California's
Propositions 65. businesses are required to warn Californians of the presence and danger of asbestos in
products, home, workplace and environment.
California Brake Friction Material Requirements (Effective 2017)
Division 4.5. California Code of Regulations. Title 22 Chapter 30
42	Alabama, Alaska, Arkansas, California, Colorado, Connecticut, Delaware, Florida, Illinois, Indiana, Kentucky, Louisiana,
Maine, Maryland, Massachusetts, Michigan, Minnesota, Mississippi, Missouri, Montana, Nebraska, New Hampshire, New
Jersey, New York, North Carolina, North Dakota, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Carolina, South
Dakota, Texas, Utah, Vermont, Virginia, Washington, West Virginia, and Wisconsin.
43	Connecticut, Colorado, Illinois, Kentucky, Louisiana, Massachusetts, Maine, New Hampshire, Oklahoma, Rhode Island,
Texas, and Utah.
44	California, Colorado, Connecticut, Delaware, Florida, Georgia, Hawaii, Iowa, Kansas, Maine, Maryland, Massachusetts,
Michigan, Minnesota, Nebraska, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, North
Dakota, Oregon, Pennsylvania, Utah, Vermont, Virginia, Washington, West Virginia, and Wisconsin.
45	Colorado, Connecticut, Georgia, Maine, Massachusetts, New York, Oregon, Vermont, and West Virginia.
Page 267 of 352

-------
Sale of any motor vehicle brake friction materials containing more than 0.1% asbestiform fibers by
weight is prohibited. All brake pads for sale in the state of California must be laboratory tested, certified
and labeled by the manufacturer.
Massachusetts
Massachusetts Toxics Use Reduction Act (TURA)
Requires companies in Massachusetts to provide annual pollution reports and to evaluate and implement
pollution prevention plans. Asbestos is included on the Complete List of TURA. Chemicals - March
2016.
Minnesota
Toxic Free Kids Act Minn. Sta	l01	116.9407
Asbestos is included on the 2016 Minnesota Chemicals of High Concern List as a known carcinogen.
New Jersey
New Jersey Right to Know Hazardous Substances
The state of New Jersey identifies hazardous chemicals and products. Asbestos is listed as a known
carcinogen and talc containing asbestos is identified on the Right to Know Hazardous Substances list.
Rhode Island
Rhode Island Air Resources - Air Toxics Air Pollution Control Reflation No. 22
Establishes acceptable ambient air levels for asbestos.
Washington
Better Brakes Law (Effective 2015) Chapter 70.285 RCW Brake Friction Material
Prohibits the sale of brake pads containing more than 0.1% asbestiform fibers (by weight) in the state of
Washington and requires manufacturer certification and package/product labelling.
Requirement to Label Building Materials that Contain Asbestos Chapter 70.310 RC W
Building materials that contain asbestos must be clearly labeled as such by manufacturers, wholesalers,
and distributors.
Page 268 of 352

-------
A,3 International Laws and Regulations
Asbestos is also regulated internationally. Nearly 60 nations have some sort of asbestos ban. The
European Union (EU) will prohibit the use of asbestos in the chlor-alkali industry by 2025 (Regulation
(EC) No 1907/2006 of the European Parliament and of the Council. 18 December 2006).
Canada banned asbestos in 2018
Prohibition of Asbestos and Products Containing Asbestos Regulations: SOR/2018-196
Canada Gazette. Pan ji \ olur i: s _ 1 v umber 21
In addition, the Rotterdam Convention is considering adding chrvsotile asbestos to Annex ILL and the
World Health Organization (WHO) has a global campaign to eliminate asbestos-related diseases (WHO
Resolution 60.2.6).
Page 269 of 352

-------
Appendix B List of Supplemental Documents
Associated Supplemental Systematic Review Data Quality Evaluation and Date Extraction
Documents - Provides additional detail and information on individual study evaluations and data
extractions including criteria and scoring results.
Physical-Chemical Properties, Fate and Transport
a.	Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Quality Evaluation of Physical-Chemical Properties Studies (U.S.
201)
b.	Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Extraction of Environmental Fate and Transport Studies (U.S. EPA.
20206)
Occupational Exposures and Releases
c.	Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Quality Evaluation of Environmental Releases and Occupational
Exposure (I _S TP \ 2020O
d.	Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Quality Evaluation of Environmental Releases and Occupational
Exposure Data Common Sources (U.S. EPA. 2020e)
Consumer and Environmental Exposures
e.	Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Quality Evaluation of Consumer Exposure (	202.0c)
f Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Quality Extraction Tables for Consumer Exposure (	201)
Environmental Hazard
g.	Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Quality Evaluation of Ecological Hazard Studies (	E020d)
Human Health Hazard
h.	Final Risk Evaluation for Asbestos: Part 1 Chrystotile Asbestos, Systematic Review
Supplemental File: Data Quality Evaluation of Human Health Hazard Studies: Mesothelioma
and Lung Cancer Studies (U.S. EPA. 2020b.)
i.	Final Risk Evaluation for Asbestos Part 1: Chrysotile Asbestos, Systematic Review
Supplemental File: Data Quality Evaluation for Epidemiological Studies of Ovarian and
Laryngeal Cancers	)20k)
Page 270 of 352

-------
Associated Supplemental Information Documents - Provides additional details and information on
exposure.
Occupational Exposures
i. Final Risk Evaluation for Asbestos: Part 1 Chrysotile Asbestos, Supplemental File:
Occupational Exposure Calculations (Chlor-Alkali)] (U.S. EPA. 2020b)
Consumer Exposures
j. Final Risk Evaluation for Asbestos: Part 1 Chrysotile Asbestos, Supplemental File: Consumer
Exposure Calculations (U.S. EPA. 2020a)
Page 271 of 352

-------
Appendix C Conditions of Use Supplementary Information
EPA identified and verified uses of asbestos throughout the scoping, PF, and risk evaluation stages. As
explained in the PF document, EPA believes that most asbestos imports listed by Harmonized Tariff
Schedule (HTS) code in government and commercial trade databases are likely misreported and are not
ongoing COU. EPA has been working with federal partners to better understand the asbestos-containing
product import information. In coordination with Customs and Border Protection (CBP), EPA has
reviewed available import information for the following asbestos Harmonized Tariff Schedule (HTS)
codes:
2524.90.0045 Chrysotile Milled Fibers, Group 4 And 5 Grades
2524.90.0055 Chrysotile Milled Fibers, Other
6812.92.0000 Asbestos, Fibers, Fabricated, Paper, Millboard and Felt
6812.93.0000 Asbestos, Fiber, Compressed, Jointing, in Sheets or Rolls
6812.99.0003 Asbestos, Fabricated, Cords and String, whether or not Plaited
6812.99.0020 Asbestos, Fibers, Fabricated, Gaskets, Packing and Seals
6812.99.0055 Asbestos, Fibers, Fabricated, Other
6813.20.0010 Asbestos, Mineral Subst, Friction Mat, Brake Lin/Pad, Civil Air
6813.20.0015 Asbestos, Mineral Subst, Friction Mat, Brake Linings and Pads
6813.20.0025 Asbestos, Mineral Subst, Friction Mat, Other
CBP provided import data for the above asbestos HTS codes in CBP's Automated Commercial
Environment (ACE) system, which provided information for 26 companies that reported the import of
asbestos-containing products between 2016 and 2018. EPA contacted these 26 companies in order to
verify the accuracy of the data reported in ACE. Of these 26 companies, 22 companies confirmed that
the HTS codes were incorrectly entered and one company could not be reached. Three companies
confirmed that the HTS codes entered in ACE are correct. EPA received confirmation that the following
asbestos-containing products are imported into the United States:
•	Gaskets for use in the exhaust for off-road utility vehicles
o 6812.99.0020 Asbestos, Fibers, Fabricated, Gaskets, Packing and Seals
•	Gaskets for sealing pipes and flanges
o 6812.93.0000 Asbestos, Fiber, Compressed, Jointing, in Sheets or Rolls
•	Brake linings for use in automobiles that are manufactured and then exported (not sold
domestically)
o 6813.20.0015 Asbestos, Mineral Subst, Friction Mat, Brake Linings and Pads
Page 272 of 352

-------
Regarding the two HTS codes that represent raw chrysotile asbestos, one company imported asbestos as
waste but reported it in ACE under the HTS code 2524.90.0055 (Chrysotile Milled Fibers, Other). EPA
did not contact the two facilities that reported under HTS code 2524.90.0045 (Chrysotile Milled Fibers,
Group 4 And 5 Grades) because these entries were from a chlor-alkali company, which has already
confirmed import and use of raw chrysotile asbestos.
Page 273 of 352

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Appendix D Releases and Exposure to the Environment
Supplementary Information
Toxics Release Inventory Data
A source of information that EPA considered in evaluating exposure is data reported under the Toxics
Release Inventory (TRI) program. TRI reporting by subject facilities is required by law to provide
information on releases and other waste management activities of Emergency Planning and Community
Right-to-Know Act (EPCRA) Section 313 chemicals (i.e., TRI chemicals) to the public for informed
decision making and to assist the EPA in determining the need for future regulations. Section 313 of
EPCRA and Section 6607 of the Pollution Prevention Act (PPA) require certain facilities to report
release and other waste management quantities of TRI-listed chemicals annually when a reporting
threshold is triggered, but these statutes do not impose any monitoring burden for determining the
quantities.
TRI data are self-reported by the subject facility where some facilities are required to measure or
monitor emission or other waste management quantities due to regulations unrelated to the TRI
program, or due to company policies. These existing, readily available data are often used by facilities
for TRI reporting purposes. When measured (e.g., monitoring) data are not "readily available," or are
known to be non-representative for TRI reporting purposes, the TRI regulations require that facilities
determine release and other waste management quantities of TRI-listed chemicals by making
"reasonable estimates." Such reasonable estimates include a variety of different approaches ranging
from published or site-specific emission factors (e.g., AP-42), mass balance calculations, or other
engineering estimation methods or best engineering judgement. TRI reports are then submitted directly
to EPA on an annual basis and must be certified by a facility's senior management official that the
quantities reported to TRI are reasonable estimates as required by law.
Under EPCRA Section 313, asbestos (friable) is a TRI-reportable substance effective January 1, 1987.
For TRI reporting, facilities in covered sectors are required to report releases or other waste management
of only the friable form of asbestos, under the general CASRN 1332-21-4. TRI interprets "friable" under
EPCRA Section 313, referring to the physical characteristic of being able to be crumbled, pulverized or
reducible to a powder with hand pressure, and "asbestos" to include the six types of asbestos as defined
under Title II of TSCA46. Facilities are required to report if they are in a covered industrial code or
federal facility and manufacture (including import) or process more than 25,000 pounds of friable
asbestos, or if they otherwise use more than 10,000 pounds of friable asbestos during a calendar year.
Table APX D-l provides production-related waste management data for friable asbestos reported by
subject facilities to the TRI program from reporting years 2015 to 201847. This is an updated table from
that reported in the PF document. In reporting year 2018, 44 facilities reported a total of approximately
34 million pounds of friable asbestos waste managed. Of this total, zero pounds were recovered for
46	According to 53FR4519 (VII)C(5), "The listing for asbestos is qualified by the term "friable." This term refers to a physical
characteristic of asbestos. EPA interprets "friable" as being crumbled, pulverized, or reducible to a powder with hand
pressure. Again, only manufacturing, processing, or use of asbestos in the friable form triggers reporting. Similarly, supplier
notification applies only to distribution of friable asbestos."
47	Data presented were queried using TRI Explorer and uses the 2019 National Analysis data set (released to the public in
October 2020). This dataset includes revisions for the years 1988 to 2018 processed by EPA.
Page 274 of 352

-------
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
energy or recycled, approximately 46,000 pounds were treated, and over 33 million pounds were
disposed of or otherwise released into the environment.
Table_APX D-2 provides a summary of asbestos TRI releases to the environment for the same
reporting years as Table_APX D-l. There were zero pounds of friable asbestos reported as released to
water via surface water discharges, and a total of 171 pounds of air releases from collective fugitive and
stack air emissions reported in 2018. The vast majority of friable asbestos was disposed of to Resource
Conservation and Recovery Act (RCRA) Subtitle C landfills and to landfills other than RCRA Subtitle
C. Of the 153,947 pounds of friable asbestos reported in 2018 as "other releases", 90,640 pounds were
sent off-site to a waste broker for disposal, 14,760 pounds were sent off-site for storage only, and 48,547
pounds were sent off-site for other off-site management.
TableAPX D-l. Summary of Asbestos TRI Production-Related Waste Managed from 2015-2018
	Qbs)	
Year
Number
of
l-'acilities
Recycling
Knergy
Recovery
Treatment
Releases :l h l
Total
Production
Related
Waste
2015
39
0
0
188,437
38,197,608
38,386,044
2016
41
2
0
31,993
26,748,379
26,780,375
2017
39
0
0
179,814
30,796,283
30,976,097
2018
44
0
0
46,106
33,888,979
33,935,085
Data source: 2015-2018 TRI Data (Undated October 2020*) (U.S. EPA. 2017e).
a Terminology used in these columns may not match the more detailed data element names used in the TRI public data
and analysis access points.
b Does not include releases due to one-time events not associated with production such as remedial actions or
earthquakes.
0 Counts all releases including release quantities transferred and release quantities disposed of by a receiving facility
reporting to TRI.
While production-related waste managed shown in Table_APX D-l excludes any quantities reported as
catastrophic or one-time releases (TRI section 8 data), release quantities shown in Table_APX D-2
include both production-related and non-routine quantities (TRI section 5 and 6 data) for 2015-2018. As
a result, release quantities may differ slightly and may further reflect differences in TRI calculation
methods for reported release range estimates (U.S. EPA. 2017e).
Page 275 of 352

-------
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
Table APX D-2. Summary of Asbestos TRI Releases to the Environment from 2015-2018 (lbs)
Year
Number
ol
l-'iieililies
Air Re
Slack Air
Releases
leases
l-'u$*ili\c
Air
Releases
\\;i(er
Releases
(hiss 1
I ndcr-
liround
Injection
Land Dispns.
R( RA
Subtitle (
landfills
il
All oilier
1.iinri
Disposal;|
Oilier Releases ฆ'
1 (Iliil On-
iind OIT-Sile
Disposal or
Oilier
Keleiises 1,1
Totals
2015
39
101
208
0
0
9,623,957
28,780,780
0
38,405,047
310
38,404,737
Totals
2016
41
178
106
0
0
8,759,578
18,603,892
0
27,363,755
285
27,363,470
Totals
2017
39
80
67
0
0
6,199,224
25,162,328
0
31,361,700
147
31,361,552
Totals
2018
44
96
75
0
0
10,599,587
23,216,673
153,947
33,970,378
171
33,816,260
Data source: 2015-2018 TRI Data (Updated October 2020) (U.S. EPA. 2017e).
a Terminology used in these columns may not match the more detailed data element names used in the TRI public data and
analysis access points.
b These release quantities do include releases due to one-time events not associated with production such as remedial actions
or earthquakes.
0 Counts release quantities once at final disposition, accounting for transfers to other TRI reporting facilities that ultimately
dispose of the chemical waste.
The Clean Water Act and the Safe Drinking Water Act
Background (Numeric Criteria and Reportable Levels)
The Clean Water Act (CWA) requires that states adopt numeric criteria for priority pollutants for which
EPA has published recommended criteria under section 304(a). States may adopt criteria that EPA
approves as part of the state's regulatory water quality standards. Once states adopt criteria as water
quality standards, the CWA requires that National Pollutant Discharge Elimination System (NPDES)
discharge permits include effluent limits as stringent as necessary to meet the standards [CWA section
301(b)(1)(C)], If state permit writers determine that permit limits are needed, they will determine the
level of pollutant allowed to ensure protection of the receiving water for a designated use. This is the
process used under the CWA to address risk to human health and aquatic life from exposure to a
pollutant in ambient waters.
EPA develops recommended ambient water quality criteria for pollutants in surface water that are
protective of aquatic life or human health designated uses with specific recommendations on the
duration and frequency of those concentrations under section 304(a) of the CWA. These criteria are
based on priorities of states and others, and a subset of chemicals are identified as "priority pollutants".
EPA has identified asbestos as a priority pollutant for which a nationally recommended human health
water quality criteria for asbestos of 7 million fibers per liter (MFL) has been developed. EPA has not
developed a nationally recommended water quality criteria for the protection of aquatic life for asbestos,
yet EPA may publish aquatic life criteria for asbestos in the future if it is identified as a priority under
the CWA.
Page 276 of 352

-------
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
EPA's National Primary Drinking Water Regulations (NPDWR), established under the Safe Drinking
Water Act (SDWA), are legally enforceable primary standards and treatment techniques that apply to
public water systems. Primary standards and treatment techniques protect public health by limiting the
levels of contaminants in drinking water. The Maximum Contaminant Level (MCL) for asbestos under
the Safe Drinking Water Act is 7 MFL, for fibers > 10 micrometers. EPA has set this level of protection
based on the best available science at the time the NPDWR was promulgated to prevent potential health
problems and considering any limitations in both the feasible treatment methods to remove a
contaminant and availability of analytical methods to reliably measure the occurrence of the
contaminant in water. In the case of asbestos, the MCL was set based entirely on the health goal since
feasible treatment methods and analytical methods were available to achieve the protective level of 7
MFL. Public water systems are required to sample each entry point into the distribution system for
asbestos at least once every 9 years. Transmission electron microscopy (TEM) is used for detection
(EPA 800/4-83-043). The detection limit is 0.01 MFL. Here are links to the analytical standards and the
drinking water regulations.
The Phase II Rule, the regulation for asbestos, became effective in 1992. The Safe Drinking Water Act
requires EPA to review the national primary drinking water regulation for each contaminant every six
years and determine if the NPDWR is a candidate for revision, at that time. EPA reviewed asbestos as
part of the Six Year Review and determined that the 7 MFL for asbestos is still protective of human
health.
As discussed in the PF document, because the drinking water exposure pathway for asbestos is currently
addressed in the SDWA regulatory analytical process for public water systems, this pathway (drinking
water for human health) was not evaluated in this Part 1 of the risk evaluation for asbestos.
EPA issues Effluent Limitations Guidelines and Pretreatment Standards which are national regulatory
standards for industrial wastewater discharges to surface waters and publicly owned treatment works, or
POTWs (municipal sewage treatment plants). EPA issues Effluent Limitations Guidelines and
Pretreatment Standards for categories of existing sources and new sources under Title III of the Clean
Water Act. The standards are technology-based {i.e., they are based on the performance of treatment and
control technologies); they are not based on risk or impacts upon receiving waters. (See effluent
guidelines).
The Effluent Limitations Guidelines and Pretreatment Standards for the Asbestos Manufacturing Point
Source Category (40 CFR Part 427) do not require that industrial facilities monitor asbestos
concentrations in discharges. Rather, the regulations contain either a zero discharge of pollutants
standard or require that the discharger not exceed a specified release amount of pollutants including total
suspended solids (TSS), chemical oxygen demand (COD) and pH. These guidelines were originally
developed in 1974 and 1975 and were revised in 1995. These guidelines cover legacy uses such as
manufacture of asbestos cement pipe, asbestos cement sheet, roofing, paper, etc. and may not be
particularly useful to the COU of asbestos. Additionally, there are effluent guidelines for the chlor-alkali
industry under 40 CFR Part 415 that cover pollutants such as chlorine, mercury, and lead, but they are
not specific to asbestos. The EPA Industrial Waste Water Treatment Technology Database does not
currently include any data for asbestos._
Reasonably Available Data from Water Release Databases and Other Information
EPA investigated industry sector, facility, operational, and permit information regulated by NPDES
under the Clean Water Act to identify any permit limits, monitoring and reporting requirements, and any
discharge provisions related to asbestos and its COU. The Clean Water Act section 402 specifies that
Page 277 of 352

-------
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
point source pollutant dischargers into waters of the United States must obtain a permit to regulate that
facility's discharge. NPDES permits are issued by states, tribes, or territories that have obtained EPA
approval to issue permits or by EPA Regions in areas without such approval. Effluent limitations serve
as the primary mechanism in NPDES permits for controlling discharges of pollutants to receiving waters
and the NPDES permit data are cataloged into the Integrated Compliance Information System (ICIS) to
track permit compliance and enforcement status. NPDES permittees must then submit Discharge
Monitoring Reports (DMRs) to the appropriate permitting authority on a periodic basis to ensure
compliance with discharge standards for water quality and human health. Note that EPA does not
currently have data available on facilities that indirectly discharge wastewater to POTWs.
Available discharge data and permit information was accessed through EPA's Envirofacts and
Enforcement Compliance History Online (ECHO) database systems. EPA then investigated these data
sources for information pertinent to asbestos COU (chlor-alkali plants, sheet gasket stamping and
titanium dioxide plants) to identify if there is evidence of asbestos discharges or concentrations and/or
violations of their wastewater permits.
ICIS-NPDES information. ICIS-NPDES is an information management system maintained by EPA to
track permit compliance and enforcement status of facilities regulated by the NPDES under the Clean
Water Act. ICIS-NPDES is designed to support the NPDES program at the state, regional, and national
levels, and contains discharge monitoring and permit data from facilities in all point source categories
who discharge directly to receiving streams.
EPA identified pollutant parameter codes in ICIS-NPDES specific to asbestos (such as asbestos, fibrous
asbestos, asbestos (chrysotile), asbestos (amphibole), asbestos fibers (ambiguous asbestos), and non-
chrysotile, non-amphibole asbestos fibers) and identified unique NPDES-permitted facilities, outfalls,
and locations for those asbestos parameters. EPA then cross-checked their identified standard industrial
codes (SIC) with SIC codes associated with the current asbestos users and COU. The results were that
none of these identified SIC codes were associated with current chrysotile asbestos COUs in this Part 1
of the risk evaluation for asbestos.
EPA next did a specific NPDES permit search for facilities that may release asbestos (chlor-alkali and
sheet gasket facilities) based on gathered location and addresses for these sites. It was found that most
chlor-alkali facilities do have issued NPDES permits for industrial (major and minor permit status)
operations and for general stormwater and construction stormwater projects. Yet for the identified
permits for these industrial subcategories, none of the NPDES limits/monitoring requirements contained
asbestos or asbestos-related parameters codes or any direct effluent screening information for asbestos.
Based on the analysis, EPA found no current surface water releases of asbestos or exceedances in the
ICIS-NPDES database.
EPA's Water Pollutant Loading Tool. EPA's Water Pollutant Loading Tool calculates pollutant
loadings from NPDES permit and Discharge Monitoring Report (DMR) data from EPA's ICIS-NPDES
for industrial and municipal point source dischargers. Data are available from the year 2007 to the
present and also include wastewater pollutant discharge data from EPA's Toxics Release Inventory
(TRI). The Loading Tool was transitioned into ECHO to increase user access to data and streamline site
maintenance and EPA retired the legacy site (the Discharge Monitoring Report Loading Tool) on
January 24, 2018. DMR data identifies the permit conditions or limits for each water discharge location,
the actual values, identified by the permittee, for each monitored pollutant that was discharged, and
whether or not the amounts discharged exceeded the permit limits.
Page 278 of 352

-------
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
111
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
DMR was used to help identify facilities with current uses that discharge asbestos to surface water.
Information was obtained from the DMR Pollutant loading tool accessed on December 1, 2017.
Facilities were identified using two different search methods: 1) "EZ Search" which identifies facilities
that submit Discharge Monitoring Reports (DMRs) and 2) "Toxics Release Inventory (TRI) Search"
which identifies facilities that report releases to the TRI. Searches were conducted for the two most
current (and complete) years in the tool: 2015 and 2016 for DMR facilities, and 2014 and 2015 for TRI
facilities.
TRI data indicate no releases of asbestos in 2014 and 2015 (only friable asbestos is subject to reporting).
The DMR database reported just one facility reporting a discharge in 2014 and 2015 (accessed on
December 1, 2017) and this facility has been identified as a mining facility in Duluth, Minnesota. Later,
in a subsequent search (October 10, 2018) this facility was no longer identified on the DMR. The DMR
reported a total of zero pounds released in 2014 and 2015 but did provide maximum and average
effluent concentrations (mg/L) of allowable asbestos. It is assumed that the entry referred to mining
runoff, since asbestos has not been mined or otherwise produced in the United States since 2002. EPA
has currently not identified in the existing literature or through consultation with industry any evidence
of discharge to surface water from DMR or TRI database as to any current uses of asbestos (release
from sheet gaskets, release from working on industrial friction products and/or release from chrysotile
asbestos diaphragms from chlor-alkali facilities). Based on this database no water dischargers were
established.
EPA did a search of the database for the parameter description of asbestos and identified three facilities
reporting actual limit values of discharge of asbestos to surface water. One of facilities was the mining
facility identified earlier on DMR and the other was a quarry. The third was an electric facility. Two
other electric facilities were also reported. These facilities were not directly related to the current uses of
asbestos mentioned earlier.
STORET. STORET refers overall to "STORage and RETrieval", an electronic data system for water
quality monitoring data developed by EPA. Since about 2000, STORET has referred to a local data
management system ("Modernized STORET") as well as data repository ("STORET Data Warehouse")
developed for purposes of assisting data owners to manage data locally and share data nationally. Until
September 2009, the distributed STORET database has been used to compile data at the national level in
the STORET Data Warehouse. As of September 2009, the Water Quality Exchange, or WQX
framework, provides the main mechanism for submitting data to the STORET Data Warehouse.
EPA did not identify in STORET any evidence of discharge to surface water for the COUs of chrysotile
asbestos. EPA also did not identify in the existing literature or through consultation with industry any
evidence of discharge to surface water.
Page 279 of 352

-------
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
Appendix E Ecological Data Extraction Tables
The EPA has reviewed acceptable ecotoxicity studies for Chrysotile Asbestos according to the data
quality evaluation criteria found in the Application of Systematic Review in TSCA Risk Evaluations (U.S.
). The ten "on-topic" ecotoxicity studies for asbestos included data from aquatic organisms
{i.e., vertebrates, invertebrates, and plants) and terrestrial species {i.e., fungi and plants). Following the
data quality evaluation, EPA determined that four "on-topic" aquatic vertebrates and invertebrate studies
were acceptable while the two "on-topic" aquatic plants studies were unacceptable as summarized in the
Table APX E-l below. In the PF, it was determined that the terrestrial exposure pathways, including
biosolids, to environmental receptors was not within the scope of this assessment. As a result, EPA
excluded three studies on terrestrial species from further analysis as terrestrial exposures were not
expected under the conditions of use for asbestos. One amphibian study was excluded from further
review because it was not conducted on chrysotile asbestos. Ultimately four aquatic toxicity studies
were used to characterize the effects of chronic exposure of chrysotile asbestos to aquatic vertebrates
and invertebrates, as summarized in Table 3-1 Environmental Hazard Characterization of Chrysotile
Asbestos.
The results of these ecotoxicity study evaluations can be found in Final Risk Evaluation for Asbestos
Part 1: Chrysotile Asbestos, Systematic Review Supplemental File: Data Quality Evaluation of
Ecological Hazard Studies. The data quality evaluation indicated these studies are of high confidence
and are used to characterize the environmental hazards of Chrysotile Asbestos. The results of these
studies indicate that there are adverse effects to aquatic organisms following exposure to chrysotile
asbestos.
TableAPX E-l. Summary Table On-topic Aquatic Toxicity Studies That Were Evaluated for
				Chrysotile Asbestos.			
Species
Freshwater
/ Salt
Water
Duration
End-point
Concentration(s)
(MFL= Millions
of fibers per
liter)
Effect(s)
Reference
Data
Quality
Evaluation
Rating
Asiatic
Clams
Freshwater
30d
LOEC <
10s fibers/L
10s fibers/L
Gill Tissue
Altered
Bel anger et
al. (1986b)
High
(Corbicula
sp.)


(100 MFL)
100 MFL





30d
Reproducti
ve LOEC =
104 fibers/L
(0.01 MFL)
104-108 fibers/L
0.01-100 MFL
Increase in
Larvae
mortality/
decrease in
larvae released




96hr-30d
No
mortality
observed;
NOEC
>108
fibers/L
(>100
MFL)
102-108 fibers/L
0.0001-100 MFL
Mortality


Page 280 of 352

-------
Species
Freshwater
/ Salt
Water
Duration
End-point
Concentration(s)
(MFL= Millions
of fibers per
liter)
Effect(s)
Reference
Data
Quality
Evaluation
Rating


30d
LOEC= 10s
fibers/L
(100 MFL)
102-108 fibers/L
0.0001-100 MFL
Growth




30d
NOEC <
10s fibers/L
(<100
MFL)
LOEC =
108 fibers/L
(100 MFL)
102-108 fibers/L
0.0001-100 MFL
Fiber
Accumulation




96hr-30d
LOEC =
102 fibers/L
(0.0001
MFL)
102-108 fibers/L
0.0001-100 MFL
Siphoning
Activity


Asiatic
Clams
(Corbicula
fluminea)
Freshwater
30d
LOEC <
102 fibers/L
(<0.0001
MFL)
102-108 fibers/L
0.0001-100 MFL
Reduction in
siphoning
activity
Bel anger et
al. (1986a)
High


30d
LOEC <
108 fibers/L
(< 100
MFL)
108 fibers/L
100 MFL
Presence of
asbestos in
tissues


Coho
Salmon
(Onchorhy
nchus
kisutch)
Saltwater
and
freshwater
40-86d
NOEC =
1.5xl06
fibers/L
(1.5 MFL)
LOEC =
3.0xl06
fibers/L
(3 MFL)
1.5xl06 fibers/L,
3.0xl06 fibers/L
1.5 MFL, 3 MFL
Behavioral
stress
(aberrant
swimming,
loss of
equilibrium)
Sublethal
effects
including
epidermal
hypertrophy
superimposed
on
hyperplasia,
necrotic
epidermis,
Bel anger et
al. (1986c)
High
Page 281 of 352

-------
Species
Freshwater
/ Salt
Water
Duration
End-point
Concentration(s)
(MFL= Millions
of fibers per
liter)
Effect(s)
Reference
Data
Quality
Evaluation
Rating





lateral line
degradation,
and lesions
near the
branchial
region




40-86d
No
significant
Mortality;
NOEC
>3.0xl06
fibers/L
(>3 MFL)
1.5xl06 fibers/L,
3.0xl06 fibers/L
1.5 MFL, 3 MFL
Mortality




40-86d
No
Significant
effect;
NOEC
>3.0xl06
fibers/L
(>3 MFL)
1.5x10 6 fibers/L,
3.0xl06 fibers/L
1.5 MFL, 3 MFL
Growth


Green
Sunfish
(Lepomis
cyanellus)
Freshwater
52-67d
NOEC
<1.5xl06
fibers/L
(<1.5 MFL)
LOEC =
1.5x10 6
fibers/L
(1.5 MFL)
1.5xl06 fibers/L,
Behavioral
stress
(aberrant
swimming,
loss of
equilibrium)
Sublethal
effects
including:
epidermal
hypertrophy
superimposed
on
hyperplasia,
necrotic
epidermis,
lateral line
degradation,
and lesions
near the


Page 282 of 352

-------
Species
Freshwater
/ Salt
Water
Duration
End-point
Concentration(s)
(MFL= Millions
of fibers per
liter)
Effect(s)
Reference
Data
Quality
Evaluation
Rating





branchial
region




40-86d
No
significant
Mortality;
NOEC
>3.0xl06
fibers/L
(3 MFL)
1.5xl06 fibers/L,
3.0xl06 fibers/L
1.5 MFL, 3 MFL
Mortality


Japanese
Saltwater
13 -21 d
No
106-1010 fibers/L
Egg
Bel anger et
High
Medaka
(iOryzias
latipes)
and
freshwater

significant
effects;
NOEC
>106
fibers/L
(>1 MFL)
1 MFL-10,000
MFL
development,
hatchability,
survival.
al. (1990)



28d
LOEC =
106 fibers/L
(1 MFL)
NOEC =
104 fibers/L
(0.01 MFL)
106-1010 fibers/L
1 MFL-10,000
MFL
Significant
reduction in
growth of
larval
individuals




7w
Not
statistically
analyzed
104-108 fibers/L
0.01-100 MFL
Reproductive
performance
(viable
eggs/day,
nonviable
eggs/day)




49d
LCioo=1010
fibers/L
1010 fibers/L
10,000 MFL
100% Larval
mortality


Duckweed
(Lemna
Freshwater
28d
LOEC =
0.5jxg
0.5-5.0 (ig
chrysotile/frond
Decreased #
fronds
Trivedi, 2007;
Trivedi et al.
Unacceptabl
e
gibba)



0.5-5.0 (ig
chrysotile/frond
Decreased
Root length
(2004)

Page 283 of 352

-------
Species
Freshwater
/ Salt
Water
Duration
End-point
Concentration(s)
(MFL= Millions
of fibers per
liter)
Effect(s)
Reference
Data
Quality
Evaluation
Rating



chrysotile/f
rond
0.5-5.0 (ig
chrysotile/frond
Decreased
Chlorophyll
Content





NOEC <
0.5jxg
chrysotile/f
rond
0.5-5.0 (ig
chrysotile/frond
Decreased
Carotenoid
content





0.5-5.0 (ig
chrysotile/frond
Decrease in
biomass/ frond






0.5-5.0 (ig
chrysotile/frond
Decreased
Protein
content (mg/g
fresh wt)






0.5-5.0 (ig
chrysotile/frond
Decreased
Free sugar
(mg/g fresh
wt)






0.5-5.0 (ig
chrysotile/frond
Decreased
Starch (mg/g
fresh wt)






0.5-5.0 (ig
chrysotile/frond
Decreased
photosynthetic
pigments






0.5-5.0 (ig
chrysotile/frond
Increased lipid
peroxidation






0.5-5.0 (ig
chrysotile/frond
Increased
cellular
hydrogen
peroxide
levels






0.5-5.0 (ig
chrysotile/mL
Increase in
catalase
activity


219
Page 284 of 352

-------
220	Appendix F Environmental Fate Data Extraction Table
221	Environmental Fate Study Summary for Chrysotile Asbestos
222		Table APX F-l. Other Fate Endpoints Study Summary for Chrysotile Asbestos
Sjslem
Sinclj Tjpe (jesir)
Resnlls
( <1111 mollis
AITiliiilcri
Diilii
Qu;ili(>
r.\iilu;ilion
Reference
Results oi-
l-nil Sinclj
Report
Non guideline,
experimental study;
the effect of lichen
colonization on
chrysotile asbestos
structure is
investigated by
analyzing the
composition of both
colonized and
uncolonized field
samples. The effect of
oxalic acid exposure
on chrysotile asbestos
structure is also
investigated at various
concentrations.

In the three asbestos



Chrysotile fibers were
incubated in oxalic acid
solutions for 35 days to
observe its effect on MgO
content. Chrysotile (both
uncolonized or colonized
by lichens) from 3
serpentinite outcrops and
one asbestos cement roof
were collected.
outcrops and asbestos-
cement roof, MgO content
(wt %) was lower by 15-
20% in lichen colonized
chrysotile than in
uncolonized chrysotile.
Incubation in 50 mM oxalic
acid transformed chrysotile
fibers into "an amorphous
powdery material,
consisting mainly of pure
silica", and without fibrous
nature.
The
reviewer
agreed with
this study's
overall
quality
level.
Favero-
Lomgo et
al. (2.005)
High
Non guideline,
experimental study;
oxalic acid and citric
acid leaching of
asbestos rich sediment
Asbestos rich sediment
and a serpentine bedrock
sample underwent
leaching in 0.025 M
oxalic acid and 0.017 M
citric acid. Total
elemental analysis was
performed using
inductively coupled
plasma spectrometry
(ICPS), individual fiber
analysis was done using
energy dispersive x-ray
analysis (EDX) and a
scanning and
transmission electron
microscope (STEM).
ICPS results showed citric
acid was slightly more
effective at removing most
metals from the sediment
samples than oxalic acid;
however, EDX analysis of
individual fibers showed
Mg/Si ratios were reduced
from 0.68-0.69 to 0.07 by
oxalic acid and only to 0.38
by citric acid.
The
reviewer
agreed with
this study's
overall
quality
level.
Schreier et
!7)
High
Non-guideline,
experimental study;
decomposition study
of asbestos in 25%
acid or caustic
solutions
Chrysotile, crocidolite,
amosite, anthophyllite,
actinolite, and tremolite
asbestos fibers were
dissolved in 25% acid or
NaOH solution
Degradation in 25% HC1,
acetic acid, H3PO4, H2SO4
and NaOH, respectively
was reported for
Chrysotile (55.69, 23.42,
55.18, 55.75 and 0.99%),
Crocidolite (4.38, 0.91,
4.37, 3.69 and 1.35%),
Due to
limited
information
assessing
the results
were
challenging.
Soeil and
Leineweber
(1969)
Unacceptable
Page 285 of 352

-------
Sjsk'in
Siudj Tj|K' ijcsir)
Results
( (UllllH'lKS
AITiliiilcd
Reference
Diilii
Qu;ili(\
r.\iiiiiiiiioii
Kesulis ol-
l-iil 1 Sluclj
Report


Amosite (12.84, 2.63,
11.67, 11.35 and 6.97%),
Anthophyllite (2.66, 0.60,
3.16,2.73 and 1.22%),
Actinolite (20.31, 12.28,
20.19, 20.38 and 9.25%)
and Tremolite (4.77, 1.99,
4.99, 4.58 and 1.80%).



223
Page 286 of 352

-------
224 	Table APX F-2. Hydrolysis Study Summary for Chrysotile Asbestos
S(ihI\ Tjpe
(jesir)
pll
Temper;iliirc
Diii'iiliun
Kesnlls
(o m im-nls
AITiliiiU'ri
Reference
Diilii
Qii;ili(\
l-'\illllill ifill
Results oi-
l-nil SiihIj
Report
Non-guideline,
experimental
study;
dissolution of
chrysotile and
crocidolite
asbestos in water
at various pH
and
temperatures.
7, 7, 7, 9,
and 4 for
experiments
1-5,
respectively
44, 6, 25, 25,
and 25ฐC for
experiments
1-5,
respectively
170 or
1024
hours
170-hour study results
evaluating Mg removal
from Chrysotile
(proportion of 1 layer):
Experiments 1-4: 0.32-
0.94.
Experiment 5 (pH 4,
25ฐC): 8.84
170-hour study results
evaluating Si removal
from Chrysotile
(proportion of 1 layer):
Experiments 1-4: 0.5-0.25.
Experiment 5: 5.05.
170-hour study results
evaluating Mg removal
from Crocidolite
(proportion of 1 layer):
Experiments 1-5: 0.42-
1.80.
170-hour study results
evaluating Si removal
from Crocidolite
(proportion of 1 layer):
0.03-0.56.
1024-hour results
(proportion of one layer
removed) for experiment 3
only:
Chrysotile, Mg: 0.94; Si:
0.36 Crocidolite, Mg:
1.42; Si: 0.37
The
reviewer
agreed
with this
study's
overall
quality
level.
10W
87)
High
Non-guideline;
dissolution
study; sample
size, temperature
andpH
evaluated; pH
change over time
compared for
asbestos
minerals,
amosite and
crocidolite and
chrysotile
5.9-6.1
(initial)
5 to 45 ฐC
20 min;
1000
hours
Rate of dissolution is a
function of surface area
and temperature. Mg2+
may be continuously
liberated from fibers
leaving a silica skeleton.
The rate-controlling step
was determined to be
removal of brucite layer.
Smaller particles liberated
more magnesium
The
reviewer
agreed
with this
study's
overall
quality
level.
Choi
and
Smith
ซ I'Vj)
High
Non guideline;
experimental
study; a particle
Not
reported but
Not reported
but held
constant
3-5 days
Chrysotile in natural water
acquires a negative
surface charge by rapid
The
reviewer
agreed
Bales
and
High
Page 287 of 352

-------
Suier;iliirc
Dunilioii
Results
(o m im-nls
AITiliiiU'ri
Reference
Diilii
Qu:ilil>
l-'\iiliisil iซui
Results of
l ull Sluclj
Report
electrophoresis
apparatus was
used to monitor
absorption
properties of
chrysotile
asbestos aging in
water
held
constant


adsorption of natural
organic matter (<1 day).
Positively charged >Mg-
OH2+ sites are removed by
dissolution in the outer
brucite sheet resulting in
exposure of underlying
>SiO" sites.
with this
study's
overall
quality
level.
Morgan.
85)

225
Page 288 of 352

-------
226 	Table APX F-3. Aquatic Bioconcentration Study Summary for Chrysotile Asbestos
Siuclj Tjpe
(je.ir)
Iniliiil
( onccn(r;i(ion
Species
Dui'iilioii
Result
( ommcnls
AITiliiiled
Reference
Diilii
Qu:ilil>
l-'\iilnsil iซui
Resulis oi-
l-nil Sluclj
Report
Non-guideline;
experimental
study; uptake
monitoring of
chrysotile
asbestos in
Coho and
juvenile green
sunfish
1.5><106and
3.0x10s
fibers/L
Coho salmon
(Oncorhynchus
kisutch) and
juvenile green
sunfish
(Lepomis
cyanellus)
Coho
salmon: 86
and 40
days;
Green
sunfish: 67
and 52
days
Asbestos fibers were
found in the asbestos-
treated fish by
transmission
electron microscopy
(TEM); however total
body burdens were
not calculated.
Sunfish lost scales
and had epidermal
tissue erosion.
Asbestos fibers were
not identified in
control or blank
samples.
The
reviewer
agreed
with this
study's
overall
quality
level.
Bel anger
et al.
i 1986c)
High
Non-guideline;
experimental
study; uptake
monitoring of
chrysotile by
Asiatic clams
2.5x10s-
8.8xl09
fibers/L
Asiatic clams
('Corbicula sp.)
96-hours
and 30-
days
Chrysotile asbestos
was detected in clams
at 69.1ฑ17.1
fibers/mg whole body
homogenate after 96
hours of exposure to
108 fibers/L and food.
Chrysotile asbestos
was detected in clams
after 30 days of
exposure to 108
fibers/L at 147.3ฑ52.6
fibers/mg dry weight
gill tissue and
903.7ฑ122.9
fibers/mg dry weight
visceral tissue.
Chrysotile asbestos
was not detected in
clams after 96 hours
at all asbestos
exposure
concentrations tested
with no food.
The
reviewer
agreed
with this
study's
overall
quality
level.
Bel anger
et al.
il^86b)
High
Non-guideline;
experimental
study;
measuring
uptake of
chrysotile
asbestos by
Asiatic clams
0, 104, and 108
fibers/L
Asiatic clams
{Corbicula sp.,
collected in
winter and
summer)
30-days
Fibers were not
detected in clams
from blank control
groups and after
exposure to 104
fiber/L groups for 30
days.
Asbestos
concentration in tissue
after exposure to 108
fiber/L for 30 days
The
reviewer
agreed
with this
study's
overall
quality
level.
Bel anger
et al.
16a)
High
Page 289 of 352

-------




(fibers/mg dry weight
tissue) in winter
samples: Gills:
132.1ฑ36.4; Viscera:
1055.1ฑ235.9 and
summer samples:
Gill: 147.5ฑ30.9;
Viscera:
1127.4ฑ190.2.



Non-guideline;
experimental
study; BCF
determination
of asbestos in
the Asiatic
clam
0, 104, and 108
fibers/L
Asiatic clam
(corbicula sp.)
30 day and
field
exposed
BCF = 0.308 in gill
tissue, 1.89 in viscera
tissue, and 1.91 in
whole clam
homogenates after 30-
days exposure to 108
fibers/L.
Field exposed BCFs =
0.16-0.19 in gills,
64.9-102 in viscera,
1,442-5,222 in whole
clams.
The
reviewer
agreed
with this
study's
overall
quality
level.
Bel anger
et al.
(1987)
High
Non-guideline;
experimental
study;
chrysotile
asbestos
uptake study in
Japanese
Medaka
5.1ฑ2.8xl06,
7.6ฑ8.1xio8
fibers/L
Japanese
Medaka
(Oryzias
latipes)
13 weeks
After 28 days of
exposure to chrysotile
asbestos at 1010
fibers/L
concentrations, fish
total body burden was
375.7 fibers/mg. After
3 months of exposure
to chrysotile asbestos
at 108 fibers/L
concentrations, fish
total body burden was
486.4ฑ47.9 fibers/mg.
The
reviewer
agreed
with this
study's
overall
quality
level.
Bel anger
et al.
(1990)
High
227
Page 290 of 352

-------
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
Appendix G SAS Codes for Estimating Kl and Km from
Grouped Data
/*This SAS code estimates a value for lung cancer potency (KL) using Poisson maximum likelihood
estimation (MLE), along with the 90% confidence interval (CI) generated using the likelihood profile
method. The basic model is RR = 1+ CE10 * KL.
This code was created by Rebekha Shaw and Bill Thayer at SRC Inc. This is version 1.0 /*
/*This is where the code begins execution. */
/*The first step is to create a data table */
data Data_Table;
input CE10_min CE10_max CE10_mid Observed Expected RR;
/*enter data here */
datalines;
0 20 10.0 6 5.75 1.04
20 100 60.0 12 2.82 4.25
100 450 275.0 17 1.57 10.82
450 1097 773.5 21 1.23 17.07
/* Enter text string to identify data source */
title "Wang et al 2013";
/*model*/
proc nlmixed data=Data_Table;
parms KLE2 10; /* KLE2 = KL*lE+02. The initial guess is 10. This can be changed if a solution is not
found (unlikely). */
Predicted = (l+CE10_mid*KLE2/100)* Expected; /^equation to calculate predicted number of lung cancer
cases*/
LL=LogPDF("POISSON",Observed,Predicted); /*LogPDF function Returns the logarithm of a probability
density (mass) function. Poisson distribution is specified. */
model Observed ~ general(LL);
estimate 'KLE2' KLE2 ALPHA=0.1;/^generates "Additional Estimates" table in the Results tab with Wald 90%
CI's*/
predict Predicted out=Predicted alpha=0.1; /^generates SAS data table with predicted values and CI's
titled "Predicted"*/
ods output FitStatistics = FitStats;
ods output ParameterEstimates = ModelParams;
Proc print data=Predicted;/*Prints the "Predicted" table in the Results tab*/
run;
data _null_;
set Fitstats;
if _n_ =1;
LLTarget = (Value/-2)-1.353;/*calculates LL_target - needed to run macro PoissonLLBounds*/
call symputx("LLTarget",LLTarget);/*creates macro variable*/
run;
data _null_;
set ModelParams;
KLMLE = Estimate*le-02; /*variable KL_MLE in macro PoissonLLBounds*/
KLINITLB= Estimate*le-02/10; /*Calculates the initial guess for the lower
macro poissonLLBounds*/
KLINITUB= Estimate*le-02*10; /*Calculates the initial guess for the upper
macro PoissonLLBounds*/
call symputx("KLMLE", KLMLE);/*creates macro variable*/
call symputx("KLINITLB", KLINITLB);/*creates macro variable*/
call symputx("KLINITUB", KLINITUB);/*creates macro variable*/
run;
/*This is the macro which calculates the 90% confidence interval using the likelihood profile method. It
is executed after the MLE solution has been found */
%macro PoissonLLBounds(inputData=, KL_MLE=, KL_Init_LB=, KL_Init_UB=,
conv_criterion=, LL_target=, max_iteration=);
bound - variable KL_itit_LB in
bound - variable KL itit LB in
Page 291 of 352

-------
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
%Let dsid=%sysfunc(open(&inputdata));	* open the input data file;
%Let NumSamples=%sysfunc(attrn(&dsid,nobs)); * get the number of observations;
%Let rc=%sysfunc(close(&dsid) ) ;	* close the data file;
%Do j=l %To 2; * one for upper bound and one for lower bound;
%If %eval(&J=l) %then %Let KL=&KL_init_LB;
%If %eval(&J=2) %then %Let KL=&KL_Init_UB;
%Let i=l; * first time through loop;
%Let ConvFactor = 10;
%let ConvRate = %sysevalf(((&KL_MLE-&KL)/&KL_MLE)/10);
%Let ConvDirect = -1;
/* negative=from the left and positive=from the right. For lower bound, the initial guess is less than
the target LL so the initial value of convdirect is -1 */
%Let KLAdjust=%Sysevalf(-l*&ConvDirect*&KL*&ConvRate);
%Do %Until (%sysevalf(&DeltaLL < &conv_criterion) OR %sysevalf(&i > &max_iteration));
Data tempDataLLBound; Set &InputData;
Predicted = (1 + CE10_Mid * &KL) * Expected;
LL=(LogPDF("POISSON",Observed,Predicted) ) ; * likelihood for each
observation;
LL_sum+LL;
output;
Run;
Data TempDataLLBound2; Set tempDataLLBound;
If _N_= &NumSamples;
NumLoops=&i;
thisKL=&KL;
ConvRateVar=&ConvRate;
ConvFactorVar=&ConvFactor;
ConvDirectVar= %eval(&ConvDirect);
KLAdjustVar=(-l*ConvDirectVar)*thisKL*ConvRateVar;
If &ConvDirect=-l then DiffLL=abs(LL_sum)-abs(&LL_Target);
Else DiffLL=abs(&LL_Target)-abs(LL_Sum);
/*	Test if we have changed direction on the convergence. If we have, change direction
(subtract from current value if we were adding before...) and decrease the convergence rate
(ConvRate) by a factor = ConvFactor. */
if DiffLL<0 then
do; /* need to change directions and make conv rate more gradual */
ConvDirectVar= %eval(-l*&ConvDirect);
ConvRateVar=%sysevalf(&convRate/&ConvFactor) ;
KLAdjustVar=(-l*ConvDirectVar)*thisKL*ConvRateVar;
call symput('KLAdjust',KLAdjustVar);
call symput('ConvDirect',ConvDirectVar);
call symput('convRate',ConvRateVar);
end;
AbsDiffLL=abs(DiffLL) ;
call symput('DeltaLL',ABsDiffLL);
output;
Run;
Data tempAllOutput; if _N_=1 then Set TempDataLLBound2; Set tempDataLLBound; Run;
%If %eval(&i=l) %then %do; Data AllOutput; Set tempAllOutput; Run; %end;
%If %eval(&i>l) %then %do; Proc Append base=A110utput data=tempA110utput; Run;
%End;
%Let i=%eval(&i+l);
%Let KL=%sysevalf (&KL + ScKLAdjust) ;
%End;
Page 292 of 352

-------
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
%If %eval(&J=l) %then
%Do;
Data tempoutl; length limit $5;
estimate=thisKL; LogLikelihood=LL_sum; loops=numloops; Run;
%End;
%If %eval(&J=2) %then
%Do;
Data tempout2; length limit $5;
estimate=thisKL; LogLikelihood=LL_sum; loops=numloops; Run;
%End;
%End;
Data PrntOutput; Set tempoutl tempout2; run;
Proc print data=PrntOutput; var limit estimate LogLikelihood Loops ; Run;
%Mend;
/*run macro PoissonLLBounds*/
%PoฑssonLLBounds(i nput Da t a=Da t a_T ab1e,
KL_MLE=&KLMLE,
KL_In i t_LB=&KLINIT LB,
KL_Init_UB=&KLINITUB,
conv_criterion= 0.001,
LL_target=&LLTarget,
max_iteration=100);
run;
/*the following code creates a summary table with the MLE KLE and confidence bounds*/
PROC SQL;
CREATE TABLE WORK.MLEKL AS
SELECT ("MLE KLE") AS Parameter,
(tl.Estimate*le-2) AS Value
FROM WORK.MODELPARAMS tl;
QUIT;
PROC SQL;
CREATE TABLE WORK.LBKLUBKL AS
SELECT (case
when tl.limit="lowern then "5% LB KL"
else "95% UB KL"
end) AS Parameter,
tl.estimate AS Value
FROM WORK.PRNTOUTPUT tl;
QUIT;
PROC SQL;
CREATE TABLE WORK.Parameter_Values AS
SELECT * FROM WORK.MLEKL
OUTER UNION CORR
SELECT * FROM WORK.LBKLUBKL
Quit;
Proc print data=Work.Parameter_values;
run;
Set TempDataLLBound2; limit='lower';
Set TempDataLLBound2; limit='upper';
Page 293 of 352

-------
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
/*This SAS code estimates a value for mesothelioma potency (KM) using Poisson maximum likelihood estimation (MLE), along with the 90% confidence interval (CI)
generated using the likelihood profile method.
This code was created by Rebekha Shaw and Bill Thayer at SRC Inc.
This is version 1.0*/
/*This is where the code begins execution. */
data Data_Table;
input TSFE_Min TSFE_Max TSFE_Mid Duration Cone PY Obs ;
/*The values of TSFE_Mid and Duration are used to calculate a parameter called Q. */
if	TSFE_Mid=. then Q = . ;
else if TSFE_Mid<10 then Q = 0;
else if TSFE_Mid>(10+duration) then
Q = (TSFE_Mid-10)**3-(TSFE_Mid-10-duration)**3;
else Q =(TSFE_Mid-10)**3;
/*enter data here. The contents of the columns are as follows:
TSFE_Min (years)
SFE_Max (years)
TSFE_Mid (years)
Duration (years)
Cone (f/cc)
Person Years (PY)
Observed cases(Obs)
*/
datalines;
20 30 27.7 1.00 6.5 1926 0
30 40 33.9 2.10 8.7 6454 0
40 50 43.1 3.00 14.6 3558 2
50 72 53.56 5.78 31.4 1080 2
/*enter the name of the data set*/
title "North Carolina Sub Co-hort (1999-2003;4 groups)";
/*model*/
proc nlmixed data= Data Table;
parms KME8 10; /*KME8 is equal to KM*lE+08. The starting guess is 10. This can be changed in the unexpected case where a solution is not found*/
Pred = Conc*Q*PY*KME8/le+08; /*equation to calculate predicted values*/
LL=LogPDF("POISSON",Obs,Pred); /*LogPDF function Returns the logarithm of a probability density (mass) function. Poisson distribution is specified.*/
model Obs ~ general(11);
estimate 'KME8' KME8 ALPHA=0.1;/*generates "Additional Estimates" table in the Results tab with 90% Wald CI's - this can be deleted if we do not want the Wald CIs
displayed in the SAS output */
predict Pred out=Predicted alpha=0.1; /*generates SAS data table with predicted values and CI's titled "Predicted"*/
ods output FitStatistics = FitStats;
ods output ParameterEstimates = ModelParams;
Proc print data=Predicted;/*Prints the "Predicted" table in the Results tab*/
OPTIONS MPRINT SYMBOLGEN ;/*this prints in the log what value is used for each variable in the macro*/
data nul1 ;
set Fitstats;
if _n_ =1;
LLTarget = (Value/-2)-1.353;/*calculates LL_target - needed to run macro PoissonLLBounds*/
call symputx("LLTarget",LLTarget);/*creates macro variable*/
data nul1 ;
set ModelParams;
KMMLE = Estimate*le-8; /*scales back the KM MLE value generated by Proc nlmixed - variable KM_MLE in macro PoissonLLBounds*/
KMINITLB= Estimate*le-8/l0; /*Calculates the initial guess for the lower bound - variable KM_itit_LB in macro poissonLLBounds*/
KMINITUB= Estimate*le-8*10; /*Calculates the initial guess for the upper bound - variable KM_itit_LB in macro PoissonLLBounds*/
call symputx("KMMLE", KMMLE);/*creates macro variable*/
call symputx("KMINITLB", KMINITLB);/*creates macro variable*/
call symputx("KMINITUB", KMINITUB);/*creates macro variable*/
/*This is the macro which calculates the 90% confidence interval using the likelihood profile method. It is executed after the MLE solution has been found */
PoissonLLBounds(inputData=, KM_MLE=, KM_Init_LB=, KM_Init_UB=,
conv_criterion=, LL_target=, max_iteratii
Page 294 of 352

-------
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
%Let dsid=%sysfunc(open(Sinputdata));	* open the input data file;
%Let Num.Sam.ples=%sysfunc (attrn( &dsid, nobs)) ; * get the number of observations;
%Let rc=%sysfunc(close(&dsid));	* close the data file
%Do j=l %To 2; * one for upper bound and i
%If %eval(&J=l) %then %Let KM=&KM_init_LB;
%If %eval(&J=2) %then %Let KM=&KM_Init_UB;
%Let i=l; * first time through loop;
%Let ConvFactor = 10;
%let ConvRate = %sysevalf(((&KM_MLE-&KM)/&KM_MLE)/10);
%Let ConvDirect = -1;
,egative=from the left and positive=from the right. For lower bound, the initial guess is less than the target LL so the initial value of convdirect .
%Let KMAdjust=%Sysevalf(-1*SConvDirect*&KM*SConvRate);
%Do %Until (%sysevalf(SDeltaLL < &conv_criterion) OR %sysevalf(&i > &max_iteration));
Data tempDataLLBound; Set SlnputData;
E = Cone * Q * PY * &KM;
LL=(LogPDF("POISSON", Obs, E)); * likelihood for each observation;
LL_sum+LL;
output;
Run;
Data TempDataLLBound2; Set tempDataLLBound;
I f _N_= StNumS ampl e s ;
NumLoops=&i;
t hi sKM= &KM;
ConvRateVar=&ConvRate;
ConvFactorVar=&ConvFactor;
ConvDirectVar= %eval(SConvDirect);
KMAdjus tVar =(-1*C onvDirectVar)*thisKM*C onvRat eVar;
If &ConvDirect=-l then DiffLL=abs(LL_sum)-abs(&LL_Target);
Else DiffLL=abs(&LL_Target)-abs(LL_Sum);
/*	Test if we have changed direction on the convergence. If we have, change direction (subtract from current value if we were addi:
gence rate (ConvRate) by a factor = ConvFactor.*/
if DiffLL<0 then
.eed to change directions and make conv rate more gradua
ConvDirectVar= %eval(-l*&ConvDirect);
ConvRateVar=%sysevaIf(&convRate/&ConvFactor) ;
KMAdjustVar=(-l*C onvDirectVar)*thisKM*C onvRat eVar;
call symput('KMAdjust', KMAdjustVar);
call symput('ConvDirect',ConvDirectVar);
call symput('convRate',ConvRateVar);
AbsDiffLL=abs(DiffLL) ,
call symput('DeltaLL',ABsDiffLL)
Data tempAHOutput; if _N_=1 then Set TempDataLLBound2; Set tempDataLLBound; Run;
%If %eval(&i=l) %then %do; Data AllOutput; Set tempAHOutput; Run; %end;
%If %eval(&i>l) %then %do; Proc Append base=AllOutput data=tempAllOutput; Run; %End;
%Let KM=%sysevalf(&KM + SKMAdjust)
%If %eval(&J=l) %then
Page 295 of 352

-------
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
Data tempoutl; length limit $5; Set TempDataLLBound2; limit='lower'; estimate=thisKM; LogLikelihood=LL_sum;
1 oops =numloops; Run;
% End;
%If %eval(&J=2) %then
%Do;
Data tempout2; length limit $5; Set TempDataLLBound2; limit='upper'; estimate=thisKM; LogLikelihood=LL_sum;
1oops=numloops; Run;
Data PrntOutput; Set tempoutl tempout2; run;
Proc print data=PrntOutput; var limit estimate LogLikelihood Loops ; Run;
PoissonLLBounds*/
%PoissoziLLBovi2ids C inputData=Data Ta!ble,
KM_MLE= &KMMLE,
KM_Init_LB=&KMINITLB,
KM_Init_UB=&KMINITUB,
conv_criterion=0.001,
LL_target=&LLTarget,
max iteration=100);
Page 296 of 352

-------
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
Appendix H BEIRIV Equations for Life Table Analysis
Lung Cancer
Let ei be the calculated excess relative risk of lung cancer in an exposed individual at age i.
Then:
Excess Lifetime Risk = Rc/; — R0lt
110
m„=Yrฐ,
i-1
110
RC/( = 2 Rc/
i-1
*o,=ฃ-su(i -h)
he
Re, =-r-rSeli(l-qei)
hei
hej = ht (1 + et)
he* = h* + hiei
It =exp(-/?;>
qet = exp (—he*)
> II ฆ
7=1
i-1
Sen =YlveJ
j=i
where:
i and j	=	Year index (1 = year 0-1, 2 = year 1-2, etc.)
ROit	=	Lifetime risk of lung cancer in the absence of exposure
Reit	=	Lifetime risk of lung cancer in the presence of exposure
R0i	=	Risk of lung cancer in the absence of exposure in year i
Rei	=	Risk of lung cancer the presence of exposure in year i
hi	=	Lung cancer incidence rate in the absence of exposure in year i
hi*	=	All-cause mortality rate in the absence of exposure in year i
qi	=	Probability of surviving year i, all causes acting (no exposure)
qei	=	Probability of surviving year i, all causes acting (with exposure)
Si,i	=	Probability of surviving up to start of year i, all causes acting (no exposure)
Sei,i	=	Probability of surviving up to start of year i, all causes acting (with exposure)
Page 297 of 352

-------
641	Mesothelioma
642
643	The same basic approach is followed for calculating lifetime risk of mesothelioma, except that the
644	baseline (un-exposed) risk is so small that it is generally assumed to be zero. Thus, the equations for
645	calculating lifetime mesothelioma risk are the same as above, except as follows:
646
647 m; = risk of mesothelioma in an exposed individual at age i
648
110
650
m
651 Re, = —rSeu(l-qei)
hej
Page 298 of 352

-------
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
Appendix I SAS Code for Life Table Analysis
There are three SAS programs in this appendix:
•	Lung Cancer Lifetable for Linear Models
•	Lung Cancer Lifetable for Non-Linear Models
•	Mesothelioma Lifetable
SAS Lung Cancer Lifetable for Linear Models
OPTIONS NODATE NONUMBER orientation=landscape linesize=max; *BT added 7/3/19;
/*
This program calculates the risk of lung cancer from inhalation exposure to asbestos,
using a lifetable approach based on BEIR IV. The basic exposure-response model is RR = 1 +
CE10*KL.
The basic code for the lifetable calculations were developed and provided to EPA
by Randall Smith at NIOSH. The code from NIOSH calculates the baseline risk (R0) and the exposed
risk (Rx)
from exposure to an exposure concentration of X_Level using NIOSH Model 2: Rx = R0 * (1 + COEF *
X) .
EPA has modified the NIOSH code as follows:
1)	The all-cause mortality and cause-specific (lung cancer) incidence data tables have been
updated based on CDC Wonder 2017.
2)	An equation has been added to calculate extra risk: Extra_Risk = (Rx - R0) / ( 1 - R0)
3)	A macro has been added to find the exposure level (X_Level) that yields an extra risk of 0.01
(1%) •
This is referred to as EC1%, which may then be used to calculate the unit risk: UR = 0.01 /
EC1%
*/
/* .\Beta Version.sas 19jan00, 26jul00, 25oct01, 06dec05, 30novl8
Experimental version
	*/
title "Lifetable calculation of lung cancer risk";
title2 "under a linear relative rate model";
Compute excess risk by the BEIR IV method using SAS datasteps.
These programs compute the risk of a cause-specific
death in the presence of competing risks, where the cause-
specific death-rate is modeled either as a relative rate
[h=h0*f(Coef*X)] or as an absolute rate [h=h0+f(Coef*X)]
where
h denotes the cause-specific death-rate,
X denotes cumulative occupational exposure (with Lag)
Coef denotes the coefficient for the effect of exposure and
hO is the corresponding rate at baseline (X=0).
(Except for Coef, these are functions of age.)
A few simple models of f(Coef*X) are easily specified as
Page 299 of 352

-------
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
I	described below. More complicated models can be specified with	|
I	a little more work. (For a more complicated example,	|
|	see \_GENERAL.LIB\PROGRAMS\SAS\BEIR-4.Method\BEIR4ex2.SAS).	|
I	I
+Reference:	+
I	Health Risks of Radon and Other Internally Deposited Alpha-	|
I	Emitters (BEIR IV). Commitee on the Biologic Effects of	|
I	Ionizing Radiations. National Academy Press. Wash. DC (1988).	|
I	See especially pages 131-136.	I
+USER-SUPPLIED ASSIGNMENTS:	+
>	The following macro variables are assigned using "%LET" state-
ments: MODEL, COEF, LAG, AGE1ST_X, DURATION, LASTAGE.
Further information appears below.
>	Exposure concentrations for computing risk are defined
in the datastep "X_LEVELS."
>	All-cause mortality information is entered as a life-table in
the data step "ALLCAUSE," and converted to rates per individual.
>	Cause-specific incidence information for unexposed referents is
entered as rates per 100,000 and converted to rates per
individual in the data step "CAUSE."
+NOTES:	+
|> Datastep "EX_RISK" is where the desired risks are computed.	|
1 ....	1
|> If the unexposed(referent) cause-specific incidence rate is from|
I a model then datastep "CAUSE" with variables AGE and RATE as	|
I modeled can be modified to incorporate this. However, care	|
I must be taken in calculating confidence limits since imprecision|
I in the estimates of all of the parameters of the model	|
I contributes to the imprecision of excess risk estimates.	|
I	I
|> This program is currently set up to apply the Linear Rel. Rate	|
I model (Lag= 0) and accumulation of excess risk is over the	|
I rates in ALLCAUSE and CAUSE unless truncated at a younger age.	|
| (See LASTAGE below.)	I
+	SAS Programmer: Randall Smith +
I	The Nat'1 Inst, for Occupational Safety & Health |
|	2 6 j ul2000, 23j ul2001, 25oct2001, 18nov2018 |
+	Modifications: +
I	26jul00 Fix the procedure bug causing it to report incorrectly |
I	the age at which accumulation of risk was stopped |
I	whenever the age-specific rates included ages |
I	before the value of &Agelst_X. (&Agelst_X is a macro|
I	expression defining the age exposure begins.) |
I	I
I	23jul01 Make changes to facilitate multiple applications of |
I	BEIR4 algorithm, i.e., MLE(Excess Risk), UCL(ExcessRisk),|
I	searching for concentrations for a fixed risk. These |
I	changes involve defining Macros named BEIR4 and SEARCH |
I	given below with code illustrating these uses for the |
I	linear relative rate model. I
I	I
I	25oct01 Modified to add Macro variable EnvAdj for whether to |
I	increase inhaled dose from intermittent occupational |
I	exposures to continuous environmental exposures |
I	and update US rates for Gibb et al. cohort. |
I	I
I	30novl8 A bug that prevented the calculation of excess risks |
I	after incorporating an adjustment from intermittent |
I	occupational exposures to continuous exposures is fixed. |
I
|	March 2019: BT (SRC) Added maxro CONVERGE_BEIR4 which iteratively
|	runs macro BEIR4 until the EXPOSURE_CONCENTRATION corresponds to an
Page 300 of 352

-------
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
extra_risk=0.01 (the point of departure [POD])
Macro CONVERGE_BEIR4 works with one value for the exposure
variable XLevel (i.e., when the data C Levels includes one record.)
The intent was to make as few changes to BEIR4 as possible. The data
X_LEVELS and variable XLevel are retained but the initial value of
XLevel is provided in the call to macro CONVERGE_BEIR4 (the value
of Xlevel in the cards statement is not used in the calculations.
Changes to the BEIR4 macro are in Part III and Part IV, and are
indicated by the letters BT.
I In addition to the parameter values that are specified by the user |
I in PART 1, and the user-provided data entered in Part II, parameters |
I for the new macro CONVERGE_BEIR4 are specified in the call to the |
| macro CONVERGE_BEIR4 (see end of this SAS program file below).
+	*/
/* PART I. USER-SUPPLIED ASSIGNMENTS (Macro variables):
	+
exposure effects:	|
/ +
Model of cumulative
1	=> Loglinear Relative rate
R=R0*exp(COEF*X)
2	=> Linear Relative rate,
R=R0*(l+COEF*X)
3	=> Absolute rate,
R=R0+COEF*X
4	=> Power relative rate
R=R0*(1+X)ACOEF
0 => User Defined & programmed
in datastep Ex_Risk below
kf %Let Model
2;
Cumulative exposure parameter:	*/ %Let COEF = 1.2e-3;
Lag or delay between exposure and effect: */ %Let Lag	= 10;
Age exposure begins:
Exposure duration (years):
Adjust dose from occupational to
continuous environmental exposures (Y/N)?
Age to stop accumulating excess risk
(supposing rates are available for
ages >= &LastAge); otherwise use all of
the supplied rate information:
%Let Agelst_x
%Let Duration
16;
40;
kf %Let EnvAdj
%Let LastAge
Yes;
= 85;
/* PART II. USER-SUPPLIED ASSIGNMENTS (Datesets AllCause, Cause, X Levels ): */
data AllCause (label="Unexposeds' age-spec mortalty rates (all)"
drop=Lx rename=(BLx=Lx) );
/*	
I Input lifetable and calculate the corresponding age-specific
I (all-causes) mortality rate (AllCause) and conditional survival
I probability for each year of age (qi) together with
I the corresponding values of age (Age).
Label Age
BLx
Lx
CndPrDth
qi
AllCause
'Age at start of year (Age=i)"
'Number alive at start of year"
'Number alive at end of year"
'Pr[Death before age i+1 | alive at age
'Pr[Survive to age i+1 | Alive at age i]
'Age-spec mortality rate (all causes)";
Page 301 of 352

-------
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
if _n_=l then input age //// 61 BLx /* ////
input Lx
CndPrDth = (BLx - Lx)/BLx;
skip next 4 lines */
qi	= 1-CndPrDth;
if qi <= 0 then AllCause
else AllCause
le+50;
- log(qi);
if age
BLx=Lx;
age+1;
retain age BLx;
&LastAge then output; else STOP;
cards;
0
Life-table starting age. (Required: Values must begin 4 lines down!)
The following are 2017 Life-table values of US population
starting at birth and ending at age 85.
(Source: Nat.Vital Statistics Reports 2019 Vol 68 No 7, Table 1,
https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_07-50 8.pdf)
100000 99422 99384 99360 99341 99326 99312 99299 99288 99278
99268 99259 99249 99236 99217 99191 99158 99116 99066 99006
98937 98858 98770 98674 98573 98466 98355 98241 98122 97999
97872 97740 97603 97461 97314 97163 97006 96843 96674 96501
96321 96135 95939 95732 95511 95275 95023 94753 94461 94144
93797 93419 93008 92560 92070 91538 90963 90345 89684 88978
88226 87424 86570 85664 84706 83696 82632 81507 80315 79048
77697 76265 74715 73064 71296 69418 67402 65245 62933 60462
57839 55053 52123 49035 45771 42382
*run;*BT 7/3/19 added Run statement here;
data CAUSE (label="Unexposeds' age-cause-spec mortalty rates");
/ *	+
I Specify unexposeds' age-specific mortality rates (per year) |
I from specific cause.	I
_i	-k /
label Age	= "Age"
Rate_e5 = "Age,cause-specific rate per 100,000"
Rate	= "Age,cause-specific rate per individual";
if _n_ = 1 then input age
///;
input Rate_e5 @@;
/* input starting age	*/
/* III => skip next 3 lines */
Rate
Rate e5
le-5; /* Convert to rate per individual */
if age <= 4
then DO; output; age+1; END;
else DO i = 0,1,2,3,4;	/*	*/
if age < &LastAge /* Fill out into yearly intervals from */
then output; /* inputted five year intervals after age 4*/
age+1;	/*	*/
END;
cards;
0 = Start age of cause-specific rate (Required: Rates begin 3 lines down!)
The following are 2017 cancer site code 22030 lung and bronchus incidence rates per
100,000 for US pop'n starting at birth.
For ages 5 and above, each rate holds for the age thru age+4 years.
Source: CDC Wonder, https://wonder.cdc.gov/cancer-v2 017.HTML
0.205 0.100 0.100 0.100 0.100 0.039 0.039 0.104 0.299 0.553 1.267 2.600 6.534 16.528 44.403
96.098 149.112 223.906 319.322 391.202 395.215
''run; *BT 7/3/19 added Run statement here;
data X_LEVELS (label= "Exposure levels (e.g., concentrations)" );
I*	+
I Specify environmental exposure levels	I
I and update label for the variable, XLevel, if necessary:	|
+	*/
/*	+
Page 302 of 352

-------
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
BT 3/8/19: Add maxro CONVERGE_BEIR4 which iteratively runs macro	|
BEIR4 until the EXPOSURE_CONCENTRATION corresponds to extra risk=0.01|
I The intent was to make as few changes to BEIR4 as possible. The data |
I X_LEVELS and variable XLevel are retained but the initial value of |
I XLevel is provided in the call to macro CONVERGE_BEIR4 (the value
I of Xlevel in the cards statement is not used in the calculations. |
+	*/
input XLevel
label XLevel= "Asbestos exposure (F/ml)";
cards;
0.0383
run;
%Macro BEIR4;
/* March 2019 - BT (SRC): Macro BEIR4 is now called by macro CONVERGE_BEIR4.
*/
/* 23jul01 modification */
/* Enclose the actual calculations and printed results in a macro	*/
/* to facilitate multiple applications of the algorithm.	*/
/* PART III. Perform calculations:	*/
data EX_RISK (label = "Estimated excess risks [Method=BEIR IV]"
/*keep = XLevel Rx ex_risk RskRatio R0 extra_Risk */
rename= (Rx=Risk));
/*	+
I Calculate risk and excess risk for each exposure concentration!
I in work.X_Level by BEIR IV method using information in	|
I work.AllCause and work.Cause to define referent population: |
_i	* f
format rate F15.8 hi F15.8; *BT 7/3/19: added the format statement;
length XLevel 8.;
Age at start of year (i)"
Exposure duration midway between i & i+1"
CElO(adj) (f/cc-yrs)"
label Age
XTime
(qi) '
XDose
R0
Rx
Ex_Risk
RskRatio
hi
hix
hstari
hstarix
qi
S_li
S lix
Cumulative Risk of lung cancer (unexposed) (R0)"
Cumulative risk of lung cancer (exposed) (Re)"
Excess risk (Rx-Ro)"
Ratio of risks (Rx/Ro)"
Lung Cancer hazard (unexposed) (hi)"
Lung Cancer hazard (exposed) (hei)"
All cause hazard (unexposed) (h*i)"
All cause hazard (exposed) (he*i)"
Probability of surviving year i assuming alive at start (unexposed)
Probability of surviving to end of year i (unexposed) (Sl,i)"
Probability of surviving to end of yeari (exposed) (Sel,i)";
/* BT 3/8/19: Calculation of unexposed's risk (following DO LOOP) could be omitted
from the iteration
but may require further changes to BEIR4(?).
*e.g.f %if i=l %then %do;*/
if _n_=l then DO;
/* Calculate unexposed's risk (R0) to be retained	*/
/* based on equation 2A-21 (pg. 131) of BEIR IV:	*/
/* Initialize: */ S_li =1; R0 = 0;
DO pointer = 1 to min(n_all,n_cause) until (age>=&LastAge-l);
set allcause (keep=age AllCause rename=(AllCause=hstari))
point=pointer nobs=n_all;
set cause (keep=age Rate rename=(age=ageCause Rate=hi))
point=pointer nobs=n_cause;
Page 303 of 352

-------
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
conform
if Age NE AgeCause then
put "** WARNING: Age values in datasets ALLCAUSE and CAUSE don't
/	@13 "Rates misaligned on age could give incorrect results'
/	@13 Pointer=
+2 "Age(ALLCAUSE)=" Age +2 "Age(CAUSE)=" AgeCause /;
qi = exp(-hstari);
R0 = R0 + ( hi/hstari * S_li * (1-qi) );
S_li = S_li * qi;
END;
END;	/* End of 'if _n_=l then DO;' stmt */
retain R0;
/* Calculate exposed's risk (Rx, renamed to Risk) for each exposure level
*/
/* ultimately based on equation 2A-22 (pg. 132) of BEIR IV	*/
/* but re-expressed in a form similar to equation 2A-21:	*/
* BT 3/20/19. This version of CONVERGE_BEIR4 will work when there
one concentration in data set x_levels -
i.e., one value for xlevel.
The Do loop for X_levels is commented out;
*DO pointX = 1 to No_of_Xs;
* set x_levels point=pointX nobs=No_of_Xs; /* BT 3/8/19: determines
Nobs is set at compilation,
so the value of nobs is available at first run through loop -
just one record and one variable (XLevel) in dataset x_levels. */
/* BT 3/20/19: added the next line to set the exposure
concentration = current value of &exposure_conc. */
xlevel = &exposure_conc;
is
when to end the loop.
/* Initialize : */ S_lix =1; Rx = 0;S_li=l; R0=0;
DO pointer = 1 to min(n_all,n_cause) until (age>=&LastAge-l);
set allcause (keep=age AllCause rename=(AllCause=hstari))
point=pointer nobs=n_all;
set cause (keep=Rate rename=(Rate=hi))
point=pointer nobs=n_cause;
XTime = min( max(0,(age+0.5-&Agelst_x-&Lag))
, ^Duration);
to Environmental Conversion */
then
if UpCase("&EnvAdj") = "YES" /* Occupational
XDose = XLevel
365/240	/* Days per year	*/
20/10	/* Ventilation (L) per day */
XTime;
ELSE if UpCase("&EnvAdj") = "NO" /* 3 0nov2 018 ('ELSE') */
then XDose = XLevel*XTime;
else DO; put //"Macro variable ENVADJ incorrectly specified."
/"It should be either YES or NO. Value specified
is: &ENVADJ"
/;
STOP;
END;
hix=.;
if &Model = 1 then hix
if &Model = 2 then hix
if &Model = 3 then hix
= hi * exp(&COEF*XDose); else
= hi * (1 + &COEF*XDose); else
= hi + &COEF*XDose;	else
Page 304 of 352

-------
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
if &Model = 4 then hix = hi * (1 + XDose)**&COEF; else
if &Model = 0 then DO;
hix = -99999; /* Code for user-defined model goes here. */
END;
hstarix = hstari	/* hi=backgrd rate is included in hstari
+ (hix - hi); /* so that adding in the excess
/* from exposure (hix-hi) gives the
/* total rate of the exposed.
qix = exp(-hstarix);
Rx = Rx + ( hix/hstarix * S_lix * ( 1-qix ) );
S_lix = S_lix * qix;
qi = exp(-hstari);
R0 = R0 + ( hi/hstari * S_li * (1-qi) );
S_li = S_li * qi;
output;
END;
Ex_Risk = Rx - R0;* Rx = risk in exposed population;
RskRatio = Rx / R0;	* R0 = from cancer;
Extra_risk = Ex_Risk/(1-R0) ;
/* BT 3/20/19 added:*/
call symput('Extra_Riskm',Extra_Risk);
/*BT 4/24/19 replaced the next line
Diff_Ex_Risk = abs(&ex_risk_target-Ex_Risk); */
Diff_Ex_Risk = abs(&ex_risk_target-Extra_Risk);
call symput('Delta_Ex_Risk',Diff_Ex_Risk);
output;
END; * corresponds to X_Levels;
STOP;
%Mend BEIR4;
BT: March 2019: parameters for the convergence that are used
in the modified version of the BEIR4 macro.
	*/
%macro Converge_BEIR4 (init_exposure_conc=, ex_risk_target=, conv_criterion=, max_iteration=);
%Let Delta_Ex_Risk = 1; * initial high value to make sure loop is run at least once
(i.e., macro BEIR4 is called
at least once);
/* BT 4/15/19: added next line to avoid error during compiling of BEIR4*/
%Let Extra_Riskm = 1;
%Let i=l; * first time through loop;
%Do %Until (%sysevalf(&Delta_Ex_risk < &conv_criterion) OR %sysevalf(&i >
&max_iteration)) ;
* first time through loop, set expsosure_conc=init_exposure_conc;
%If &i=l %Then
%Do;
%Let exposure_conc=&init_exposure_conc;
Page 305 of 352

-------
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
%End;
%If &i>l %Then
%Do;
data tempBEIRCONVERGE;
exposure_conc
BEIR4 (=Extra Risk);
^BEIR4 has run at least once. Adjust
Extra Riskm is created in
the same code that we used
^update the concentration;
NumLoops=&i;
thisExposureConc=&exposure_conc;
/* BT 4/15/19: replaced all of the convergence code with
in the meso code.*/
numvar=&ex_risk_target;
denvar=&Extra_Riskm;
thisexposureconc = thisexposureconc * (numvar/denvar);
call symput('exposure_conc',thisexposureconc);
output;
Run;
%End; ^Corresponds to If i>l statement;
%BEIR4;
%Let i=%eval(&i+l);
%End;
%Let EC_lPercent = &exposure_conc;
/ *	+
I Report results if convergence criterion met:
+	*/
%If %sysevalf(&Delta_Ex_risk < &conv_criterion) %then %do;
title5 "based on KL = &COEF, Concentration = &EC_lPercent, and LastAge = &LastAge";
data _null_;	/* Modified 26-july-00 */
pointer=l;
set allcause (keep=age
rename=(age=ageall0)) point=pointer nobs=n_all;
set cause (keep=age
rename=(age=ageCsO)) point=pointer nobs=n_cause;
pointer=n_all;
set allcause (keep=age
rename=(age=agealll)) point=pointer nobs=n_all;
pointer=n_cause;
set cause (keep=age
rename=(age=ageCsl)) point=pointer nobs=n_cause;
Tmp = sum(min(AgeAlll,AgeCsl,(&Lastage-l)),1);
file PRINT;
if ageallO NE ageCsO then DO;
put /"ERROR: The initial age for all-causes rate differs from the"
/"	initial age for the cause-specific rate.";
END;
else DO;
put / "Values of macro variables used in this computation:	"
// @3 "Value"	@17 "Macro_Var" @29 "Description"
/ @3 "	"	@17 "	" @29 "	'
// @3 "&Model " @17 "MODEL"	@29 "1 = Loglinear Relative Rate,1
Page 306 of 352

-------
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
/



@29
"2 = Linear Relative Rate, "
/



@29
"3 = Linear Absolute Rate, "
/



@29
"4 = 'Power' Relative Rate, "
/



@29
"0 = User defined. "
/
@3
"SCoef
@17
"COEF" @29
"Exposure parameter estimate"
//
@3
" & Lag
@17
"LAG" @29
"Exposure Lag "
//
@3
"SAgelst
@17
"AGE1ST X" @29
"Age exposure begins"
/
@3
"^Duration"
@17
"DURATION" @29
"Duration of exposure"
/
@3
"SEnvAdj"
@17
"ENVADJ" @29
"Adjust dose from intermittent"
/



@29
"occupational exposures to "
/



@29
"continuous environmental exposures"


/
@3
@17
	.. @29 "	
// M	„
11 @3 "EC1% = " @10 " &EC_1Percent" @25 "(f/ml); Ft: = " @39
"&Extra_Ris km"
// M	
/"The risks are calculated from age " ageallO " up to age " Tmp "."
// ;
if agealll NE ageCsl then
put /"WARNING: The last age for the all-causes rates differs from"
/"	the last age for the cause-specific rates, suggesting"
/"	the possibility that the rates weren't entered as desired."
/;
END;
Stop;
run;
/* BT 7/5/19: Start of code that was added to merge variables for unexposed risk
(S	1i and S_lix) to the rest of the output, by age;
*/
Data newSRCData(keep=SRC_age SRC_S_li SRC_S_lix);
set ex_Risk;
SRC_age=0; SRC_S_li=l; SRC_S_lix=l;
output;
do obsnum=l to last-1;
set ex_Risk point=obsnum nobs=last;
if _error_ then abort;
SRC_age=age+l; SRC_S_li=S_li; SRC_S_lix=S_lix;
output;
end;
stop;
run;
*	rename variables to enable overwriting the values of S_li and S_lix in ex_risk with the values
in newSRCData;
* Data file tempSRCData has age=0-85 while the ex_Risk file has age 0-84, with last two
records
both having age=84.;
Data tempSRCData; Set newSRCData(rename=(SRC_Age=age SRC_S_li=S_li SRC_S_lix=S_lix));
if age=&LastAge then age=%sysevalf(&Lastage-l); Else age=age;
Run;
*	there are duplicate values for age in both ex_risk and tempSRCData
which may produce too many records, if that happens, then we use two set
statements;
Data ex_risk; merge ex_risk tempSRCData; By Age; Run;
/* BT 7/5/19: End of code that was added to merge variables for unexposed risk
(S	1i and S_lix) to the rest of the output, by age;
*/
*BT 7/3/19: made the these changes to the following Proc Print procedure:
-	commented out the label option and added the split, uniform and
width= options
-	included all variables to the format statement;
Page 307 of 352

-------
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
proc print data=ex_risk /*label*/ noobs split='/' width=FULL;
format age F4. Xdose Ell. hi Ell. hstari Ell. hix Ell. hstarix Ell. qi Ell. qix Ell.
S_li Ell. S_lix Ell. R0 Ell. Risk Ell. Ex_Risk Ell. ;
label Age	= "Age at start of year (i) "
XDose = "CElO(adj) (f\cc-yrs)"
R0	=	"Cumulative Risk of lung cancer (unexposed) (R0)"
Risk	= "Cumulative risk of lung cancer (exposed) (Re)"
Ex_Risk	=	"Excess risk/[Rx-Ro]/ /(Ex_Risk)"
hi	=	"Lung Cancer hazard (unexposed) (hi)"
hix	=	"Lung Cancer hazard (exposed) (hei)"
hstari	=	"All cause hazard (unexposed) (h*i)"
hstarix	=	"All cause hazard (exposed) (he*i)"
qi	=	"Probability of surviving year i assuming alive at start
(unexposed) (qi)"
qix	= "Probability of surviving year i assuming alive at start
(exposed) (qei)"
S_li	=	"Probability of surviving to end of year i (unexposed) (Sl,i)"
S_lix	=	"Probability of surviving to end of yeari (exposed) (Sel,i)";
Var Age Xdose hi hstari hix hstarix qi qix S_li S_lix R0 Risk Extra_risk; *BT
7/3/19: Var statement added;
label Extra_risk="Extra Risk (Re - R0)\(1 - R0)";
run;
%End; *end of the If statement that tests if convergence was met;
%Mend Converge_BEIR4;
/* 	+
| March 2019: BT (SRC) Added maxro CONVERGE_BEIR4 which iteratively |
| runs macro BEIR4 until the EXPOSURE_CONCENTRATION corresponds to an |
I extra_risk=0.01 (the point of departure [POD]).	I
I In addition to the parameter for CONVERGE_BEIR4, the user should also|
I review parameters and data that are assigned/entered in Part 1 and |
I Part II (see above). Parameters for CONVERGE_BEIR4 are defined below |
+	*/
*%BEIR4; * originally called macr BEIR4 directly. Now BEIR4 is called by Converge_BEIR4;
%Converge_BEIR4(init_exposure_conc=l,	/* initial exposure concentration (initial
guess) */
ex_risk_target=0.01000000, /* the point of departure
(POD) - the target extra risk */
conv_criterion= 0.00000001,
max_iteration=200);	/* to avoid excessively long
run times */
Run;
SAS Lung Cancer Lifetable for Non-Linear Models
OPTIONS NODATE NONUMBER orientation=landscape linesize=max; *BT added 7/3/19;
/*
This program calculates the risk of lung cancer from inhalation exposure to asbestos,
using a lifetable approach based on BEIR IV. The basic exposure-response model is RR = exp(beta
* CE10).
The basic code for the lifetable calculations were developed and provided to EPA
by Randall Smith at NIOSH. The code from NIOSH calculates the baseline risk (R0) and the exposed
risk (Rx)
from exposure to an exposure concentration of X_Level using NIOSH Model 1: Rx = R0 * exp(COEF *
X_Level).
EPA has modified the NIOSH as follows:
1) The all-cause mortality and cause-specific (lung cancer) incidence data tables have been
updated based on CDC Wonder 2017.
Page 308 of 352

-------
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
2)	An equation has been added to calculate extra risk: Extra_Risk = (Rx - R0) / ( 1 - R0)
3)	A macro has been added to find the exposure level (X_Level) that yields an extra risk of 0.01
(1%).
This is referred to as EC1%, which may then be used to calculate the unit risk: UR = 0.01 /
EC1%
*/
/* .\Beta Version.sas 19jan00, 26jul00, 25oct01, 06dec05, 30novl8
Experimental version
	*/
title "Lifetable calculation of lung cancer risk";
title2 "under a non-linear relative rate model";
f-k	+
I	Compute excess risk by the BEIR IV method using SAS datasteps.	|
I	These programs compute the risk of a cause-specific	|
I	death in the presence of competing risks, where the cause-	|
I	specific death-rate is modeled either as a relative rate	|
I	[h=h0*f(Coef*X)] or as an absolute rate [h=h0+f(Coef*X)]	I
I	where	I
I h denotes the cause-specific death-rate,	I
I X denotes cumulative occupational exposure (with Lag)	|
I Coef denotes the coefficient for the effect of exposure and	|
I hO is the corresponding rate at baseline (X=0).	|
I (Except for Coef, these are functions of age.)	|
I	A few simple models of f(Coef*X) are easily specified as	|
I	described below. More complicated models can be specified with	|
I	a little more work. (For a more complicated example,	|
|	see \_GENERAL.LIB\PROGRAMS\SAS\BEIR-4.Method\BEIR4ex2.SAS).	|
+Reference:	+
I	Health Risks of Radon and Other Internally Deposited Alpha-	|
I	Emitters (BEIR IV). Commitee on the Biologic Effects of	|
I	Ionizing Radiations. National Academy Press. Wash. DC (1988).	|
I	See especially pages 131-136.	I
+USER-SUPPLIED ASSIGNMENTS:
|> The following macro variables are assigned using "%LET" state-
| ments: MODEL, COEF, LAG, AGE1ST_X, DURATION, LASTAGE.
I Further information appears below.
|> Exposure concentrations for computing risk are defined
| in the datastep "X_LEVELS."
|> All-cause mortality information is entered as a life-table in
I the data step "ALLCAUSE," and converted to rates per individual.
|> Cause-specific mortality information for unexposed referents is
I entered as rates per 100,000 and converted to rates per
I individual in the data step "CAUSE."
+NOTES:	+
|> Datastep "EX_RISK" is where the desired risks are computed.	|
|> If the unexposed(referent) cause-specific mortality rate is from|
I a model then datastep "CAUSE" with variables AGE and RATE as |
I modeled can be modified to incorporate this. However, care	|
I must be taken in calculating confidence limits since imprecision|
I in the estimates of all of the parameters of the model	|
I contributes to the imprecision of excess risk estimates.	|
|> This program is currently set up to apply the Linear Rel. Rate	|
I model (Lag= 0) and accumulation of excess risk is over the	|
I rates in ALLCAUSE and CAUSE unless truncated at a younger age.	|
| (See LASTAGE below.)	I
Page 309 of 352

-------
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
+	SAS Programmer: Randall Smith +
I	The Nat'1 Inst, for Occupational Safety & Health |
|	26jul2000, 23jul2001, 25oct2001, 18nov2018 |
+	Modifications: +
I	26jul00 Fix the procedure bug causing it to report incorrectly |
I	the age at which accumulation of risk was stopped |
I	whenever the age-specific rates included ages |
I	before the value of &Agelst_X. (&Agelst_X is a macro|
I	expression defining the age exposure begins.) |
1	. . ... 1
I	23jul01 Make changes to facilitate multiple applications of |
I	BEIR4 algorithm, i.e., MLE(Excess Risk), UCL(ExcessRisk),|
I	searching for concentrations for a fixed risk. These |
I	changes involve defining Macros named BEIR4 and SEARCH |
I	given below with code illustrating these uses for the |
I	linear relative rate model. I
I	I
I	25oct01 Modified to add Macro variable EnvAdj for whether to |
I	increase inhaled dose from intermittent occupational |
I	exposures to continuous environmental exposures |
I	and update US rates for Gibb et al. cohort. |
I	I
I	30novl8 A bug that prevented the calculation of excess risks |
I	after incorporating an adjustment from intermittent |
I	occupational exposures to continuous exposures is fixed. |
I
|	March 2019: BT (SRC) Added maxro CONVERGE_BEIR4 which iteratively
|	runs macro BEIR4 until the EXPOSURE_CONCENTRATION corresponds to an
Iextra_risk=0.01 (the point of departure [POD]).
Macro CONVERGE_BEIR4 works with one value for the exposure
variable XLevel (i.e., when the data C_Levels includes one record.)
The intent was to make as few changes to BEIR4 as possible. The data
X_LEVELS and variable XLevel are retained but the initial value of
XLevel is provided in the call to macro CONVERGE_BEIR4 (the value
of Xlevel in the cards statement is not used in the calculations.
Changes to the BEIR4 macro are in Part III and Part IV, and are
indicated by the letters BT.
I In addition to the parameter values that are specified by the user |
I in PART 1, and the user-provided data entered in Part II, parameters |
I for the new macro CONVERGE_BEIR4 are specified in the call to the |
| macro CONVERGE_BEIR4 (see end of this SAS program file below).
+	*/
/* PART I. USER-SUPPLIED ASSIGNMENTS (Macro variables):
f-k	f
I	Model of cumulative exposure effects:	|
I	1 => Loglinear Relative rate	|
|	R=R0*exp(COEF*X)	|
I	2 => Linear Relative rate,	|
|	R=R0*(l+COEF*X)	|
I	3 => Absolute rate,	|
|	R=R0+COEF*X	|
I	4 => Power relative rate	|
|	R=R0*(1+X)ACOEF	|
I	0 => User Defined & programmed	|
I	in datastep Ex_Risk below	|
Cumulative exposure parameter:
Lag or delay between exposure and effect:
%Let Model
%Let COEF
%Let Lag
1;
le-2;
10;
Page 310 of 352

-------
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
Age exposure begins:
/*ฆ Exposure duration (years) :
/*ฆ Adjust dose from occupational to
continuous environmental exposures (Y/N)?
/*ฆ Age to stop> accumulating excess risk
(supposing rates are available for
ages >= SLastAge); otherwise use all of
the supplied rate information:
%Let Agelst_x
%Let Duration
40;
20;
V %Let EnvAdj = Yes;
V	%Let LastAge = 85;
V
/- PART II. USER-SUPPLIEE' ASSIGNMENTS (Datesets AllCause, Cause, X_Levels ): V
data AllCause (label="Unexposeds' age-spec mortalty rates (all)"
drop=Lx rename=(BLx=Lx) );
/-	
Input lifetable and calculate the corresponding age-specific
(all-causes) mortality rate (AllCause) and conditional survival
probability for each year of age (qi) together with
the corresponding values of age (Age).
Label Age
BLx
Lx
CndPrDth
qi
AllCause
'Age at start of year (Age=i)"
'Number alive at start of year"
'Number alive at end of year"
'Pr[Death before age i+1 | alive at age
'Pr[Survive to age i+1 | Alive at age i]
'Age-spec mortality rate (all causes)";
if
=1 then input age //// @1 BLx @; I'1, 1111
input Lx @@;
CndPrDth
(BLx - Lx)/BLx;
ship) next 4 lines "V
qi	= 1-CndPrDth;
if qi <:= 0 then AllCause = le+50;
else AllCause = - log(qi);
if age < SLastAge then output; else STOP;
BLx=Lx;
age+1;
retain age BLx;
cards;
0 = Life-table starting age. (Required: Values must begin 4 lines down!)
The following are 2017 Life-table values of US population
starting at birth and ending at age 85.
(Source: Nat.Vital Statistics Reports 2019 Vol 68 No 7, Table 1,
https://www.cdc.gov/nchs/data/nvsr/nvsr 68/nvsr 68_07-508.pdf)
100000
99268
98 937
97872
99422
9 9259
98858
977 4 0
99384 99360
99249 99236
96321 96135
93797 93419
88226 87424
98770
97 603
95 93 9
93 0 0 8
8 657 0
98 67 4
97 4 61
95732
925 60
85664
9 9341
9 9217
98573
97314
95511
92070
99326
99191
984 66
97163
95275
91538
9 9312
99158
9 9 2 9 9
99116
99288 99278
9 90 6 6 9 9 0 0 6
98355 98241
97006 96843
95023
90 9 63
98122
9 6 674
9 7 9 9 9
9 6501
94753 94461 94144
90345 89684 88978
84706 83696 82632
77697 76265 74715 73064 71296 69418 67402
57839 55053 52123 49035 45771 42382
81507
652 4 5
80315 79048
62933 60462
'runj^BT 7/3/19 added Run statement here;
data CAUSE (label="Unexposeds' age-cause-spec mortalty rates");
/*	1	
Specify unexposeds' age-specific mortality rates (per year)
from specific cause.
+	
label Age	= "Age"
Rate_e5 = "Age,cause-specific rate per 100,000"
Rate	= "Age,cause-specific rate per individual";
if n
1 then input age
/* input starting age
Page 311 of 352

-------
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
///; /* /// => skip next 3 lines */
input Rate_e5
Rate = Rate_e5 * le-5; /* Convert to rate per individual */
if age <= 4
then DO; output; age+1; END;
else DO i = 0,1,2,3,4;	/*	*/
if age < &LastAge /* Fill out into yearly intervals from */
then output; /* inputted five year intervals after age 4*/
age+1;	/*	*/
END;
cards;
0 = Start age of cause-specific rate (Required: Rates begin 3 lines down!)
The following are 2017 cancer site code 22030 lung and bronchus incidence rates per
100,000 for US pop'n starting at birth.
For ages 5 and above, each rate holds for the age thru age+4 years.
Source: CDC Wonder, https://wonder.cdc.gov/cancer-v2 017.HTML
0.205 0.100 0.100 0.100 0.100 0.039 0.039 0.104 0.299 0.553 1.267 2.600 6.534 16.528 44.403
96.098 149.112 223.906 319.322 391.202 395.215
*run; *BT 7/3/19 added Run statement here;
data X_LEVELS (label= "Exposure levels (e.g., concentrations)" );
/ *	+
I Specify environmental exposure levels	I
I and update label for the variable, XLevel, if necessary:	|
+	*/
j *	+
| BT 3/8/19: Add maxro CONVERGE_BEIR4 which iteratively runs macro	|
| BEIR4 until the EXPOSURE_CONCENTRATION corresponds to extra_risk=0.01|
I The intent was to make as few changes to BEIR4 as possible. The data |
I X_LEVELS and variable XLevel are retained but the initial value of |
I XLevel is provided in the call to macro CONVERGE_BEIR4 (the value
I of Xlevel in the cards statement is not used in the calculations. |
+	*/
input XLevel @@;
label XLevel= "Asbestos exposure (F/ml)";
cards;
0.0383
%Macro BEIR4;
/* March 2019 - BT (SRC): Macro BEIR4 is now called by macro CONVERGE_BEIR4.
*/
/* 23jul01 modification */
/* Enclose the actual calculations and printed results in a macro	*/
/* to facilitate multiple applications of the algorithm.	*/
/* PART III. Perform calculations:	*/
data EX_RISK (label = "Estimated excess risks [Method=BEIR IV]"
/*keep = XLevel Rx ex_risk RskRatio R0 extra_Risk */
rename= (Rx=Risk));
I*	+
I Calculate risk and excess risk for each exposure concentration!
I in work.X_Level by BEIR IV method using information in	|
I work.AllCause and work.Cause to define referent population: |
_i	* f
format rate F15.8 hi F15.8; *BT 7/3/19: added the format statement;
length XLevel 8.;
label Age	= "Age at start of year (i)"
XTime = "Exposure duration midway between i & i+1"
XDose = "CElO(adj) (f/cc-yrs)"
R0	= "Cumulative Risk of lung cancer (unexposed) (R0)"
Page 312 of 352

-------
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
(qi)
Ft:
Ex_Risk
RskRatio
hi
hi::
hstari
hstarix
qi
S_li
S lix
Cumulative risk of lung cancer (exposed) (Re)"
Excess risk (Rx-Ro)"
Ratio of risks (Rx/Ro)"
Lung Cancer hazard (unexposed) (hi)"
Lung Cancer hazard (exposed) (hei)"
All cause hazard (unexposed) (fVi)"
All cause hazard (exposed) (he^i)"
Probability of surviving year i assuming alive at start (unexposed)
Probability of surviving to end of year i (unexposed) (Sl,i)"
Probability of surviving to end of yeari (exposed) (Sel,i)";
/*ฆ BT 3/8/19: Calculation of unexposed's risk (following DO LOOP) could be omitted
from the iteration
but may require further changes to BEIR4(?).
4e.g., %if i=l %then %do;V
if _n_=l then DO;
/* Calculate unexposed's risk (RO) to be retained
/*ฆ based on equation 2A-21 (pg. 131) of BEIR IV:
/* Initialize: "V S li
RO
conform
DO pointer = 1 to rnin (n_all, n_cause) until (age>=&LastAge-l);
set allcause (keep=age AllCause rename=(AllCause=hstari))
point=pointer nobs=n_all;
set cause (keep=age Rate rename=(age=ageCause Rate=hi))
point=pointer nobs=n_cause;
if Age NE AgeCause then
put "ฆ" WARNING: Age values in datasets ALLCAUSE and CAUSE don't
@13 "Rates misaligned on age could give incorrect results'
@13 Pointer=
+2 "Age(ALLCAUSE)=" Age +2 "Age(CAUSE)=" AgeCause /;
qi = exp(-hstari);
RO = RO + ( hi/hstari
S_li = S_li ^ qi;
END;
END;
S_li - (1-qi) );
/'*ฆ End of 'if n =1 then DO;' stmt V
retain RO;
/* Calculate exposed's risk (Rx, renamed to Risk) for each	exposure level
V
/*ฆ ultimately based on equation 2A-22 (pg. 132) of BEIR IV	V
/* but re-expressed in a form similar to equation 2A-21:	"V
BT 3/20/19. This version of CONVERGE_BEIR4 will work when there
one concentration in data set x_levels -
i.e., one value for xlevel.
The Do loop> for X_levels is commented out;
ฆ"ฆDO pointX = 1 to No_of_Xs;
^ set x_levels point=pointX nobs=No_of_Xs; /^ BT 3/8/19: determines
Nobs is set at compilation,
so the value of nobs is available at first run through loop -
just one record and one variable (XLevel) in dataset x_levels. "V
/* BT 3/20/19: added the next lint to set the exposure
concentration = current value of &exposure_conc. "V
xlevel = Sexposure_conc;
is
when to end the loop.
Page 313 of 352

-------
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
/* Initialize
V S lix = 1; Rx = 0;S li=l; R0=0;
DO pointer = 1 to min(n_all,n_cause) until (age>=&LastAge-l);
set allcause (keep=age AllCause rename=(AllCause=hstari))
point=pointer nobs=n_all;
set cause (keep=Rate rename=(Rate=hi))
point=pointer nobs=n_cause;
XTime = min( max(0,(age+0.5-&Agelst_x-&Lag))
, &Duration );
Conversion */
is: &ENVADJ"
if UpCase ( "ScEnvAdj ")
then XDose
"YES"
/* Occupational to Environmental
/* Days per year	*/
/* Ventilation (L) per day */
/* 30nov2018 ('ELSE') */
XLevel
*	365/240
*	20/10
*	XTime;
ELSE if UpCase("&EnvAdj") = "NO"
then XDose = XLevel*XTime;
else DO; put //"Macro variable ENVADJ incorrectly specified."
/"It should be either YES or NO. Value specified
/;
STOP;
END;
hix=.;
if &Model = 1 then hix
if &Model = 2 then hix
if &Model = 3 then hix
if &Model = 4 then hix
if &Model = 0 then DO;
hix = -99999; /* Code for user-defined model goes here. */
END;
hi * exp(&COEF*XDose);
hi * (1 + &COEF*XDose);
hi + &COEF*XDose;
hi * (1 + XDose)**&COEF;
else
else
else
else
hstarix = hstari
+ (hix - hi),
/*	hi=backgrd rate is included in hstari
/*	so that adding in the excess
/*	from exposure (hix-hi) gives the
/*	total rate of the exposed.
qix = exp(-hstarix);
Rx = Rx + ( hix/hstarix
S_lix = S_lix * qix;
S lix
( 1-qix ) );
qi = exp(-hstari);
R0 = R0 + ( hi/hstari
S_li = S_li * qi;
S li
(1-qi) );
output;
END;
Ex_Risk = Rx - R0;* Rx = risk in exposed population;
RskRatio = Rx / R0;	* R0 = from cancer;
Extra_risk = Ex_Risk/(1-R0) ;
/* BT 3/20/19 added:*/
call symput('Extra_Riskm',Extra_Risk);
/*BT 4/24/19 replaced the next line
Diff_Ex_Risk = abs(&ex_risk_target-Ex_Risk); */
Diff_Ex_Risk = abs(&ex_risk_target-Extra_Risk);
call symput('Delta_Ex_Risk',Diff_Ex_Risk);
output;
* END; * corresponds to X_Levels;
STOP;
Page 314 of 352

-------
\m
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
run;
%Mend BEIR4;
BT: March 2019: parameters for the convergence that are used
in the modified version of the BEIR4 macro.
	*/
%macro Converge_BEIR4 (init_exposure_conc=, ex_risk_target=, conv_criterion=, max_iteration=);
%Let Delta_Ex_Risk = 1; * initial high value to make sure loop is run at least once
(i.e., macro BEIR4 is called
at least once);
/* BT 4/15/19: added next line to avoid error during compiling of BEIR4*/
%Let Extra_Riskm = 1;
%Let i=l; * first time through loop;
%Do %Until (%sysevalf(&Delta_Ex_risk < &conv_criterion) OR %sysevalf(&i >
&max_iteration)) ;
* first time through loop, set expsosure_conc=init_exposure_conc;
%If &i=l %Then
%Do;
%Let exposure_conc=&init_exposure_conc;
%End;
%If &i>l %Then
%Do;
exposure_conc
BEIR4 (=Extra Risk);
data tempBEIRCONVERGE;
^BEIR4 has run at least once. Adjust
Extra Riskm is created in
the same code that we used
^update the concentration;
NumLoops=&i;
thisExposureConc=&exposure_conc;
/* BT 4/15/19: replaced all of the convergence code with
in the meso code.*/
numvar=&ex_risk_target;
denvar=&Extra_Riskm;
thisexposureconc = thisexposureconc * (numvar/denvar);
call symput('exposure_conc',thisexposureconc);
output;
Run;
%End; ^Corresponds to If i>l statement;
%BEIR4;
%Let i=%eval(&i+l);
%End;
%Let EC_lPercent = &exposure_conc;
Page 315 of 352

-------
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
Report results if convergence criterion met:
%If %sysevalf(&Delta_Ex_risk < &conv_criterion) %then %do;
title5 "based on beta=&COEF, Concentration=&EC_lPercent, and LastAge=&LastAge";
data _null_;	/* Modified 26-july-00 */
pointer=l;
set allcause (keep=age
rename=(age=ageall0)) point=pointer nobs=n_all;
set cause (keep=age
rename=(age=ageCsO)) point=pointer nobs=n_cause;
pointer=n_all;
set allcause (keep=age
rename=(age=agealll)) point=pointer nobs=n_all;
pointer=n_cause;
set cause (keep=age
rename=(age=ageCsl)) point=pointer nobs=n_cause;
Tmp = sum(min(AgeAlll,AgeCsl,(&Lastage-l)),1);
file PRINT;
if ageallO NE ageCsO then DO;
put /"ERROR: The initial ag
END;
else DO;
e for all-causes rate differs from the"
initial age for the cause-specific rate.";
this computation:
Description"
put /
"Values of macro variables
used in
11
@3
"Value"
@17
"Macro Var" @29
/
@3
II II
@17
II
" I3 0Q


IS z y
//
@3
"&Model
@17
"MODEL"
@29
/




@29
/




@29
/




@29
/




@29
/
@3
"&Coef
@17
"COEF"
@29
//
@3
"&Lag
@17
"LAG"
@29
//
@3
"&Agelst x'
@17
"AGE1ST
X" @29
/
@3
"^Duration'
@17
"DURATION" @29
/
@3
"&EnvAdj"
@17
"ENVADJ
@29
/




@29
/

/
@3
..
@29
@17 "
1	= Loglinear Relative Rate,
2	= Linear Relative Rate,
3	= Linear Absolute Rate,
4	= 'Power' Relative Rate,
0 = User defined.
Exposure parameter estimate"
Exposure Lag "
Age exposure begins"
Duration of exposure"
Adjust dose from intermittent"
occupational exposures to "
continuous environmental exposures'
	" @29 "	
''&Extra Riskm"
//
// @3 "EC1% = " @10 "&EC IPercent" @25 "(f/ml); Rx = " @39
//
/"The risks are calculated from age " ageallO " up to age " Tmp "."
// ;
if agealll NE ageCsl then
put /"WARNING: The last age for the all-causes rates differs from"
/"	the last age for the cause-specific rates, suggesting"
/"	the possibility that the rates weren't entered as desired."
/;
END;
Stop;
run;
/* BT 7/5/19: Start of code that was added to merge variables for unexposed risk
(S	1i and S_lix) to the rest of the output, by age;
*/
Data newSRCData(keep=SRC_age SRC_S_li SRC_S_lix);
Page 316 of 352

-------
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
set ex_Risk;
SRC_age=0; SRC_S_li=l; SRC_S_lix=l;
output;
do obsnum=l to last-1;
set ex_Risk point=obsnum nobs=last;
if _error_ then abort;
SRC_age=age+l; SRC_S_li=S_li; SRC_S_lix=S_lix;
output;
end;
stop;
run;
*	rename variables to enable overwriting the values of S_li and S_lix in ex_risk with the values
in newSRCData;
* Data file tempSRCData has age=0-85 while the ex_Risk file has age 0-84, with last two
records
both having age=84.;
Data tempSRCData; Set newSRCData(rename=(SRC_Age=age SRC_S_li=S_li SRC_S_lix=S_lix));
if age=&LastAge then age=%sysevalf(&Lastage-l); Else age=age;
Run;
*	there are duplicate values for age in both ex_risk and tempSRCData
which may produce too many records, if that happens, then we use two set
statements;
Data ex_risk; merge ex_risk tempSRCData; By Age; Run;
/* BT 7/5/19: End of code that was added to merge variables for unexposed risk
(S	1i and S_lix) to the rest of the output, by age;
*/
*BT 7/3/19: made the these changes to the following Proc Print procedure:
-	commented out the label option and added the split, uniform and
width= options
-	included all variables to the format statement;
proc print data=ex_risk /*label*/ noobs split='/' width=FULL;
format age F4. Xdose Ell. hi Ell. hstari Ell. hix Ell. hstarix Ell. qi Ell. qix Ell.
S_li Ell. S_lix Ell. R0 Ell. Risk Ell. Ex_Risk Ell. ;
label Age	= "Age at start of year (i) "
XDose = "CElO(adj) (f\cc-yrs)"
R0
Risk
Ex_Risk
hi
hix
hstari
hstarix
qi
(unexposed) (qi)"
(exposed) (qei) '
S_li
S lix
= "Cumulative Risk of lung cancer (unexposed) (R0)"
= "Cumulative risk of lung cancer (exposed) (Re)"
= "Excess risk/[Rx-Ro]/ /(Ex_Risk)"
= "Lung Cancer hazard (unexposed) (hi) "
= "Lung Cancer hazard (exposed) (hei)"
= "All cause hazard (unexposed) (h*i)"
= "All cause hazard (exposed) (he*i)"
= "Probability of surviving year i assuming alive at start
qix	= "Probability of surviving year i assuming alive at start
= "Probability of surviving to end of year i (unexposed) (Sl,i)"
= "Probability of surviving to end of yeari (exposed) (Sel,i)";
Var Age Xdose hi hstari hix hstarix qi qix S_li S_lix R0 Risk Extra_risk; *BT
7/3/19: Var statement added;
label Extra_risk="Extra Risk (Re - R0)\(1 - R0)";
run;
%End; *end of the If statement that tests if convergence was met;
%Mend Converge_BEIR4;
/* 	+
| March 2019: BT (SRC) Added maxro CONVERGE_BEIR4 which iteratively |
| runs macro BEIR4 until the EXPOSURE_CONCENTRATION corresponds to an |
I extra_risk=0.01 (the point of departure [POD]).
Page 317 of 352

-------
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
I In addition to the parameter for CONVERGE_BEIR4, the user should also|
I review parameters and data that are assigned/entered in Part 1 and |
I Part II (see above). Parameters for CONVERGE_BEIR4 are defined below |
+	*/
*%BEIR4; * originally called macr BEIR4 directly. Now BEIR4 is called by Converge_BEIR4;
%Converge_BEIR4(init_exposure_conc=l,	/* initial exposure concentration (initial
guess) */
ex_risk_target=0.01000000, /* the point of departure
(POD) - the target extra risk */
conv_criterion= 0.00000001,
max_iteration=200);	/* to avoid excessively long
run times */
Run;
Page 318 of 352

-------
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
SAS Mesothelioma Lifetable
OPTIONS NODATE NONUMBER orientation=landscape papersize=legal;
/*
This program calculates the risk of mesothelioma from inhalation exposure to asbestos,
using a lifetable approach. The basic model is Im = C * KM * Q.
The basic code for the lifetable calculations were developed and provided to EPA
by Randall Smith at NIOSH.
For mesothelioma, calculations are based on NIOSH Model 3: Rx = R0 + COEF * X_Dose
For mesothelioma, R0 is assumed to be zero.
EPA has modified the NIOSH as follows:
1)	The all-cause and cause-specific (mesothelioma) mortality data tables have been updated.
2)	Code has been asdded to calculate X_Dose = X_Level * Q, where Q is a function of TSFE and
exposure duration.
2)	An equation has been added to calculate extra risk: Extra_Risk = (Rx - R0) / ( 1 - R0)
3)	A macro has been added to find the exposure concentration (X_Level) that yields an extra risk
of 1%. This is referred to as EC.
This value may then be used to calculate the unit risk: UR = 0.01 / EC
*/
/* .\Beta Version.sas 19jan00, 26jul00, 25oct01, 06dec05, 30novl8
Experimental version
title "Lifetable calculation of mesothelioma risk"
title2 "under a linear absolute rate model";
j *	+
I	Compute excess risk by the BEIR IV method using SAS datasteps.	|
I	These programs compute the risk of a cause-specific	|
I	death in the presence of competing risks, where the cause-	|
I	specific death-rate is modeled either as a relative rate	|
I	[h=h0*f(Coef*X)] or as an absolute rate [h=h0+f(Coef*X)]	I
I	where	I
I h denotes the cause-specific death-rate,	I
I X denotes cumulative occupational exposure (with Lag)	|
I Coef denotes the coefficient for the effect of exposure and	|
I hO is the corresponding rate at baseline (X=0).	|
I (Except for Coef, these are functions of age.)	|
I	A few simple models of f(Coef*X) are easily specified as	|
I	described below. More complicated models can be specified with	|
I	a little more work. (For a more complicated example,	|
|	see \_GENERAL.LIB\PROGRAMS\SAS\BEIR-4.Method\BEIR4ex2.SAS).	|
+Reference:	+
I	Health Risks of Radon and Other Internally Deposited Alpha-	|
I	Emitters (BEIR IV). Commitee on the Biologic Effects of	|
I	Ionizing Radiations. National Academy Press. Wash. DC (1988).	|
I	See especially pages 131-136.	I
+USER-SUPPLIED ASSIGNMENTS:
|> The following macro variables are assigned using "%LET" state-
| ments: MODEL, COEF, LAG, AGE1ST_X, DURATION, LASTAGE.
I Further information appears below.
|> Exposure concentrations for computing risk are defined
| in the datastep "X_LEVELS."
Page 319 of 352

-------
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
>	All-cause mortality information is entered as a life-table in
the data step "ALLCAUSE," and converted to rates per individual.
>	Cause-specific mortality information for unexposed referents is
entered as rates per 100,000 and converted to rates per
individual in the data step "CAUSE."
+NOTES:	+
|> Datastep "EX_RISK" is where the desired risks are computed.	|
I	I
|> If the unexposed(referent) cause-specific mortality rate is from|
I a model then datastep "CAUSE" with variables AGE and RATE as	|
I modeled can be modified to incorporate this. However, care	|
I must be taken in calculating confidence limits since imprecision|
I in the estimates of all of the parameters of the model	|
I contributes to the imprecision of excess risk estimates.	|
I	I
|> This program is currently set up to apply the Linear Rel. Rate	|
I model (Lag= 0) and accumulation of excess risk is over the	|
I rates in ALLCAUSE and CAUSE unless truncated at a younger age.	|
| (See LASTAGE below.)	I
+	SAS Programmer: Randall Smith +
I	The Nat'1 Inst, for Occupational Safety & Health |
|	26jul2000, 23jul2001, 25oct2001, 18nov2018 |
+	Modifications: +
I	26jul00 Fix the procedure bug causing it to report incorrectly |
I	the age at which accumulation of risk was stopped |
I	whenever the age-specific rates included ages |
I	before the value of &Agelst_X. (&Agelst_X is a macro|
I	expression defining the age exposure begins.) |
I	I
I	23jul01 Make changes to facilitate multiple applications of |
I	BEIR4 algorithm, i.e., MLE(Excess Risk), UCL(ExcessRisk),|
I	searching for concentrations for a fixed risk. These |
I	changes involve defining Macros named BEIR4 and SEARCH |
I	given below with code illustrating these uses for the |
I	linear relative rate model. I
I	I
I	25oct01 Modified to add Macro variable EnvAdj for whether to |
I	increase inhaled dose from intermittent occupational |
I	exposures to continuous environmental exposures |
I	and update US rates for Gibb et al. cohort. |
I	I
I	30novl8 A bug that prevented the calculation of excess risks |
I	after incorporating an adjustment from intermittent |
I	occupational exposures to continuous exposures is fixed. |
I
|	April 2019: BT (SRC) Added maxro CONVERGE_BEIR4 which iteratively
|	runs macro BEIR4 until the EXPOSURE_CONCENTRATION corresponds to an
Iextra_risk=0.01 (the point of departure [POD]).
Macro CONVERGE_BEIR4 works with one value for the exposure
variable XLevel (i.e., when the data C Levels includes one record.)
The intent was to make as few changes to BEIR4 as possible. The data
X_LEVELS and variable XLevel are retained but the initial value of
XLevel is provided in the call to macro CONVERGE_BEIR4 (the value
of Xlevel in the cards statement is not used in the calculations.
Changes to the BEIR4 macro are in Part III and Part IV, and are
indicated by the letters BT.
In addition to the parameter values that are specified by the user
in PART 1, and the user-provided data entered in Part II, parameters
Page 320 of 352

-------
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
I for the new macro CONVERGE_BEIR4 are specified in the call to the
| macro CONVERGE_BEIR4 (see end of this SAS program file below).
+	
/* PART I
/*	
I Model of
USER-SUPPLIED ASSIGNMENTS (Macro variables)
	+
exposure effects:	|
cumulative
1	=> Loglinear Relative rate
R=R0*exp(COEF*X)
2	=> Linear Relative rate,
R=R0*(l+COEF*X)
3	=> Absolute rate,
R=R0+COEF*X
4	=> Power relative rate
R=R0*(1+X)ACOEF
0 => User Defined & programmed
in datastep Ex_Risk below
Cumulative exposure parameter:
%Let Model
%Let COEF
2.9 61e-9;
this
/*
/*
/*
Lag or delay between exposure and effect: */ %Let Lag
value is ignired */
Age exposure begins:
Exposure duration (years):
Adjust dose from occupational to
continuous environmental exposures (Y/N)?
Age to stop accumulating excess risk
(supposing rates are available for
ages >= &LastAge); otherwise use all of
the supplied rate information:
%Let Agelst_x
%Let Duration
10; /* Lag is built into Q,
20;
20;
kf %Let EnvAdj
%Let LastAge
Yes;
= 85;
/* PART II. USER-SUPPLIED ASSIGNMENTS (Datesets AllCause, Cause, X Levels ): */
data AllCause (label="Unxposeds' age-spec mortalty rates (all)"
drop=Lx rename=(BLx=Lx) );
/*	
I Input lifetable and calculate the corresponding age-specific
I (all-causes) mortality rate (AllCause) and conditional survival
I probability for each year of age (qi) together with
I the corresponding values of age (Age).
Label Age
BLx
Lx
CndPrDth
qi
AllCause
'Age at start of year (Age=i)"
'Number alive at start of year"
'Number alive at end of year"
'Pr[Death before age i+1 | alive at age
'Pr[Survive to age i+1 | Alive at age i]
'Age-spec mortality rate (all causes)";
if _n_=l then input age //// 61 BLx @;
input Lx @@;
CndPrDth = (BLx - Lx)/BLx;
qi	= 1-CndPrDth;
if qi <= 0 then AllCause = le+50;
else AllCause = - log(qi);
if age < &LastAge then output; else STOP;
BLx=Lx;
age+1;
retain age BLx;
cards;
0
Life-table starting age. (Required: Values must begin 4 lines down!)
Page 321 of 352

-------
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
The following are 2017 Life-table values of US population
starting at birth and ending at age 85.
(Source: Nat.Vital Statistics Reports 2019 Vol 68 No 7, Table 1,
https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_07-50 8.pdf)
100000 99422 99384 99360 99341 99326 99312 99299 99288 99278
99268 99259 99249 99236 99217 99191 99158 99116 99066 99006
98937 98858 98770 98674 98573 98466 98355 98241 98122 97999
97872 97740 97603 97461 97314 97163 97006 96843 96674 96501
96321 96135 95939 95732 95511 95275 95023 94753 94461 94144
93797 93419 93008 92560 92070 91538 90963 90345 89684 88978
88226 87424 86570 85664 84706 83696 82632 81507 80315 79048
77697 76265 74715 73064 71296 69418 67402 65245 62933 60462
57839 55053 52123 49035 45771 42382
data CAUSE (label="Unxposeds' age-cause-spec mortalty rates");
/*	+
I Specify unexposeds' age-specific mortality rates (per year) |
I from specific cause. I
+	*/
label Age	= "Age"
Rate_e5 = "Age,cause-specific rate per 100,000"
Rate	= "Age,cause-specific rate per individual";
if _n_ = 1 then input age /* input starting age	*/
///; /* // => skip next 3 lines	*/
input Rate_e5 @@;
Rate = Rate_e5 * le-5; /* Convert to rate per individual */
if age <= 4
then DO; output; age+1; END;
else DO i = 0,1,2,3,4;	/*	*/
if age < &LastAge /* Fill out into yearly intervals from */
then output; /* inputted five year intervals after age 4*/
age+1;	/*	*/
END;
cards;
0 = Start age of cause-specific rate (Required: Rates begin 3 lines down!)
The following are 2013 ICD10 = 113 death rates per 100,000 for US pop'n starting at
birth.
For ages 5 and above, each rate holds for the age thru age+4 years.
Source: CDC Wonder
0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0
data X_LEVELS (label= "Exposure levels (e.g., concentrations)" );
/ *	+
I Specify environmental exposure levels I
I and update label for the variable, XLevel, if necessary: |
+	*/
input XLevel @@;
label XLevel= "Asbestos exposure (F/ml)";
cards;
0.001
%Macro BEIR4;
/* April 2 2019 - BT (SRC): Macro BEIR4 is now called by macro CONVERGE_BEIR4.*/
/* 23jul01 modification */
/* Enclose the actual calculations and printed results in a macro	*/
/* to facilitate multiple applications of the algorithm.	*/
Page 322 of 352

-------
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
/* PART III. Perform calculations:
data EX_RISK (label = "Estimated excess risks [Method=BEIR IV] '
/*keep = XLevel Rx ex_risk RskRatio */
rename= (Rx=Risk));
j *	+
I Calculate risk and excess risk for each exposure concentration!
I in work.X_Leve1 by BEIR IV method using information in	|
I work.AllCause and work.Cause to define referent population: |
_i	-k /
length XLevel 8.;
(qi) '
label Age
XTime
XDose
R0
Rx
Ex_Risk
RskRatio
hi
hix
hstari
hstarix
qi
S_li
S lix
Age at start of year (i)"
Exposure duration midway between i & i+l[Xtime]"
Cumulative exposure midway thru year (C*Q*Adj)[XDose]
Cumulative risk of mesothelioma
Cumulative risk of mesothelioma
Excess risk (Rx-Ro)"
Ratio of risks (Rx/Ro)"
(unexposed) (R0) '
(exposed) (Re)"
Mesothelioma hazard (unexposed) (hi)"
Mesothelioma hazard (exposed) (hei)"
All cause hazard (unexposed) (h*i)"
All cause hazard (exposed) (he*i)"
Probability of surviving year i assuming alive at start (unexposed)
Probability of surviving to end of year i
Probability of surviving to end of yeari
XLevel = nECl%";
(unexposed) (Sl,i)'
(exposed) (Sel,i)"
S lix */
/* BT 7/5/19:add arrays for writing out the values for Array S_li and
*	ARRAY A_S_li[0:85]; *0 corresponds to age=0;
*	ARRAY A S lix[0:85];
*A S li[0]=1;* A S lix[0]=1;
/* BT 3/8/19: Calculation of unexposed's risk (following DO LOOP) could be omitted
from the iteration
but may require further changes to BEIR4(?).
*e.g.f %if i=l %then %do;*/
if _n_=l then DO;
/* Calculate unexposed's risk (R0) to be retained	*/
/* based on equation 2A-21 (pg. 131) of BEIR IV:	*/
/* Initialize: */ S_li =1; R0 = 0; R0rs=0; S_lirs=l;
DO pointer = 1 to min(n_all,n_cause) until (age>=&LastAge-l);
set allcause (keep=age AllCause rename=(AllCause=hstari))
point=pointer nobs=n_all;
set cause (keep=age Rate rename=(age=ageCause Rate=hi))
point=pointer nobs=n_cause;
if Age NE AgeCause then
put WARNING: Age values in datasets ALLCAUSE and CAUSE don't conform
/	@13 "Rates misaligned on age could give incorrect results"
/	@13 Pointer=
+2 "Age(ALLCAUSE)=" Age +2 "Age(CAUSE)=" AgeCause /;
qi = exp(-hstari);
R0 = R0 + ( hi/hstari * S_li * (1—qi) );
S li = S li * qi;
END;
Page 323 of 352

-------
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
END;
retain R0;
/* End of 'if
=1 then DO;' stmt */
/* Calculate exposed's risk (Rx) for each exposure level	*/
/* ultimately based on equation 2A-22 (pg. 132) of BEIR IV	*/
/* but re-expressed in a form similar to equation 2A-21:	*/
* BT 3/20/19. This version of CONVERGE_BEIR4 will work when there is
one concentration in data set x_levels - i.e., one value
for xlevel.
The Do loop for X_levels is commented out;
*DO pointX = 1 to No_of_Xs;
* set x_levels point=pointX nobs=No_of_Xs; /* BT 3/8/19: determines when to
end the loop. Nobs is set at compilation,
so the value of nobs is available at first run through loop -
just one record and one variable (XLevel) in dataset x_levels. */
xlevel = &exposure_conc;
/* Initialize : */ S_lix =1; Rx = 0; S_li=l;
DO pointer = 1 to min(n_all,n_cause) until (age>=&LastAge-l);
set allcause (keep=age AllCause rename=(AllCause=hstari))
point=pointer nobs=n_all;
set cause (keep=Rate rename=(Rate=hi))
point=pointer nobs=n_cause;
/*
XTime = min( max(0,(age+0.5-&Agelst_x-&Lag))
, ^Duration );
Q = •;
If Age < 10 then Q = 0;
If Age >= (XTime +10) then Q = ((Age-10)**3)-((-10-XTime)**3);
Else Q = (XTime-10)**3;
*/
TSFE=.;
If Age < &Agelst_x then TSFE = 0;
Else TSFE = Age - &Agelst_x + 0.5;
d = . ;
If Age < &Agelst_x then d = 0; else
If Age >= &Agelst_x + ^Duration then d = ^Duration - 0.5;
Else d = Age-&Agelst_x + 0.5;
Q=.;
If TSFE < 10 then Q = 0; else
If TSFE >= d+10 then Q = (TSFE-10)**3-(TSFE-10-d)**3;
Else Q = (TSFE-10)**3;
if UpCase("&EnvAdj") = "YES" /* Occupational to Environmental Conversion */
then XDose = XLevel
*	365/240	/* Days per year	*/
*	20/10	/* Ventilation (L) per day */
*	Q;	/* BT: in lung cancer program, this line has
just XTime (instead of Q) */
ELSE if UpCase("&EnvAdj") = "NO" /* 3 0nov2 018 ('ELSE') */
then XDose = XLevel*XTime;
else DO; put //"Macro variable ENVADJ incorrectly specified."
/"It should be either YES or NO. Value specified is: &ENVADJ"
/;
STOP;
Page 324 of 352

-------
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
END;
hix=
&Model
&Model
&Model
&Model
if &Model
hix =
END;
then hix
then hix
then hix
then hix
then DO;
hi
hi
hi
exp(&C0EF*XDose);
(1 + &C0EF*XDose)j
hi + &C0EF*XDose;
(1 + XDose)** &C0EF;
else
else
else
else
-99999; /* Code for user-defined model goes here. */
/*start of what RS added */
qi = exp(-hstari);
R0 = R0 + ( hi/hstari * S_li
S li = S li * qi;
(1-qi) ).
/*end of what RS added */
hstarix
qix
Rx
S lix
hstari	f*
+ (hix - hi); /*
/*
/*
exp(-hstarix);
Rx + ( hix/hstarix
S_lix * qix;
hi=backgrd rate is included in hstari */
so that adding in the excess	*/
from exposure (hix-hi) gives the */
total rate of the exposed.	*/
S lix
( 1-qix ) );
output;
END;
Ex_Risk = Rx - R0; /* BT 4/2/19: was Ex_Risk = Rx - R0; */
* RskRatio = Rx / R0;
output;
/* BT 4/14/19: the macro variables for risk and difference between
the calculated risk
and the target risk were moved from Converge_BEIR4
to BEIR4 */
call symput('Extra_Riskm',Ex_Risk);
Diff_Ex_Risk = abs(&ex_risk_target-Ex_Risk);
call symput('Delta_Ex_Risk',Diff_Ex_Risk);
END; * corresponds to X_Levels;
STOP;
run;
%Mend BEIR4;
/*	
BT: March 2019: parameters for the convergence that are used
in the modified version of the BEIR4 macro.
	*/
%macro Converge_BEIR4 (init_exposure_conc=, ex_risk_target=, conv_criterion=, max_iteration=);
%Let Extra_Riskm = 1;
%Let Delta_Ex_Risk = 1; * initial high value to make sure loop is run at least once
(i.e., macro BEIR4 is called
at least once);
%Let i=l; * first time through loop;
Page 325 of 352

-------
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
%Do %Until (%sysevalf(&Delta_Ex_risk < &conv_criterion) OR %sysevalf(&i >
&max_iteration)) ;
* first time through loop, set expsosure_conc=init_exposure_conc;
%If &i=l %Then
%Do;
%Let exposure_conc=&init_exposure_conc;
%End;
%If &i>l %Then
%Do;
data tempBEIRCONVERGE;
/* BT March 2019: BEIR4 has run at
least once. Adjust exposure_conc
Extra_Riskm is created in
BEIR4 (=Ex_Risk)*/
concentration in loop i-1;
^update the concentration;
NumLoops=&i;
thisExposureConc=&exposure_conc; *set equal to
numvar=&ex_risk_target;
denvar=&Extra_Riskm;
thisexposureconc = thisexposureconc * (numvar/denvar);
call symput('exposure_conc',thisexposureconc);
output;
Run;
%End; ^Corresponds to If i>l statement;
%BEIR4;
%Let i=%eval(&i+l);
%End;
%Let EC_lPercent = &exposure_conc;
/ *	+
I Report results if convergence criterion met:
+	*/
%If %sysevalf(&Delta_Ex_risk < &conv_criterion) %then %do;
title5 "based on KM=&COEF, Concentration=&EC_lPercent, and LastAge=&LastAge"
data _null_;	/* Modified 26-july-00 */
pointer=l;
set allcause (keep=age
rename=(age=ageall0)) point=pointer nobs=n_all;
set cause (keep=age
rename=(age=ageCsO)) point=pointer nobs=n_cause;
pointer=n_all;
set allcause (keep=age
rename=(age=agealll)) point=pointer nobs=n_all;
pointer=n_cause;
set cause (keep=age
rename=(age=ageCsl)) point=pointer nobs=n_cause;
Tmp = sum(min(AgeAlll,AgeCsl,(&Lastage-l)),1);
file PRINT;
if ageallO NE ageCsO then DO;
put /"ERROR: The initial age for all-causes rate differs from the"
/"	initial age for the cause-specific rate.";
Page 326 of 352

-------
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
END;
else DO;
put /
"Values of
macro variables
used in
this computation:
//
@3
"Value"

@17
"Macro Var" @29
"Description"
/
@3
II TT

@17
II
" I3 0Q




IS z y

//
@3
"&Model

@17
"MODEL"
@29
"1 = Loglinear Relative Rate,"
/





@29
"2 = Linear Relative Rate, "
/





@29
"3 = Linear Absolute Rate, "
/





@29
"4 = 'Power' Relative Rate, "
/





@29
"0 = User defined. "
/
@3
"&Coef

@17
"COEF"
@29
"Exposure parameter estimate"
//
@3
"&Lag

@17
"LAG"
@29
"Exposure Lag "
//
@3
"&Agelst
x'
@17
"AGE1ST
X" @29
"Age exposure begins"
/
@3
"^Duration*
@17
"DURATION" @29
"E'uration of exposure"
/
@3
"&EnvAdj
"
@17
"ENVADJ
@29
"Adjust dose from intermittent"
/





@29
"occupational exposures to "
/


/
@3
„
@29
@17
"continuous environmental exposures"
	.. @29 "	
// n	„
// @3 "EC1% = " @10 "&EC_1Percent" @20 " (f/ml); Rx = " @34
"&Extra_Ris km"
// M	
/"The risks are calculated from age " ageallO " up to age " Tmp "."
// ;
if agealll NE ageCsl then
put /"WARNING: The last age for the all-causes rates differs from"
/"	the last age for the cause-specific rates, suggesting"
/"	the possibility that the rates weren't entered as desired."
/;
END;
Stop;
run;
/* BT 7/5/19: Start of code that was added to merge variables for unexposed risk
(S	1i and S_lix) to the rest of the output, by age;
*/
Data newSRCData(keep=SRC_age SRC_S_li SRC_S_lix);
set ex_Risk;
SRC_age=0; SRC_S_li=l; SRC_S_lix=l;
output;
do obsnum=l to last-1;
set ex_Risk point=obsnum nobs=last;
if _error_ then abort;
SRC_age=age+l; SRC_S_li=S_li; SRC_S_lix=S_lix;
output;
end;
stop;
run;
*	rename variables to enable overwriting the values of S_li and S_lix in ex_risk with the values
in newSRCData;
* Data file tempSRCData has age=0-85 while the ex_Risk file has age 0-84, with last two
records
both having age=84.;
Data tempSRCData; Set newSRCData(rename=(SRC_Age=age SRC_S_li=S_li SRC_S_lix=S_lix));
if age=&LastAge then age=%sysevalf(&Lastage-l); Else age=age;
Run;
*	there are duplicate values for age in both ex_risk and tempSRCData
which may produce too many records, if that happens, then we use two set
statements;
Data ex_risk; merge ex_risk tempSRCData; By Age; Run;
/* BT 7/5/19: End of code that was added to merge variables for unexposed risk
Page 327 of 352

-------
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
(S	1i and S_lix) to the rest of the output, by age;
*/
proc print data=ex_risk label noobs;
var Age TSFE d Q hix hi hstari hstarix qi qix S_li S_lix R0 Risk;
label d="Exp. duration midway thru year i (yrs)"
TSFE="TSFE midway thru year i (yrs)"
Q="Q (yrs3)"
qix="Probability of surviving year i assuming alive at start (exposed)
(qei)";
run;
%End; *end of the If statement that tests if convergence was met;
%Mend Converge_BEIR4;
/* the following options are for debugging - comment out after code is running as expected*/
Options mlogic mprint symbolgen;
%Let LastAge =85;
%LET LAG = 10;
%Let MODEL = 3;
%Let COEF = 0.000000015;
April 2019: BT (SRC) Added maxro CONVERGE_BEIR4 which iteratively
runs macro BEIR4 until the EXPOSURE_CONCENTRATION corresponds to an
extra_risk=0.01 (the point of departure [POD]).
At the second iteration of the Converge_BEIR4 macro, the exposure
concentration is adjusted by a factor equal to the initial
concentration x ConvRate. It is recommended to use a convrate equal
to 0.1, which produces an adjustment of approximately 10% of the
initial concentration value. The conversion rate is adjusted in
later iterations (to smaller adustments) as needed to converge.
I In addition to the parameter for CONVERGE_BEIR4, the user should also|
I review parameters and data that are assigned/entered in Part 1 and |
I Part II (see above). Parameters for CONVERGE_BEIR4 are defined below |
+	*/
^%BEIR4;
originally called macr BEIR4 directly. Now BEIR4 is called by Converge_BEIR4;
bConverge_BEIR4(init_exposure_conc=0.1,
/* initial exposure concentration (initial
guess) */
(POD) - the target extra risk */
run times */
ex_risk_target= 0.0100,
conv_criterion= 0.00000001,
max iteration=300);
/* the point of departure
/* to avoid excessively long
Run;
Page 328 of 352

-------
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
Appendix J Results of Modeling for IUR Derivation
Section 1
Hein et al. (2007)
SOUTH CAROLINA LUNG CANCER KL FITTING
Citation: Hein et a I 2007
Cohort: South Carolina
Data: NIOSH
CE10 (PCM s/cc-yrs)
Lung Cancer Deaths
Win
Max
Mid
Obs
Exp
RR
0
1.5
0.8
34
22.10
1.54
1.5
5
2.9
33
25.30
1.30
5
15
8.7
34
21.70
1.57
15
60
31.3
35
18.80
1.86
80
120
85.3
37
9.20
4.02
120

189.6
25
4.70
5.32
198
101.3D
1.94
AJpfct Feted at 1.00
10
5 -
4
3 -
2 -
a = 1.0 (fixed)
KL = 2.&4E-02

50
100	150
CE10 (PCM 5/cc-yr)
200
Alpha = Fitted
250
ID
7 -
6
5
4 -
3 -
2 -r
a = 1.35 (fitted)
KL = 1.76E-02

100	150
CE10 (PCM s/cc-yr)
200
250
Value
Alpha
KL
AIC
MLE
1.00
2.94E-02
42.41
UB
1.00
3.88E-02
-
Value
Alpha
KL
AIC
MLE
1.35
1.76E-02
30.48
UB
1.17
2.64E-02
-
Page 329 of 352

-------
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
Section 2
Loomis et al. (2009)
NORTH CAROLINA LUNG CANCER KL FITTING
Citation: Loomis et al 2009
Cohort: North Carolina
Data: Table 5
CE10 (PCM sftttfs)
Lunq Cancer Deaths
Min
Max
Mid
Obs
Exp
RR
0
2.3
1.2
37
37.00
1.00
2.3
11.5
6.9
37
32.74
1.13
11.5
34.8
23.2
35
22.15
1.58
34.6
152.7
93.8
37
29.60
1.25
152.7
2194
1173.4
35
18.62
1.88
181 140.11 1
29
Alpha Fixed at 1.00
3.0
2.5 ฆ
1 20
I 15 -
I
1.0
0.5 ฆ
0.0
a = 1.0 (fixed)
KL = 8.08E-C4
200
43C
500 0OD
CE1Q(PCM sJoo-yr}
10CO
12G0
Value
Alpha
KL
AIC
MLE
1.00
8.08E-04
35.33
UB
1.00
1.31E-Q3
-
Alpha = Fitted
1430
3.0
2.5 ฆ
2.0 -
1.5
1.0
0.5
0.0
a = 1.18 (fitted)
KL = 5.15E-34
200 400 600 800 1000
CEI0 (PCM a'cc-yr}
1200
14C0
Value
Alpha
KL
AIC
MLE
1.18
5.15E-04
32.83
UB
1.09
1.02E-03
-
Page 330 of 352

-------
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
Section 3
Wang et al. (2013b)
QINGHAI, CHINA LUNG CANCER KL FITTING
Citation: Wang et al, 2013
Cohort: Chinese miners (all)
Data: Table 5 + 6
CE10 (PCM s/cc-yre)
Lunq Cancer Deaths
Min
Max
Mid
Obs
Exp
RR
a
20
10.0
6
5.75
1.04
20
100
60.0
12
2.82
4.25
100
450
275.0
17
1.57
10.32
450
1097
773,5
21
1.23
17.07
56 11.37 4.92
Alpha Fixed at 1.00	Alpha = Fitted
35
3D
25
I 2D
i is
I
to
5
0
o = 1.21 (fitted!
Kl = 2.16E-02
Value
Alpha
KL
AlC
MLE
1.00
2.72E-02
23.62
UB
1.00
3.51E-02
-
Value
Alpha
KL
AlC
MLE
1.21
2.16E-02
24.44
UB
0.46
6.47E-02
-
Page 331 of 352

-------
2747
2748
2749
2750
Age at 1st
exposure
(years)
Duration of exposure (years)
1
5
10
15
20
25
30
35
40
62
78
0
4.14E-
03
3.34E-
02
6.34E-
02
8.71E-
02
1.06E-
01
1.20E-
01
1.31E-
01
1.38E-
01
1.44E-
01
1.55E-
01
1.56E-
01
2
3.82E-
03
3.06E-
02
5.80E-
02
7.96E-
02
9.63E-
02
1.09E-
01
1.18E-
01
1.25E-
01
1.30E-
01
1.40E-
01
1.41E-
01
4
3.52E-
03
2.81E-
02
5.30E-
02
7.25E-
02
8.75E-
02
9.88E-
02
1.07E-
01
1.13E-
01
1.18E-
01
1.26E-
01
1.26E-
01
6
3.23E-
03
2.56E-
02
4.83E-
02
6.59E-
02
7.93E-
02
8.94E-
02
9.68E-
02
1.02E-
01
1.06E-
01
1.13E-
01
1.13E-
01
8
2.97E-
03
2.34E-
02
4.39E-
02
5.97E-
02
7.17E-
02
8.07E-
02
8.73E-
02
9.20E-
02
9.54E-
02
1.01E-
01

10
2.72E-
03
2.13E-
02
3.98E-
02
5.40E-
02
6.47E-
02
7.26E-
02
7.84E-
02
8.26E-
02
8.56E-
02
9.08E-
02

12
248E-
03
1.93E-
02
3.60E-
02
4.87E-
02
5.82E-
02
6.53E-
02
7.04E-
02
7.41E-
02
7.67E-
02
8.10E-
02

14
2.27E-
03
1.75E-
02
3.25E-
02
4.38E-
02
5.23E-
02
5.85E-
02
6.30E-
02
6.63E-
02
6.86E-
02
7.21E-
02

16
2.07E-
03
1.58E-
02
2.92E-
02
3.93E-
02
4.68E-
02
5.23E-
02
5.63E-
02
5.92E-
02
6.12E-
02
6.41E-
02

18
1.88E-
03
142E-
02
2.62E-
02
3.52E-
02
4.19E-
02
4.67E-
02
5.02E-
02
5.28E-
02
5.46E-
02
5.69E-
02

20
1.71E-
03
1.28E-
02
2.35E-
02
3.15E-
02
3.74E-
02
4.17E-
02
4.48E-
02
4.70E-
02
4.86E-
02
5.04E-
02

22
1.56E-
03
1.14E-
02
2.10E-
02
2.81E-
02
3.33E-
02
3.71E-
02
3.99E-
02
4.19E-
02
4.33E-
02
4.46E-
02

24
142E-
03
1.02E-
02
1.87E-
02
2.50E-
02
2.96E-
02
3.30E-
02
3.55E-
02
3.73E-
02
3.85E-
02


26
1.29E-
03
9.15E-
03
1.67E-
02
2.23E-
02
2.64E-
02
2.94E-
02
3.16E-
02
3.32E-
02
3.42E-
02


28
1.17E-
03
8.16E-
03
1.49E-
02
1.98E-
02
2.34E-
02
2.62E-
02
2.81E-
02
2.95E-
02
3.04E-
02


30
1.07E-
03
7.27E-
03
1.32E-
02
1.76E-
02
2.09E-
02
2.33E-
02
2.51E-
02
2.63E-
02
2.69E-
02


32
9.76E-
04
648E-
03
1.18E-
02
1.57E-
02
1.86E-
02
2.08E-
02
2.24E-
02
2.34E-
02
2.38E-
02


34
8.95E-
04
5.78E-
03
1.05E-
02
1.40E-
02
1.66E-
02
1.86E-
02
2.00E-
02
2.07E-
02
2.10E-
02


36
8.24E-
04
5.17E-
03
9.37E-
03
1.25E-
02
1.49E-
02
1.66E-
02
1.78E-
02
1.84E-
02
1.85E-
02


38
7.62E-
04
4.63E-
03
8.39E-
03
1.12E-
02
1.34E-
02
1.49E-
02
1.58E-
02
1.62E-
02
1.62E-
02


40
7.08E-
04
4.16E-
03
7.54E-
03
1.01E-
02
1.20E-
02
1.33E-
02
1.40E-
02
1.42E-
02
1.42E-
02


Page 332 of 352
Appendix K Less Than Lifetime (or Partial lifetime)
IUR
TableApx K-l. (LTL) Chrysotile Asbestos Inhalation Unit Risk Values for Less Than
Lifetime Condition of Use

-------
42
6.58E-
04
3.75E-
03
6.79E-
03
9.09E-
03
1.08E-
02
1.18E-
02
1.23E-
02
1.24E-
02
1.24E-
02


44
6.16E-
04
3.39E-
03
6.12E-
03
8.18E-
03
9.63E-
03
1.05E-
02
1.08E-
02
1.08E-
02
1.08E-
02


46
5.74E-
04
3.07E-
03
5.53E-
03
7.35E-
03
8.56E-
03
9.17E-
03
9.31E-
03
9.31E-
03



48
5.35E-
04
2.78E-
03
4.98E-
03
6.57E-
03
7.54E-
03
7.95E-
03
7.98E-
03
7.98E-
03



50
4.99E-
04
2.52E-
03
4.48E-
03
5.82E-
03
6.56E-
03
6.77E-
03
6.77E-
03
6.77E-
03



52
4.61E-
04
2.28E-
03
3.99E-
03
5.09E-
03
5.60E-
03
5.68E-
03
5.68E-
03




54
4.26E-
04
2.04E-
03
3.51E-
03
4.37E-
03
4.68E-
03
4.69E-
03
4.69E-
03




56
3.87E-
04
1.81E-
03
3.03E-
03
3.66E-
03
3.80E-
03
3.80E-
03





58
3.47E-
04
1.57E-
03
2.56E-
03
2.97E-
03
3.00E-
03
3.00E-
03





60
3.07E-
04
1.34E-
03
2.08E-
03
2.30E-
03
2.30E-
03
2.30E-
03





62
2.61E-
04
1.10E-
03
1.61E-
03
1.69E-
03
1.69E-
03






64
2.17E-
04
8.58E-
04
1.17E-
03
1.18E-
03
1.18E-
03






66
1.69E-
04
6.25E-
04
7.61E-
04
7.61E-
04







68
1.23E-
04
4.10E-
04
4.43E-
04
4.43E-
04







70
8.08E-
05
2.17E-
04
2.17E-
04
2.17E-
04







72
4.22E-
05
7.47E-
05
7.47E-
05








74
7.97E-
06
7.97E-
06
7.97E-
06








2751	For calculation of Table Apx K-l, the following procedure was used. For each cell of the table,
2752	the lung cancer and mesothelioma partial lifetime risk corresponding to the age at first exposure
2753	and duration of exposure was calculated using selected models for lung cancer and mesothelioma
2754	and potency factors from Table 3-10 and 3-11, Then lung cancer and mesothelioma risks were
2755	statistically combined using the same procedure as described in Section 3.2.3.8.2.
Page 333 of 352

-------
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
Appendix L Sensitivity Analysis of Exposures for
DIY/Bystander Episodic Exposure Scenarios
As presented in Section 4.3.5, there are some uncertainties pertaining to the assumptions made
for exposure durations for both DIY users and bystanders for the brake repair/replacement
scenarios. This Appendix provides a more detailed analyses using various combinations of age
at start of first exposure and duration of exposure for both the DIYers and the bystanders for both
the brake repair/replacement and the UTV gasket repair/replacement scenarios.
In Table L-l, the assumption is that DIY brake/repair replacement with compressed air begins at
age 16 years and continues for 20 years instead of for 62 years.
Here, the unit risk for Users is:
The unit risk for Bystanders is:
IURltl(DIY Brakes) = IUR(16,20) = 0.0468 per f/cc
IURltl(DIY Bystanders) = IUR(0,20) = 0.1057 per f/cc
TableApx L-l. Excess Lifetime Cancer Risk for Indoor DIY Brake/Repair Replacement
with Compressed Air Use for Consumers for 20 year duration (exposures from Table 2-27
Consumer
Exposure
Scenario
Exposure Levels (f/cc)
ELCR (20 yr
exposure starting at
age 16 years)
ELCR ((20 yr
exposure starting at
age 0 years))
DIY User
DIY Bystander
DIY User
DIY Bystander
Central
Tendency
High-
end
Central
Tendency
High-end
Central
Tendency
High-
end
Central
Tendency
High-
end
Aftennarket
automotive parts -
brakes (3-hour
TWA indoors
every 3 years with
compressed air)
0.0445
0.4368
0.0130
0.0296
2.6 E-5
2.6 E-4
1.7 E-5
3.9 E-5
TWFConcomitant Exposures (1 hour per day every day) (l/24)*(365/365) = 0.04167
DIY User: ELCR (Central Tendency) 0.0445 f/cc • 0.0001142 • 0.0468 per f/cc + 0.0445 • 0.3 • 0.04167 • 0.0468
DIY User: ELCR,High-endi = 0.4368 f/cc • 0.0001142 • 0.0468 per f/cc + 0.4368 • 0.3 • 0.04167 • 0.0468
DIY Bystander: ELCR (Central Tendency) 0.013 f/cc • 0.0001142 • 0.1057 per f/cc + 0.013 • 0.3 • 0.04167 • 0.1057
DIY Bystander: ELCR,High-endi = 0.0296 f/cc • 0.0001142 • 0.1057 per f/cc + 0.0296 • 0.3 • 0.04167 • 0.1057
Exposure values from Table 2-27 were used to represent indoor brake work (with compressed
air) and are the basis for the exposure levels used in Table Apx L-l. EPA then assumed that the
concentration of chrysotile asbestos in the interval between brake work (every 3 years) is 30% of
that during measured active use. Consumers were assumed to spend one hour per day in their
garages based on the 50th percentile estimate in the EPA Exposure Factors Handbook. Based on
these assumptions, the consumer risk estimates were exceeded for central tendency and high-end
exposures (L-l). Estimates exceeding the benchmark are shaded in pink and bolded. Comparing
these results with those of Table 4-39, we see that the ratio of the risks for the DIY User based
on 20 years exposure compared to 40 years of exposures is equal to the ratio of the less than
lifetime inhalation unit risks:
DIY Users: [IUR(16,20) = 0.0468 per f/cc] / [IUR(16,62) = 0.0641 per f/cc] = 0.73
DIY Users: [20 yr risk (Central) = 2.63 E-5] / [62 yr risk (Central) = 3.60 E-5] = 0.73
Page 334 of 352

-------
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
DIY Users: [20 yr risk (High) = 2.58 E-4] / [62 yr risk (High) = 3.50 E-4]
= 0.73
Similarly for bystanders, the ratio of the risk based on 20 years exposure compared to 62 years
exposure is equal to the ratio of the 20-year less than lifetime risk to the lifetime unit risk:
DIY Bystanders: [IUR(0,20) = 0.1057 per f/cc] / [IUR(Lifetime) = 0.16 per f/cc]	= 0.66
DIY Bystanders: [20 yr risk (Central) = 1.73 E-5] / [78 yr risk (Central) = 2.62 E-5] = 0.66
DIY Bystanders: [20 yr risk (High) = 3.95 E-5] / [78 yr risk (High) = 5.97 E-5]	= 0.66
Using this approach, and relying on the ratios presented in Table 4-49, TableApx L-2 provides
and ratios for five different sensitivity pairings.
Table Apx L-2. Ratios of risk for alternative exposure scenarios compared to DIY User
and Bystander exposure scenario assuming DIY User is first exposed at age 16 years for 62
	years duration and DIY Bystander is exposed from age 0-78 years.	
Kxposure scenario
Age at
Ill-Si
exposure
(years)
Duration
(years)
liasclinc
partial
lifetime Il k
Kxposure
scenario
partial
lifetime Il k
Uatio of
risks for
exposure
scenario
Baseline
DIY User
16
62
0.0641
0.0641
1
Bystander
0
78
0.16
0.16
1







Sensitivity #1
DIY User
16
20
0.0641
0.0468
0.73
Bystander
0
20
0.16
0.1057
0.66







Sensitivity #2
DIY User
20
40
0.0641
0.0486
0.76
Bystander
0
40
0.16
0.144
0.90







Sensitivity #3
DIY User
20
20
0.0641
0.0374
0.58
Bystander
0
20
0.16
0.1057
0.63







Sensitivity #4
DIY User
30
40
0.0641
0.0269
0.42
Bystander
0
40
0.16
0.144
0.90







Sensitivity #5
DIY User
30
20
0.0641
0.0209
0.33
Bystander
0
20
0.16
0.1057
0.63
Table Apx L-3 through Table Apx L-7 below show the results of applying these ratios to all of
the possible scenarios presented in Table 4-48 using the five sensitivity analyses pairings in
Table Apx L-2. Table Apx L-8 at the end summarizes the results to show how only one of 24
scenarios changes from an exceedance to no exceedance for all five sensitivity analyses (DIY
user, Brakes Repair/ replacement, Outdoor, once every 3 years, 30 min/d in driveway, high-end
only).
Page 335 of 352

-------
2809
2810
2811
2812
TableApx L-3. Sensitivity Analysis #1: Summary of Risk Estimates for Inhalation
Exposures to Consumers and Bystanders by COU (Cancer benchmark is 10"6) Comparing
the Baseline Exposure Scenario from Table 4-48 with Risks Assuming DIY Users Are
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
16-36 (*0.73)
and
Bystanders
0-20 (*0.66
Imported
Asbestos
Products
Brakes
Repair/replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 1
hour/d in garage
Section
4.2.3.1
DIY
Central
Tendency
3.6 E-5
2.6 E-5
High-end
3.5 E-4
2.6 E-4
Bystander
Central
Tendency
2.6 E-5
1.7 E-5
High-end
6.0 E-5
4.0 E-5
Brakes Repair/
replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 8
hours/d in garage
Section
4.2.3.1
DIY
Central
Tendency
2.6 E-4
1.9 E-4
High-end
2.6 E-3
1.9 E-3
Bystander
Central
Tendency
1.7 E-5
1.1 E-5
High-end
3.9 E-5
2.6 E-5
Brakes
Repair/replacement
Indoor,
compressed air,
once at 16 years,
staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.1
DIY
Central
Tendency
5.4 E-6
3.9 E-6
High End
5.3 E-5
3.9 E-5
Bystander
Central
Tendency
3.4 E-6
2.2 E-6
High-end
7.8 E-6
5.1 E-6
Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 5
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
8.2 E-8
6.0 E-8
High-end
4.4 E-7
3.2 E-7
Bystander
Central
Tendency
2.1 E-8
1.4 E-8
High-end
1.1 E-7
7.3 E-8
Page 336 of 352

-------
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
16-36 (*0.73)
and
Bystanders
0-20 (*0.66

Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 30
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
2.4 E-7
1.8 E-7
High-end
1.3 E-6
9.5 E-7
Bystander
Central
Tendency
5.9 E-8
3.9 E-8
High-end
3.2 E-7
2.1 E-7
Imported
Asbestos
Products
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 1
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.9 E-5
1.4 E-5
High-end
5.3 E-5
3.9 E-5
Bystander
Central
Tendency
2.4 E-5
1.6 E-5
High-end
6.1 E-5
4.0 E-5
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 8
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.5 E-4
1.1 E-4
High-end
4.2 E-4
3.1 E-4
Bystander
Central
Tendency
2.4 E-5
1.6 E-5
High-end
6.1 E-5
4.0 E-5
Gasket Repair
Repair/replacement
Indoor, once at 16
years, staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.2
DIY
Central
Tendency
2.9 E-6
2.1 E-6
High end
8.0 E-6
5.8 E-6
Bystander
Central
Tendency
3.2 E-6
2.1 E-6
High-end
7.9 E-6
5.2 E-6
2813
Page 337 of 352

-------
2814
2815
2816
2817
TableApx L-4. Sensitivity Analysis #2: Summary of Risk Estimates for Inhalation
Exposures to Consumers and Bystanders by COU (Cancer benchmark is 10"6) Comparing
the Baseline Exposure Scenario from Table 4-48 with Risks Assuming DIY Users Are
Exposed From Age 20-6
0 years anc
Bystanders Are Exposed Age 0-40 years.
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
20-40 (*0.76)
and
Bystanders
0-40 (*0.90)
Imported
Asbestos
Products
Brakes
Repair/replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 1
hour/d in garage
Section
4.2.3.1
DIY
Central
Tendency
3.6 E-5
2.7 E-5
High-end
3.5 E-4
2.7 E-4
Bystander
Central
Tendency
2.6 E-5
2.3 E-5
High-end
6.0 E-5
5.4 E-5
Brakes Repair/
replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 8
hours/d in garage
Section
4.2.3.1
DIY
Central
Tendency
2.6 E-4
2.0 E-4
High-end
2.6 E-3
2.0 E-3
Bystander
Central
Tendency
1.7 E-5
1.5 E-5
High-end
3.9 E-5
3.5 E-5
Brakes
Repair/replacement
Indoor,
compressed air,
once at 16 years,
staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.1
DIY
Central
Tendency
5.4 E-6
4.1 E-6
High End
5.3 E-5
4.0 E-5
Bystander
Central
Tendency
3.4 E-6
3.1 E-6
High-end
7.8 E-6
7.0 E-6
Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 5
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
8.2 E-8
6.2 E-8
High-end
4.4 E-7
3.3 E-7
Bystander
Central
Tendency
2.1 E-8
1.9 E-8
High-end
1.1 E-7
9.9 E-8
Page 338 of 352

-------
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
20-40 (*0.76)
and
Bystanders
0-40 (*0.90)

Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 30
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
2.4 E-7
1.8 E-7
High-end
1.3 E-6
9.9 E-7
Bystander
Central
Tendency
5.9 E-8
5.3 E-8
High-end
3.2 E-7
2.9 E-7
Imported
Asbestos
Products
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 1
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.9 E-5
1.4 E-5
High-end
5.3 E-5
4.0 E-5
Bystander
Central
Tendency
2.4 E-5
2.2 E-5
High-end
6.1 E-5
5.5 E-5
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 8
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.5 E-4
1.1 E-4
High-end
4.2 E-4
3.2 E-4
Bystander
Central
Tendency
2.4 E-5
2.2 E-5
High-end
6.1 E-5
5.5 E-5
Gasket Repair
Repair/replacement
Indoor, once at 16
years, staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.2
DIY
Central
Tendency
2.9 E-6
2.2 E-6
High end
8.0 E-6
6.1 E-6
Bystander
Central
Tendency
3.2 E-6
2.9 E-6
High-end
7.9 E-6
7.1 E-6
2818
Page 339 of 352

-------
2819
2820
2821
2822
TableApx L-5. Sensitivity Analysis #3: Summary of Risk Estimates for Inhalation
Exposures to Consumers and Bystanders by COU (Cancer benchmark is 10"6) Comparing
the Baseline Exposure Scenario from Table 4-48 with Risks Assuming DIY Users Are
Exposed From Age 20-4
0 years anc
Bystanders Are Exposed Age 0-20 years.
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
20-40 (*0.58)
and
Bystanders
0-20 (*0.63)
Imported
Asbestos
Products
Brakes
Repair/replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 1
hour/d in garage
Section
4.2.3.1
DIY
Central
Tendency
3.6 E-5
2.1 E-5
High-end
3.5 E-4
2.0 E-4
Bystander
Central
Tendency
2.6 E-5
1.6 E-5
High-end
6.0 E-5
3.8 E-5
Brakes Repair/
replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 8
hours/d in garage
Section
4.2.3.1
DIY
Central
Tendency
2.6 E-4
1.5 E-4
High-end
2.6 E-3
1.5 E-3
Bystander
Central
Tendency
1.7 E-5
1.1 E-5
High-end
3.9 E-5
2.5 E-5
Brakes
Repair/replacement
Indoor,
compressed air,
once at 16 years,
staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.1
DIY
Central
Tendency
5.4 E-6
3.1 E-6
High End
5.3 E-5
3.1 E-5
Bystander
Central
Tendency
3.4 E-6
2.1 E-6
High-end
7.8 E-6
4.9 E-6
Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 5
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
8.2 E-8
4.8 E-8
High-end
4.4 E-7
2.6 E-7
Bystander
Central
Tendency
2.1 E-8
1.3 E-8
High-end
1.1 E-7
6.9 E-8
Page 340 of 352

-------
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
20-40 (*0.58)
and
Bystanders
0-20 (*0.63)

Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 30
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
2.4 E-7
1.4 E-7
High-end
1.3 E-6
7.5 E-7
Bystander
Central
Tendency
5.9 E-8
3.7 E-8
High-end
3.2 E-7
2.0 E-7
Imported
Asbestos
Products
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 1
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.9 E-5
1.1 E-5
High-end
5.3 E-5
3.1 E-5
Bystander
Central
Tendency
2.4 E-5
1.5 E-5
High-end
6.1 E-5
3.8 E-5
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 8
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.5 E-4
8.7 E-5
High-end
4.2 E-4
2.4 E-4
Bystander
Central
Tendency
2.4 E-5
1.5 E-5
High-end
6.1 E-5
3.8 E-5
Gasket Repair
Repair/replacement
Indoor, once at 16
years, staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.2
DIY
Central
Tendency
2.9 E-6
1.7 E-6
High end
8.0 E-6
4.6 E-6
Bystander
Central
Tendency
3.2 E-6
2.0 E-6
High-end
7.9 E-6
5.0 E-6
2823
Page 341 of 352

-------
2824
2825
2826
2827
TableApx L-6. Sensitivity Analysis #4: Summary of Risk Estimates for Inhalation
Exposures to Consumers and Bystanders by COU (Cancer benchmark is 10"6) Comparing
the Baseline Exposure Scenario from Table 4-48 with Risks Assuming DIY Users Are
Exposed From Age 30-7
0 years anc
Bystanders Are Exposed Age 0-40 years.
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
30-70 (*0.42)
and
Bystanders
0-40 (*0.90)
Imported
Asbestos
Products
Brakes
Repair/replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 1
hour/d in garage
Section
4.2.3.1
DIY
Central
Tendency
3.6 E-5
1.5 E-5
High-end
3.5 E-4
1.5 E-4
Bystander
Central
Tendency
2.6 E-5
2.3 E-5
High-end
6.0 E-5
5.4 E-5
Brakes Repair/
replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 8
hours/d in garage
Section
4.2.3.1
DIY
Central
Tendency
2.6 E-4
1.1 E-4
High-end
2.6 E-3
1.1 E-3
Bystander
Central
Tendency
1.7 E-5
1.5 E-5
High-end
3.9 E-5
3.5 E-5
Brakes
Repair/replacement
Indoor,
compressed air,
once at 16 years,
staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.1
DIY
Central
Tendency
5.4 E-6
2.3 E-6
High End
5.3 E-5
2.2 E-5
Bystander
Central
Tendency
3.4 E-6
3.1 E-6
High-end
7.8 E-6
7.0 E-6
Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 5
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
8.2 E-8
3.4 E-8
High-end
4.4 E-7
1.8 E-7
Bystander
Central
Tendency
2.1 E-8
1.9 E-8
High-end
1.1 E-7
9.9 E-8
Page 342 of 352

-------
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
30-70 (*0.42)
and
Bystanders
0-40 (*0.90)

Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 30
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
2.4 E-7
1.0 E-7
High-end
1.3 E-6
5.5 E-7
Bystander
Central
Tendency
5.9 E-8
5.3 E-8
High-end
3.2 E-7
2.9 E-7
Imported
Asbestos
Products
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 1
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.9 E-5
8.0 E-6
High-end
5.3 E-5
2.2 E-5
Bystander
Central
Tendency
2.4 E-5
2.2 E-5
High-end
6.1 E-5
5.5 E-5
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 8
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.5 E-4
6.3 E-5
High-end
4.2 E-4
1.8 E-4
Bystander
Central
Tendency
2.4 E-5
2.2 E-5
High-end
6.1 E-5
5.5 E-5
Gasket Repair
Repair/replacement
Indoor, once at 16
years, staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.2
DIY
Central
Tendency
2.9 E-6
1.2 E-6
High end
8.0 E-6
3.4 E-6
Bystander
Central
Tendency
3.2 E-6
2.9 E-6
High-end
7.9 E-6
7.1 E-6
2828
Page 343 of 352

-------
2829
2830
2831
2832
TableApx L-7. Sensitivity Analysis #5: Summary of Risk Estimates for Inhalation
Exposures to Consumers and Bystanders by COU (Cancer benchmark is 10"6) Comparing
the Baseline Exposure Scenario from Table 4-48 with Risks Assuming DIY Users Are
Exposed From Age 30-5
0 years anc
Bystanders Are Exposed Age 0-20 years.
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
16-36 (*0.33)
and
Bystanders
0-20 (*0.63)
Imported
Asbestos
products
Brakes
Repair/replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 1
hour/d in garage
Section
4.2.3.1
DIY
Central
Tendency
3.6 E-5
1.2 E-5
High-end
3.5 E-4
1.2 E-4
Bystander
Central
Tendency
2.6 E-5
1.6 E-5
High-end
6.0 E-5
3.8 E-5
Brakes Repair/
replacement
Indoor,
compressed air,
once every 3 years
for 62 years
starting at 16
years, exposures at
30% of active used
between uses, 8
hours/d in garage
Section
4.2.3.1
DIY
Central
Tendency
2.6 E-4
8.6 E-5
High-end
2.6 E-3
8.6 E-4
Bystander
Central
Tendency
1.7 E-5
1.1 E-5
High-end
3.9 E-5
2.5 E-5
Brakes
Repair/replacement
Indoor,
compressed air,
once at 16 years,
staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.1
DIY
Central
Tendency
5.4 E-6
1.8 E-6
High End
5.3 E-5
1.7 E-5
Bystander
Central
Tendency
3.4 E-6
2.1 E-6
High-end
7.8 E-6
4.9 E-6
Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 5
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
8.2 E-8
2.7 E-8
High-end
4.4 E-7
1.5 E-7
Bystander
Central
Tendency
2.1 E-8
1.3 E-8
High-end
1.1 E-7
6.9 E-8
Page 344 of 352

-------
Life Cycle
Stage/Category
Subcategory
Consumer
Exposure
Scenario
Population
Exposure
Duration
and Level
Cancer Risk
Estimates
(from Table
4-48)
Cancer Risk
Estimates
Users age
16-36 (*0.33)
and
Bystanders
0-20 (*0.63)

Brakes Repair/
replacement
Outdoor, once
every 3 years for
62 years starting at
16 years,
exposures at 2% of
active used
between uses, 30
min/d in driveway
Section
4.2.3.1
DIY
Central
Tendency
2.4 E-7
7.9 E-8
High-end
1.3 E-6
4.3 E-7
Bystander
Central
Tendency
5.9 E-8
3.7 E-8
High-end
3.2 E-7
2.0 E-7
Imported
Asbestos
Products
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 1
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.9 E-5
6.3 E-6
High-end
5.3 E-5
1.7 E-5
Bystander
Central
Tendency
2.4 E-5
1.5 E-5
High-end
6.1 E-5
3.8 E-5
Gaskets Repair/
replacement in
UTVs
Indoor, 1 hour/d,
once every 3 years
for 62 years
starting at 16 years
exposures at 30%
of active used
between uses, 8
hour/d in garage
Section
4.2.3.2
DIY
Central
Tendency
1.5 E-4
5.0 E-5
High-end
4.2 E-4
1.4 E-4
Bystander
Central
Tendency
2.4 E-5
1.5 E-5
High-end
6.1 E-5
3.8 E-5
Gasket Repair
Repair/replacement
Indoor, once at 16
years, staying in
residence for 10
years, 1 hour/d in
garage
Section
4.2.3.2
DIY
Central
Tendency
2.9 E-6
9.6 E-7
High end
8.0 E-6
2.6 E-6
Bystander
Central
Tendency
3.2 E-6
2.0 E-6
High-end
7.9 E-6
5.0 E-6
2833
Page 345 of 352

-------
2834	TableApx L-4: Results of 24 Sensitivity Analysis of Exposure Assumptions for Consumer
2835		DIY/Bystander Episodic Exposure Scenarios	
Sensitivity
Analysis
l)IY (age at start
and age at end of
duration)
Bystander (age at
start and age at
end of duration)
Change in Kisk
from Kxceedance
to No
Kxceedance
Scenario
A (Tected
Baseline
16-78
0-78
None
17/24 Exceed
Benchmarks
1
16-36
0-20
1/24
DIY user, Brake
repair, 30 min/day,
high-end
2
20-60
0-40
1/24
DIY user, Brake
repair, 30 min/day,
high-end
3
20-40
0-40
1/24
DIY user, Brake
repair, 30 min/day,
high-end
4
30-70
0-40
1/24
DIY user, Brake
repair, 30 min/day,
high-end
5
30-50
0-20
1/24
DIY user, Brake
repair, 30 min/day,
high-end
2836
Page 346 of 352

-------
2837	Appendix M Adjustment Factors to Correct for Bias
2838	in Cancer Risk Estimation
2839	This appendix presents estimates for an "adjustment factor" used to correct for bias in the
2840	estimation of the risk of cancer due to the lack of inclusion of cancers other than lung cancer and
2841	mesothelioma in the estimation of the inhalation risk.
2842	Biases in the Cancer Risk Values
2843	The initial analysis did not include the risk from other cancers that have been associated with
2844	exposure to chrysotile asbestos. The reason for these shortcomings in the analysis is simply that
2845	there were no studies that could be used to model the exposure-response relationship for
2846	incidence or for other causes of cancer. However, EPA has developed "adjustment factors"
2847	which are used to correct for the negative bias in the risk values derived from lung cancer and
2848	mesothelioma, which are described below.
2849	Biases Related to Not Including Other Cancer Sites
2850	The inhalation cancer risk estimates originally derived by EPA in the DRE were only based upon
2851	lung cancer and mesothelioma mortality. Other cancers that have been recognized as being
2852	causally associated with asbestos sites by the International Agency for Research on Cancer
2853	(1ARC) include laryngeal and ovarian cancer (IARC. 2009). The I ARC also noted that 'positive
2854	associations have been observed between exposure to all forms of asbestos and cancer of the
2855	pharynx, stomach, and colorectum'. However, the evidence for an association between these
2856	cancers and asbestos exposure is mixed and IARC did not view it as sufficient for a
2857	determination of causality. The EPA concurs with the IARC's evaluation and has limited its
2858	effort to estimating the additional risk of ovarian and laryngeal cancer from inhalation exposure
2859	to chrysotile asbestos.
2860	A direct estimate of the risk of ovarian and laryngeal cancer cannot be made since none of the
2861	published studies have reported exposure-response results for these sites. An indirect estimate of
2862	additional risk can be determined by developing an adjustment factor by comparing the excess
2863	deaths from lung cancer with the number of excess deaths from the other cancer sites using the
2864	following formula:
2865	Adjustment factor = 1 + (excess other cancer)/(excess lung cancer)
2866	This approach has been applied to estimate adjustment factors for ovarian and laryngeal cancers
2867	using data from studies of chrysotile asbestos-only exposed workers or had minimal exposures to
2868	other forms of asbestos and reported findings for these cancer sites. The results from these
2869	analyses are presented in Tables 1 and 2. Based on this analysis, the adjustment factors are 1.04
2870	and 1.02 for ovarian and laryngeal cancer, respectively. A combined adjustment factor can be
2871	estimated by summing the individual adjustment factors and subtracting one, which results in an
2872	overall adjustment factor of 1.06.
2873	The studies included in Tables 1 and 2 were the only published studies that were of workers who
2874	were only exposed to chrysotile asbestos or had only minor exposures to other forms of asbestos,
2875	and that reported results for laryngeal or ovarian cancer. They were identified by reviewing the
Page 347 of 352

-------
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
I ARC report, other reviews (Berman and Crump. 2008a; Hodgson and Darnton. 2000) and
Stayner et al. (1996). and published meta-analyses for ovarian (Camargo et al.. 2011; Reid et al..
2011) and laryngeal cancer (Peng et al.. 2.015).
Following is a brief description of the studies that were included in the estimation of the
adjustment factors with the exception of the studies by Hein et al. (2007). Loom is et al. (2009)
and Wang et al. (2013a). which were previously described in Section 3.2.4. It is noteworthy that
the number of cases of laryngeal and ovarian cancer observed in these studies are small, and the
results for these sites are generally statistically unstable {i.e., wide confidence intervals). For this
reason, we have pooled the results from these studies in order to estimate the adjustment factors
rather than relying on the results from individual studies alone.
Acheson et al. (1982) conducted a cohort mortality study of women who were exposed to
asbestos in manufacturing gas masks in 1939. One group of women manufactured masks
containing chrysotile asbestos at a facility in Blackburn, England (n=570), and the other
containing crocidolite at a facility in Leyland, England (n=757). Follow-up of these cohorts for
vital status ascertainment was from 1951 to 1980. Mortality rates from England and Wales were
used in a life-table analysis to compute expected numbers of death and standardized mortality
ratios (SMRs). A statistically non-significant increase in ovarian cancer (SMR=1.48,
95%CI=0.48-3.44) was observed among the women who manufactured masks using only
chrysotile asbestos. Lung cancer mortality was also not significantly elevated in the chrysotile
asbestos group (SMR=1.25, 95%CI=0.46-2.7248). This study did not report results for laryngeal
cancer.
Gardner et al. (1986) conducted a cohort study of 2167 workers (1510 men and 657 women) who
were employed sometime between 1941 and 1983 in a chrysotile asbestos cement products
factory in England. The factory only used chrysotile asbestos except for a "small" amount of
amosite during 4 months in 1976. Follow-up for ascertainment of vital status was through
December 31, 1984. SMRs were estimated using a life-table with rates for England and Wales
for the referent. There was little evidence of an increased risk of ovarian (SMR=1.11, 95%CI=
0.23-3.2549), laryngeal (SMR=0.91, 95%CI=0.02-5.0750) or lung cancer mortality (SMR =0.97,
95%CI=0.69-1.31) in this study.
Newhouse and Sullivan (1989) conducted a cohort mortality study of workers employed between
1941 and 1979 at a factory that produced friction products {i.e., brake blocks, and brake and
clutch linings). A total of 13,450 workers (9104 men and 4346 women) were followed for vital
status ascertainment until 1986. The factory only used chrysotile asbestos except during two
short periods when crocidolite was used. A slight excess of ovarian cancer (SMR=1.08, 90%CI=
61-179), and a deficit of laryngeal cancer mortality (SMR=0.64, 90%CI=0.28-1.26) was
observed in this study neither of which were statistically significant. Lung cancer mortality was
not elevated in this cohort (SMR= 0.99, 95%CI=0.87-1.13).
Tarchi et al. (1994) conducted a cohort mortality study of rock salt workers in Italy who were
exposed to chrysotile asbestos. The study included 487 workers (367 men and 120 women) who
48	Confidence interval was estimated using Fisher Exact method
49	Confidence interval was estimated using Fisher Exact method
50	Confidence interval was estimated using Fisher Exact method
Page 348 of 352

-------
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
were employed in the mine sometime between 1965 and 1989 and followed for vital status
ascertainment until the end of 1989. SMRs were estimated using lifetable methods with rates
from the Tuscany region as the referent. An increase in ovarian cancer (SMR=4.76, 95%
CI=0.57-15.7351) and laryngeal cancer mortality (SMR=1.35, 95%CI=0.03-7.5352) but these
findings were based on small numbers (2 cases of ovary and 1 case of laryngeal cancer) and were
statistically non-significant. Lung cancer mortality was also increased but not statistically
significant (SMR=1.46, 90%CI=0.79-2.48).
Germani et al. (1999) conducted a cohort mortality study of 631 Italian women who were
compensated for asbestosis and alive on December 31, 1979. The women were followed up for
the ascertainment of vital status until October 30, 1997. SMRs were estimated using lifetable
methods and national rates as the referent population. A statistically significant increase in
ovarian cancer mortality was observed (SMR=4.77, 95%CI=2.18-9.06). Only one case of
laryngeal cancer was observed in this study, which was an excess, but this was a highly unstable
(SMR=8.09, 95%CI=0.21-45.0853). A large and statistically significant excess of lung cancer
mortality (SMR=4.83, 95%CI=2.76-4.84) was observed.
Liddell et al. (1997) conducted a retrospective cohort mortality study of 11788 men who were
born between 1891-1920 and had worked for at least one month in the chrysotile asbestos mines
and mills in Quebec. Follow-up of the cohort for vital status ascertainment was through 1992. A
small excess of laryngeal (SMR=1.11, 95%CI=0.78-1.5454), and a modest but statistically
significant increase in lung cancer mortality (SMR=1.37, 95%CI=1.27-1.4855) was observed in
this study. The study only included men and thus did not provide any results for ovarian cancer.
Malmo and Costa (2004) conducted a cohort mortality study of 1653 former workers who were
hired before 1971 in an Italian textile plant that only used chrysotile asbestos. The cohort was
followed for vital status ascertainment through January 1, 1981. SMRs were estimated using life-
table analyses and rates from residents of Turin who listed manual employment during a census
in 1981 were used as the referent. Only one case of ovarian cancer was observed in this study
which represented a small and statistically non-significant excess (SMR=1.28, 95%CI=.02-7.12).
A statistically non-significant excess of laryngeal cancer was observed among males
(SMR=4.44, 95%CI=0.90-12.97), but no cases were observed among females (the expected
number was not reported). A statistically significant and relatively large increase in lung cancer
mortality was observed in both males (SMR=3.02, 95%CI=1.89-4.57) and females (SMR=5.23,
95%CI=2.10-10.79).
Pira et al. (2017) conducted a retrospective cohort mortality study of 1056 men who were
employed for at least one year in an Italian chrysotile asbestos mine between 1930 and 1990.
Follow-up of the cohort for vital status ascertainment was through 2014. SMRs were estimated
using lifetable methods and national rates for before 1981 and rates from the Piedmont region
where the mine was located from 1981 onward were used as the referent. A statistically non-
significant excess of laryngeal (SMR=1.58, 95%CI=0.68-3.11) and of lung cancer (SMR=1.16,
51	Confidence interval was estimated using Fisher Exact method.
52	Confidence interval was estimated using Fisher Exact method.
53	Confidence interval was estimated using Fisher Exact method.
54	Confidence interval was estimated using Fisher Exact method.
55	Confidence interval was estimated using Fisher Exact method.
Page 349 of 352

-------
2953	95%CI=0.87-1.52) was observed. The study only included men and thus did not provide any
2954	results for ovarian cancer.
2955	Uncertainty in Approach
2956	An uncertainty related to this approach is that these adjustment factors are treated as a constant
2957	when in fact the ratio varies substantially between the studies. This variation may be just random
2958	but may be likely due to differences in study design or to levels of exposure to chrysotile
2959	asbestos.
2960	Conclusions
2961	The adjustment factor may be used to upwardly adjust the unit risk estimates for lung cancer to
2962	take into account the biases resulting from not including cancer sites other than lung cancer and
2963	mesothelioma.
Page 350 of 352

-------
Table Apx M-l: Estimate of adjustment factor for ovarian cancer
1st Author (Year)
Lung Cancer56
Ovarian Cancer
SMR
Observed
Expected
Obs-Exp
SMR
Observed
Expected
Obs-Exp
Acheson et al.








(i r>S2)
1.25
6
4.8
1.2
1.48
5
3.4
1.6
Gardner and








Powell (1986")
1.42
6
4.2
1.8
1.11
3
2.7
0.3
Newhouse and








Sullivan (1989}
0.57
12
21.1
-9.1
1.08
11
10.1
0.9
Tarchi et al.








14)
4.14
2
0.48
1.52
4.76
2
0.42
1.58
Germani et al.








>9)
4.83
16
3.31
12.69
5.26
4
0.76
3.24
Malmo and Costa








(2004)
5.23
7
1.34
5.66
1.28
1
0.78
0.22
Hein et al. (2007}
2.22
61
27.48
33.52
0.62
6
9.68
-3.68
* i' mis et al.








(2009)
1.96
277
141.66
135.34
1.23
9
7.34
1.66
Wane et al.








( )
1.23
2
1.62
0.38
7.69
1
0.13
0.87










Sum=
389
205.99
183.01
Sum=
42
35.31
6.69












Adjustment Factor =1.04




56 Lung cancer results are for women
Page 351 of 352

-------
Table Apx M-2: Adjustment factor for laryngeal cancer57
1st Author
(Year)
Lun
g Cancer58
Laryngeal Cancer
SMR
Observed
Expected
Obs-Exp
SMR
Observed
Expected
Obs-Exp
Gardner and
Powell (1986)
0.97
41
42.4
-1.4
0.91
1
1.1
-0.1
Newhouse and
Sullivan CI989)
0.99
241
242.5
-1.5
0.64
6
9.4
-3.4
Tarchi et al.
>4)
1.46
10
6.84
3.16
1.35
1
0.74
0.26
Liddell et al.
)
1.37
646
471.5
174.5
1.11
36
32.43
3.57
German! et al.
)
4.83
16
3.31
12.69
8.09
1
0.12
0.88
Malmo and Costa
(2004)
3.36
29
8.62
20.38
4.44
3
0.68
2.32
Hein et al. (2007)
1.95
198
101.7
96.3
1.68
6
3.6
2.4
Loom is et al.
(2009)
1.96
277
141.66
135.44
1.15
6
5.21
0.79
Wane et al.
( )
3.76
55
14.62
40.38
4.08
2
0.49
1.51
Pira et al. (2017)
1.16
53
45.5
7.5
1.58
8
5.1
2.9










Sum=
1653
1101.07
552.03
Sum=
71
59.86
11.14












Adjustment Factor
=1.02




57	Foreign language study by Sun et al. (2003) that was not fully translated and was included as a sensitivity anlysis, but the adjustment factor did not change. Only one
case of laryngeal cancer was reported which was close to the expected value (SMR=1.01, 95%CI=0.14-7.17). Results for ovarian cancer were not reported. A statistically
significant excess of lung cancer was observed in this study (SMR=3.88, 95%CI=3.14-4.79).
58	Lung cancer rates are for men and women combined
Page 352 of 352

-------